Champagne, Elaine T - Rice - Chemistry and Technology-Amer Assn of Cereal Chemists (2006) PDF

Cover photographs courtesy of U.S. Department of Agriculture, Agricultural

Research Service. Front: U.S. long-grain rice; photo by Keith Weller. Back: rice
harvesting in Fort Bend County, TX; photo by David Nance.
Reference in this publication to a trademark, proprietary product,

or company name by personnel of the U.S. Department of Agriculture
or anyone else is intended for explicit description only and does not
imply approval or recommendation to the exclusion of others
that may be suitable.
Library of Congress Catalog Card Number: 2003115879

International Standard Book Number: 1-891127-34-9
© 1972,1985, 2004 by the American Association of Cereal Chemists, Inc.

Published 1972. Third Edition 2004
All rights reserved.

No part of this book may be reproduced in any form, including photocopy,
microfilm, information storage and retrieval system, computer database or
software, or by any other means, including electronic or mechanical,
without written permission from the publisher.
Copyright is not claimed in any portion of this work written by United States
government employees as part of their official duties.
Printed in the United States of America on acid-free paper.
American Association of Cereal Chemists, Inc.

3340 Pilot Knob Road
St. Paul, Minnesota 55121-2097, U.S.A.

CONTRIBUTORS
Donald B. Bechtel, U.S. Department of Agriculture–Agricultural Research Service,

Grain Marketing Research Laboratory, Manhattan, KS
C. J. Bergman, U.S. Department of Agriculture–Agricultural Research Service,
Rice Research Unit, Beaumont, TX
Kshirod R. Bhattacharya, Rice Research and Development Centre, Mysore, India
Norris Bond, Satake USA, Inc., Houston, TX
David L. Calderwood (retired), U.S. Department of Agriculture–Agricultural
Research Service, Beaumont, TX
Elaine T. Champagne, U.S. Department of Agriculture–Agricultural Research
Service, Southern Regional Research Center, New Orleans, LA
Nathan W. Childs, U.S. Department of Agriculture–Economic Research Service,
Washington, DC
Robert R. Cogburn (retired), Hamshire, TX
Jennifer Eastman, ACH Food and Nutrition, Memphis, TN
Melissa Fitzgerald, NSW Agriculture, Yanco, NSW, Australia
J. H. Gibbons, Department of Agronomy, Rice Research and Extension Center,
University of Arkansas, Stuttgart, AR
J. Samuel Godber, Department of Food Science, Louisiana State University
Agricultural Center, Baton Rouge, LA
Diane Wright Hoffpauer, Light Heart, LLC, Crowley, LA
Terry A. Howell, Jr., McKee Foods Corporation, Collegedale, TN
Bienvenido O. Juliano, Philippine Rice Research Institute Los Baños, Laguna,
Philippines
Otto R. Kunze (retired), Texas Agricultural Experiment Station and Texas A&M
University, College Station, TX
Yubin Lan, Agricultural Engineering Technology, Fort Valley State University,
Fort Valley, GA
Wayne E. Marshall, U.S. Department of Agriculture–Agricultural Research
Service, Southern Regional Research Center, New Orleans, LA
Anna Myers McClung, U.S. Department of Agriculture–Agricultural Research
Service, Rice Research Unit, Beaumont, TX
K. S. McKenzie, California Cooperative Rice Research Foundation, Biggs, CA
Jean-Francois Meullenet, Rice Processing Program, University of Arkansas,
Fayetteville, AR
K. A. K. Moldenhauer, Department of Agronomy, Rice Research and Extension
Center, University of Arkansas, Stuttgart, AR
Yutaka Ogawa, Kirin Beer Co. Ltd., Yokohama, Japan
K. Ohtsubo, National Food Research Institute, Tsukuba Science City, Ibaraki,
Japan
Frank T. Orthoefer, Food Sciences and Technologies, LLC, Germantown, TN
Frederick F. Shih, U.S. Department of Agriculture–Agricultural Research Service,
Southern Regional Research Center, New Orleans, LA
Terry J. Siebenmorgen, Rice Processing Program, University of Arkansas,
Fayetteville, AR
Heidi C. Wilkinson, USA Rice Federation, Houston, TX
Delilah F. Wood, U.S. Department of Agriculture–Agricultural Research Service,
Western Regional Research Center, Albany, CA
Finis T. Wratten (deceased), Agricultural Engineering Department, Louisiana
State University, Baton Rouge, LA
An-I Yeh, Graduate Institute of Food Science & Technology, National Taiwan
University, Taipei, Taiwan
Wallace Yokoyama, U.S. Department of Agriculture–Agricultural Research
Service, Western Regional Research Center, Albany, CA
Kiyoshi Yoshizawa, Tokyo University of Agriculture, Tokyo, Japan
PREFACE TO THE THIRD EDITION
Rice has taken center stage this last decade, not only as an important provider of
nourishment for the world’s population, but as a grain now recognized as having
many unique nutritional and functional attributes with potential to be captured in a
multitude of value-added food and nonfood applications. Basic, up-to-date
knowledge of rice chemistry and technology is needed to guide the research that
will develop new applications and lead rice into the coming decades.
The third edition of Rice: Chemistry and Technology updates the 1985
monograph, with emphasis on current developments. The book presents, in a single
work, comprehensive overviews covering topics ranging from the rice plant and
varieties to rice structure and composition and the functionality of its components.
Postharvest processing technologies for drying, storage, and milling and those for
making traditional and new value-added products are discussed in detail. New
nutritional findings are presented.
A multiauthored work, each chapter of the monograph has been written by one
or more authorities on the subject. The authors have styled their chapters as
overviews, with extensive bibliographies directing the reader to the primary
literature. This monograph is intended to be an addition to your collection and not a
replacement for the second edition, which covers the older literature in more detail.
I wish to thank the authors for their time and talent in writing the chapters. The
authors and I acknowledge and dedicate this monograph to the godfather of the rice
world, Bienvenido O. Juliano, whose contributions to the field have guided our
research. You paved the way for us, Ben, and we thank you.
Elaine T. Champagne
PREFACE TO THE SECOND EDITION
Rice is the principal food cereal in tropical Asia, where 90% of the world’s rice
crop is grown and consumed. Significant progress in the chemistry and technology
of rice in the last decade has prompted the Publications Committee of the American
Association of Cereal Chemistry to revise its Monograph No. 4, Rice: Chemistry
and Technology, edited by Dave Houston and published in 1972.
The present monograph updates the 1972 monograph, with emphasis on
developments in the 1970s and 1980s. Chapters on parboiling and milling
emphasize the Asian situation, since the status in the United States and Europe is
adequately discussed in published chapters. Extrusion-cooked rice foods and rice
noodles are emphasized, as well as Japanese convenience foods. Rice straw is
included because of the current interest in biomass utilization. All chapters on
technology and processing were contributed by authorities on the subject.
Contributors were encouraged to emphasize varietal differences and possible topics
for future research. As editor, I elected to write most of the chemistry chapters and
those on by-products and residue, to minimize overlap and maximize coverage of
the interfaces between the chemistry and structure of the rice grain and its
technological properties. This monograph was mainly planned, written, and edited
in 1983-1984 during my sabbatical leave from the International Rice Research
Institute (IRRI).
I wish to thank the contributors for their time and talent in writing the chapters;
R. Don Sullins, then chairman of the AACC Publications Committee, for his
assistance and support; M. S. Swaminathan, Director General of IRRI, for
approving my leave; the Southeast Asian Research Center for Agriculture for
providing me a Visiting Professorship and office space; the IRRI Department of
Communications and Publications for graphics and photography; the IRRI library
staff for rechecking literature citations; Daisy Herrero for typing the manuscripts;
and my research colleagues in the Cereal Chemistry Department for reviewing my
chapters. Finally, I wish to thank my wife, Linda, and my children for their patience
and understanding during the preparation of the monograph.
B. O. J.
PREFACE TO THE FIRST EDITION
Rice, as one of the two major food cereals, provides—together with a

comparable amount of wheat—a large proportion of the total nourishment of the
world’s population. However, reported research on rice chemistry and technology
has lagged markedly behind that reported for wheat. Moreover, the available
information on rice other than on its culture has remained largely scattered in
contrast to that for wheat, which has been well collected and summarized.
The present monograph, the fourth in the Monograph Series sponsored by the
American Association of Cereal Chemists, attempts to minimize this disparity. The
book has two major aims: to collect and present for the first time in a single work an
ordered, coherent, and informative series of reviews on rice chemistry and
technology; and to provide an extensive bibliography that will permit direct access
to the primary literature. This combination offers useful data to all connected with
the handling, processing, or sale of rice and its by-products, as well as to any
individual seeking information on rice composition or technology. A third, minor,
aim is to use predominantly the metric system of measurement in accord with its
extensive scientific acceptance and the worldwide trend towards its general
adoption.
As a multiauthored work, the monograph has the advantage that each chapter is
presented by an authority on the subject—and some disadvantages that inevitably
accompany this type of publication. The dedicated efforts of the authors have
provided the values to be found in this volume; errors and omissions must be
attributed to the editor, who welcomes all corrections and suggestions for improving
any possible later edition.
Credit for initiating this work belongs to the Monograph Committee of the
Association and to Past President Byron Miller and Executive Vice-President
Raymond J. Tarleton, who put the plan into effect. My particular thanks go to
Director of Publications Merrill J. Busch, and to Assistant Editor Carolyn M. Light
and her able proofreading and typesetting co-workers, who diligently, patiently, and
cooperatively handled the multitude of details in preparing this work for
publication. I thank also my wife, Twylla, not only for her aid, but especially for her
sustained forebearance during preparation of the monograph.
D. F. H.
CONTENTS
1. Production and Utilization of Rice. NATHAN W. CHILDS, 1

Rice Compared with Other Cereal Grains, 1
Rice Ecosystems, 2
World Rice Production, 3
Production in Major Rice-Growing Regions, 5
Southeast Asia South Asia China and Northeast Asia Latin America North
America Africa Europe and the Middle East Other Regions
World Rice Trade, 12
Overview Market Structure Major Exporters Major Importers
Global Rice Consumption, 18
Overview Regional and Country Comparisons Per-Capita Analysis
Rice in the Twenty-First Century, 20
2. The Rice Plant: Growth, Development, and Genetic Improvement.

ANNA MYERS McCLUNG, 25
Growth and Development of the Rice Plant, 25
Planting, Germination, and Emergence Vegetative Growth Stage Reproductive
Stage Grainfill Stage Grain Harvest
Factors Influencing Grain Quality, 31
Cultural Management Environmental Factors
Genetic Improvement of Rice, 33
U.S. Breeding History Genetic Variability Population Structure and Breeding Line
Development Breeding Goals and Selection Methods
3. Rice Varieties. K. A. K. MOLDENHAUER, J. H. GIBBONS,

and K. S. McKENZIE, 49
Variety Origin and Classification, 50
Southern Region California
Leading Varieties, 55
Varietal Quality Characteristics, 63
4. The Rice Grain and Its Gross Composition. ELAINE T. CHAMPAGNE,
DELILAH F. WOOD, BIENVENIDO O. JULIANO, and DONALD B.
BECHTEL, 77
Gross Structure of the Rice Grain, 77
Overall Structure Hull Caryopsis Pericarp Seed
Composition of the Rice Grain and Its Milling Fractions, 88
Proximate Analysis of Rough Rice and Its Fractions Hull Bran Milled Rice
Brown Rice
Factors Affecting Composition, 96
Management and Cultural Practices Soil Climate Location on Panicle
Genotypic Differences Processing
Summary, 100
5. Starch. MELISSA FITZGERALD, 109

The Starch Granule, 109
Macrostructure of the Starch Granule Microstructure of the Components
Functional Properties of Starch, 115
Glass Transition Gelatinization Swelling and Pasting Retrogradation
The Enzymes of Starch Synthesis, 118
Adenosine 5-Diphosphatase Glucose Pyrophosphorylase The Starch Synthases
The Branching Enzymes The Debranching Enzymes The Disproportionating
Enzyme
Conclusion, 133
6. Rice Proteins. FREDERICK F. SHIH, 143

Protein Distribution and Composition, 143
Protein Bodies Protein Fractions Identification of Rice Varieties Protein Mutants
Processing of Protein Products, 149
Bran Proteins Endosperm Proteins Functional Properties Utilization
Nutritional Properties, 153
Effect of Processing Digestibility Protein Quality Hypoallergenicity
Conclusion, 157
7. Rice Lipids. J. SAMUEL GODBER and BIENVENIDO O. JULIANO,

163
Extraction of Lipids, 164
Major Lipid Categories, 164
Starch Lipids Influence of Starch Purification Nonstarch Lipids
Specific Classes Of Lipids, 173
Ferulate Esters: “Oryzanol Fraction” Glycosyl Glycerides Sphingolipids Sterol
Lipids Tocopherols and Tocotrienols Wax
Factors Affecting Lipid Composition, 183
Summary, 186
8. Physical and Mechanical Properties of Rice. OTTO R. KUNZE,
YUBIN LAN, and FINIS T. WRATTEN, 191
Grain Dimensions, 193
Mass Properties, 195
Density and Porosity Coefficient of Thermal Expansion Coefficient of
Hygroscopic Expansion Static and Dynamic Coefficients of Friction Resistance to
Airflow Angle of Repose
Thermal Properties, 199
Specific Heat Thermal Conductivity and Diffusivity Film Heat Transfer
Coefficient Glass Transition Temperature
Hygroscopic Properties, 201
Equilibrium Moisture Content Moisture Movement into the Grain Hygroscopic
Conductivity and Diffusivity
Mechanical Properties, 206
Tensile Strength Compressive Strength Modulus of Elasticity Modulus of
Toughness Hardness
Moisture Adsorption Properties—Consequences, 211
Before Harvest During Harvest During Drying After Drying
Weight–Volume Relationships, 218
Summary and Conclusions, 218
9. Rough-Rice Drying—Moisture Adsorption and Desorption. OTTO R.

KUNZE and DAVID L. CALDERWOOD, 223
Selected Historical Aspects, 224
Rice in the Field, 229
Optimum Harvest Moisture, 236
Fundamentals of Rice Drying, 238
Modeling Grain Moisture, 245
Gradients Stresses
Commercial Rice Drying, 248
Handling Undried Rice Storage of Undried Rice Continuous-Flow Dryers
Multipass Drying Tempering Supplemental Drying with Aeration Dryer
Adjustments Preheat Drying Combination System
Bin Drying, 256
Equipment Airflow Requirements Depth of Rice Fan-Operating Schedule Air
Delivery and Fan Power Supplemental Heat Stirring Augers Handling of
Materials
Corn and Small-Grain Dryers, 261
Other Drying Methods, 261
Fluid-Bed Drying Infrared Drying Dielectric Drying Sack Drying
Drying in Countries Other than the United States, 262
Conclusions, 263
10. Rough-Rice Storage. TERRY A. HOWELL, JR., and ROBERT R.
COGBURN, 269
Storage Technology, 269
Structures Management Practices
Maintenance of Rice Quality, 275
Aging Protection of Rice from Degradation
Summary, 282
11. Rice Milling. NORRIS BOND, 283

Definition, 283
Relationship to Other Processes, 283
Stages of the Milling Process, 284
Cleaning Brown Rice Production Bran Removal Classification—Removal of
Broken Grains
Controlling the Rice Milling Process, 300
12. Impact of Drying, Storage, and Milling on Rice Quality and

Functionality. TERRY J. SIEBENMORGEN and JEAN-FRANCOIS
MEULLENET, 301
Pre-Drying Issues, 301
Respiration Factors Affecting Respiration Quality Effects Resulting from
Respiration Wet-Holding Effects on Rice Functionality
Drying, 307
Effects of Drying on Milling Quality Effects of Drying on Sensory Quality Effects
of Drying on Other Functional Characteristics
Storage, 311
Effects of Storage Conditions on Functional Properties of Rice Effects of Storage
Conditions on Sensory Properties of Rice
Milling, 313
Kernel Size Uniformity Milling System Effects on Kernel Quality Degree of
Milling Effects on Functionality Milling Effects on Sensory Quality Postmilling
Physical Quality Reduction
Summary, 325
13. Parboiling of Rice. KSHIROD R. BHATTACHARYA, 329

History, 329
Origin of Parboiling Discovery of the Nutritional Benefit of Parboiling
Modernization of the Parboiling Industry Present Status of the Parboiling Industry in
the World
The Process: Production of Parboiled Rice, 338
Processing Conditions Production Systems
Changes During Parboiling, 359
Changes During Soaking Changes in Organized Constituents Changes in Other
Constituents Other Changes
The Product and Its Characteristics, 367
Physical Properties Qualities of Parboiled Rice Relevant to Consumers
Fundamental Nature of Parboiled Rice, 375
Physicochemical Properties State of Starch in Parboiled Rice Effect of Rice
Variety on Properties of Parboiled Rice Tests for Parboiled Rice
Products from Parboiled Rice, 392
Canned Rice Puffed Rice Flaked Rice
The Future, 394
14. Enrichment and Fortification of Rice. DIANE WRIGHT HOFFPAUER,

405
History of Regulations, 405
Current Regulations, 406
Impact of Folic Acid Fortification, 408
Rice Enrichment Processes, 408
Powder Enrichment Coated-Kernel Enrichment Extruded Kernels Application of
Folic Acid
Testing Enrichment and Fortification Levels, 410
Global Rice Fortification, 411
Conclusion, 413
15. Rice End-Use Quality Analysis. C. J. BERGMAN, K. R.

BHATTACHARYA, and K. OHTSUBO, 415
Physical Properties, 418
Milling Quality Kernel Dimensions Grain Color
Functional Properties, 423
Hydration During Cooking Loss of Solids During Cooking Gelatinization
Temperature Gel Consistency Pasting Characteristics Elongation Ratio
Instrumental Measurement of Cooked-Rice Texture Cooking of Rice for Sensory or
Instrumental Testing
Biochemical Properties, 441
Amylose Content Amylopectin and Amylose Characteristics Protein and Lipid
Content Aroma
General Techniques, 453
Near-Infrared Spectroscopy Genetic Markers
Future Research Needs, 460
16. Value-Added Rice Products. HEIDI C. WILKINSON

and ELAINE T. CHAMPAGNE, 473
Grain Types, 473
Brown Rice Regular Milled White Rice Sweet or Waxy Rice Aromatic Rice
Arborio Rice
Forms of Rice, 476
Unprocessed Forms Processed Forms
Components and Co-Products, 480
Rice in Processed Foods, 482
Breakfast Cereals Snacks Side Dishes and Packaged Mixes Entrées Soups
Desserts Baby Food Beer Pet Foods
Unusual Applications, 489
Meat and Analogs Milk and Dairy Alternatives
Conclusion, 490
17. Preparation and Applications of Rice Flour. AN-I YEH, 495

Preparation, 497
High-Protein Rice Flour Brown Rice Flour Japanese Specialty Flours
Functional Properties, 502
Effect of Inherent Properties Storage Effect Milling Effect
Applications, 514
Noodles Cakes Crackers Baked Products Breakfast Cereal and Baby Foods
Rice Milk and Beverages Miscellaneous Products
Summary, 534
18. Rice in Brewing. KIYOSHI YOSHIZAWA and YUTAKA OGAWA, 541

Rice in Sake Production, 541
Manufacture of Sake Rice as a Principal Ingredient Nature of Rice for Sake
Brewing Estimation of Quality of Rice as an Ingredient Rice Processing
Alternatives in Rice Processing
Rice in Beer Production, 557
Raw Materials for Beer Production Manufacture of Beer Rice as an Adjunct
19. Rice Bran and Oil. FRANK T. ORTHOEFER and JENNIFER

EASTMAN, 569
Milling, 569
Rice Bran, 570
Chemical Composition Composition of Rice Germ Factors Affecting the
Composition of Bran Antinutritional Factors in Bran Stabilization Nutritional
Properties of Rice Bran
Rice Bran Oil, 581
Oil Extraction Rice Bran Oil Processing Refined Rice Bran Oil Rice Bran Oil
Nutrition Rice Bran Oil Utilization
Summary, 591
20. Nutritional Properties of Rice and Rice Bran. WALLACE

YOKOYAMA, 595
Glycemic Properties, 595
Milled Rice Rice Bran
Lipemic Properties, 597
Milled Rice Rice Bran
Hypocholesterolemic Properties, 598
Rice Bran Rice Bran Oil
Other Components of Rice, 605
Effects on Cholesterol Effects on Cancer
Conclusions, 606
21. Utilization of Rice Hull and Rice Straw as Adsorbents. WAYNE E.

MARSHALL, 611
Utilization of Rice Hulls and Straw, 613
As a Char or an Activated Carbon As Noncarbonized Material As Rice Hull Ash
Summary, 627
Index, 631
CHAPTER 1
PRODUCTION AND UTILIZATION OF RICE
Nathan W. Childs
U.S. Department of Agriculture
Economic Research Service
Washington, DC
Rice (Oryza sativa L.) is one of the leading food crops in the world and the sta-
ple food for more than half the world’s population. It is generally considered a
semiaquatic, annual grass plant. Cultivars of the two cultivated species, O. sativa L.
and O. glaberrima Steud., can grow in a wide range of water-soil regimes, from
deeply flooded land to dry, hilly slopes (Luh, 1991b). Because of its long history of
cultivation and selection under diverse environments, remarkable diversity exists in
rice. The grain is grown in more than 100 countries on every continent except Ant-
arctica, extending from 53° north to 40° south and from sea level to 3,000 m above
sea level. However, O. glaberrima is grown only in Africa and only on a limited
scale. The production practices for rice in various countries range from extremely
primitive to highly mechanized (Luh, 1991b).
Detailed descriptions of various aspects of rice production and utilization have
been published by Grist (1975), De Datta (1981), Juliano (1985), and Luh
(1991a,b).
RICE COMPARED WITH OTHER CEREAL GRAINS
Annual production of rough rice approaches that of wheat and corn (USDA,
2001a) (Table 1). Rice and wheat are the two most important food grains, since corn
is used mainly as feed except in Latin America and parts of Asia. Rice is the staple
food of East, Southeast, and South Asia, with the region as a whole accounting for
more than 90% of global production (Table 1) and more than 88% of consumption
(Luh, 1991b).
Mean paddy yields of rice are second only to those of corn (Table 2). Cereal
yields are generally higher in temperate areas such as Japan, Korea, Europe, the
United States, Australia, Egypt, Argentina, and Uruguay than in the tropical rice-
growing countries such as Indonesia, Burma, Thailand, much of India, and the
Philippines.
When the actual extraction rates of the cereals (the fraction of each grain utilized
as food) are considered (Luh, 1991b), rice is calculated to produce more food
energy per hectare than the other cereals (Eggum, 1979; FAO, 2001). Total food
protein production per hectare is also high for rice, second only to that for wheat.
1
2 / Rice: Chemistry and Technology, 3rd ed.
TABLE 1
Annual Production of Cereal Crops by Region: Average 1998–2000a
Cereal Grains (1,000 t)
Rough Sor-
Area Wheat Rice Corn ghum Millet Barley Rye Oats Total
North Americab 93.3 9.2 273.4 19.6 0.0 20.4 0.6 6.0 422.5
Latin Americac 19.4 21.5 59.5 4.9 0.1 1.7 0.1 0.9 107.9
Western
Europe 102.5 2.6 37.2 0.6 0.1 51.6 5.8 6.8 207.2
Eastern Europe
and former
Soviet Union 92.9 1.2 30.2 0.0 1.4 32.9 13.0 9.2 180.7
Middle East 32.9 2.7 3.3 0.6 0.1 10.8 0.2 0.3 50.8
Africa 16.9 16.5 40.1 17.4 14.5 3.9 0.0 0.1 109.5
Asia 204.2 538.8 155.4 12.7 12.4 5.4 0.0 0.6 929.5
Oceania 22.8 1.4 0.5 1.9 0.0 5.9 0.0 1.5 34.0
World total 584.8 593.9 599.7 57.6 28.4 132.4 19.7 25.3 2,041.9
a Data from USDA (2001b).
b Canada, United States, and Mexico.
c Includes Central America, South America, and the Caribbean.
TABLE 2
Mean Grain Yield (t/ha) of Cereal Crops by Region: Mean for 1998–2000a
Cereal Grains
Rough Sor-
Wheat Rice Corn ghum Millet Barley Rye Oats
North America b 2.76 6.53 7.24 3.79 N/Ac 3.08 1.99 2.36
Latin Americad 2.52 3.61 2.92 3.00 1.07 1.84 1.21 1.36
Western Europe 5.85 6.52 8.86 3.12 1.00 4.59 4.51 3.31
Eastern Europe and
former Soviet
Union 1.77 2.78 3.28 1.00 0.71 4.23 1.66 1.27
Middle East 1.80 3.85 3.54 1.01 1.21 1.59 1.46 1.75
Africa 1.84 2.29 1.65 0.85 0.66 0.79 0.15 0.17
Asia 2.83 3.94 3.74 1.08 0.88 2.46 N/A 1.20
Oceania 1.89 8.98 6.13 3.06 1.00 2.00 0.57 1.91
World average 2.66 3.88 4.33 1.43 0.75 2.33 2.03 1.72
a Data from USDA, (2001b).
b Canada, United States, and Mexico.
c Not applicable.
d Includes Central America, South America, and the Caribbean.
When the superior quality of rice protein is considered, the yield of utilizable pro-
tein is actually higher for rice than for wheat.
RICE ECOSYSTEMS
The ecosystems within which rice is grown are characterized by elevation,

rainfall pattern, depth of flooding and drainage, and adaptation of rice to the
agroecological factors. Riceland ecosystems can be divided into four broad
categories—irrigated, rainfed lowland, upland (or dryland), and flood-prone
(IRRI, 1997).
Production and Utilization / 3
IRRIGATED RICE
In 1991, irrigated rice ecosystems (in which water is added to supplement water
supplied by natural processes) accounted for almost 55% of the total rice-harvested
area of 147.2 million hectares and contributed more than 75% of global rice pro-
duction. Irrigated rice production is concentrated mostly in the humid and subhumid
subtropics and in the humid tropics. Generally, farmers of irrigated land use more
purchased inputs than farmers of nonirrigated lands, and they typically achieve
higher yields.
RAINFED LOWLAND RICE

Rainfed lowland rice is transplanted or direct-seeded in puddled soil on level to
slightly sloping, bunded (diked) fields that are flooded for at least part of the crop-
ping season. Production is characterized by a lack of water control, with floods and
drought being potential problems. Farmers grow traditional, photoperiod-sensitive
varieties and rely heavily on labor instead of purchasing inputs. Nearly 25% of the
global rice area was rainfed lowlands in 1991, accounting for 17% of world pro-
duction. Almost 90% of rainfed rice production occurs in South and Southeast Asia.
UPLAND RICE
Upland rice (rice grown without standing water) is grown in Asia, Africa, and
Latin America and accounted for almost 13% of the total harvested rice land in
1991. Although upland rice constitutes a relatively small share of the total rice area,
it is the dominant rice culture in Latin America and West Africa. Because of
extremely low yields, upland rice accounted for just 4% of global rice production.
FLOOD-PRONE RICE
Flood-prone rice ecosystems (which include both deep-water rice and floating
rice) accounted for less than 9% of the total rice area and about 4% of the world
production in 1991. About 11.4 million hectares of rice land in South and Southeast
Asia are subject to various types of uncontrolled flooding. West Africa and Latin
America also have some flood-prone rice land. Rice yields from flood-prone eco-
systems are low, about 1.5 tonnes (t) per hectare, and extremely variable because of
problem soils and unpredictable combinations of drought and flood. About 90% of
the flood-prone rice area is located in Asia.
WORLD RICE PRODUCTION
Global rice production was estimated at 588.8 million tonnes (paddy basis) in
2000/01, an increase of 13% from 1990/91 but nearly 2 percentage points behind
the growth in consumption during the decade (Fig. 1) (USDA, 2001b). Average
rough-rice yields rose more than 10% in the 1990s to a record 3.9 t/ha by 2000/01,
accounting for the bulk of the production increase. Harvested area expanded slightly
more than 3% to 151.9 million hectares.
The rate of production growth in the 1990s was well below rates achieved in
each of the previous three decades, primarily due to slower yield growth. Global
rice production expanded 31% in the 1980s, almost entirely due to rising yields; the
harvested area expanded only fractionally. Production growth outpaced consump-
tion in the 1980s, leading to a significant increase in global stocks. In the 1970s,
global rice production rose 27%, almost 4 percentage points behind the rate of con-
sumption growth. The rice-growing area expanded nearly 9%, and average yields
rose almost 17%. In the 1960s, global production rose more than a third, a result of
an almost 11% increase in harvested area and a 21% boost in average yield. How-
ever, production growth barely kept pace with rising demand in the 1960s, even
with the large expansion in area and higher yields.
The strong yield growth achieved from the mid-1960s through the 1980s was
largely due to the rapid rate of adoption of high-yielding varieties in much of Asia,
as well as increased plantings of hybrid rice in China. By the 1990s, most of the
irrigated rice lands in Asia were growing high-yielding varieties. Development of
high-yielding varieties for less-favorable ecosystems has lagged behind the devel-
opment for irrigated systems (IRRI, 1997).
India, Thailand, and Vietnam, all major rice exporters, accounted for the bulk of
the expansion in global rice area and production in the 1990s. India increased rice
production 17% in the 1990s, well behind a 39% increase in the 1980s. Higher
yields accounted for the bulk of the increase in both decades. Vietnam’s rice pro-
duction expanded 65% and Thailand’s 46% during the 1990s, with both area and
yield rising in each country. Production grew at a faster rate in the 1990s than in the
1980s for these two exporters.
In contrast, rice area and production declined in some major rice-producing
countries in the 1990s. Among major producers, Japan reported the sharpest drop in
area and production in that decade, a result of policies aimed at diverting rice land
to other uses. Rice area continued to decline in China during the 1990s as well,
although rising yields boosted production. Outside of Asia, rice plantings continued
to decline in Brazil and Mexico during the 1990s, a result of policies aimed at
shifting land to more profitable crops or alternative enterprises.
Four types of rice—indica, japonica, aromatic, and glutinous—account for
nearly all of global rice production. Indica is estimated to be grown on more than
80% of rice land and is the dominant type of rice produced globally. It is mostly
grown in the tropics and subtropics. It cooks fluffy, dry, and separate and is usually
more slender than japonica rice. Japonica rice is typically grown in more temperate
Fig. 1. Global rice production, consumption, and ending stocks, 1961/62 to 2000/01. (Data from
USDA, 2001b)
climates (such as Japan, northern China, Europe, and Australia) and typically cooks
moist and clingy. It accounts for about 15% of global rice production and typically
achieves higher yields than indica. Aromatic rices (primarily basmati and jasmine)
are noted for their fragrant taste and smell. They account for about 1% of total rice
production. Glutinous rice, which accounts for most of the remainder, is grown
mostly in Southeast Asia. It is used in desserts, pastes, and ceremonial dishes.
Rice production is heavily concentrated in Asia, with just four countries (China,
India, Indonesia, and Bangladesh) accounting for nearly 70% of global production.
These four countries are the largest rice-consuming countries as well. While China
produces both indica (mostly in the south) and japonica (mostly in the north), the
other three grow primarily indica rice. In addition to China, the other major produc-
ers of japonica rice are Japan, South Korea, North Korea, Taiwan, the European
Union (EU), Australia, the United States (mostly in California), and Egypt. Other
major producers of indica rice include Vietnam, Thailand, Burma, the Philippines,
Brazil, the southern United States, Pakistan, and Malaysia.
China is the largest rice-producing country, accounting for 32% of the global
production from 20% of the world’s rice-growing area. India, the second-largest
producer, has the largest rice-growing area, 29% of the total, and accounts for 22%
of global rice production (USDA, 2001b). The rest of Asia, the Americas, and sub-
Saharan Africa account for 37, 5, and 3% of global rice production, respectively. In
terms of cultivated rice area, their respective shares are 40, 5, and 5%.
PRODUCTION IN MAJOR RICE-GROWING REGIONS
Southeast Asia
Rice production in Southeast Asia increased more than a third during the 1990s,
well ahead of the global average, with higher yields and expanded area each con-
tributing about equally (Table 3). In the 1980s, production expanded 28%, mostly
due to higher yields; area rose only slightly. The region accounts for almost a fourth
of global rice production and is the largest source of exports. Production growth was
stronger in the 1990s for the region’s exporters than importers. Thailand and Viet-
nam, the world’s largest rice exporters, reported the strongest production growth in
the region and accounted for most of the area expansion.
Thailand expanded production more than 46% during the 1990s, with both area
and yield rising (Fig. 2). Rising exports was a main reason for Thailand’s area
expansion. In the 1980s, Thailand expanded production only 9%, with area nearly
stagnant. Vietnam’s production rose 65% from 1990 to 2000, with both area and
yield rising significantly. Vietnam reemerged as a major rice exporter in the late
1980s after decades of war and political upheaval that had kept it out of the global
export market. Production growth exceeded consumption in both countries during
the 1990s, allowing rice exports to expand over the decade.
Among major importers in the region, the Philippines reported the strongest
increase in production in the 1990s. Production rose 40%, barely ahead of con-
sumption, with yields rising slightly faster than area. This was well ahead of the
28% production expansion achieved during the 1980s. Yield growth accounted for
all of the expansion; area was virtually stagnant. Indonesia, the world’s largest rice-
importing country, increased production just 17% during the 1990s, a major factor
behind the record amount of rice imported during the decade. In contrast to most

8QLWHG
(J\SW
8QLWHG 6WDWHV
6WDWHV

H
U
D
W
F
H
K
U ,QGLD
H
S

V

H
Q &KLQD
Q
R
7

7KDLODQG :RUOG

Fig. 2. Rough rice yields: global and selected countries, 1961/62 to 2000/01. (Data from USDA,
2001b)
other countries, expanded area accounted for most of Indonesia’s rising production;
yields rose only slightly in the 1990s. Most of the area expansion occurred on
islands with lower average yields than on the main island of Java. In the 1980s, ris-
ing yields and larger plantings were responsible for a 3% increase in Indonesia’s
rice production, allowing the country to briefly achieve self-sufficiency.
Among countries producing smaller amounts of rice in the region, Cambodia
expanded production nearly 43% in the 1990s, with production almost matching its
1970 record by the end of the decade. War and political turmoil severely hurt Cam-
bodia’s rice sector in the 1970s and 1980s. In Laos and Malaysia, production grew
at a slower rate in the 1990s than the 1980s, with both countries having to rely more
on imports. Burma, the world’s largest rice-exporting country before World War II,
expanded production 25% in the 1990s yet failed to keep pace with consumption.
Larger area accounted for nearly all of the Burma’s production growth; yields were
virtually stagnant.
South Asia
South Asia accounts for about 30% of global rice production, and the region is a
net exporter, accounting for around 14% of total exports (Table 3). Rice production
expanded about 20% in South Asia from 1990 to 2000, with rising yields accounting
for three-fourths of the expansion. However, this was well below the 34% growth in
production achieved in the 1980s, primarily due to much weaker yield growth. India
and Pakistan are major exporters, while Bangladesh accounts for the bulk of the
region’s imports. Sri Lanka, Nepal, and Afghanistan import smaller amounts of
rice.
India accounted for the largest share of the region’s production growth in the
1990s, with India’s total rice production rising 17%, outpacing consumption. Yield
growth accounted for two-thirds of the production expansion. By the end of the
1990s, India had accumulated huge stocks of rice. In the 1980s, India’s production
expanded more than 38%, slightly ahead of domestic use, allowing India to expand
exports. India’s yield growth was much stronger in the 1980s than in the 1990s.
(India was an early adopter of high-yielding varieties.) Pakistan expanded rice pro-
duction more than 32% during the 1990s, with yields rising more than 24% and
harvested area up 7%. Production rose faster than consumption in Pakistan, allow-
ing it to expand exports substantially. In the 1980s, production growth was much
weaker for Pakistan, as yields were virtually stagnant during the decade.
Strong yield growth allowed Bangladesh to expand rice production 34% from
1990 to 2000, even as area expanded only fractionally. Production growth outpaced
consumption over the decade, allowing Bangladesh to reduce imports substantially
by 2000. In the 1980s, strong yield growth was responsible for a 40% increase in
Bangladesh’s rice production. The country’s rice area has expanded only slightly
since the late 1960s. In Sri Lanka, rice production grew just 8% from 1990 to 2000,
nearly all due to rising yields. Its rice-producing area has not expanded much since
the late 1960s. Nepal reported nearly stagnant production during the 1990s, a factor
behind rising imports. In contrast, area, yield, and production in Nepal all rose sub-
stantially in the 1980s. In Afghanistan, severe political and economic problems
caused rice area, yield, and production to drop sharply in the 1980s and 1990s. Area
has declined even further since 2000.
China and Northeast Asia
Excluding China, rice production declined 11% in Northeast Asia from 1990 to
2000, solely due to a steady decline in area (Table 3). Yields, already the highest in
Asia, rose slightly. Rice-growing area and production declined in every country in
the region. Northeast Asia currently produces less than 4% of global rice produc-
tion. Strong yield growth had pushed production up 9% in the 1980s, despite an
almost 11% reduction in area.
In the 1990s, rice production dropped almost 10% in Japan, the largest rice-
producing country in Northeast Asia (excluding China), the result of a 15% decline
in rice-growing area. Average yields rose about 6%. The government of Japan has
employed supply control measures since 1971 to reduce supplies because very small
population growth and declining per capita rice consumption have caused total use
to decline. Despite a 13% decline in harvested area, Japan’s rice production rose
nearly 8% during the 1980s, a major factor behind a large stock buildup during the
decade.
In South Korea, rice production declined nearly 9% in the 1990s, matching the
decline in consumption, as a 14% drop in plantings more than offset a higher yield.
After steadily declining since 1987, rice area started rising in South Korea in the
late 1990s. Rice stocks became extremely tight in South Korea in the mid-1990s.
Rice area dropped nearly 20% in Taiwan from 1990 to 2000, primarily in
response to government programs designed to reduce area planted to rice. Produc-
tion contracted only 13%, as yields rose slightly. Like Japan and South Korea,
Taiwan has declining per capita consumption, a result of diet diversification.
In the 1990s, rice production dropped 25% in North Korea, with both area and
yield declining. Production has continued to decline in the twenty-first century.
Adverse weather, inadequate infrastructure, government policies, and lack of inputs
are largely responsible for severe food shortages in North Korea. Despite contract-
ing rice area, rising yields allowed North Korea to increase rice production slightly
in the 1980s.
Total rice production in China rose 5% during the 1990s as reduced plantings
were offset by rising yields. Consumption expanded at a slower rate, a major factor
behind a huge buildup of stocks during the decade. Production actually declined
during the first half of the 1990s, a result of smaller plantings. However, production
rose substantially from 1995 to 1999, with record or near-record crops harvested
every year from 1997 to 1999. China’s rice area peaked in the mid-1970s and
declined for nearly 20 years before increasing again after 1995. In 1999, a major
policy change reduced or eliminated support prices for low-quality early-season
(first-crop) indica rice. This caused a sharp drop in plantings in 2000 and 2001,
resulting in a major decline in total rice production. In the 1980s, rising yields
pushed China’s rice production up 35%. Greater plantings of hybrid rice account for
much of China’s strong yield growth.
Latin America
Latin America accounts for more than 3% of global rice production and 4% of
the rice-growing area. Despite a 7% contraction in rice area, rice production in
Latin America rose more than 19% from 1990 to 2000 (Table 3), slightly exceeding
consumption growth. In the 1980s, production expanded just 12% as area contracted
more than 23%. Consumption growth outpaced production, a major factor behind
the substantial rise in imports during the 1980s. Rice plantings in Latin America
peaked in the late 1970s and have declined since, primarily due to policy reforms in
the region that reduced government support for grain production and opened mar-
kets to imports.
Latin America comprises South America, Central America, the Caribbean, and
Mexico. However, for purposes of this chapter, discussion of Mexico’s rice
sector is included in the section on North America. South America accounts for
about 92% of the rice produced in Latin America, with Central America con-
tributing more than 4% and the Caribbean more than 3%. From 1990 to 2000,
rice production rose 22% in South America, even with a 7% drop in rice area. In
the 1980s, production rose only 13%, despite strong growth in yields, due to a
22% drop in area.
Brazil, the largest non-Asian rice-producing and -consuming country, accounted
for most of the decline in rice area in the 1980s and 1990s. Area also declined in
Colombia (the second-largest rice-producing country in South America) and in
Ecuador during the 1990s. Most of the area contraction in Brazil was for low-
yielding upland rice, produced mostly in the northeast. Policy reforms in Brazil
encouraged the shift of acreage to more profitable crops in the region. Brazil’s rice
yields, which had been stagnant for decades, climbed sharply after 1980 as the
higher-yielding irrigated rice began to account for a larger share of production and
newer, higher-yielding varieties were introduced. Despite a contraction in area dur-
ing the 1990s, Brazil’s rice production rose slightly, a result of higher yields. In
Colombia, both area and yield rose about 20% during the 1990s, boosting produc-
tion more than 40%. The region’s two largest exporters (Argentina and Uruguay)
substantially increased area and production in the 1980s and 1990s, a major factor
behind greater shipments from both.
A 25% drop in area and weaker yields caused rice production to decline 29% in
the Caribbean in the 1990s. Cuba—once the largest producer in the Caribbean—saw
area and production drop substantially in the 1990s as input subsidies from the for-
mer Soviet Union ended. Rice area and production dropped in Haiti as well, a result
of policy reforms, low prices, and a lack of competitiveness.
Rice production in Central America rose 23% from 1990 to 2000, with area ris-
ing 6% and yields nearly 17%. Panama, Costa Rica, and Nicaragua, the largest rice
producers in the region, accounted for all of the increase in area and yield during the
1990s. In contrast, area and production declined in Honduras and Guatemala. Pro-
duction was stagnant in El Salvador as rising yields offset declining area. However,
both area and production in the region have declined in the twenty-first century,
with Costa Rica and Nicaragua accounting for the bulk of the decline. In the 1980s,
area dropped more than 8% in Central America, although rising yield boosted pro-
duction 25%. Costa Rica accounted for the bulk of the decline in the 1980s.
North America
North America accounts for about 1.5% of global rice production, from less
than 1% of the cultivated area. The United States accounts for the bulk of rice
production in North America. Mexico is the only other producing country. Rice
production in North America expanded 23% from 1990 to 2000, a result of a 14%
increase in average yields and a 9% increase in harvested area. The United States
accounted for all of the increase in both area and production. In the 1980s, rice
production in North America increased just 8%. Plantings dropped 17%,
offsetting much of the 27% increase in yields. Rice production was virtually
stagnant in the United States in the 1980s, as reduced plantings offset rising
yields. Policies limiting rice acreage prevented U.S. production from rising
sharply during the 1980s, even as farmers increasingly adopted higher yielding,
semidwarf varieties.
After declining sharply for well over a decade, Mexico’s rice area and produc-
tion stabilized in the late 1990s. Its rice area peaked in 1975 and its production in
1985. Policy reforms begun in the 1980s, which reduced producer support and
opened Mexico to imports, were responsible for the area decline. Much of Mexico’s
rice area, especially in the north, shifted to more profitable cropping alternatives
such as horticulture.
Africa
Africa accounts for about 3% of global rice production, and the region is a sub-
stantial net importer. Rice production in Africa grew more than 27% from 1990 to
2000 to a record 17.5 million tonnes, due almost entirely to expanded area. Yield
growth was negligible. In sub-Saharan Africa, yield growth was actually negative
over the decade. Lack of means to purchase inputs, inadequate infrastructure, and
lack of varietal development are the main factors behind the decline in average
yields. Production growth trailed consumption by a wide margin in the 1990s,
leading to record imports. In the 1980s, greater plantings and rising yields boosted
rice production in Africa more than 72%. West Africa accounts for the bulk of rice
production and consumption in sub-Saharan Africa.
Egypt, Nigeria, and Madagascar are the largest rice-producing countries in
Africa. Egypt’s rough-rice production rose nearly 90% during the 1990s (to a record
6 million tonnes) due to a large area expansion and higher yields. Egypt is the only
rice-surplus country in Africa. Production expanded by a third during the 1980s,
with higher yields accounting for most of the expansion. Egypt achieves the highest
rice yields in the world due to the varieties grown, climate, and farming practices.
Nigeria’s rice production dropped 20% from 1990 to 2000 as yields plummeted.
Harvested area actually rose 37%. Nigeria is the largest rice importer in Africa.
Rough-rice production expanded slightly in Madagascar in the 1990s, to a record
2.7 million tonnes, as higher yields were nearly offset by a decline in area. Produc-
tion growth was slightly stronger in the 1980s due to a small expansion in area.
Other major rice-producing countries in Africa include Cote d’Ivoire, Guinea, Mali,
and Zaire. While Cote d’Ivoire has the largest rice area among these four countries,
it achieves extremely low yields.
Europe and the Middle East
Producing less than 1% of global rice production, the Middle East is one of the
largest import markets. Rice production in the Middle East dropped more than
3% from 1990 to 2000, with both rice area and yield declining at about the same
rate. In the late 1990s and early 2000s, a severe drought reduced rice production
in parts of the Middle East—especially in Iran and Iraq. In contrast, consumption
in the region expanded 64% during the 1990s, a result of strong population
growth and rising incomes. In the 1980s, rising yields and greater plantings
boosted rice production 30% in the Middle East, about 7 percentage points behind
the growth in consumption. Iran accounts for about half the region’s rice
production; Iraq and Turkey produce most of the remainder. Turkey produces
mostly japonica rice; Iran and Iraq grow indica rice. Virtually all rice production
is irrigated in the Middle East.
Europe (excluding the former Soviet Union) produces less than 1% of global rice
production, with nearly all of the crop grown in southern Europe. Italy is the largest
producer, accounting for almost half of Western Europe’s rice production. Spain,
Greece, Portugal, and France account for most of the rest; extremely small amounts of
rice are grown in the Balkans. More than two-thirds of Europe’s rice production is
japonica. Indica accounts for nearly all the remainder. In the 1990s, rice production
rose slightly, with both area and yield higher by the end of the decade. Rice area
peaked in 1996 and production in 1999. Rice production in the EU began to increase
after 2000, a result of rising yields. Producer prices are well above global levels in the
EU, a result of high tariffs and government intervention buying. Virtually all the crop
is grown in irrigated fields, achieving some of the highest yields in the world.
Other Regions
The former Soviet Union and Australia produce small amounts of rice. Rice pro-
duction in the former Soviet Union has declined for 20 years, with both area and
yield contracting sharply. Lack of inputs, decaying infrastructure, major political
changes, and government policies have all contributed to the decline. In the 1990s,
rice production dropped more than 50%, primarily due to a big drop in plantings;
yields declined only slightly. Production dropped almost 14% in the 1980s, with
both area and yield declining at about the same rate. In contrast, rice production in
the former Soviet Union more than doubled during the 1970s. Larger plantings
accounted for most of the increase; average yields rose slightly. The country pro-
duces both indica and japonica rice.
Russia accounts for about half the region’s rice production. Although average
yields in Russia rose slightly from 1990 to 2000, production dropped more than a
third. Uzbekistan is the second-largest rice-producing country in the region;

Kazakhstan and Ukraine account for most of the remainder. A severe drought in
much of Central Asia that began in the late 1990s reduced area and yields.
Australia began growing rice early in the twentieth century. Most of the crop is
exported. All of the crop is grown under irrigated systems, achieving extremely
high yields. Rice production in Australia expanded more than 120% in the 1990s,
reaching nearly 1.4 million tonnes in 1999. Area expansion accounted for most of
the growth; yields, already quite high, rose slightly. In 2001, production was nearly
1.8 million tonnes, a record, as both area and yield rose substantially. Nearly all of
Australia’s crop is grown in New South Wales. Rice plantings rose sharply in
response to expanding export markets for japonica rice. Japan and Turkey substan-
tially increased imports in the second half of the 1990s. Beginning in 1995, the
Uruguay Round Agreement on Agriculture opened Japan’s rice market to annual
minimum access imports. In the 1980s, Australia’s rice production rose about 25%
as higher yields more than offset a decline in plantings. Area declined in response to
a very weak export market for japonica rice. Australia’s rice industry is very
dependent on the availability of water.
WORLD RICE TRADE
Overview
Global rice trade ranged from 16.5 million to 27.7 million tonnes (milled basis)
from 1991 to 2000, with the 1997/98 “El Niño” a major factor behind the 1998
record of 27.7 million tonnes. Global rice trade in the 1990s was almost 60% larger
than a decade earlier. Trade liberalization and rising populations were the main
factors behind the strong growth in global import demand. Trade accounted for
more than 5% of global production during the 1990s, up from less than 4% in the
1980s.
In the 1980s, total trade ranged from 11.2 million to 13.9 million tonnes,
expanding less than 10%. Self-sufficiency programs in several major importing
countries (primarily in Asia) severely limited trade growth in the 1980s. In
contrast, global trade expanded nearly 53% in the 1970s, with Africa and the
Middle East accounting for most of the growth. Higher incomes, often a result of
higher oil prices, and rising populations were responsible for much of the import
growth in the 1970s.
In the 1990s, policy changes promoted greater rice trade. The signing of the
Uruguay Round of the General Agreement on Tariffs and Trade (GATT) in 1994,
which led to the creation of the World Trade Organization (WTO), partially opened
Japan and South Korea to imports, boosting global trade. Taiwan’s rice market was
partially opened to imported rice after it joined the WTO in late 2001. In the 1980s,
many countries, especially in Latin America, began opening their markets as a
requirement of joining the WTO. In addition, numerous regional trade agreements,
such as the North American Free Trade Agreement, have encouraged greater rice
trade. Much of the momentum for recent trade liberalization had its roots in the
Uruguay Round of the GATT that began in 1986.
Even though rice trade expanded substantially in the 1990s, trade as a share of
production still remains quite small compared with those of wheat (more than 20%),
coarse grain (more than 10%), and soybeans (almost 30%). Policies aimed at self-
sufficiency, as well as import restrictions, are the main factors behind the low level
of trade for rice.
Although a net exporter, Asia is the largest import market for rice, taking 6.2
million tonnes in 2000 and accounting for more than 28% of global imports. South-
east Asia accounted for nearly two-thirds of the region’s rice imports. Asia’s rice
imports have steadily declined since the El Niño-driven 1998 record of 13.1 million
tonnes, primarily due to strong production growth by the region’s importers. Asia’s
share of global imports is substantially below its share in the 1960s and 1970s. In
the early 1960s, Asia accounted for more than 60% of global rice imports.
Africa is the second-largest import market for rice, taking a record 5.3 million
tonnes in 2000 and accounting for 23% of global imports. Sub-Saharan Africa
accounts for the bulk of the region’s imports. Rice imports by sub-Saharan Africa
nearly doubled during the 1990s, a result of strong population growth and only a
fractional increase in production. The Middle East ranks third, importing almost 4.5
million tonnes of rice in 2000, a record and more than twice the amount imported in
1990. Rising incomes and strong population growth are responsible for the import
growth. Africa and the Middle East accounted for a much larger share of global rice
imports in the 1980s and 1990s than during the previous two decades, a result of
strong import growth in both regions.
Latin America, North America, and Western Europe are the only other large
import markets for rice. In 2000, Latin America imported almost 2.5 million tonnes
of rice, or more than 11% of global imports. In 1998, severe crop damage from El
Niño drove imports to a record 3.7 million tonnes. Today, Latin America’s share of
global imports is substantially higher than during the 1960s and 1970s. Imports by
North America were nearly a million tonnes in 2000, a record and almost double
those of a decade earlier. Mexico accounts for more than half the region’s rice
imports; the United States accounts for the bulk of the remainder, with Canada
importing much smaller amounts.
Western Europe’s share of global imports averaged almost 9% during the 1990s,
down from more than 12% a decade earlier. Import growth in Western Europe is
extremely weak, a result of fractional population growth and protectionist trade
policies. The Caribbean and Central America import smaller amounts of rice.
Imports by the Caribbean more than doubled during the 1990s, exceeding 800,000 t
by 1999. Imports have continued to increase in the twenty-first century. Import
growth by Central America was even stronger in the 1990s; they reached 270,000 t
in 1999, more than doubling in just seven years. Imports have continued to expand
in the twenty-first century.
Asia accounts for the bulk of global rice exports, shipping 16.9 million tonnes, or
more than 70% of total rice exports in 2000. Asia’s share was 63% in 1990, up from
60% in 1980. Southeast Asia accounts for more than 60% of the region’s exports,
and South Asia accounts for most of the remainder. North America is the second-
largest exporting region, shipping nearly 2.8 million tonnes or nearly 12% of total
rice exports. The United States is the only rice exporter in North America. The
region showed no long-term export growth from 1980 through 2000. South America
is the third-largest exporting region, shipping about 7% of global exports, up from
less than 5% in the 1980s. Including intra-EU trade (i.e., EU-15), Western Europe
accounts for nearly 6% of global exports. However, more than two-thirds of this
rice is shipped within Europe. Egypt and Australia account for most of the remain-
der of global rice exports.
Market Structure
The international rice market is described as thin, volatile, and risky. Less than
6% of global production is currently traded annually, compared with about 18% for
wheat, more than 25% for soybeans, and almost 13% for corn. With only a small
share of production traded in global markets, a production shortfall in a major rice-
importing country can cause a large swing in international prices. Much of this
thinness is due to self-sufficiency policies in many countries that bar imports and
protect local producers (Barker, 1985). About 40% of global production is depend-
ent on the timing of the Asian monsoon, a major source of production risk and price
volatility. A major weather disturbance in South or Southeast Asia can have a very
large impact on global production and trade.
The global rice market is severely segregated by type and quality, with little sub-
stitution among buyers. In fact, tastes and preferences are so strong that prices for
various types of rice can move in opposite directions. There is little substitution in
production among the various types of rice either, as soil and climate often dictate
the type of rice that can be grown economically in any particular area. As a result,
global rice prices are typically more volatile than prices for other grains.
Indica rice is the dominant type of rice traded worldwide, accounting for nearly
75% of global trade. Japonica and aromatic rices each account for about 12%. Glu-
tinous rice accounts for the remainder. Thailand, Vietnam, China, the United States,
Pakistan, and India are the primary exporters of indica rice. Argentina, Uruguay,
Guyana, Burma, and Surinam export smaller amounts of indica as well. Australia,
Egypt, China, the EU, and the United States are the primary exporters of japonica
rice. Thailand, India, and Pakistan export the bulk of the aromatic rices, with the
United States exporting a very small amount. Thailand accounts for most of the
glutinous rice traded. In addition, the United States exports a very small amount of
glutinous rice (grown mostly in California) to Japan.
Southeast Asia, South Asia, sub-Saharan Africa, and Latin America are the pri-
mary import markets for indica rice. Northeast Asia and the Eastern Mediterranean
are the major import markets for japonica rice. Europe, the Middle East, and the
United States account for the bulk of basmati imports. China, the United States,
Hong Kong, and Singapore are the primary markets for jasmine rice. Southeast Asia
and Japan are the major import markets for glutinous rice.
Major Exporters
The global rice export market is rather concentrated. The top six exporters
(Thailand, Vietnam, the United States, China, Pakistan, and India) account for more
than 80% of world trade (Table 4). Thailand is the largest rice exporter, shipping
more than 7 million tonnes a year and accounting for 25–30% of the global rice
trade. Thailand’s exports rose almost 50% during the 1990s, and the country has the
potential to expand rice exports even more. Vietnam is typically the second-largest
exporter, with around 17% of global trade. Vietnam has expanded exports substan-
tially over the past decade, with shipments exceeding 4.5 million tonnes in 1999, a
result of strong growth in production. Vietnam, a major exporter before World War
II, was virtually out of the market from the 1950s until the late 1980s, a result of
decades of war and political turmoil. The United States is currently the third-largest
rice exporter, with about 12% of the global trade. It was the largest rice-exporting
country during most of the late 1970s and early 1980s, accounting for 20–25% of
global exports. In the 1990s, the U.S. share declined as global trade expanded, and
U.S. exports showed no long-term growth. Higher production costs in the United
States and strong growth in domestic consumption limit any expansion in U.S. rice
exports.
China is currently the fourth-largest exporting country with 9–10% of total
exports. China expanded exports during the 1990s, shipping a record 3.7 million
tonnes in 1998, a response to record global import demand and huge supplies.
China’s exports have dropped since 2000, a result of weaker global trade and
declining production. Pakistan, the fifth-largest exporter, has slowly increased rice
exports since the late 1960s. While a relatively small producer, Pakistan annually
exports more than 40% of its crop. Pakistan is one of the few Asian countries in
which rice is not a staple. It currently accounts for 9–10% of the global rice trade.
Pakistan has little ability to significantly expand rice exports. India, typically the
sixth-largest exporter, was a minor exporter of rice until the mid-1990s, when the
country exported 4.1 million tonnes in 1995. Bumper crops and strong global trade
allowed India to remain a major exporter for the remainder of the decade. After
1999, India’s exports sharply declined as domestic support prices exceeded interna-
tional levels, pricing India out of most markets. In 2001 and 2002, India regained its
market share through substantial export subsidies, currently accounts for 15–20% of
global rice trade, and now ranks second or third largest.
TABLE 4
Global Rice Trade (million tones, milled basis), 1980–2000a
Calendar Years
Country 1980 1985 1990 1995 2000
Major exporters
Thailand 2,681 3,993 3,938 5,946 6,549
Vietnam 33 60 1,670 2,314 3,370
United States 2,977 1,902 220 2,973 2,756
China 1,116 1,019 326 32 2,951
India 428 245 514 1,462 1,449
Pakistan 971 962 904 1,764 2,026
Burma 675 450 192 645 159
Other 3,800 2,784 3,936 5,664 3,612
World total 12,681 11,415 11,700 20,800 22,872
Major importers
Indonesia 2,040 53 77 3,081 1,500
Philippines 0 528 538 277 900
Bangladesh 187 226 123 1,567 638
Malaysia 167 420 298 402 596
Brazil 239 497 493 578 612
Iran 616 722 888 1,759 1,100
Iraq 379 405 388 99 1,274
Saudi Arabia 356 492 547 618 992
Nigeria 394 569 24 300 1,200
EU-15b N/A N/A 500 762 800
Other 8,303 7,503 7,824 11,357 13,260
World total 12,681 11,415 11,700 20,800 22,872
a Data from USDA (2001a,b).
b European Union. Does not include intra-EU trade.
Australia, Uruguay, Argentina, Egypt, the EU, Japan, Guyana, Surinam, and
Taiwan account for most of the rest of global exports. Uruguay, Australia, Egypt,
and Argentina are the largest. Uruguay and Australia have shipped 600,000–
700,000 t a year since the late 1990s, while Egypt and Argentina have averaged
500,000 t. All four countries exhibited strong export growth in the 1990s. The bulk
of Argentina’s and Uruguay’s exports are shipped within South America. Excluding
intra-EU trade, the EU exports 300,000–400,000 t of rice, nearly all japonica. Italy
accounts for the bulk of these shipments. Most the rice is exported as food aid or
with a subsidy.
Guyana and Surinam export smaller quantities of rice. Guyana’s exports peaked
in the late 1990s and Surinam’s a decade earlier. For both exporters, shipment lev-
els have substantially declined since the late 1990s. Quality concerns and a lack of
competitiveness in international markets are major reasons. Japan and Taiwan
export rice under food aid programs. Japan’s export levels vary substantially each
year, exceeding 500,000 t in some years. Taiwan currently exports about 100,000 t
of rice a year, down from more than 200,000 t in the early 1990s. Taiwan had
severe stock buildups in the 1980s and early 1990s. Both Japan and Taiwan are very
high-cost rice producers.
Major Importers
The import market is much less concentrated. Indonesia, Nigeria, Iran, Iraq, the
Philippines, and the EU are currently the largest importers. Since 1999, they have
accounted for 25–35% of global rice trade. Nearly all of their rice imports are
indica. Other large importers include Saudi Arabia, Senegal, Cote d’Ivoire, Brazil,
Bangladesh, Japan, Malaysia, and South Africa. Except for Japan, which is the larg-
est importer of japonica rice, these countries import primarily indica rice. These two
groups of importers together account for 40–50% of global rice imports.
Indonesia is the country that imports the largest amount of rice, taking nearly 5.8
million tonnes in 1998, the largest amount of rice ever imported by a single country.
The country has been a major importer of rice for more than 50 years, averaging
more than a million tonnes a year during the 1950s. From 1976 to 1980, Indonesia’s
imports averaged more than 1.8 million tonnes a year, or nearly 18% of global
trade. However, in the 1980s, Indonesia became nearly self-sufficient in rice, a
result of policies that supported production and limited imports. By the early 1990s,
Indonesia was again a major rice importer. Imports declined in the late 1990s, a
result of bumper crops and a large buildup in stocks. Imports began to pickup again
in 2002 and are projected to exceed 3 million tonnes annually.
Nigeria, Iran, and Iraq currently import more than a million tonnes of rice a year.
Nigeria takes mostly parboiled rice; Iran and Iraq import mostly high-quality indica.
All three began importing large amounts of rice in the mid-1970s, a result of high
oil prices and strong population growth. Bangladesh and Brazil were also major rice
importers during most of the 1990s, often taking a million tonnes or more each year.
In fact, Bangladesh was India’s top export market, and Brazil accounted for the
bulk of South America’s exports. Both Bangladesh and Brazil typically import huge
amounts of rice in years of production shortfalls and substantially reduce imports in
years of normal harvest. Since 1999, imports have dropped substantially for both
countries, primarily the result of stagnant to declining consumption in Brazil and
record harvests in Bangladesh.
The Philippines has been a steady rice importer since the mid-1990s, importing a
record 2.2 million tonnes in 1998. During the late 1970s and early 1980s, the Phil-
ippines was nearly self-sufficient in rice, a result of policies aimed at boosting pro-
duction. Imports have declined since 1999 because of bumper crops and a buildup
of stocks. The EU imports about 800,000 t a year (excluding intra-EU trade), mostly
indica rice. Import growth is very slow and primarily driven by population growth.
Extremely high tariffs protect high-cost producers from international competition.
The EU imports large amounts of brown rice to be further processed in the EU.
Saudi Arabia, Senegal, and Cote d’Ivoire have steadily expanded imports since
the 1960s. Saudi Arabia, which does not grow rice, imports more than 800,000 t a
year, mostly high-quality parboiled and aromatic rice. Senegal and Cote d’Ivoire
import mostly low-quality rice, including 100% brokens. Both countries are very
price-conscious buyers. Cote d’Ivoire takes nearly 600,000 t and Senegal about
700,000 t a year.
Japan, the largest rice-importing country before World War II, purchases around
650,000 t of rice a year, virtually all under a minimum-access agreement with the
WTO. In 1969, Japan became the first Asian rice-importing country to become self-
sufficient, a result of high domestic prices and import controls. Japan’s rice produc-
ers have the highest production costs in the world. As a requirement for member-
ship in the WTO, Japan agreed to a minimum import level each year starting in
1995 and gradually rising through 2001. Although Japan changed its policy in 1999,
allowing over-quota imports, the tariff rates were set so high that over-quota
imports are unlikely.
Malaysia imports around 600,000 t of rice a year, mostly medium- and high-
quality long-grain. Slight annual increases in production combined with declining
per capita use make Malaysia a steady import market. The Republic of South
Africa, which grows no rice, imports more than 500,000 t of rice a year, mostly
high-quality parboiled rice. Imports have grown substantially since the late 1970s, a
result of population growth, higher incomes, and greater political freedom.
In addition to these top markets, several other countries regularly import smaller
amounts of rice. Cuba, Turkey, Russia, Singapore, China, Mexico, North Korea, and
the United States each import more than 300,000 t of rice a year. All are indica
markets except Turkey and North Korea, which typically take japonica. Cuba, the
largest import market in the Caribbean, takes 450,000 t a year, a result of declining
production and growing population. Mexico is the largest importer in North America,
taking nearly 500,000 t annually, mostly rough rice from the United States. Russia
became a major import market in the 1990s, as domestic production contracted
sharply. Russia’s imports have averaged 375,000 t since 1996, with shipment levels
depending on domestic production, food aid, and foreign exchange. Singapore, which
grows no rice, imports around 350,000 t a year of high-quality rice, mostly fragrant
rice. Turkey, the second largest import market for japonica rice, imports more than
300,000 t annually, as rising consumption exceeds production growth.
Although they are net exporters, both China and the United States import rice
each year. China imports around 300,000 t, nearly all high-quality Thai jasmine rice
for high-income urban consumers. The United States imports more than 400,000 t
of rice each year, mostly aromatic rice from Thailand, India, and Pakistan. North
Korea’s imports have averaged more than 350,000 t a year since 1995, all food aid,
with Japan the largest supplier. North Korea was virtually self-sufficient in rice
before the late 1980s.
GLOBAL RICE CONSUMPTION
Overview
Rice is one of the world’s most important cereals for human consumption. In the
densely populated countries of Asia, especially Bangladesh, Burma, Cambodia,
China, India, Indonesia, Korea, Laos, the Philippines, Sri Lanka, Thailand, and
Vietnam, rice is a staple. As much as 75% of the daily caloric intake of people in
these Asian countries is derived from rice. More than 50% of the world’s population
depend on rice as their primary caloric source (FAO, 2001).
The bulk of rice consumed globally is fully milled white rice, both regular milled
and parboiled (placed under severe steam pressure before milling). Brown rice
accounts for a much smaller share. Rice is also consumed in the form of noodles,
puffed rice, fermented sweet rice, and snack foods made by extrusion cooking. Rice
is used in making beer, rice wine, sake, and vinegar. Some oriental desserts require
the use of glutinous or sweet rice, which consists entirely of amylopectin in the
starch, in contrast to the nonglutinous rice that contains both amylopectin and
amylose (10–26%). In the developed countries (such as the United States, Canada,
Japan, and the EU), rice is used in numerous processed foods including breakfast
cereals, candy, package mixes, crackers and other snacks, soups, and baby food
(Luh, 1991b). Since the 1990s, the United States has used rice, mostly brokens, in
premium pet foods.
Global rice consumption, including feed, seed, and processed use, has doubled
over the past 30 years. For 2001/02 (using an aggregate of local market years),
global consumption on a milled basis is estimated at a record 403.9 million tonnes.
However, this represents an increase of less than 2% from a year earlier. The rate of
growth in global consumption has steadily declined over the past four decades.
From 1990 to 2000, global rice consumption rose just 16%, down from 27% a dec-
ade earlier, 29% in the 1970s, and more than 31% during the 1960s (USDA, 2001b).
The slower expansion is due to weaker population growth and to diet diversification
in several higher-income countries that are major consumers of rice.
Developing countries have accounted for the bulk of the expansion in rice con-
sumption over the past 40 years. From 1980 to 2000, annual rice consumption
increased, on average, 2% a year in developing countries, compared with less than
1% in developed countries. Rice consumption in developed countries actually
declined during the 1960s and 1970s, with Japan accounting for almost all of the
drop. With stronger growth, the developing countries’ share of global rice con-
sumption has steadily expanded. In 1999/00 developing countries accounted for
more than 95% of global rice consumption, up fractionally from a decade earlier. In
1960, developing countries were responsible for 90% of global consumption. Their
share grew to 93% by 1970 and to 94% by 1980. Strong population growth and ris-
ing incomes accounted for the much of the growth from 1960 through 1980.
Regional and Country Comparisons
Asia accounts for the bulk of global rice consumption. In 1999/00, Asia
accounted for more than 88% of the total world rice consumption of 398.3 million
tonnes (milled equivalent). Asia’s share of total consumption has declined over the
last four decades; from 92% in 1960, to 91% in 1970 and 89% by 1980. Africa is
the second-largest rice-consuming continent, accounting for almost 4% of global

rice consumption in 1999/00. The region’s share has steadily increased over the past
40 years, with population growth a major factor. In the 1960s, Africa accounted for
less than 2.5% of global rice consumption. South America is the third-largest
domestic market for rice, accounting for more than 3% of global consumption in
1999/00. The region’s share of global rice consumption has changed little during the
past 40 years.
North America accounts for 1.7% of global rice consumption, up from 1.5% a
decade earlier. Its share has steadily expanded since the 1960s, with United States
being the largest market. U.S. consumption has grown rapidly, largely due to the
fact that a growing share of the population is from Asia and Latin America. Europe,
including the former Soviet Union, accounts for slightly more than 1% of global
rice consumption. Central America, the Caribbean, and Oceania together account
for less than 1% of global rice consumption.
China, India, Indonesia, Bangladesh, and Vietnam are the largest rice-consuming
countries in the world. In 1999/00, they accounted for more than 73% of global rice
consumption. China is the largest rice-consuming country in the world. In 1999/00,
its total rice consumption exceeded 133.7 million tones, or more than a third of the
global rice consumption. India ranked second, with total consumption estimated at
82.5 million tonnes, accounting for nearly 21% of global consumption. Indonesia
consumed 35.4 million tonnes, Bangladesh almost 23.8 million, and Vietnam 16.8
million. Except for China, rice consumption continues to rise in each of these coun-
tries. China’s total rice consumption is nearly flat, as declining per-capita consump-
tion nearly offsets increasing population. The other major rice-consuming countries
in Asia are Thailand (9.3 million tonnes), the Philippines (8.4 million), Japan (9
million) Burma (9.7 million), and South Korea (5 million). Except for Japan, rice
consumption is rising in each of these major rice-consuming countries.
Outside Asia, Brazil is the largest rice-consuming country, with 1999/00 con-
sumption estimated at more than 8.0 million tonnes. Like China, total rice con-
sumption in Brazil is rising at a fractional rate due to declining per-capita consump-
tion. In 1999/00, the United States consumed 3.8 million tonnes, Iran 3 million,
Egypt 2.9 million, Nigeria 2.8 million, and the European Union 2.2 million.
Per-Capita Analysis
Rice provided more than 20% of the per-capita calories and 14% of the protein
for human consumption worldwide (excluding use of cereal grains in beer use)
during 1997 to 1999 (FAO, 2001). Rice is especially important to diets in Asia and
South America. In Asia, rice provided 32% of the total calories consumed and 23%
of the protein. In South America, rice provided 11% of the total calories and 12% of
the protein. While rice accounted for less than 8% of total daily food consumption
in Africa, it is critical to diets in some African countries.
Rice is the dominant food in many developing countries, including several in
Africa. From 1997 to 1999, rice provided more than 1,500 kcal per capita per day in
Burma, Bangladesh, Laos, Cambodia, and Vietnam. In these five countries, rice
accounted for up to 75% of total calories consumed per day. In addition, the relative
importance of rice and other foods in diets in these countries has not changed much
in recent years. And rice is only slightly less critical to diets in several other coun-
tries. Rice provided more than 900 cal per day—up to 40% of total daily con-
sumption—in Madagascar, Sierra Leone, China, the Philippines, Sri Lanka, and
Thailand. Compared with a decade earlier, rice has declined in importance in
China and Thailand. Finally, rice provided more than a fourth of daily caloric
consumption in Gambia, Guyana, India, Malaysia, and Nepal. Over the past
decade, rice has declined in importance in India, Malaysia, and Nepal as diets
have diversified.
Worldwide rice consumption from 1997 to 1999, based on a mean world popula-
tion of 5.88 billion and mean milled rice production of 391.8 million tonnes, was
57.8 kg per person (not including beer use) (Table 5), up about 1 kg from a decade
earlier. In developing countries, per-capita consumption was 71 kg in 1997–1999,
compared with just 11.1 kg for developed countries. Asia reported the highest per-
capita consumption, 86.4 kg, down from 87.8 in 1987–1989, a result of declining
per-capita rice consumption in higher-income countries. Japan, Taiwan, and South
Korea accounted for the bulk of the decline. In South America, 1997–1999 per-
capita consumption was 30.9 kg, down from more than 32 kg a decade earlier. As in
Asia, the decline was due to diet diversification in higher-income countries. Brazil
was responsible for most of the region’s decline.
In Africa, per-capita rice consumption in 1997–1999 is estimated at 17.8 kg, up
nearly 2.4 kg from 1987–1989. In contrast to Asia and South America, rising
incomes are typically associated with increasing rice consumption in much of
Africa, particularly sub-Saharan. Per-capita consumption was much lower in other
continents: 8.5 kg in North America (excluding the Caribbean), 9.4 in Central
America, 4.8 in Western Europe, and 15.1 in Oceania (primarily Australia and
Papua New Guinea).
Burma has the highest per-capita rice consumption of any country in the world,
estimated at almost 211 kg per year in 1997–1999, up 7 kg from a decade earlier.
Bangladesh, Cambodia, Indonesia, Laos, and Vietnam reported the next highest per-
capita consumption, 150–170 kg. Except for Laos, per-capita consumption was up
substantially from a decade earlier in each of these countries. Per-capita con-
sumption in 1997–1999 was 90 kg or more in Gambia, Guinea-Bissau, Madagascar,
Sierra Leone, China (including feed use), South Korea, Malaysia, Nepal, the Philip-
pines, Sri Lanka, and Thailand. Many countries reported per-capita consumption
exceeding 50 kg per year, including Guinea, Senegal, Cote d’Ivoire, Costa Rica,
Guyana, Surinam, India, Japan, and North Korea. By comparison, per-capita con-
sumption was just 3.6 kg in Argentina, 4.3 in Russia, 5.6 in Mexico, 8.0 in Austra-
lia, and 8.6 in the United States.
RICE IN THE TWENTY-FIRST CENTURY
The global rice market faces several challenges and opportunities as it enters the
twenty-first century.
POPULATION AND PRODUCTION

First, from the late 1960s through 1990, global rice production grew at a faster
pace than population. However, production growth has slowed since the early 1990s
and has declined every year since 1999/00. By contrast, annual global rice con-
sumption continues to increase. In recent years, abundant stocks have allowed
global consumption to climb to record levels without raising prices and have
avoided any widespread famine resulting from smaller rice harvests. However,
global rice production will eventually have to increase to avoid major food short-
ages and substantially higher consumer rice prices.
This situation will become even more critical over the next decade as developing
countries, which depend on rice as a food staple to a much higher degree than
wealthier countries, become a larger share of the global population. The United
Nations Population Division (2003) projects global population in 2025 at nearly 7.9
billion (medium variant), up almost 1.8 billion from 2000, with developing coun-
tries accounting for nearly all of the growth.
By 2025, global rough-rice production will have to increase by 56% (IRRI, 1997)
just to meet the higher demand resulting from population growth. The bulk of this new
rice will have to come from higher field yields, as rice area is unlikely to expand
beyond the 1999/00 record of 155 million hectares without causing serious environ-
mental problems. In fact, with growing urban and industrial competition for land, it
is not certain that rice land can even remain at current levels, especially in Asia.
With global paddy yields estimated at 3.9 t/ha in 2001/02, average yields will need
to exceed 6.0 t by 2025 to satisfy projected global demand without raising area.
To boost yields to this level, researchers will need to focus on raising yields not
only for irrigated production, but for rain-fed areas and for marginal ecosystems such
as flood-prone or dryland areas. To date, the bulk of the success in boosting yields
since the late 1960s has been attained in irrigated rice systems. And land will not be
the only constraint (IRRI, 1997). Rice farmers will also be under substantial pressure
to reduce water and pesticide use. Since 1990, the International Rice Research Institute
(IRRI) has been working through conventional breeding methods on a new rice plant
called “super rice” that can produce yields 20% higher than current high-yielding
varieties. The rice will also contain greater resistance to disease and insects.
BIOTECHNOLOGY
The development and consumer acceptance of transgenic (genetically enhanced)
rice has the potential to greatly influence the global rice market, providing benefits
to producers and consumers alike. In early 2001, a consortium of public and private
research organizations—including IRRI—developed research samples of a geneti-
cally enhanced rice that contains b-carotene, the precursor of vitamin A. The rice,
called “Golden Rice” because of its color, is designed to fight vitamin A deficiency,
which is responsible for almost half a million cases of blindness each year in devel-
oping countries. Worldwide, about 125 million children, predominantly in devel-
oping countries, suffer from vitamin A deficiency. IRRI’s biotechnologists hope to
eventually develop more rice plants that deliver not only vitamin A but also zinc
and iron, and later, increased levels of protein (IRRI, 2001).
In addition to improving the nutritional qualities of rice, biotechnology has the
potential to make big contributions to reducing world hunger and reverse the recent
trend of stagnant crop yields. The United Nation’s Food and Agricultural Organiza-
tion (FAO, 2000) reported that the use of genetic engineering can overcome
obstacles (such as land degradation, strained water resources, and reduced irrigation
investment) to achieving higher yields in developing countries. In the United States,
researchers have developed herbicide-resistant rice that is expected to be commer-
cially available in a few years.
However, consumer acceptance of genetically enhanced rice will be critical to
successful adoption of these varieties. Already, the several large rice-importing
countries (i.e., Japan, the EU, and South Korea) have adopted mandatory biotech
food labeling regulations. In addition, the EU is finalizing its traceability and

labeling regulations for genetically enhanced food or feed products. Consumer
confidence will be critical to the widespread adoption of genetically enhanced rice.
Also, the cost-effectiveness of this rice, vis-à-vis other sources of a specific nutrient
(e.g., vitamin A in Golden Rice) will be a critical factor in determining market
acceptance.
DIVERSIFICATION
Higher incomes and greater commercialization in Asia will allow many rice
farmers to move away from monoculture rice production to a more diversified—and
profitable—farming system. While rice would remain the primary wet-season crop in
much of tropical Asia, the dry-season crop could shift to more profitable horticultural
products. This would likely reduce rural poverty and lead to less environmental
damage than intensive rice cultivation. Rice prices might increase slightly, helping
farmers and rural communities, as producers face alternative cropping options.
TRADE LIBERALIZATION
Finally, greater trade liberalization will promote expanded rice trade, likely con-
tributing to more stable international rice prices. With a larger and more stable
global rice market, importing countries will be more confident of acquiring greater
supplies when domestic production falls short of expectations. Expanded rice trade
will promote greater efficiency—as production shifts to the most competitive
areas—and slightly higher global income.
REFERENCES
Barker, R., Herdt , R. W., and Rose, B. 1985. The IRRI. 2001. Golden Rice: The Eyes of the World
Rice Economy of Asia. Resources for the Are Watching. International Rice Research
Future, Washington, DC. Institute, Los Baños, Laguana, Philippines.
De Datta, S. K. 1981. Principles and Practices of Juliano, B. O., Ed. 1985. Rice: Chemistry and
Rice Production. J. Wiley & Sons, New York. Technology, 2nd ed. Am. Assoc. Cereal
Eggum, B. O. 1979. The nutritional value of rice in Chem., St. Paul, MN.
comparison with other cereals. Pages 91-111 in: Luh, B. S., ed. 1991a. Rice Production, Volume
Proc. Workshop on Chemical Aspects of Rice I. AVI, New York.
Grain Quality. International Rice Research Luh, B. S., ed. 1991b. Rice Utilization, Volume
Institute, Los Baños, Laguana, Philippines. II. AVI, New York.
FAO. 2000. FAO Statement on Biotechnology. United Nations Population Division. 2003. World
Food and Agriculture Organization of the Population Prospects: The 2002 Revision.
United Nations, Rome. (website: www.fao. Forthcoming. (Website: www.un.org/esa/
org) population)
FAO. 2001. Food Balance Sheet. Food and Agri- USDA. 2001a. Grain: World markets and trade.
cultural Organization of the United Nations, Circ. Ser. FG 09_01. U.S. Dep. Agric., Foreign
Rome. Agric. Serv., Washington, DC.
Grist, D. H. 1975. Rice. Longmans, Green, and USDA. 2001b. Production, Supply, and Distribu-
Co., London. tion, electronic database. September, 2001.
IRRI. 1997. Rice Almanac, 2nd ed. International U.S. Dep. Agric., Foreign Agric. Serv.,
Rice Research Institute, Los Baños, Laguana, Washington, DC. (Website:
Philippines. www.fas.usda.gov/grain_arc.html)
CHAPTER 2
THE RICE PLANT: GROWTH, DEVELOPMENT,

AND GENETIC IMPROVEMENT
Anna Myers McClung

U.S. Department of Agriculture–Agricultural Research Service
Rice Research Unit
Beaumont, Texas
GROWTH AND DEVELOPMENT OF THE RICE PLANT
From seeding to harvest, the growth and development of the rice plant occurs
over a span of approximately four months under typical growing conditions in the
United States. The primary developmental phases of the rice plant include germina-
tion, vegetative, reproductive, and grainfill stages. A detailed description of the
morphological changes that occur within each of these phases has been given by
Counce et al (2000). Weather, field conditions, cultural management methods, and
genetics of the crop influence the physiological processes that occur within the plant
and thus affect the duration of each phase as well as the composition and dimen-
sions of the grain that ultimately develops. For example, in the southern rice-
growing region of the United States, planting rice at a relatively early date (e.g.,
February) under cool conditions can result in slow seedling emergence and initial
vegetative growth, with grainfill occurring under very warm conditions (e.g., July).
Planting in this same region at a later date (e.g., April) and under warmer field
conditions generally results in rapid seedling emergence and vegetative growth,
with grainfill occurring in relatively cool temperatures. Because of this, breeders
test new genetic materials over multiple years and locations to get a better
indication of the general performance of a cultivar in response to a wide range of
environments.
Planting, Germination, and Emergence
In the United States, rice is usually planted in March along the Gulf Coast, during
April in Arkansas, Mississippi and Missouri, and into June in California. The planting
date depends upon several factors, including weather conditions, field preparation, and
availability of water in situations where regulated canals provide irrigation.
In conventional tillage systems, field preparation usually begins in the late sum-
mer or fall with cultivation to remove weeds. In the following spring, further culti-
vation takes place to prepare the seedbed for sowing. Levees (small, parallel berms)
are established at varying distances throughout the field, depending on the slope of
25
the land. The area between the levees has a slope of no more than 0.2%, which is
used to maintain a uniform flood depth of about 10 cm and facilitate draining of the
field (Fig. 1). In the United States, rice is grown under flood irrigation, and no
transplanted or upland rice is commercially produced. Farmers plant dry or
presoaked seed with a grain drill on dry ground or drop pregerminated seed into
flooded fields with an airplane. Some farmers use no-till or reduced-tillage methods
in combination with preplant herbicides that reduce seedbed preparation time and
allow earlier planting.
Producers use the best cultural management practices for their location in an
effort to establish uniformly spaced plants. The seeding rate depends upon seed size
(i.e., how many kernels per kilogram of seed), germination rate, and field conditions
at planting. Essentially all commercial rice cultivars in the United States lack seed
dormancy. However, dormancy in red rice, a weed species closely related to culti-
vated rice, allows it to persist in soils for decades. Many of the field-management
methods used in rice cultivation were developed because no selective herbicides are
available for control of red rice. Following drill seeding, fields may be flushed
several times to encourage germination and emergence of the rice. Ultimately, U.S.
farmers desire to have a uniform plant stand of approximately 150–200 plants per
square meter. Plant stands influence the number and uniformity of tiller production.
These factors, among others, influence grain development and thus yield and
milling quality.
Fig. 1. Water depth in the rice field is controlled by levees (dark lines through the field), which are
established at specified distances to maintain a uniform slope across the field that facilitates flooding
and draining. (Courtesy USDA-ARS)
The Plant / 27
Vegetative Growth Stage
After the shoot has emerged from the soil, a series of leaves emerges as the main
culm (stem) elongates. Stem growth occurs as a result of elongation of hollow
internodes that are separated by solid cross walls called nodes. Leaves, additional
culms (tillers), and independent root systems can form at these nodal areas. Tillers that
develop from the main culm are called primary tillers, whereas those arising from later
tillers are called secondary, tertiary, etc. The number of seed-bearing tillers depends
on genetic factors, climatic conditions, and cultural management. The plant’s “factory
size,” or production capacity, is established during the vegetative phase, and therefore
it is critical that physiological stress be avoided during this time.
Historically, Asian rice cultivars have been developed for production under
transplanted conditions, in which groups of seedlings are planted several centime-
ters apart. These cultivars have been developed to have an extended vegetative
phase and produce large numbers of tillers to take advantage of the space between
transplants. In contrast, U.S. cultivars, which have been developed for use with high
seeding rates, produce relatively few tillers over a short period of time. A compari-
son by Wu et al (1998) showed that the average maximum tiller number for two
U.S. cultivars ranged from 285 to 426/m2 across different plant densities, whereas
the tiller number for Te Qing, a cultivar from China, ranged from 721 to 810/m2.
The higher yield potential of Te Qing, relative to the U.S. cultivars, resulted from
an extended tillering phase, a higher tiller number, and more grain per panicle. The
Fig. 2. The site of panicle differentiation (arrow) within the main stem of the rice plant.
relatively low number of tillers found in U.S. cultivars results in uniform grain rip-
ening and enhanced milling yield. U.S. cultivars having a long vegetative phase
may produce more tillers, but yield and grain quality can be reduced if the harvest
season is cut short by poor weather conditions in the fall.
Reproductive Stage
The shift from the vegetative to the reproductive stage occurs at panicle differ-
entiation (PD), when all of the nodes in the main culm have formed and the panicle
(seed head) begins to develop. This critical growth stage determines the number of
potential kernels on the panicle (Fig. 2). In U.S. cultivars, the initiation of the
reproductive phase is determined by the number of heat units (a function of number
of days of warm temperatures) accumulated during the season. In contrast, many
indica cultivars from Asia are called “photoperiod sensitive” because they undergo
similar phase shifts in response to changing daylength during the season.
Photoperiod-sensitive cultivars (e.g., jasmine rice from Thailand or basmati rice
from India) cannot be grown commercially in the United States because they do not
shift to the reproductive stage during the U.S. summer field season. With the
initiation of PD, carbohydrate that has been stored in leaves and stems begins to
remobilize to the developing panicle. Late-forming tillers may not form seed heads
Fig. 3. Panicle florets open for a short period of time during flowering to facilitate pollen shed from
anthers.
The Plant / 29
because their carbohydrate reserves supply tillers that already have panicles
developing. To prevent physiological stress at this stage, farmers scout fields for PD
to assure proper timing of fertilizer applications.
During the two to three weeks following PD, the panicle doubles in size every
three days (Nagai, 1958), and the male and female parts of the panicle flowers (flo-
rets) develop. The developing panicle moves up through the interior of the plant by
elongation of the culm internodes. At the “boot” stage, a swollen area at the top of
the stem indicates that the enlarged panicle will soon emerge from the surrounding
leaf sheath. “Heading” describes the stage at which 50% of the panicles in the field
have fully emerged from the leaf sheath. Within one day after the panicle has
emerged, the florets begin to open and expose the sexual parts (anthesis) (Fig. 3).
Pollination begins with the florets at the top of the panicle and, within three to five
days, reaches the bottom (Fig. 4). Pollination occurs when the anthers dehisce and
pollen is shed onto the stigma, which connects to the ovary. The pollen grain forms
a tube that grows down through the stigma, resulting in fertilization of the ovary
within a few hours. Because the pollination process occurs between the anthers and
stigma within a floret as it is opening, rice is predominantly self-fertilized.
However, cross-fertilization can occur with neighboring plants due to insect activity
or wind. This is an important problem in seed-production fields when cross-
Fig. 4. Pollen is shed from the tip of the branches to the base of the panicle during flowering.
pollination occurs with a red rice plant or another cultivar. High winds, extreme
temperatures, low humidity, heavy rains, or physiological stress during pollination
can reduce successful fertilization and result in sterility, poor seed set, and
nonuniform grain development.
Grainfill Stage
Within three days after embryo formation, nutrients (i.e., carbohydrates, pro-
teins, lipids, and minerals) from the vegetative parts of the plant remobilize to the
developing embryo and endosperm. Grainfill progresses from the top of the panicle
downward, in the same pattern as pollination, and takes place over a period of 30–
50 days, depending on the cultivar and environmental conditions. Amylose contents
vary slightly between kernels on the top and bottom of the panicle due to differ-
ences in activity of the granule-bound starch synthase (GBSS) enzyme located in
the endosperm of these grains (Umemoto et al, 1994). Dry matter accumulation
occurs at a rapid rate at first and then slows at the end of the grainfill period due to
the feedback inhibition of certain physiological processes and the senescence of
leaves and stems. The grain becomes hardened and translucent as it loses moisture.
Physiological stress that occurs during the grainfill phase influences the number of
grains filled, the ultimate size of the grain, and its composition.
Grain Harvest
Timely grain harvest is critical to preserving high milling yields. When the grain
has reached the hard dough stage and has about 25–30% grain moisture, the fields are
drained. This aids grain drying and improves field conditions for mechanical harvest
with a combine. However, draining the fields too early can reduce yields and milling
quality. If lodging occurs (i.e., the culms bend and cause the panicles to fall to the
ground) (Fig. 5) due to high winds or weak-stemmed cultivars, drained fields prevent
the panicles from falling into the water and help preserve grain quality.
Fig. 5. Rice cultivars with weak stems can lodge, resulting in losses in yield and quality.
The Plant / 31
As rice dries in the field, the grain reaches a critical point at which it is very sus-
ceptible to formation of internal cracks called fissures. The degree of fissuring
relates to the genetics of the cultivar, grain moisture, air temperature, and humidity
(Siebenmorgan et al, 1992; Lan and Kunze, 1996). Jodari and Linscombe (1996)
compared the two cultivars Cypress and Lemont and found that Cypress could be
harvested over a broad range of harvest moistures without fissuring, whereas
Lemont had to be harvested at relatively high grain moisture to maintain high head-
rice yields. Although low moisture diffusivity was associated with fissure resistance
in Cypress (Jodari and Linscombe, 1998), it can also result in increased drying costs
before grain storage.
FACTORS INFLUENCING GRAIN QUALITY
Cultural Management
In the late 1970s and early 1980s, U.S. breeding programs were successful in
releasing short-statured, semidwarf cultivars that produced high yields, resisted
lodging, and showed a dramatic yield response to high levels of fertilizer. Rutger
and Bollich (1991) called the development of semidwarf rice cultivars one of the
most important influences on rice improvement because it allowed increased
fertilizer use and higher yields without inducing lodging. Although semidwarf and
tall cultivars can produce the same amount of aboveground biomass, semidwarf
cultivars have a higher harvest index, producing a greater portion of dry matter as
grain instead of straw (Roberts et al, 1993).
The amount and timing of fertilizer applications for the rice crop depends upon
soil type, cultivar growth and development, and whether a ratoon crop (i.e., grain
produced as a result of regrowth of tillers from stubble left in the field after the first
crop is harvested) will be harvested. Jongkaewwattana et al (1993) observed that the
amount of nitrogen required for optimizing the main crop yield tended to be higher
than that required for optimizing milling yield. The amount of nitrogen required for
maximum yield and milling quality depended on the cultivar. Seetanum and De
Datta (1973) determined that milling quality was generally improved when nitrogen
was applied in split (multiple) applications, rather than all at planting. A nitrogen
application at heading was associated with an increase in grain protein content,
grain hardness, and milling quality. Higher nitrogen applications have been reported
to decrease chalkiness (i.e., white belly), grain cracking, and breakage during
milling (Bhashyam et al, 1985). Moreover, increased grain protein content as a
result of high amounts of nitrogen fertilizer influences the viscosity profile of rice
flour paste as it is heated and cooled (Fitzgerald, 2002).
Management of insect and disease pests of rice also affects yield and grain qual-
ity. A summary of important diseases and insect pests of rice in the United States
has been presented elsewhere (Groth and Lee, 2002; Way, 2002), so only a few
pests that directly affect grain quality are highlighted here. Nymphs of the rice stink
bug, Oebalus pugnax (F.), feed on developing grains and cause the introduction of
various pathogens, which results in a discoloration of the kernels called “peck.”
Incidence of peck greatly reduces the economic value of the crop. Similarly, the
primary concern for controlling rice diseases is to prevent loss of grain production.
However, even if the rice pathogen does not directly affect the grain, diseases like
sheath blight (caused by Rhizoctonia solani Kühn), rice blast (caused by Pyricularia
grisea (Cooke) Sacc.), and panicle blight (caused by Burkholderia glumae) can
cause poorly filled grains and chalky kernels that shatter during milling. The fungus
that causes kernel smut, Tilletia barclayana (Bref.) Sacc. & Syd., produces black
spores on panicle florets that can discolor grain during parboiling. Another disease
found on the rice panicle is false smut caused by Ustilaginoidea virens (Cooke)
Takah. The large green or orange spore balls that form on the panicle from this
disease can discolor seedlots. Narrow brown leaf spot caused by Cercospora
janseana (Racib) O. O. Const. is a disease that can occur on the leaves or panicles.
It causes the grain to rapidly lose moisture in the field, making it more susceptible
to fissuring, which decreases milling yield (Castro et al, 1994).
Environmental Factors
Temperatures that occur during the preflowering through grainfill phases can
dramatically affect sexual fertilization, grain development, and grain composition.
Low temperatures may occur during these stages in temperate areas like California
but are uncommon in the southern rice-growing region of the United States except
in the ratoon crop, which flowers in September or October. Extreme temperatures
during reproduction can cause floret sterility and influence the rate of grainfill. As
carbohydrates are translocated to the grain, the grain length fills first, followed by
grain width and thickness (Tashiro and Wardlaw, 1991). Thus, the timing of tem-
perature stress can affect the shape of the grain.
High temperature stress during grainfill can disrupt starch synthesis and the
development of starch grains, resulting in grain chalkiness (Tashiro and Wardlaw,
1991). In well-filled, translucent grains, the starch granules are densely packed in a
protein matrix. In chalky endosperm areas, the starch granules are loosely packed
and can be irregular in shape. Air spaces between the starch granules change how
light is refracted through the grain. The location and amount of chalkiness in the
grain is related to the timing and degree of the temperature stress (Tashiro and
Wardlaw, 1991). Lisle et al (2002) have shown that the structures of amylose and
amylopectin differ between chalky and translucent grains, and this has an impact on
cooking quality.
The primary constituent of the rice endosperm is starch, which consists of long,
linear molecules of amylose and highly branched molecules of amylopectin. Tem-
peratures during grainfill affect the relative proportion of these two molecules in the
endosperm. It has been commonly observed that cultivars typically having low
amylose content (e.g., medium- or short-grain cultivars) end up with relatively high
grain amylose content when grown under cool temperatures. However, cultivars
typically having high to intermediate amylose content (e.g., long-grain cultivars) are
much less responsive to temperature changes during grainfill. Umemoto et al (1995)
reported that japonica rice, when grown under relatively low temperatures, has
higher grain amylose content due to the increased activity of the GBSS enzyme.
Hirano and Sano (1998) have suggested that differences in the promoter regions of
the GBSS gene in indica and japonica rice cultivars may explain why the amylose
content of some cultivars is strongly influenced by temperature. Wang et al (1995)
reported that amylose contents differ among japonica, indica, and waxy rice culti-
vars because of inefficient processing of the mRNA associated with the gene that
controls GBSS. Indica cultivars produced completely spliced mRNAs that generate
large amounts of the GBSS enzyme, which results in increased contents of amylose.
The Plant / 33
Larkin and Park (1999) determined that a single nucleotide polymorphism in the
leader intron 5 splice site of the GBSS gene determines the degree of pre-mRNA
processing. One version of this polymorphism results in relatively high levels of
amylose being produced regardless of temperature. The other version of the poly-
morphism results in transcript stability and translational efficiency that are
temperature sensitive.
GENETIC IMPROVEMENT OF RICE
U.S. Breeding History
Rice breeding efforts in the United States began in the early 1900s, using culti-
vars that had been introduced from other parts of the world for commercial produc-
tion (Rutger and Bollich, 1991). Initial efforts were focused on selecting within
these cultivars for plants that were better adapted to U.S. growing conditions. By
the 1930s, breeding programs were established in several rice-producing states, and
selections from the original introductions were crossed with newer introductions.
Over time, the U.S. germ plasm pool became relatively narrow-based (Dilday,
1990) and distinct from those used in other parts of the world (Mackill and
McKenzie, 2003). Rice in the United States is cultivated in relatively small, uniform
regions in contrast to the diverse and widespread environments found in other rice-
producing countries. This has facilitated the development of numerous cultivars
having similar growth and developmental characteristics that can be used in highly
mechanized production systems. During the 1950s, the U.S. rice industry estab-
lished criteria for grain cooking quality that must be met before release of a cultivar.
As a result, the United States has developed long-, medium-, and short-grain culti-
vars that have been used as breeding materials in many countries around the world
because of their excellent milling yields and uniform cooking quality (Mackill and
McKenzie, 2003). One of the reasons for the success of the U.S. breeding efforts
has been the selection for cereal quality traits throughout the breeding process,
whereas in many other countries, this evaluation is done at the end of the cultivar
development process.
Genetic Variability
Fundamental to the success of rice-improvement programs is access to germ

plasm that is genetically diverse for traits of interest. The USDA-ARS National
Plant Germplasm System maintains more than 17,000 germ plasm sources, which
are available for use by breeders and researchers. However, most breeders have
been unsuccessful in incorporating foreign germ plasm into their breeding programs
because of the negative effect it has on milling and cooking quality. Exceptions to
this have occurred—for incorporating semidwarf plant height, parboiling and can-
ning stability, disease resistance, and specialty grain traits (Rutger and Bollich,
1991; Mackill and McKenzie, 2003). More recently, researchers have begun to
explore wild species of rice as a resource for identifying useful genes that can
improve cultivated rice (Xiao et al, 1998; McCouch et al, 2001; Moncada et al,
2001; Eizenga et al, 2002; Thomson et al, 2003). With these populations, molecular
marker technology has facilitated the introgression of alleles for desirable traits into
breeding materials that can be used for the development of new commercial culti-
vars. In some cases, tissue culture techniques have been employed to rescue
embryos that would have otherwise aborted due to large differences in the genetic
makeup of the two mated species (Eizenga et al, 2002).
Mutation breeding has been utilized in the rice research community for decades.
Chemicals or radiation are used to induce genetic changes in one or a few genes.
Because many genetic mutations are lethal, thousands of seeds are mutagenized so
that the number of survivors is sufficient for evaluation. Mutation breeding has been
commonly used in rice to induce short-statured cultivars (semidwarfs), early matur-
ity, and waxy endosperm. More recently, rice mutagenesis has been used to induce
herbicide resistance (Johnson et al, 2002; Sandhu et al, 2002), influence cooking
properties (Bao et al, 2001), and improve nutritional aspects (Larson et al, 2000).
In the 1970s, tissue culture techniques were developed as a means to modify or
propagate plants. Genetic changes (called somaclonal variation) that occur during
tissue culture are due to aberrant segregation of chromosomes or induced mutations
and are frequently lethal or genetically unstable. However, examples of heritable
somaclonal variation in height and plant color (Xie et al, 1996) and sheath blight
resistance (Xie et al, 1992) have been reported.
Genetically modified plants are those that have genes integrated from unrelated
plants or organisms using particle bombardment or bacteria-mediated transforma-
tion techniques. This technology has been used to improve traits commonly found in
rice as well as to introduce traits that are rare or unknown in rice. Zhang et al (1998)
used particle bombardment to successfully introduce a bacterial leaf-blight-resis-
tance gene from a wild species of rice into two rice cultivars that are widely grown
throughout Asia. The bacterium, Agrobacterium tumefaciens, has been used to
introduce a modified (antisense) form of the gene that controls glutelin A in rice
(Maruta et al, 2001). The presence of the antisense gene resulted in decreased levels
of glutelins in the grain. In brewing tests, the transgenic rice produced Japanese rice
wine (sake) with improved quality because of reduced amounts of amino acids,
which can give undesirable flavors. Tinjuangjun et al (2000) used particle bom-
bardment to introduce a gene from an herb, snowdrop, into two commercial rice
varieties, which produced resistance to the brown plant hopper, a common yield-
limiting insect in Asia. Rice has also been transformed with a gene that produces a
protein that inactivates a fungal toxin in corn. Transformation of rice with this gene
was found to confer resistance to rice blast disease (Uchimiya et al, 2002). Genes
from a bacterium and daffodil were used to transform rice, resulting in increased
levels of vitamin A precursors (Potrykus, 2001). Other genes have been used to
confer resistance to nonselective herbicides (Oard et al, 1996). These examples
demonstrate that transgenic technology greatly expands the possibilities for crop
improvement in rice.
Population Structure and Breeding Line Development
The first step in initiating a breeding program is to generate plant populations

that genetically segregate for traits of interest. Parental lines that complement each
other are selected for use in crosses. For example, a cultivar with high yield and
good disease resistance may be crossed with another cultivar having high milling
yield and resistance to lodging, with the goal of developing a new cultivar that is
improved for all traits. In many cases, improved but unreleased breeding lines are
crossed with released cultivars to develop segregating populations.
The Plant / 35
Crosses can be made in the field but more typically are made under controlled
greenhouse conditions. When cultivars with divergent maturities are crossed, they
are planted at different dates to increase the likelihood that both will flower at the
same time. For example, a late-maturing cultivar may be planted one or two weeks
before its early-maturing mate. Because the rice plant produces panicle-bearing
stems over a period of at least a week, crosses can be made between early and late
tillers on different plants. Using vacuum suction, anthers from the plant chosen to
be the female in the cross are removed (emasculation) before the pollen grains are
shed (Fig. 6). A tiller of the male parent in the cross is placed next to and slightly
above the emasculated tiller, and then both are covered by a small bag. When florets
Fig. 6. To facilitate making crosses for a breeding program, anthers are removed from rice florets,
using vacuum suction.
Fig. 7. A few days after making a cross, F1 hybrid seed can be seen developing. (Courtesy Jim Medley)
of the two tillers open, pollen from the male falls down onto the stigma of the
female flower. Within a few days, the developing seed of the cross pollination can
be seen on the female plant. Once the F1 hybrid seed is mature (Fig. 7), it is planted,
and undergoes natural self-pollination at flowering.
Seed produced from the hybrid plant will segregate for all traits that differ
between the two parents. With each subsequent generation of self pollination that
occurs (i.e., each field season), more and more genes (loci) become fixed and no
longer segregate. Thus, in early generations, when abundant segregation occurs,
selection is made among individual plants within a family. In later breeding genera-
tions, selection is made primarily among different families that are relatively uni-
form. Traits that are highly heritable (strongly affected by genes and weakly
affected by the environment) and controlled by a few genes are selected at the
beginning of the breeding process, using individual plants or single rows evaluated
in one environment. Traits controlled by many genes and sensitive to fluctuations in
environmental conditions (quantitatively inherited traits) are selected in later
breeding generations, when enough seed can be produced for testing in replicated
trials and at multiple locations.
BREEDING STATEGIES
Various kinds of population structures and selection strategies are employed in
rice breeding (Mckenzie et al, 1994); however, most breeders in the United States
use pedigree- or bulk-breeding schemes. In the pedigree-breeding method, each
segregant is handled separately, and the breeder makes selections within and among
families of progeny. This method is used when the breeder is able to select for traits
that are highly heritable (e.g., semidwarfism, maturity, amylose content) and sig-
nificant improvement can be accomplished during early-breeding generations. With
the bulk-breeding method, undesirable progeny are discarded, and seed of selected
offspring are combined together and planted as a bulk population in subsequent
generations. This method is used when the breeder desires to apply only minimal
selection during early generations and waits until later generations when consider-
able recombination has occurred among multiple genes that influence complexly
inherited traits (e.g., yield). Backcross breeding is used when a particular trait is
being transferred to an adapted cultivar. With this method, the donor parent (i.e.,
possessing the desired gene) is crossed to a leading commercial cultivar (the recur-
rent parent), and then the offspring are repeatedly backcrossed to the recurrent par-
ent. After each generation of backcrossing, selection is made for progeny that pos-
sess the desired trait. The objective of this breeding method is to essentially recover
the same genotype and phenotype of the recurrent parent except for the new trait
that has been introduced. Recurrent-selection techniques, which are commonly used
in cross-pollinated species, have also been used for rice improvement. Moldenhauer
and Lee (1994) used this method to develop a germ plasm pool incorporating
numerous genes tolerant to sheath blight disease. With this method, paired matings
are made using several donor parents, and then their hybrid offspring are
intercrossed. This cycle, repeated several times, results in novel combinations of
genes at multiple locations in the genome. This labor-intensive and time-consuming
method is used only when conventional methods have not been successful, the trait
is controlled by several genes, and strong donor sources are not available.
Anther culture is another method for generating breeding lines for selection.
Anthers obtained from F1 hybrid plants or later generations are placed on culture
The Plant / 37
media that induce individual pollen grains to undergo cell division and form callus
tissue. Subsequently, the callus tissue is transferred to different culture media that
induce root and shoot growth. The plantlets are then transferred to another medium
that encourages the young plant to further develop before it is transferred to soil
(Fig. 8). The process from plating the initial anthers to harvesting seed from a
regenerated plant takes about 12–18 months. The plants derived from this method
are genetically stable and equivalent to the homozygous (fixed) lines developed
from conventional breeding methods that may take 8–10 field seasons to produce.
Thus, the primary interest in using anther culture technology is the ability to rapidly
produce nonsegregating, breeding lines, resulting in savings in time and labor. Most
of the rice produced in the United States is long-grain and comes from a “javanica”
germ plasm base whereas medium-grain cultivars are derived from a “japonica”
Fig. 8. Anther culture involves initiation of callus from a single pollen grain in an anther (A),
development of calli from plated anthers (B), initiation of shoot and root development from callus (C),
and transfer of plantlets to a growth medium before placement in soil (D). (Courtesy Shannon Pinson)
background. Because javanica and indica germ plasms are less responsive to anther
culture techniques than japonica materials, much of the success in using anther
culture techniques in U.S. long-grain germ plasm has come from modifying culture
media and developing “bridging parents” (Chu et al, 2002). Martinez et al (1996)
demonstrated that anther culture technology was similar to the pedigree-breeding
method for developing disease-resistant rice germ plasm, but pedigree breeding
allowed for greater recombination among minor genes and thus was more useful for
complexly inherited traits like partial resistance to blast.
For rice cultivars developed in the United States since the 1980s using
conventional breeding methods, it has taken about 10 years from the time the
original cross was made to the release of the variety. For most programs, both
winter and summer breeding nurseries are used to allow two to three growing
seasons each year. Thus, without access to winter nurseries, the development time
for a new cultivar would be even longer. Examples of exceptions to this include M-
401 (Carnahan et al, 1981) and M-203 (Carnahan et al, 1989), which were devel-
oped in seven years through mutation breeding; Dellmont (Bollich et al, 1993a),
developed through backcrossing in five years; Texmont (Bollich et al, 1993b),
developed in six years using anther culture technology; and Cadet and Jacinto
(McClung, 2002), developed in five years using molecular markers as a selection
tool. Even with technologies that expedite the breeding process, it is important for
breeders to have a clear understanding of the traits that are desired by rice produc-
ers, millers, and processors because cultivar development takes such a long time.
HYBRID RICE
Interest in developing hybrid rice has existed since 1926, when heterosis in rice
was first reported by Jones (Li and Yuan, 2000). Hybrid technology is the basis of
the corn industry, since corn is a cross-pollinated species, but it has also been suc-
cessfully used in self-pollinated grain crops like sorghum and wheat. Hybrids are
heterozygous, having half of their alleles from one parent and the other half from a
second parent. Heterosis is the increased plant vigor and yield that result from the
heterozygous condition at many genetic loci in the plant. Rice hybrids were first
developed in China, where male sterility and male restorer systems were first dis-
covered. Male-sterile plants produce anthers having no pollen, sterile pollen, or a
limited amount of viable pollen. In all cases, the anthers do not dehisce, rendering
the plant male-sterile, which facilitates hybridization with a plant having normal
fertility. Seed harvested from the male-sterile plant is a result of cross-pollination.
In 1976, the first rice hybrid was produced commercially in China (Li and Yuan,
2000). Rice hybrids can produce 20–30% higher yield than conventional cultivars
but generally have lower milling quality. They are currently being grown on a wide
scale in China and India, and many other countries have established research pro-
grams for their development. In 2000, the first rice hybrid was commercialized in
the United States by RiceTec, Inc. (Alvin, TX) (Fig. 9). Most rice hybrids have
been developed using a three-line system: a cytoplasmic male-sterile line, a main-
tainer line, and a restorer line. Because it is difficult and costly to produce three
lines for each hybrid, most programs are developing a two-line system in which
male sterility is induced by temperature or photoperiod sensitivity. Within the
United States, much of the germ plasm currently used in hybrids has come from
China and has poor milling quality. In addition, the cooking quality of hybrid rice is
affected by complex genetic interactions of the cytoplasmic male-sterile plant, the
The Plant / 39
Fig. 9. A helicopter is used to disperse pollen from strips of male plants (light colored) to strips of
male-sterile, female plants (dark colored) for production of hybrid rice seed. (Courtesy RiceTec, Inc.)
maternal plant, and the triploid nature of rice endosperm, which together affect
protein content, amino acid constituents (Shi et al, 1996), and flour pasting
properties (Bao and Xia, 1999). As a result, there has been only limited acceptance
of rice hybrids within the United States, but it is just a matter of time before
improvements in grain quality make rice hybrid production more common.
Breeding Goals and Selection Methods
During the course of development of a rice cultivar, a breeder selects for numer-
ous traits (e.g., height, maturity, yield, milling yield, cooking quality, disease resis-
tance, etc.), each of which may be controlled by a few or by many genes. The
appearance of a trait is known as its phenotype and is a result of the presence of an
allele (a form of a gene) under particular environmental conditions. Some genes are
more sensitive to environmental conditions (e.g., genes for milling yield) than
others (e.g., genes for pigmentation). In addition, traits can be influenced by
interaction with other genes that are in the background of the cultivar (called
“epistasis”). With each generation of self-pollination that occurs in rice, 50% of the
segregating genes become fixed as one allele or another. Table 1 gives a hypotheti-
cal example of the genetic makeup of six offspring from a cross between two par-
ents, A and B. The two parents have different alleles for 14 of the 15 genes shown
but are identical for gene 11, perhaps due to a common ancestor. Traits 1–4, char-
acteristics that the breeder wants to improve, are controlled by one (e.g., Trait 3) or
more (e.g., Traits 1, 2, and 4) genes. However, there are other traits controlled by
genes having subtle effects that the breeder cannot perceive through available
evaluation methods (e.g., genes 11–15). Progeny 1 and 2 are genetically identical
for all but Trait 3. Progeny 3 and 4 are different for all traits except Trait 4. Progeny
5 and 6 are different from each other at all genetic loci except gene 11, where the
parents did not differ. Since rice is believed to possess more than 25,000 genes, it is
clear that making significant genetic improvement can be difficult with so many
The Plant / 41
“moving pieces.” This is one of the reasons that breeders tend to focus on using
narrow, improved germ plasm in their crosses so as to not disrupt the pyramid of
desirable traits that they have spent years putting together. Even using a narrow
germ plasm base, it can take 10 generations of selection to develop a cultivar that no
longer segregates. Thus, the ability to make progress from selection is a result of the
expression of the trait under different environments, the number of genes involved,
the interaction of alleles (additive, dominance, and epistatic effects), and the meth-
ods required to detect phenotypic differences.
The key to having an effective breeding program is the ability to evaluate large
numbers of progeny to identify novel recombinants. Most breeding programs make
100–300 crosses and evaluate 20,000–40,000 breeding lines each year. Screening
methods that are rapid, give clear results, and can be used on large numbers of
progeny greatly enhance the progress that can be achieved from selection. Some
traits, like height and maturity (days to heading), can be easily measured. Although
these traits may be influenced by the growing environment to some extent, breeders
typically include well-characterized varieties as repeated standards throughout the
field to provide a quick visual reference. For some pathogens (e.g., P. grisea
Cooke) Sacc., causal agent of rice blast disease, and R. solani Kühn, causal agent of
sheath blight disease) for which propagation and inoculation methods are well
established, screening nurseries are planted. For other diseases, like narrow brown
leaf spot and kernel smut, the breeder takes the opportunity to make selections when
there is a natural incidence of the disease or by planting nurseries at sites that
frequently have endemics. Screening for tolerance to insect damage is usually based
on the latter method since there are no successful rearing methods or efficient
screening techniques for insect pests common to the U.S. rice-growing regions.
BREEDING GOALS FOR AGRONOMIC TRAITS

Agronomic traits, which are associated with the production capacity of the culti-
var, make up the majority of the breeding effort in most varietal development pro-
grams. Development of cultivars with strong seedling vigor allows the plant to be
established quickly and the field to be flooded earlier for better weed control. Culti-
vars with poor stand establishment as a result of weak seedling vigor have dimin-
ished yield potential due to production of a limited number of panicle-bearing stems
per hectare. Seedling vigor is generally related to the viability of the grain (germi-
nation percentage) and the ability of the seedling to elongate under cool tempera-
tures and in soils having different textures. Most cultivated rices have been selected
to have low seed dormancy, a major difference from red rice biotypes, which can
remain dormant for decades. Selection for seedling vigor can be conducted under
field conditions (Fig. 10) or by using growth chambers in which the temperature can
be controlled. Turner et al (1982) determined that both coleoptile and mesocotyl
elongation were important for seedling vigor. Redona and Mackill (1996) found that
greenhouse or field screening conditions were effective for selection for mesocotyl
elongation but that using a slantboard test under controlled environmental condi-
tions was more effective for selecting shoot weight and shoot length. Currently,
most U.S. breeding programs select cultivars that are less than 110 cm tall or have
the semidwarfing (sd1) gene. Reduced plant height from sd1 is due to shortened
internodes, which result in shorter, stronger stems that are less susceptible to
lodging. U.S. varieties are selected to mature in about 120–130 days, which takes
full advantage of the growing season while reducing the risk of a late-season harvest
that may reduce yield and quality due to inclement weather. Varieties that mature
too early may have reduced main-crop yield potential because of a shortened
vegetative period, whereas late-maturing cultivars may have reduced grainfill under
cool fall temperatures.
Developing rice cultivars with high yield potential is the top breeding objective
since the majority of the government support payments for rice are based on pro-
duction. Breeders spend most of their effort during the early stages of cultivar
development culling out materials with unacceptable agronomic traits that can be
easily observed. Some components of yield, like number of kernels per panicle,
number of tillers, and grain weight, are heritable and can be selected during early
generations (Samonte et al, 1998; Wu et al, 1998), whereas other components are
complexly inherited and must be evaluated much later. Following four to five gen-
erations of selection, the breeding material segregates less and sufficient quantities
of seed are available for more extensive testing. At this point, the best breeding
lines are entered into yield trials that may be replicated at several locations. Breed-
ers include current commercial cultivars as the standards to beat in these trials.
Selection for resistance to diseases and insects is dependent on the level of eco-
nomic damage caused by the pest, whether there is a good screening method avail-
able, the difficulty of the method, and the amount of seed required for the test. Gen-
erally when there is enough seed for conducting yield trials, screening for resistance
to the blast pathogen is also performed. This is usually done in upland nurseries,
where plants are inoculated with a mixture of the pathotypes that commonly occur
in the United States. In both inoculated trials and noninoculated ones (i.e., those that
rely upon natural incidence of disease), highly susceptible cultivars are planted fre-
quently throughout the nursery and serve as spreaders of disease. Timed sprinkler
systems are used to maintain high leaf surface moisture, which encourages disease
development. Field testing for sheath blight resistance currently requires at least 30
g of seed and thus is postponed until advanced breeding generations. Sheath blight
inoculum is produced on rice hulls during the spring and then distributed over field
plots approximately 40 days after planting. Four weeks after heading, plots are rated
Fig. 10. Genetic lines having poor stand establishment are eliminated from the breeding program.
The Plant / 43
for severity of disease development. Recently a method for screening for resistance
to B. glumae, a bacterium associated with disease at the seedling, booting, and
grainfill stages (the latter known as panicle blight disease) has been developed
(Shajahan et al, 2000). A bacterial suspension culture is sprayed on breeding lines at
the booting stage, and panicle sterility ratings are made approximately four weeks
later. Evaluation of tolerance to the rice water weevil (Lissorhoptrus oryzophilus
Kuschel) is conducted on a limited basis in Louisiana, Texas, and California, where
high natural weevil infestations occur. In California, cultivars are evaluated for
stunted plant growth due to larvae feeding on the roots. In the southern states, culti-
vars are evaluated for larval infestation and yield loss by comparing control plots
with plots treated with an insecticide.
In Texas and Louisiana, ratoon crop potential is also considered an important
breeding goal. After the main crop is harvested, the field is reflooded and fertilizer
is applied. A second crop of tillers develops from the stubble left in the field, and
grain is harvested in about 60 days. A good ratoon crop produces a yield about half
that of the main crop. Since input costs are minimal, a reasonable ratoon crop can
result in significant economic value to the farmer. Cultivars that are well suited for
ratoon cropping need to be early maturing (less than 120 days) so that there is
enough time for the second crop to complete grainfill. Turner and Jund (1993)
determined that the amount of total nonstructural carbohydrates remaining in the
stems after the main crop harvest was strongly associated with yield potential of the
ratoon crop. If the main crop is stressed because of disease or insect pressure, this
can also limit the ratoon crop yield. Breeders generally reserve testing for ratoon
crop potential for only the most promising breeding lines that are within a couple of
years of commercial release.
BREEDING GOALS FOR GRAIN QUALITY TRAITS

For a rice variety to be commercially acceptable, it must meet the demands of
the producer, miller, processor, and consumer. The primary grain quality criteria for
the U.S. rice market have been described by Webb (1985). Thus, in addition to high
field yields, breeders must select for milling quality, grain appearance, and cooking
quality that meet criteria expected for long-, medium-, and short-grain market
classes. Additional traits are necessary for rices used in specialty markets. Most
breeders select for grain dimensions, grain appearance, and translucency throughout
the breeding process. During early generations of selection, this is usually done by
subjectively comparing rough-, brown-, or milled-rice samples with a commercial
cultivar that is preferred by millers. During later generations, when seed is not as
limited and field trials can be performed at multiple locations, breeding lines are
evaluated for milling yield using a small, laboratory-scale mill. This is done over
several years and locations so as to identify cultivars with good, stable milling
yields over diverse growing environments (Gravois et al, 1991). Cultivars that are
chalky or lack uniform grain size generally have reduced milling yields and can be
culled using milling results.
U.S. rice researchers have had the benefit of working closely with the rice
industry to help establish criteria for grain chemistry parameters, which have helped
to standardize U.S. rice cooking and processing quality. Methods have been devel-
oped for screening large numbers of breeding progeny to confirm that any commer-
cialized cultivar will meet end-user standards (Webb, 1985). The most important
criteria are amylose content and alkali spreading value. Screening methods for these
traits are so efficient that they can be easily performed using just 1 g of milled rice.
These two traits are controlled by genes located in the same region of chromosome
6 (He et al, 1999). Grain protein and cooking properties, as evaluated by viscosity
profiles of the rice flour (Blakeney et al, 1991), are determined just before release
of a cultivar. These few analytical tests are generally adequate to verify that the
cultivar will have acceptable cooking quality for its grain market class.
In addition to rice varieties having standard cooking properties, some cultivars
have been developed that have improved quality following parboiling and
canning processes. The first U.S. cultivar developed for these specific purposes
was Newrex (Bollich et al, 1980). The improved processing quality of Newrex
and the long-grain cultivars that were subsequently derived from it (Rexmont, L-
205, and Dixiebelle) trace to the cultivar Jojutla from Mexico. All of these
cultivars possess high amylose content (~27%), higher amylographic hot and cool
paste viscosities, reduced starch solids loss during processing, better cooked grain
appearance, and firmer cooked texture compared with standard U.S. long-grain
cultivars.
Since the 1980s, rice specialty markets have increased as the rice industry and
consumers have become more aware of the diversity that is available in rice for
flavor, texture, and appearance. Specialty rice cultivars are grown on a limited basis
in the United States but generally are produced under contracts at a premium price
and thus provide significant economic value to farmers. The predominant specialty
rice categories in the United States include aromatics, waxy rices, Japanese
premium quality rices, and Arborio types. Cultivars for any of these categories can
be developed using a combination of quality traits, including grain shape, amylose
content, alkali spreading value, aroma, and paste viscosity profiles. The primary
volatile component associated with aromatic rice is 2-acetyl-1-pyrroline, which
conveys a popcorn or buttery smell. Relative quantities of this compound can be
determined in breeding lines with less than 1 g of rice using a gas chromatograph
(Bergman et al, 2000). Although some analytical methods are available, selecting
for various texture and flavor sensory components generally requires relatively
large samples of cooked rice that are evaluated by a trained human sensory panel
(Champagne et al, 1999; Bett-Garber et al, 2001). Development of screening tools
associated with grain quality components that affect consumer preference still
requires much research.
Although these analytical methods have streamlined the breeding process for
aspects of grain quality, the development of molecular-marker technology offers
even greater efficiencies in selection. Over 3,000 DNA markers have been located
on the 12 chromosomes of rice, and some have been associated with economically
important traits. Molecular markers are excellent selection tools for crop improve-
ment because, unlike phenotypic traits, they are not influenced by the production
environment or interaction with other genes. In addition, markers can discern which
progeny are homozygous and true-breeding and which will continue to segregate in
subsequent generations. This helps the breeder to discard progeny that are geneti-
cally fixed for undesirable alleles, focus more resources on progeny with desirable
alleles, and know which progeny will continue to segregate and require additional
testing. Sufficient quantities of DNA can be extracted from a few leaves at the
seedling stage, and marker results can be made available before harvest. Although
tremendous improvements have been made in the efficiency of this technology, it is
The Plant / 45
still very expensive relative to traditional phenotypic screening methods. In addi-

tion, sometimes the same markers are not useful across diverse germ plasms, and
new markers may need to be developed for different breeding populations. Gener-
ally, markers are located near the gene of interest but are not part of the gene itself.
Thus, it is best to develop closely linked, flanking markers to reduce the likelihood
of recombination occurring between the markers and the gene of interest. Traits that
are complexly inherited require identification of multiple markers known as quan-
titative trait loci (QTL), each of which may account for only a small portion of the
genetic variance associated with the trait.
DNA markers linked with many of the key parameters associated with rice grain
quality have been developed, which allows selection for rice quality traits even
before the grain is produced on the plant. Ayres et al (1997) identified a microsatel-
lite sequence polymorphism linked to the GBSS gene that could differentiate classes
of apparent amylose content in rice. The methodology was further streamlined by
Bergman et al (2001), making it more suitable for use with the large numbers of
progeny that are typically used in breeding programs. Using 198 nonwaxy geno-
types, the method explained 88% of the variation in amylose content. A non-gel-
based system using a fluorescent probe that detects a single nucleotide polymor-
phism can differentiate rice cultivars having high to intermediate amylose content
and firm cooked texture from cultivars having low amylose content and sticky tex-
ture (Bormans et al, 2002). Recently, Larkin et al (2003) determined that pasting
properties of cooked rice were primarily associated with the GBSS marker. Thus,
cultivars like Jodon and L-202, which have high amylose contents like Newrex and
Rexmont but very different flour viscosity profiles, could be differentiated on the
basis of a simple assay for the GBSS marker. Use of this marker facilitated the
development of two specialty rices, Cadet and Jacinto, in just five years (McClung,
2002). Recently, molecular markers have been developed that are associated with
the alk gene, which is associated with alkali spreading value (M.-H. Chen, C.
Bergman, and R. Fjellstrom, personal communication). In addition, molecular
markers have been developed for genes associated with aroma (Ahn et al, 1992;
Cordeiro et al, 2002) and cooked kernel elongation (Ahn et al, 1993), traits
important in jasmine- and basmati-type specialty rices. Although milling quality is a
complex trait, QTL associated with grain shape partially explain milling yield
differences (Tan et al, 2001).
Successful development of new U.S. rice cultivars that meet the needs of the
changing global marketplace will depend upon the close interaction of researchers
with the rice industry. Identification of the chromosomal locations controlling the
physiological processes that influence growth of the rice plant and production of its
grain will lead to the development of new varieties with enhanced nutritional and
functional properties. This will result in expanded utilization of rice and will lead
the way for similar advances to occur in other grain crops.
REFERENCES
Ahn, S. N., Bollich, C. N., and Tanksley, S. D. kernel elongation in rice. Theor. Appl. Genet.
1992. RFLP tagging of a gene for aroma in 87:27-32.
rice. Theor. Appl. Genet. 84:825-828. Ayers, N. M., McClung, A. M., Larkin, P. D.,
Ahn, S. N., Bollich, C. N., McClung, A. M., and Bligh, H. F. J., Jones, C. A., and Park, W. D.
Tanksley, S. D. 1993. RFLP analysis of 1997. Microsatellites and a single-nulceotide
genomic regions associated with cooked- polymorphism differentiate apparent amylose
classes in an extended pedigree of U.S. rice Castro, E. M., McClung, A. M., and Marchetti,
germplasm. Theor. Appl. Genet. 94:773-781. M. A. 1994. Impact of narrow brown leaf spot
Bao, J. S., and Xia, Y. W. 1999. Genetic control on harvest moisture and fissuring in rice. Proc.
of paste viscosity characteristics in indica rice Rice Technical Working Group Meeting, 25th.
(Oryza sativa). Theor. Appl. Genet. 98:1120- Texas Agricultural Experiment Station, Col-
1124. lege Station.
Bao, J. S., Shu, Q., Xia, Y. W., Bergman, C., and Champagne, E. T., Bett, K. L., Vinyard, B. T.,
McClung, A. 2001. Effects of gamma irradia- McClung, A. M., Barton, F. E., II, Moldenhauer,
tion on aspects of milled rice (Oryza sativa) K., Linscombe, S., and McKenzie, K. 1999.
end-use quality. J. Food Qual. 24:327-336. Correlation between cooked rice texture and
Bhashyam, M. K., Srinivas, T., and Khan, T. A. Rapid Visco Analyser measurements. Cereal
1985. Evaluation of grain chalkiness in rice. Chem. 76:764-771.
Rice J. 88:5, 13-16. Chu, Q. R., Linscombe, S. D., and Cao, H. X.
Bergman, C. J., Delgado J. T., Bryant, R., 2002. Rice anther culture expedites variety
Grimm, C., Cadwallader, K. R., and Webb, B. improvement for Louisiana. Pages 396-401 in:
D. 2000. Rapid gas chromatographic tech- Proc. Temperate Rice Conf., 2nd. J. E. Hill
nique for quantifying 2-acetyl-1-pyrroline and and B. Hardy, Eds. Int. Rice Res. Inst., Los
hexanal in rice (Oryza sativa L.). Cereal Baños, Philippines.
Chem. 77:454-458. Cordeiro, G. M., Christopher, M. J., Henry, R.,
Bergman, C. J., Delgado J. T., McClung, A. M., and Reinke, R. F. 2002. Identification of
and Fjellstrom, R. G. 2001. An improved microsatellite markers for fragrance in rice by
method for using a microsatellite in the rice analysis of the rice genome sequence. Mol.
waxy gene to determine amylose class. Cereal Breed. 9:245-250.
Chem. 78:257-260. Counce, P. A., Kiesling, T. C., and Mitchell, A. J.
Bett-Garber, K. L., Champagne, E. T., McClung, 2000. A uniform objective and adaptive sys-
A. M., Moldenhauer, K. A., Linscombe, S. D., tem for expressing rice development. Crop
and McKenzie, K. S. 2001. Categorizing rice Sci. 40:436-443.
cultivars based on cluster analysis of amylase Dilday, R. H. 1990. Contribution of ancestral
content, protein content and sensory attributes. lines in the development of new cultivars of
Cereal Chem. 78:551-558. rice. Crop Sci. 30:905-911.
Blakeney, A. B., Welsh, L. A., and Bannon, D. R. Eizenga, G. C., Rutger, J. N., and Lee, F. N.
1991. Rice quality analysis using a computer 2002. Rice wild relatives as a source of disease
controlled RVA. Pages 180-182. in: Cereals resistance for U.S. rice cultivars. Pages 417-
International. D. J. Martin and C. W. Wrigley, 423 in: Proc. Temperate Rice Conf., 2nd. J. E.
Eds. Cereal Chemistry Division, Royal Aust. Hill and B. Hardy, Eds. Int. Rice Res. Inst.,
Chem. Inst., Melbourne, Australia. Los Baños, Philippines.
Bollich, C. N., Webb, B. D., Marchetti, M. A., Fitzgerald, M. 2002. Genotype, environment, and
and Scott, J. E. 1980. Registration of Newrex genotype x environment affect the cooking
rice. Crop Sci. 20:285-286. quality of rice. Pages 644-645 in: Proc. Tem-
Bollich, C. N., Hung, H. H., Webb, B. D., perate Rice Conf., 2nd. J. E. Hill and B.
Marchetti, M. A., and Scott, J. E. 1993a. Reg- Hardy, Eds. Int. Rice Res. Inst., Los Baños,
istration of Dellmont rice. Crop Sci. 33:1410- Philippines.
1411. Gravois, K. A., Moldenhauer, K. A. K., and
Bollich, C. N., Magill, C. W., Livore A. B., Rohman, P. C. 1991. Genetic and genotype X
Hung, H. H., Webb, B. D., Marchetti, M. A., environment effects for rough rice and head
and Scott, J. E. 1993b. Registration of Tex- rice yields. Crop Sci. 31:907-911.
mont rice. Crop Sci. 33:354. Groth, D., and Lee, F. 2002. Rice diseases. Pages
Bormans, C. A., Rhodes, R. B., Kephart, D. D., 413-436. in: Rice: Origin, History, Technol-
McClung, A. M., and Park, W. D. 2002. ogy, and Production. C. W. Smith and R. H.
Analysis of a single nucleotide polymorphism Dilday, Eds. John Wiley and Sons, Inc.,
that controls the cooking quality of rice using Hoboken, NJ.
a non-gel based assay. Euphytica 128:261- He, P., Li, S. G., Qian, Q., Ma, Y. Q., Li, J. Z.,
267. Wang, W. M., Chen, Y., and Zhu, L. H. 1999.
Carnahan, H. L., Johnson, C. W., Tseng, T. S., Genetic analysis of rice grain quality. Theor.
and Brandon, D. M. 1981. Registration of M- Appl. Genet. 98:502-508.
401 rice. Crop Sci. 21:986-987. Hirano, H.-Y., and Sano, Y. 1998. Enhancement
Carnahan, H. L., Johnson, C. W., Tseng, T. S., of Wx gene expression and the accumulation
Oster, J. J., and Brandon, D. M. 1989. Regis- of amylose in response to cool temperatures
tration of M-203 rice. Crop Sci. 29:1089- during seed development in rice. Plant Cell
1090. Physiol. 39:807-812.
The Plant / 47
Jodari, F., and Linscombe, S. D. 1996. Grain relation to blast (Pyricularia grisea Sacc.)
fissuring and milling yields of rice cultivars as resistance. Theor. Appl. Genet. 92:583-590.
influenced by environmental conditions. Crop Maruta, Y., Ueli, J., Saito, H., Nitta, N., and
Sci. 36:1496-1502. Imaseki, H. 2001. Transgenic rice with
Jodari, F., and Linscombe, S. D. 1998. Compari- reduced glutelin content by transformation
son of moisture loss rate and fissuring resis- with glutelin A antisense gene. Mol. Breed.
tance between two rice cultivars. Page 73 in: 8:273-284.
Proc. Rice Technical Working Group, 27th. McClung, A. M. 2002. Techniques for develop-
Texas Agric. Exp. Stn., College Station. ment of new cultivars. Pages 177-202 in: Rice:
Johnson, K., Elhardt, M., Smith, L., Linscombe, Origin, History, Technology, and Production.
S., Johnson, C., Abreu, A., and Mitten, D. C. W. Smith and R. H. Dilday, Eds. John
2002. Managing the potential for developing Wiley and Sons, Inc., Hoboken, NJ.
herbicide-resistant weeds in herbicide-tolerant McCouch, S. R., Thomson, M. J., Septiningsih,
rice. Pages 551-557. in: Proc. Temperate Rice E. M., Mocada, P., Li, J., Xiao, J., Ahn, S. N.,
Conf., 2nd. J. E. Hill and B. Hardy. Eds. Int. Tai, T., Martinez, C., McClung, A. M., Lai,
Rice Res. Inst., Los Baños, Philippines. X.-H., Moeljopawiro, S., Yuan, L. P., Moon,
Jongkaewwattana, S., Shu, G., Brandon, D. M., H., Guimaraes, E., and Tohme, J. 2001. Wild
and Hill, J. E. 1993. Effect of nitrogen and QTLs for rice improvement. Pages 151-169 in:
harvest grain moisture on head rice yield. Proc. Chandler Symposium: Rice Research
Agron. J. 85:1143-1146. and Production in the 21st Century. W. G.
Lan, Y., and Kunze, O. R. 1996. Fissure resis- Rockwood, Ed. Int. Rice Res. Inst., Manila,
tance of rice varieties. Appl. Eng. Agric. Philippines.
12:365-368. McKenzie, K. S., Johnson, C. J., Tseng, S. T.,
Larkin, P. D., and Park, W. D. 1999. Transcript Oster, J. J., and Brandon, D. M. 1994. Breed-
accumulation and utilization of alternate and ing improved rice cultivars for temperate
non-consensus slice sites in rice granule- regions: A case study. Aust. J. Exp. Agric.
bound starch synthase are temperature-sensi- 34:897-905.
tive and controlled by a single-nulceotide Moldenhauer, K. A. K., and Lee, F. N. 1994.
polymorphism. Plant Mol. Biol. 40:719-727. Recurrent selection for sheath blight tolerance
Larkin, P. D., McClung, A. M., Ayres, N., and in rice. Pages 25-33. in: Temperate Rice—
Park, W. D. 2003. The effect of the Waxy Achievements and Potential, Vol. 1. Proc.
locus (granule bound starch synthase) on Temperate Rice Conf. E. Humphreys, E. A.
pasting curve characteristics in specialty Murray, W. S. Clampett, and L. G. Lewin,
rices (Oryza sativa L.). Euphytica 131:243- Eds. Charles Sturt Univ., Wagga Wagga,
253. NSW, Australia.
Larson, S. R., Rutger, J. N., Young, K. A., and Moncada, P., Martinez, C. P., Borrero, J., Chatel,
Raboy, V. 2000. Isolation and genetic linkage M., Gauch, M., Guimaraees, E., Tohme, J.,
mapping of a non-lethal rice (Oryza sativa L.) and McCouch, S. R. 2001. Quantitative trait
low phytic acid 1 mutation. Crop Sci. loci for yield and yield components in an
40:1397-1405. Oryza sativa x Oryza rufipogon BCF2 popula-
Li, J., and Yuan, L. 2000. Hybrid rice: Genetics, tion evaluated in an upland environment.
breeding, and seed production. Pages 15-155 Theor. Appl. Genet. 102:41-52.
in: Plant Breeding Reviews, Vol. 17. J. Janick Nagai, I. 1958. Japonica rice—Its Breeding and
Ed. John Wiley and Sons, Inc., New York. Culture. Yokendo LTD, Tokyo.
Lisle, A. J., Martin, M., and Fitzgerald, M. A. Oard, J. H., Linscombe, S. D., Braverman, M. P.,
2002. Finding the link between chalkiness and Jodari, F., Blouin, D. C., Leech, M., Vain, P.,
starch synthesis. Pages 485-487 in: Proc. Cooley, J. C., and Christou, P. 1996. Devel-
Temperate Rice Conf., 2nd. J. E. Hill and B. opment, field evaluation, and agronomic per-
Hardy. Eds. Int. Rice Res. Inst., Los Baños, formance of transgenic herbicide resistant rice.
Philippines. Mol. Breed. 2:359-368.
Mackill, D. J., and McKenzie, K. S. 2003. Origin Potrykus, I. 2001. Golden rice and beyond. Plant
and characteristics of U.S. rice cultivars. Pages Physiol. 125:1157-1161.
87-100 in: Rice: Origin, History, Technology, Redona, E. D., and Mackill, D. J. 1996. Genetic
and Production. C. Wayne Smith and R. H. variation for seedling vigor traits in rice. Crop
Dilday, Eds. John Wiley and Sons, Inc., Sci. 36:285-290.
Hoboken, NJ. Roberts, S. R., Hill, J. E., Brandon, D. M., Miller,
Martinez, C. P., Correa Victoria, F., Amezquita, B. C., Scardaci, S. C., Wick, C. M., and
M. C., Tulande, E., Lema, G., and Zeigler, R. Williams, J. F. 1993. Biological yield and har-
S. 1996. Comparison of rice lines derived vest index in rice: Nitrogen response of tall and
through anther culture and pedigree method in semidwarf cultivars. J. Prod. Agric. 6:585-588.
Rutger, J. N., and Bollich, C. N. 1991. Use of ties generated by particle bombardment. Mol.
introduced germplasm in U.S. rice improve- Breed. 6:391-399.
ment. Pages 1-13. in: Use of Plant Introduc- Turner, F. T., and Jund, M. J. 1993. Rice ratoon
tions in Cultivar Development, Part 1. CSSA crop yield linked to main crop stem carbohy-
Special Publ. Crop Science Society of Amer- drates. Crop Sci. 33:150-153.
ica, Madison, WI. Turner, F. T., Chen, C. C., and Bollich, C. N.
Samonte, S. O. PB., Wilson, L. T., and McClung, 1982. Coleoptile and mesocotyl lengths in
A. M. 1998. Path analyses of yield and yield- semidwarf rice seedlings. Crop Sci. 22:43-46.
related traits of fifteen diverse rice genotypes. Uchimiya, H., Fujii, S., Huang, J., Fushimi, T.,
Crop Sci. 38:1130-1136. Nishioka, M., Kim, K.-M., Yamada, M. K.,
Sandhu, S. S., Bastos, C. R., Azini, L. E., Neto, Kurusu, T., Kuchitsu, K., and Tagawa, M.
A. T., and Colombo, C. 2002. RAPD analysis 2002. Transgenic rice plants conferring
of herbicide-resistant Brasilian rice lines pro- increased tolerance to rice blast and multiple
duced via mutagenesis. Genet. Mol. Res. environmental stresses. Mol. Breed. 9:25-31.
1:359-370. Umemoto, T., Nakamura, Y., and Ishikura, N.
Seetanum, W., and De Datta, S. K. 1973. Grain 1994. Effect of grain location on the panicle
yield, milling quality, and seed viability of on activities involved in starch synthesis in
rice as influenced by time of nitrogen applica- rice endosperm. Phytochemistry 36:843-847.
tion and time of harvest. Agron. J. 65:390-394. Umemoto, T., Nakamura, Y., and Ishikura, N.
Shajahan, A. K. M., Groth, D. E., Clark, C. A., 1995. Activity of starch synthase and the
Linscombe, S. D., and Rush, M. C. 2000. Epi- amylose content in rice endosperm. Phyto-
demiology studies on panicle blight of rice: chemistry 40:1613-1616.
Critical stage of infection and the effect of Wang, Z.-Y., Zheng, F.-Q., Shen, G.-Z., Gao, J.-
infected seeds on disease development and P., Snustad, D. P., Li, M.-G., Zhang, J.-L., and
yield in rice. Page 77 in: Proc. Rice Technical Hong, M.-M. 1995. The amylase content in
Working Group, 28th. Louisiana Agric. Exp. rice endosperm is related to the post-transcrip-
Stn., Crowley, LA. tional regulation of the waxy gene. Plant J.
Shi, C. H., Xue, J. M., Yu, Y. G., Yang, X. E., 7:613-622.
and Zhu, J. 1996. Analysis of genetic effects Way, M. O. 2002. Rice arthropod pests and their
on nutrient quality traits in indica rice. Theor. management in the United States. Pages 437-
Appl. Genet. 92:1099-1102. 456 in: Rice: Origin, History, Technology,
Siebenmorgan, T. J., Counce, P. A., Lu, R., and and Production. C. W. Smith and R. H.
Kocher, M. F. 1992. Correlation of head rice Dilday, Eds. John Wiley and Sons, Inc.,
yield with individual kernel moisture content Hoboken, NJ.
distribution at harvest. Trans. ASAE 35:1879- Webb, B. D. 1985. Criteria for rice quality in the
1884. United States. Pages 403-442 in: Rice: Chem-
Tan, Y. F., Sun, M., Xing, Y. Z., Hua, J. P., Sun, istry and Technology, 2nd ed. B. O. Juliano,
X. L., Zhang, Q. F., and Corke, H. 2001. Map- Ed. Am. Assoc. Cereal Chem., St. Paul, MN.
ping quantitative trait loci for milling quality, Wu, G. W., Wilson, L. T., and McClung, A.
protein content and color characteristics of rice 1998. Contribution of rice tillers to dry matter
using recombinant inbred line population accumulation and yield. Agron. J. 90:317-323.
derived from an elite rice hybrid. Theor. Appl. Xiao, J., Li, J., Grandillo, S., Ahn, S. N., Yuan,
Genet. 103:1037-1045. L., Tanksley, S. D., and McCouch, S. R. 1998.
Tashiro, T., and Wardlaw, I. F. 1991. The effect Identification of trait-improving quantitative
of high temperature on kernel dimension and trait loci alleles from a wild rice relative,
the type and occurrence of kernel damage in Oryza rufipogon. Genetics 150:899-909.
rice. Aust. J. Agric. Res. 42:485-496. Xie, Q. J., Linscombe, S. D., Rush, M. C., and
Thomson, M. J., Tai, T. H., McClung, A. M., Lai, Jodari-Karimi, F. 1992. Registration of LSBR-
X.-H., Hinga, M. E., Lobos, K. B., Xu, Y., 33 and LSBR-5 sheath blight resistance germ-
Martinez, C. P., and McCouch, S. R. 2003. plasm lines of rice. Crop Sci. 32:507.
Mapping quantitative trait loci for yield, yield Xie, Q. J., Rush, M. C., and Linscombe, S. D.
components and morphological traits in an 1996. Inheritance of homozygous somaclonal
advanced backcross population between Oryza variation in rice. Crop Sci. 36:1491-1495.
rufipogon and the Oryza sativa cultivar Jeffer- Zhang, S., Song, W.-Y., Chen, L., Taylor, N.,
son. Theor. Appl. Genet. 107:479-493. Ronald, P., Beachy, R., and Fauquet, C. 1998.
Tinjuangjun, P., Loc, N. T., Gatehouse, A. M. R., Transgenic elite India rice varieties, resistant
Gatehouse, J. A., and Christou, P. 2000. to Xanthomonas oryzae pv. oryzae. Mol.
Enhanced insect resistance in Thai rice varie- Breed. 4:551-558.
CHAPTER 3
RICE VARIETIES
K. A. K. Moldenhauer
J. H. Gibbons
Department of Agronomy
Rice Research and Extension Center
University of Arkansas
Stuttgart, Arkansas
K. S. McKenzie
California Cooperative Rice Research Foundation
Biggs, California
Rice in the United States is grown in two main production areas: the southern
region (Arkansas, Louisiana, Texas, Mississippi, and southern Missouri) and
California. Rice varieties for these regions have been developed primarily by coop-
erative public rice-breeding programs. The overall goal of these breeding programs
has been to ensure “a maximum and stable production of the types of rice required
by producers and consumers” (Adair et al, 1966). Approximately 140 public rice
varieties have been released to U.S. rice producers since 1911. This has been
achieved through the characterization of specific cooking and processing character-
istics for U.S. market classes and through the development of varieties with
improved agronomic and quality characteristics to satisfy those requirements.
Agronomic improvements have included reduced height, improved lodging
resistance, increased yield potential, early maturity, and selected stress tolerances
(disease, cool temperatures, and seedling vigor) and have led to significant improve-
ment in productivity. Kernel size, shape, and appearance were the initial quality-
selection criteria in the breeding programs. The selection for rice quality has
evolved into an increasingly complex and challenging process. Rice-breeding pro-
grams have expanded to include the introduction of new germ plasm, the utilization
of new technology and instrumentation, and consideration of consumer preferences
for characteristics ranging from processing to ethnic choice.
Throughout the last 40 years of cultivar improvement, the traditional U.S. market
classes have defined and maintained specific quality characteristics. Specific
defined physicochemical characteristics are associated with the long-, medium-, and
short-grain market classes, and new rice varieties are selected to have those proper-
ties. The marketplace further distinguishes varieties within the U.S. market classes
based on their region of production. This distinction is a reflection of more subtle
differences in quality due to the variety or its production environment. The high-
49
volume markets include the Southern Long-Grain, Southern Medium-Grain, and

Calrose Medium-Grain classes. Special-purpose or niche-market rice varieties,
grown on a much smaller acreage in the United States, include processing, aro-
matic, short-grain, and colored-bran types. These specialty rice varieties must
satisfy specific processing needs or consumer preferences in domestic or interna-
tional markets. In many cases, they are old varieties or introductions from other
rice-producing regions of the world, where they are popular for their quality char-
acteristics.
VARIETY ORIGIN AND CLASSIFICATION
All U.S. medium grains are classified as temperate japonicas (Mackill, 1995),
although they resulted from intersubspecific hybridization between long-grain tropi-
cal japonicas and short-grain temperate japonicas. The southern and California
medium grains have distinct pedigrees, although they do share common ancestors
(Dilday, 1990).
All U.S. long grains and their progenitors have been classified as tropical
japonicas, formerly known as “javanicas” (Mackill, 1995). The southern and
California long-grains have different progenitors.
Variety development in the two U.S. production areas has emphasized somewhat
separate gene pools because of differences in climate and market-class pref-
erence.
Southern Region
In the southern United States, long-grain rice that cooks “dry and fluffy” has
historically been important to production, beginning with the success of Carolina
Gold grown in South Carolina in the seventeenth and eighteenth centuries for export
to European markets. As rice production moved from North and South Carolina and
Georgia to Louisiana and Texas at the turn of the century, Carolina Gold was
replaced by Honduras, a long-grain introduction with good cooking quality and
better yield (Chambliss, 1920; Jones, 1936). Other varieties were selected from
hundreds of long- and short-grain rice introductions. Long-grain quality was pre-
ferred, but long-grain introductions at that time were low-yielding, late-maturing,
had poor grain milling yields, were of tall stature, and lodged (Jones, 1936; Adair et
al, 1973; Dethloff 1988). Short-grain introductions from Japan, on the other hand,
matured earlier, yielded better, milled well, were of short stature, and lodged.
Despite these many agronomic advantages, however, short grains did not have the
preferred “dry and fluffy” cooking quality. The solution was envisaged as the
“medium grain,” a variety that would combine the good qualities of the long and
short grain. Denthloff (1988) explained,
It was to this task that Sol Wright, a farmer near Crowley, Louisiana, applied him-
self beginning in 1909. Wright began selection trials for a variety that would provide
the yields of the short grain and the attractiveness of the long. Working from a genetic
stock of the Japanese variety selected from the Shoemaker farm, which had been used
for cooperative experiment work by the USDA in 1906, Wright’s product, released on
the market in 1911, was the Blue Rose variety, a medium, durable, and attractive grain
that seemed to outperform both the Honduras and the Kiushu in the field and in the
market.
Varieties / 51
The novel The Blue Camellia is based on the story of the development of Blue
Rose. The parentage of the cultivar remains unclear, but American breeders have
thought that it was a “natural hybrid” between a long-grain-like Honduras and a
short-grain-like Shinriki (Adair et al, 1966). Blue Rose appears to be the first U.S.
variety originating from hybridization and from a cross between temperate and
tropical japonicas.
Medium-grain improvement by the public cooperative breeding programs in the
southern region started with hybridization of Blue Rose with long and short grains.
Since 1932, 24 medium-grain varieties have been released (Table 1). The varieties
Nato and Saturn were successful medium grains developed in Louisiana; Zenith and
Mars were developed in Arkansas. The first successful southern medium-grain
semidwarf, Bengal, was released in 1993 (Linscombe et al, 1993).
The founding parent of the southern long-grain varieties was Rexoro, which was
developed into two main lineages: the bonnets and the patnas (Mackill and
McKenzie, 2003). Both lineages originated from the Texas program. The bonnets
(Bluebonnet and Bluebonnet 50) were widely grown throughout the southern region
and became well established in Arkansas, where they were the primary focus of
variety improvement for many years (resulting in Starbonnet, Newbonnet,
Tebonnet, Katy, Kaybonnet, Drew, and Lagrue). The patnas (Texas Patna, Belle
Patna, Bluebelle, and Labelle) were developed primarily in Texas. The patnas were
combined with the bonnets in Lebonnet, which became a parent for Lemont, the
first successful southern long-grain semidwarf (Mackill and McKenzie, 2003). Sec-
ond- and third-generation semidwarfs (Cypress and Cocodrie, respectively) have
been developed from Lemont by the Louisiana program. Arkansas releases since the
early 1990s combine bonnet and patna parentage without the semidwarf gene.
Recent important production varieties include Wells, Ahrent, and Francis. In all, the
southern cooperative programs released approximately 72 long-grain varieties from
1918 to 2002 (Table 1). These programs, located in Arkansas, California, Louisiana,
Texas, Mississippi, and Missouri, have involved longstanding cooperative work
between researchers from the Agricultural Research Service of the U.S. Department
of Agriculture (USDA-ARS) and the respective state agricultural experiment
stations.
California
The climate of the rice-growing region of the California Central Valley is unsuit-
able for production of tropical long-grain rice because of cool temperatures. Rice
production there was founded on the temperate short-grain introductions Colusa and
Caloro (Table 1) from China and Japan (Dilday, 1990). These varieties cooked
“moist and sticky” and established a basic quality type for rice that could be pro-
duced in California.
Plant breeders advanced from the simple introduction and selection of varieties
to hybridization. Crossing the short-grain Caloro and the long-grain Lady Wright
produced a variety with intermediate grain shape, Calady (Jones, 1936). Subsequent
crossing and backcrossing of Calady to the adapted Caloro produced the medium-
grain Calrose in 1948. This variety was superior to Colusa and Caloro in agronomic
and quality characteristics and had the cooking characteristics generally similar to
the short-grain parent. It continued to replace the short-grain varieties and was
widely grown and successfully marketed into the late 1970s. The name “rose” indi-
TABLE 1
U.S. Varieties
Program Grain Year
Variety Typea Released Qualityb Registration/Reference
Arkansas
Prelude L 1943 Typical Jones et al (1953)
Rexark L 1947 Typical Jones et al (1953)
Vegold L 1963 Typical Crop Sci. 5:286-287
Starbonnet L 1967 Typical Crop Sci. 8:399
Bonnet 73 L 1972 Typical Crop Sci. 13:772-775
Bond L 1983 Typical Crop Sci. 24:208-209
Newbonnet L 1984 Typical Crop Sci. 24:209-210
Tebonnet L 1984 Typical Crop Sci. 25:1126-1127
Katy L 1989 Typical Crop Sci. 30:747-748
Alan L 1990 Typical Crop Sci. 31:229
Millie L 1990 Typical Crop Sci. 31:229-230
Adair L 1993 Typical Crop Sci. 34:1123
Lagrue L 1993 Typical Crop Sci. 34:1124
Kaybonnet L 1994 Typical Crop Sci. 35:587-588
Drew L 1996 Typical Crop Sci. 38:896-897
Wells L 1999 Typical U.S. patent 6,281,416 B1
Ahrent L 2001 Typical Moldenhauer et al (2001)
Francis L 2002 Typical Moldenhauer et al (2002)
Zenith M 1936 Typical Johnston (1958), p. 699
Arkrose M 1942 Typical Johnston (1958), p. 694
Kamrose M 1943 Typical Jones et al (1953)
Northrose M 1962 Typical Crop Sci. 5:285
Nova M 1963 Typical Crop Sci. 5:285-286
Nova 66 M 1966 Typical Crop Sci. 8:399-401
Nova 76 M 1976 Typical Crop Sci. 19:743-745
Mars M 1977 Typical Crop Sci. 19:743-744
Orion M 1991 Typical Crop Sci. 32:495
Nortai S 1972 Typical Crop Sci. 13:774
California
L-201 L 1979 Soft cooking Crop Sci. 19:745-746
L-202 L 1984 Soft cooking Tseng et al (1984)
A-301 L 1987 Aromatic Crop Sci. 27:1310
L-203 L 1991 Soft cooking Crop Sci. 32:496
L-204 L 1996 Typical Crop Sci. 37:1390
A-201 L 1997 Aromatic Crop Sci. 37:1390-1391
Calmati-201 L 1999 Basmati Crop Sci. 41:2005
L-205 L 1999 Processing Crop Sci. 41:2004
Calady M 1934 Typical Jones (1936)
Calrose M 1948 Calrose Johnston (1958), p. 696
CS-M3 M 1968 Calrose Crop Sci. 10:728
M5 M 1973 Calrose Crop Sci. 15:887
Calrose 76 M 1976 Calrose Rutger et al (1977)
M7 M 1977 Calrose Crop Sci. 18:356-357
(continued on next page)
a L = long, M = medium, S = short.
b Typical long-grain: 21–23% apparent amylose, intermediate gelatinization. Typical medium-grain:
12–18% apparent amylose, low gelatinization. Typical short-grain: 15–19% apparent amylose; low
gelatinization. Soft cooking: 23–25% apparent amylose, intermediate gelatinization; low amylo-
graphic hot paste viscosity and setback. Toro type: long-grain kernel with typical medium-grain
chemical characteristics. Processing (canning): 24–26% amylose, intermediate gelatinization, low
breakdown, high setback, firm cooking. Aromatic: containing primary aroma compound 2-acetyl-1-
pyrroline. Basmati: aromatic with kernel elongation. Waxy: <5% amylose, low gelatinization.
Varieties / 53
TABLE 1 (continued)
Program Grain Year
California (continued)
M9 M 1977 Calrose Carnahan et al (1978)
M-101 M 1979 Calrose Crop Sci. 19:929
M-401 M 1981 Premium Crop Sci. 21:986-987
M-201 M 1982 Calrose Crop Sci. 22:1087-1088
M-203 M 1989 Premium Crop Sci. 29:1089-1090
M-402 M 1999 Premium U.S. patent 6,281,415 B1
M-104 M 2000 Calrose Johnson (2002b)
M-205 M 2000 Calrose Johnson (2002a)
M-206 M 2002 Calrose Johnson (2003)
Early Wataribune S 1913 Typical Jones (1936)
Caloro S 1921 Typical Johnston (1958), p. 696
CS-S4 S 1971 Typical Crop Sci. 12:256-257
S6 S 1974 Typical Crop Sci. 15:99
Calmochi-201 S 1979 Waxy Crop Sci. 19:746
S-201 S 1980 Typical Crop Sci. 20:551
Calmochi-202 S 1981 Waxy Crop Sci. 21:985-986
Calmochi-101 S 1986 Waxy Crop Sci. 26:197
S-101 S 1989 Typical Crop Sci. 29:1090-1091
Calhikari-201 S 1999 Premium U.S. patent 6,316,705 B1
Louisiana
Evangeline L 1918 Typical Chambliss and Jenkins (1923)
Fortuna L 1918 Typical Johnston (1958), p. 697
Tokalon L 1918 Typical Chambliss and Jenkins (1923)
Rexoro L 1928 Typical Johnston (1958), p. 698
Nira L 1932 Typical Johnston (1958), p. 700
Sunbonnet L 1953 Typical Johnston (1958), p. 698
Toro L 1955 Toro type Johnston (1958), p. 699
Leah L 1982 Typical Crop Sci. 22:1086-1087
Toro-2 L 1982 Toro type Crop Sci. 24:1212-1213
Cypress L 1992 Typical Crop Sci. 33:355
Jodon L 1994 Processing, soft Crop Sci. 35:1217
cooking
Cocodrie L 1998 Typical Linscombe et al (2000)
Delitus L 1918 Aromatic Jones et al (1941)
Della L 1973 Aromatic Crop Sci. 13:773
Dellrose L 1995 Aromatic Crop Sci. 36:1413
Dellmati L 1999 Basmati Crop Sci. 43:435-436
Shoemed M 1932 Typical Jones (1936)
Magnolia M 1945 Typical Johnston (1958), p. 697
Lacrosse M 1949 Typical Johnston (1958), p. 697
Nato M 1956 Typical Johnston (1958), p. 698
Saturn M 1964 Typical Crop Sci. 5:288
Vista M 1973 Typical Crop Sci. 13:774-775
LA110 M 1979 Atypical: high Crop Sci. 19:744-745
amy. low gel.
TABLE 1 (continued)
Program Grain Year
Louisiana (continued)
Lacassine M 1987 Typical Crop Sci. 32:494
Mercury M 1987 Typical Crop Sci. 28:193-194
Bengal M 1992 Typical Linscombe et al (1993)
Lafitte M 1996 Typical Crop Sci. 37:631
Earl M 2000 Typical Crop Sci. 41:2003-2004
Kiushu S 1899 Typical Jones (1936)
Shinriki S 1902 Typical Jones (1936)
Colusa S 1917 Typical Johnston (1958), p. 696
Acadia S 1918 Typical Jones (1936)
Missouri
Mo. R-500 L 1956 Typical Johnston (1958), p. 700
Palmyra L 1963 Typical Crop Sci. 5:287
Cody S 1944 Typical Johnston (1958), p. 699
Mississippi
Litton L 1996 Typical Rice J. 100(2):14-15
Priscilla L 1997 Typical Rice J. 102(1):24-25
Texas
Texas Patna L 1942 Typical Johnston (1958), p. 698
Bluebonnet L 1944 Typical Johnston (1958), p. 695
TP 49 L 1948 Typical Johnston (1958), p. 699
Bluebonnet 50 L 1951 Typical Johnston (1958), p. 695
Cen. Patna 231 L 1951 Atypical: low amy. Johnston (1958), p. 696
high gel.
Imp. Bluebonnet L 1951 Typical Johnston (1958), p. 697
Belle Patna L 1961 Typical Crop Sci. 5:287
Bluebelle L 1965 Typical Crop Sci. 8:399-401
Dawn L 1966 Typical Crop Sci. 8:401
Labelle L 1972 Typical Crop Sci. 13:773-774
Lebonnet L 1974 Typical Crop Sci. 15:886
Newrex L 1979 Processing Crop Sci. 20:286-287
Bellemont L 1983 Typical Crop Sci. 23:803-804
Lemont L 1983 Typical Crop Sci. 25:883-885
Skybonnet L 1983 Typical Crop Sci. 25:886-887
Gulfmont L 1986 Typical Crop Sci. 30:1159-1160
Rexmont L 1986 Processing Crop Sci. 30:1160
Jasmine 85 L 1989 Aromatic Crop Sci. 38:896
Maybelle L 1989 Typical Crop Sci. 31:1090
Dellmont L 1993 Aromatic Crop Sci. 33:1410-1411
Rosemont L 1993 Typical Crop Sci. 33:877
Texmont L 1993 Typical Crop Sci. 33:354
Jackson L 1996 Typical Crop Sci. 36:1412
Jefferson L 1996 Typical Crop Sci. 37:629-630
Dixiebelle L 1998 Processing Crop Sci. 38:898
Madison L 1998 Typical Crop Sci. 39:1256
Cadet L 1999 Soft cooking Rice J. 103(1):12-13
Jacinto L 1999 Soft cooking Rice J. 103(1):12-13
Saber L 1999 Typical Rice J. 104(1):19
Bolivar L 2002 Processing Rice J. 105(1):16-17
Lavaca L 2002 Processing Rice J. 105(1):16-17
Neches L 2002 Waxy Rice J. 105(1):16-17
Sierra L 2002 Basmati Rice J. 105(1):16-17
Varieties / 55
TABLE 1 (continued)
Program Grain Year
Variety Typea Released Qualityb Registration /Reference
Texas (continued)
Gulfrose M 1960 Typical Crop Sci. 5:288
Brazos M 1974 Typical Crop Sci. 15:887
Pecos M 1983 Typical Crop Sci. 25:885-886
Rico 1 M 1987 Typical Crop Sci. 30:1161
Private
Carolina White L 1694 Typical Jones (1936)
Carolina Gold L 1695 Typical Jones (1936)
Honduras L 1890 Typical Chambliss (1920)
Edith L 1916 Typical Jones (1936)
Lady Wright L 1920 Atypical: low amy. Jones (1936)
int. gel.
Blue Rose M 1911 Typical Johnston (1958), p. 696; Jones
(1936)
Early Prolific M 1915 Atypical: low amy. Jones (1936)
high gel.
Wataribune S 1908 Typical Chambliss (1920)
cates medium-grain shape, and “Cal” indicates California origin and production.
Specific processing and cooking properties were associated with Calrose. It was
established as a market class and is still used to identify California medium-grain
quality (Rutger et al, 2000; Johnson, 2003). In 1969, in response to the “green revo-
lution” occurring in Asia, California rice growers established a grower-funded rice
marketing order (a fee on production) to support research and intensify the rice-
improvement program (McKenzie et al, 1994). This accelerated rice-breeding pro-
gram has released 37 improved rice varieties since 1969, including medium-, short-,
and long-grain varieties, as well as specialty types with improvements in adaptation,
productivity, and quality (Table 1). Wilson (1979) presents a detailed review of
California’s rice history from its beginnings to 1980.
LEADING VARIETIES
Southern Region
From 1911 to 2002, the southern cooperative breeding programs have released
approximately 100 long-, medium-, and short-grain rice varieties (Table 1). From
the 1920s until 1984, the Rice Millers Association documented the acreage grown
for each variety. Since 1988, the Rice Technical Working Group (RTWG) has
continued this work. The RTWG is an informal group of public agencies (state
agricultural experiment stations, agricultural extension services, the USDA-ARS)
and industry groups actively engaged in rice research and extension. The group
meets biennially and publishes proceedings.
The variety acreage database since 1929 shows 58 varieties in production on
more than 1% of the southern regional rice acreage, with 5–13 varieties in produc-
tion within in a given year (Figs. 1 and 2). To select the leading varieties within this
group, University of Arkansas researchers identified those having the top three
acreage percentages for a given year. This group of leading varieties is shown in
Table 2. Taking into account overlapping production cycles (i.e., years when a
leading variety was not yet or was no longer in a leading position), they found that
groups of three to seven leading varieties have generally accounted for 80% or more
of the regional rice acreage in any given year.
Until 1945, the medium grains Blue Rose and Early Prolific dominated the
southern rice acreage (Fig. 1). The long grains Lady Wright, Edith, Fortuna, and
Nira occupied a combined acreage of less than 20%, despite higher market prices
and consumer preferences for long grains (Jones et al, 1952). Rexoro began to
emerge in the 1940s as the leading long grain with good eating quality, and long-
grain acreages began to increase. Texas Patna (1942) was the first long grain pro-
duced through hybridization, by H. M. Beachell (Johnston, 1958). However, pro-
duction of both varieties was limited to Texas and Louisiana by their late maturity.
Fig. 1. Southern varieties, 1929–1984.

Varieties / 57
The 1940s brought other changes: increased disease pressures on the “old varieties”
and the change to combines from binder-threshers, which favored long-grain varie-
ties because they were more easily combined and dried (Adair, 1956). The medium-
grain Zenith and long-grain Bluebonnet became the next generation of widely
grown rice varieties, released by the southern cooperative breeding programs.
Zenith combined earlier maturity with good grain quality. Bluebonnet’s maturity
was suitable for production in Arkansas. Bluebonnet expanded long-grain produc-
tion during the 1950s to more than 50% of the Southern acreage. It was the first
regionally successful long-grain variety.
Medium-grain production peaked again at 40–50% of the acreage at the end of
the Bluebonnet era, from 1962 to 1967, following the release of Nato. However,
Fig. 2. Southern varieties, 1988–2001.

Varieties / 59
since the late 1960s, medium-grain production has declined steadily with declining
markets and has remained below 20% since 1979. Nato, along with Saturn (intro-
duced in 1964), were the leading medium-grain production varieties until 1982,
when they were replaced by Mars (introduced in 1977) and subsequently Bengal
(introduced in 1992).
Southern long-grain production increased steadily over the period 1964–1979
and has remained at approximately 80% of the regional acreage since 1980. The 15
leading long grains since Bluebonnet 50 (Table 2) brought significant improvements
in plant height, lodging resistance, maturity, yield, milling, and disease resistance.
These improvements supported the expansion of long-grain production and mar-
keting in the South. New high-yielding long grains Cocodrie (Linscombe et al,
2000), Wells (Moldenhauer et al, 1999) and Francis (Moldenhauer et al, 2001) have
supported record production yields.
Short grains have never been a preferred grain type in the South. “Japan” rice
was grown on 3% of the acreage during the 1930s; Caloro and Colusa were grown
on 3% during the 1950s; and Nortai was grown on approximately 2% of the acreage
during the 1970s.
ACREAGE IMPACT
University of Arkansas researchers calculated an impact score for the leading
southern varieties based on the summation of their acreage percentages in the
regional database over time (Table 2). This score approximates the area under the
individual variety cures in Figures 1 and 2 and takes into account both the extent
and the duration of regional production. Impact scores have ranged from 36 to 792
(Table 2). Ten varieties have had impact scores greater than 250 and stand out as
varieties with the greatest regional significance. Two of these are Early Prolific and
Rexoro, which were not top-ranking varieties in their time but were produced over a
long period—30 years. The other eight include the region’s top-ranking varieties:
Blue Rose, Zenith, Bluebonnet, Nato, Starbonnet, Labelle, Lemont, and Cypress.
These varieties have each occupied 31–66% of the regional acreage for more than
five years and have had production cycles of 10–36 years. Each was one of the top
three varieties in at least three rice-growing states.
Leading southern varieties with impact scores less than 250 ranked second or
third in the regional database. These varieties had the highest acreage percentages
within their states of release but were not grown extensively in other rice-producing
states.
Variety impact scores have decreased over time. The top-ranking varieties have
maintained high acreage percentage values over time, but production cycles have
decreased from 36 years (Blue Rose) to approximately 10 years (Cypress). Varieties
ranked second and third have shown a decrease in acreage percentage values over
time, as well as a decrease in production cycle. Production cycles that were 10–14
years before 1990 have decreased to three to five years since then. The increased
turnover of varieties since the mid-1980s reflects the increased availability of
improved varieties and their rapid adoption by producers.
ADOPTION OF SEMIDWARFS
The southern region adopted semidwarfs on 40% of the acreage following the
release of Lemont and Gulfmont. The region has maintained a semidwarf acreage
average of 63% since the release of Cypress and Bengal in 1994. Other semidwarfs,
Varieties / 61
produced on less than 1% of the southern acreage since 1985, include Rosemont,
Texmont, Madison, Jefferson, Lacassine, Jodon, Litton, and Priscilla.
California
The short-grain varieties, predominantly Caloro and Colusa, occupied essentially

all of California’s production until the late 1950s. The state’s production shifted to
Calrose following the release of that variety in 1948. California’s short-grain acre-
age continued to decline because of the success of Calrose and its progeny, which
currently occupy more than 80% of the rice acreage. Long grains, waxy short
grains, and aromatic long grains have been developed but have never occupied a
large percentage of California’s rice production (Fig. 3).
In 1969, California initiated an accelerated rice-breeding program that began
delivering new rice varieties to growers in 1976. The successful development of
semidwarf Calrose medium grains was accomplished by Rutger et al (1977) through
induced breeding and by Carnahan et al (1978) through backcrossing. These
founding semidwarfs formed the germ plasm pools that have allowed the develop-
ment and release of 19 improved medium- and short-grain California varieties.
These semidwarf medium grains were rapidly adopted and by 1982 had replaced
their ancestral parent, Calrose, in California. The descendents of Calrose were
selected to have Calrose cooking and processing characteristics and are commonly
commingled in commercial drying, storage, and utilization.
In 1979, the California rice industry developed a uniform naming system for
new, publicly developed rice varieties based on grain type, maturity group, and
order of release. Development of this system was necessary to avoid confusing the
large number of varieties, to prevent mixing of different grain types, and to avoid
inappropriate planting dates. The name of a new variety contains a prefix letter
designating its grain type as long (L), medium (M), or short (S). Specialty rice car-
ries a descriptive word prefix, such as Calmochi for waxy or sweet rice, Calmati for
basmati-like rice, Calhikari for premium-quality short-grain rice, and A for aro-
matic rice. Immediately following the letter or name descriptor is a three-digit
number separated from the letter or name by a hyphen (-). The first digit in the
number designates the maturity group as 1 (very early), 2 (early), 3 (intermediate),
or 4 (late). The last two digits indicate the order of release of this type, from 01 to
99, starting in 1979, when this system began. For example, M-202 indicates an
early-maturing medium-grain variety that was second in order of release.
The California breeding program began to develop long grains from different
parentages adapted for California. Tseng et al (1984) released the well-adapted and
productive L-202. L-202 has been a successful parent in the development of recent
long-grain varieties Cypress and Cocodrie, developed in Louisiana. L-202 seed was
also exported to Spain and, without permission, renamed “Thaibonnet,” where it has
become the major long-grain variety grown in that region. Tseng and colleagues
released additional long grains with improvements in agronomic, milling, and cook-
ing quality; however, long-grain production still occupies less than 5% of Califor-
nia’s rice acreage (Fig. 3).
California’s traditional short-grain acreage has remained small in recent years
after losing a major market in Puerto Rico (Fig. 3). Premium-quality short grains,
primarily the Japanese varieties Koshihikari and Akitakomachi, emerged in the late
1990s in response to the opening of the Japanese market to rice. Satisfying the
Fig. 3. California varieties, 1988–2002.

Varieties / 63
quality requirement with the Japanese varieties for the Japanese market has proven
to be a significant challenge commercially. Very high standards are set for purity,
kernel appearance, and cooking and taste properties. Developing high-yielding
adapted varieties with premium-quality characteristics has proven to be an even
more difficult task. Premium short-grain production seems to have become estab-
lished in California, but the acreage fluctuates because it is subject to trade and
marketing issues.
California has an established premium-quality medium-grain production. These
types cook similarly to the Japanese premium short grains, have a similar texture,
appear very shiny, and remain soft after cooling. They trace their ancestry back to
the proprietary tall, late-maturing, medium-grain varieties Terso and Kokuhorose.
M-401, an induced semidwarf of Terso, is the predominant variety.
Specialty rice varieties occupy a small acreage. They include waxy short grains,
aromatic long grains, Mediterranean bold grains (Arborio, Calriso, and Baldo), and
varieties with colored bran. They are grown under contract and include proprietary
lines and introductions.
VARIETAL QUALITY CHARACTERISTICS
Southern Region
MILLING QUALITY
Good milling quality is important because rice is consumed as a whole kernel.
Farmers are paid on the basis of the yield of whole milled kernels, called “head” or
“whole-grain” rice and defined by U.S. millers as kernels that are at least 75% of
the length of a whole kernel. Milling yield is expressed as the percentage of head
rice relative to the percentage of total milled rice (head rice plus brokens). The
average U.S. head-rice yields reported by the Rice Millers Association have
remained constant over the last 30 years, fluctuating between 58 and 64% (USDA,
2000).
Head-rice yields of Southern cultivars have increased from 56% in Rexoro dur-
ing the 1950s (Adair et al, 1973) to a record 67% in the recent cultivar Cypress
(Arkansas Rice Performance Trial data). Leading production varieties known for
the best milling quality have included the long grains Bluebelle, Labelle, Newbon-
net, Tebonnet, Cypress, Kaybonnet, and Cocodrie and the medium-grain Nato.
Head-rice stability across locations, planting dates, or years has been variable
(Adair et al, 1966; Gravois et al, 1991).
Historically, head-rice yields were optimized by timing crop maturity to be in the
fall months (before frost). This reduced the incidence of sun checking (fissuring in
the field) and chalkiness, which can be characteristic of crops harvested during the
summer (Adair et al, 1973). Nine-year averages for date of planting experiments in
Louisiana during the 1950s showed that head-rice yields could be increased 3–5
percentage points by planting in May and harvesting during September and October,
compared with earlier plantings that matured in July and August (Adair et al, 1973).
These experiments included varieties with maturities that ranged from 160 days
(Rexoro) to 108 days (Belle Patna). It was generally concluded that “improved milling
quality is a major advantage of late seeding” (Adair et al, 1966, 1973). In Arkansas,
fall harvesting was typical until the mid-1980s, when early-maturing varieties began to
be planted earlier in the spring (N. Slaton, personal communication).
During the 1970s and early 1980s, the southern region grew primarily Labelle,
Starbonnet, and Lebonnet. Lebonnet combined large kernels (22.4-mg brown rice)
with lower head-rice yields (58%); Starbonnet and Labelle combined smaller ker-
nels (17-mg brown rice) with higher head-rice yields (61–63%). The industry
adjusted to a smaller kernel size. Parboiling the large-grained variety helped to
markedly reduce the kernel breakage in milling and improved profitability.
Agronomically, recommendations for optimizing head-rice yields at harvest are
based on grain moisture content. This determination tries to optimize several
opposing processes. As grain moisture contents decrease below 21%, head-rice
yields increase by optimizing the number of mature kernels on the panicle, thus
avoiding immature kernels. Immature kernels break during milling (Jennings et al,
1979; Bautista and Siebenmorgen, 2000). The closer harvest grain moistures are to
12%, the lower the drying costs incurred (Calderwood et al, 1980). However, the
lower the moisture content, the greater the risk of fissured kernels. Grain fissuring
in the field, or sun checking, has long been considered a primary reason for
reductions in head-rice yields (Adair et al, 1973; Juliano, 1990).
The University of Arkansas Rice Processing Program has evaluated kernel
breakage during milling due to weak mechanical strength. Three sources of kernel
weakness are considered: fissuring, thinness, and chalkiness (Lu and Siebenmorgen,
1995). Available data indicate that these factors are independent and can be
addressed by traditional breeding techniques.
Kernel Fissuring. Head-rice yield stability at harvest has been demonstrated
under field conditions (Jodari and Linscombe, 1996; Berrio and Cuevas-Perez,
1989) and under simulated laboratory screening conditions (Juliano et al, 1993;
Berrio et al, 2002). It is characterized in the field by the cultivar’s ability to main-
tain optimum head-rice yields for 7–14 days after reaching an optimum harvest
moisture content. This stability is advantageous to farmers negotiating a timely har-
vest. Lack of stability can result in a period of only three days within which to
harvest the crop while optimizing head-rice yield. Stability in these studies has been
associated with fissuring resistance, although fissuring resistance itself appears to be
independent of optimum head-rice yield values and also appears independent of
optimum harvest moisture values. Fissuring resistance is thought to be related to
“critical moisture content,” the moisture content below which fissuring occurs
(Juliano and Perez, 1993; Bautista et al, 2000a).
Southern long-grain cultivars with excellent field stability and fissuring resis-
tance are Cypress and Cocodrie; susceptibility to fissuring has been shown by
Lemont, Leah, Lacassine, Tebonnet, and Drew (Jodari and Linscombe, 1996;
Bautista et al, 2000b). The medium-grain Bengal has good head-rice stability,
despite moderate levels of fissuring (Jodari and Linscombe, 1996).
Fissuring-resistant cultivars have shown the highest head-rice yield averages
both in experiment station tests (Arkansas Rice Performance Trials and Uniform
Rice Regional Nursery) and in Arkansas milling survey data. They also show the
highest stability across planting dates (Slaton et al, 1998, 1999). During postharvest
processing, fissuring-resistant cultivars have higher critical drying temperatures
(Bautista et al, 2000b) and shorter tempering times (Cnossen and Siebenmorgen,
2000) (see Chapter 12).
Chalkiness. Chalkiness refers to opaque areas in various parts of the grain kernel;
the condition is called white belly, white center, or white back. Chalkiness occurs
when starch and protein particles become loosely packed during grain maturation
Varieties / 65
(Jennings et al, 1979; Webb, 1980; Lisle et al, 2000). Chalkiness is increased by
temperatures that fall outside the optimum range for grain maturation (Resurreccion
et al, 1977), by low light (Suzuki and Murayama, 1967; Yoshida and Hara, 1977),
and by increases in relative humidity or windiness (Torres et al, 2002).
Chalky grains break more readily than translucent ones, reducing head-rice
yields (Jennings et al, 1979; Webb, 1985; Juliano and Villareal, 1993). It also
affects the cosmetic appearance of the rice, and most international markets do not
accept rice that contains more than 2% chalky grains. The exception is the market
for Arborio, in which chalkiness seems to be desirable (Juliano and Villareal, 1993;
Lisle et al, 2000). Chalkiness also affects cooking and processing quality (Webb,
1985; Lisle et al, 2000).
The U.S. milling standards for rice define chalky kernels as whole or broken
kernels that are at least 50% chalky, and it sets limits for percentages of chalky ker-
nels (FGIS, USDA, 1989). Percentages of chalky kernels range from 1% (U.S. No.
1) to 15% (U.S. No. 6) for long grains. For medium and short grains, allowable
percentages range from 2% (No. 1) to 15% (No. 6).
U.S. rice-breeding programs discard lines with considerable chalkiness, but
chalkiness can occur in some years or environments. During the 1980s, chalkiness
was evaluated for 27 U.S. varieties, including the leading production varieties Star-
bonnet, Lebonnet, Labelle, Lemont, Newbonnet, and Tebonnet (B. D. Webb, C. N.
Bollich, B. R. Jackson, D. G. Kanter, S. D. Linscombe, K. A. K. Moldenhauer,
S. T. Tseng, and H. D. Petersen, 1989, personal communication). Chalkiness gener-
ally ranged from 5 to 13% across the southern region from 1984 through 1986, but
values peaked in 1988 at 20–55% in all leading varieties.
Chalkiness has also been shown to affect cooking quality, independently of the
effects of temperature on other chemical characteristics (amylose content, gelatini-
zation temperature) (Lisle et al, 2000). In this study, chalkiness increased at the base
of the panicle. When translucent grains were compared with chalky ones from the
same panicle, the chalky grains had slightly lower peak and final viscosities than did
the translucent ones. Analysis of starch composition and structure, however, did not
reveal any differences between chalky and translucent grains. The authors suggest
that the observed differences in pasting characteristics reflect different degrees of
crystallinity within the starch granule. Air spaces in chalky areas offer the oppor-
tunity for increased water absorption during cooking, and the presence of single
starch granules, rather than compound amyloplasts, dilutes the concentration of
starch in the gel and leads to a softer cooked rice.
CHEMICAL QUALITY
Since the 1950s, grain length in the United States has been associated with spe-
cific shape, cooking, and processing characteristics (Jones et al, 1941; Beachell and
Halick, 1957; Adair et al, 1966; Webb, 1985; Rutger et al, 2000). Long grains have
been selected for a slender shape that cooks dry and fluffy. Medium grains have
been selected for a medium, slightly bolder, short shape that cooks moist and
slightly sticky. Short grains have been selected for a bold, rounder shape that cooks
moist and slightly sticky. Three major U.S. market classes have existed since the
1950s: Southern Medium-Grain, Southern Long-Grain, and the Calrose. Varieties
released within these classes can be commingled, and the quality standards have
remained constant over time. The advantage of this is ease of identification at all
stages of processing and minimum requirements for separate storage and handling.
The disadvantage is that the system restricts various cooking types that otherwise
might be available to the industry (Webb et al, 1972).
Chemical standards for southern market classes were established during the
1950s at the USDA-ARS Rice Research Unit at Beaumont, TX, where quality
evaluations have been made ever since. Chemical characteristics of leading south-
ern varieties are shown in Table 3, along with historical data for their grain dimen-
sions.
Historical Rice Quality Characteristics. The Regional Rice Quality Laboratory
at Beaumont, TX, was established in 1955 following grain quality problems associ-
ated with Century Patna 231 (introduced in 1951). Century Patna 231 resulted from
a cross of good-quality parents: Rexoro and Supreme Blue Rose. However, it pro-
duced lower milled-rice yields and had cooking characteristics unlike those of either
parent (Beachell and Halick, 1957). It required longer cooking time than Rexoro,
and the grains were prone to split and fray following cooking. This latter character-
istic made it particularly unsuitable for canning, which was an emerging industry at
the time.
Guidelines for desirable chemical characteristics were established relative to a
group of reference varieties with known cooking and processing characteristics.
These reference varieties included Century Patna 231 (poor-quality long grain),
Early Prolific (poor-quality medium grain), Toro (moist-cooking long grain),
Rexoro and Texas Patna (excellent eating quality and canning stability), Bluebonnet
50 and Belle Patna (good eating quality, less stable processing quality).
In the 1950s, chemical tests established that Century Patna 231 had low apparent
amylose (Halick and Keneaster, 1956) and a high gelatinization temperature com-
bined with a low gelatinization time (Halick and Kelly, 1959). The variety was
excessively sticky and referred to as “glutinous.” Early Prolific also had low amy-
lose and a high gelatinization temperature; its stickiness was considered “gluti-
nous.” A high gelatinization temperature was generally found to be undesirable. It
continues to be considered an undesirable quality characteristic internationally
(Jennings et al, 1979).
The long-grain Toro was known for moist cooking quality and had amylose
content and gelatinization temperature similar to those of short- and medium-grain
types that were equally moist when cooked (Halick and Kelly, 1959).
The long grains Rexoro, Texas Patna, and Texas Patna 49 were known as supe-
rior for cooking and processing. They had intermediate gelatinization temperature,
high amylose content, and a positive setback between the peak viscosity of the hot
paste and the viscosity of paste cooled to 50°C. This was determined by amylogra-
phy using Brabender units.
The varieties Belle Patna, Bluebonnet 50, and Fortuna had good cooking quality
but were less well adapted for parboiling and canning. They had lower apparent
amylose content than Rexoro and Texas Patna (Halick and Keneaster, 1956) and a
less positive setback. The long grains most suited for parboiling and canning were
those with high amylose and maximum setbacks on cooling (Rexoro and Texas
Patna).
Quality Testing Methods. Rapid tests for apparent amylose, using a starch-
iodine-blue test (Halick and Keneaster, 1956), and relative gelatinization tempera-
ture, using alkali digestion (Little et al, 1958), were incorporated into U.S. rice-
breeding programs in 1955 (Beachell and Stansel, 1963; Webb et al, 1972). These
were considered the most important two characteristics indicative of quality. They
Varieties / 67
also were shown to have a relatively high level of heritability (Beachell and Halick,
1957; Webb et al, 1972; McKenzie and Rutger, 1983).
Breeding programs continue to evaluate cooking and processing quality in segre-
gating populations primarily by measuring apparent amylose content and alkali
spreading value. The methodology has changed over the years, as have values used
to define breeding windows. Recent data published for historical southern varieties
shows that apparent amylose content in long grains has remained between 21 and
23% for the last 50 years (Table 3). For medium grains, the historical variation is
greater: 12–18%.
For later generations, an amylograph is used to distinguish lines that have extra-
high amylose content (24–25% apparent amylose). This became prevalent in the
late 1980s with the release of the California long grain L-202. This variety had a
higher amylose content combined with softer cooking characteristics relative to
those of typical Southern long grains. This combination of characteristics was
associated with lower peak and hot paste viscosities in amylograph tests, and the
amylograph data were used to distinguish lines from southern crosses with L-202
that were atypical relative to either a “Lemont” or “Rexmont” quality type.
Other quality characteristics evaluated in later generations are protein, moisture,
and fat content. In some cases, water uptake capacity, image analysis, instrumental
texture analysis, degree of whiteness, and determinations of 2-acetyl-1-pyrolline
and hexanal are also evaluated (Champagne, 1999). Ongoing research has empha-
sized identification of quality characteristics that affect functionality, with a view to
improving the consistency of rice used in processing. Another important research
area involves identification of sensory quality characteristics that define different
ethnic preferences. An underlying need is the development of rapid and objective
methodologies for use in breeding programs.
Processing Varieties. In the early 1960s, requirements of the canning and quick-
cook rice industry included dry and flaky grains with a minimum amount of split-
ting or fraying and low solids loss following pressure cooking (Beachell and
Stansel, 1963). Varieties with low gelatinization temperature were generally not
suitable for canning or quick-cook processing, although some were suitable for par-
boiling. Thus, most short and medium grains and some long grains with low gelati-
nization temperature (despite high amylose or setback) were found unsuitable for
processing. Low gelatinization temperatures associated with medium and short
grains were considered useful to brewers and to cereal manufacturers who used
diastatic digestion in their processing (Halick and Kelly, 1959).
During the 1950s, the change to earlier-maturing long grains (Bluebonnet 50,
Belle Patna) shifted grain maturation to early fall and late summer and to higher
temperatures during the grain-maturation period. This was accompanied by a
change in variety chemical characteristics (Beachell and Stansel, 1963), with a
marked effect on gelatinization temperature. The authors commented that in
“extreme” years with cool temperatures during grain maturation, or when matura-
tion occurred very late in the season, typical long grains with intermediate gelatini-
zation temperatures could have low gelatinization values. The amylose content and
positive amylographic setback could also be higher under low-temperature condi-
tions. The late varieties Rexoro, Texas Patna, and Texas Patna 49 had better stabil-
ity of chemical and processing characteristics, whereas the earlier varieties, Blue-
bonnet 50 and Belle Patna, maturing in hotter summer months, could have lower
amylose and amylographic setback and higher gelatinization temperatures. These
Varieties / 69
fluctuations were not sufficient to affect cooking quality, but they were considered
important for processing behavior (Beachell and Stansel, 1963). The shift to low
gelatinization temperature with low growing temperature, regardless of amylose
content or setback, was not suitable for parboiling, canning, or quick cooking, based
on industry evaluation. Conversely, the shift to lower amylose (with high growing
temperature) was also associated with a reduction in parboiling and canning
stability.
Concern for processing quality set in motion the exploration for high-amylose
germ plasm to be used in the development of higher-amylose U.S. cultivars for
improved processing and canning stability (Beachell and Stansel, 1963). The Jojutla
cultivar was identified and its desirable quality characteristics incorporated into a
series of high-amylose superior-processing varieties: Newrex (1979), Rexmont
(1986), Dixiebelle (1997), and the California variety L-205 (1999).
Distinguishing quality characteristics of the Jojutla type include higher amylose,
low amylographic breakdown viscosity and high setback viscosity, stronger gel
consistency, and firmer cooked-rice texture. These quality characteristics lead to
canning stability, as determined directly by lower solids loss and fewer canned
kernel splits and fraying of edges than result from traditional long grains (Webb and
Bollich, 1980).
The cultivar Jojutla is classified as a tropical japonica (Mackill and McKenzie,
2003). Recent work with microsatellite markers to differentiate amylose classes
found that Jojutla has the Wx^a allele of the waxy gene associated with indica rice
(Ayres et al, 1997; McClung et al, 1999).
Another genetic source for superior processing quality has been recovered from
the Chinese indica variety Teqing, the parent of Bolivar (2001). Bolivar has high
amylose and processing characteristics similar to those of the traditional Jojutla
type.
Superior processing varieties are targeted for the canned-soup niche market.
Since the mid-1970s, no growth has been seen in this market. Throughout the
1990s, annual use averaged 5,000 metric tons (t) but fell to a record low (2,500 t) in
1999 (USDA, 2000). Canning today represents less than 1% of processed rice use.
New uses for rice in processed foods have emerged since the 1980s, and growth
of processed food has grown steadily. Processed food use peaked at 25% of total
domestic consumption (718,750 t) in 1995. Since then, growth has slowed, and, in
2000, processed food use represented 22% of domestic consumption (USDA, 2000).
Processed food products in 1999 included cereal (30%); packaged mixes (16%);
frozen dinners (4%); baby food (4%); crackers and snacks (3%); candy, rice cakes,
and soup together (6%); and pet food (37%).
Customer requirements in the food-processing sector tend to be the most
demanding in the market and require narrow ranges in functional properties (D. R.
McCaskill, personal communication). Rice sold for processing typically receives a
“low-grade” parboiling treatment first.
Aromatic Rices. Rice imports have increased steadily since 1980 and have aver-
aged around 450,000 t per year since 1996 (USDA, 2000). They represent 9.5% of
total U.S. rice consumption. Most of the imported rice is aromatic jasmine rice from
Thailand (80%) and basmati rice from India and Pakistan; imported rice also
includes small quantities of the Italian Arborio. The increase in imported rice
reflects “palate expansion” for directly consumed rice that is the basis of ethnic cui-
sine. This reflects both immigrant preferences and the growth of Mexican, Italian,
Chinese, Cajun, and Japanese cuisine within the general U.S. population (Meyers,
1995; Chambers and Childs, 2000).
U.S. breeding programs have developed several aromatic varieties for U.S. pro-
duction to fill this market niche. The first release was Della (1971). This variety, as
well as Dellmont (1992) and Dellrose (1995), contain the principle aroma com-
pound in aromatic rice, 2-acetyl-1-pyrroline, and have traditional U.S. long-grain
cooking quality (dry and fluffy).
The most recent releases include Dellmati (1998), Calmati 201 (1999), and
Sierra (2002). These contain the principle aroma compound and also produce elon-
gated slender kernels when cooked, as does basmati rice from India and Pakistan.
Jasmine 85 was released to provide a counterpart to imported Thai jasmine rice.
However, imported jasmine has been preferred by consumers. The reasons for this
have been unclear; however, progress has been made and acceptance by Asian con-
sumers is predicted, based on hardness and stickiness measurements (Meullenet et
al, 2001). All U.S. breeding programs are continuing their efforts to develop
aromatic cultivars with quality comparable to that of the imported varieties (Rutger
et al, 2000). Molecular markers have been identified to aid in selection for aroma
and kernel elongation (McClung et al, 1999).
California
The Calrose market type grown in California may include several medium-grain
varieties. M-202 (Johnson et al, 1986), released in 1986, has been the predominant
variety produced in the state for years; the next most widely grown Calrose medium
grains are the new releases M-205 and M-104 (Johnson 2002a,b) (Fig. 3). Table 4
contains a summary of the major physicochemical characteristics of several Calrose
medium grains. They have a low apparent amylose content and low gelatinization
temperature. The kernel size and shape are identifiable features of these varieties.
Cooking and processing characteristics, including desirability for breakfast cereals,
are recognized in the marketplace but are not well characterized by standard
laboratory testing methods. Environmental factors like climate and temperature in
the California rice-production region contribute to grain quality.
TABLE 4
Characteristics of California Medium-Grain Varietiesa
ACb % Proteinc KOH Scored GTe Brown Rice Kernelsf

Variety Type (%) Brown Milled 1.7% 1.5% (°C) Length Width L/W Weight
M-104 Calrose 17.9 8.2 7.2 6.7 6.0 68.5 6.3 2.8 2.3 24.2
M-202 Calrose 16.9 8.1 7.0 7.0 6.3 66.7 6.0 2.8 2.1 23.3
M-204 Calrose 18.4 7.7 7.0 7.0 6.3 66.6 6.2 2.8 2.2 24.7
M-205 Calrose 17.9 7.5 6.4 6.9 6.3 67.0 6.5 2.7 2.4 25.1
M-401 Premium 18.1 6.0 5.7 7.0 6.4 65.9 6.4 2.8 2.3 25.9
M-402 Premium 18.0 6.9 5.9 7.0 6.5 65.8 6.2 2.7 2.3 22.8
a Average of experiment samples (California Cooperative Research Foundation) from 2000 and 2001.
b Apparent amylose content.
c N × 5.95, dry basis.
d Dilute alkali reaction score, 1.0 to 7.0.
e Gelatinization temperature as measured with a differential scanning calorimeter.
f Kernel dimensions in millimeters, length (L)-width (W) ratio, and 1,000-kernel weight in grams.
Varieties / 71
TABLE 5
Characteristics of California Short-Grain Varietiesa
KOH
% Proteinc Scored Brown Rice Kernelsf
ACb GTe
S-102 Short 18.5 7.6 6.9 6.8 6.0 69.2 5.8 3.2 1.8 27.7
Koshihikari Premium 17.8 6.8 5.6 7.0 6.3 66.5 5.1 2.8 1.8 20.3
Akitakomachi Premium 17.2 7.4 6.5 7.0 6.0 68.9 5.3 2.9 1.8 21.6
Calhikari-201 Premium 18.0 7.0 6.0 7.0 6.0 68.3 5.1 3.0 1.7 20.5
Calmochi-101 Waxy 0.0 7.4 6.5 6.3 6.0 69.1 5.3 2.9 1.8 23.4
a Average of rice experiment station samples from 2000 and 2001.
Traditional California short grains have low amylose and low gelatinization tem-
perature (Table 5). The kernels are relatively large and may have some chalkiness.
Because this chalky spot or region is whiter than the surrounding endosperm, these
short grain types were referred to as “pearl” rice. In addition to table rice, the short
grains, such as S-102, are often used in the production of puffed rice cakes. Table 5
also contains the physicochemical characteristics for premium-quality short grains
grown in California. These short grains have a smaller, much more translucent ker-
nel than traditional short grains and produce very high whole-kernel milling yields.
Koshihikari, a Japanese short-grain variety released in the 1950s, is the established
standard for Japanese premium quality. The breeding, production, and quality of
Koshihikari have been recently reviewed by Iwate (2001). Other premium short
grains grown in California include Akitakomachi, a very early-maturing variety
developed in Japan, and Calhikari-201, a semidwarf variety developed in California.
Eating quality, considered one of the most important traits of rice in Japan, has been
the focus of extensive research as well as evaluation of rice for use and sale in the
marketplace. Near-infrared-based “Japanese taste machines” that measure compo-
nents like amylose, protein, moisture, K and Mg, and fatty acid content correlated
with taste panel results are used to analyze samples and issue a taste score for com-
merce in Japan. A review of rice grain quality from a Japanese perspective is avail-
able from Matsuo et al (1997).
Development of long grains for production in California faces both the agro-
nomic challenge of cold tolerance and the need to achieve the milling, cooking, and
processing properties found in long grains grown in the southern United States.
Because of the soft cooking tendency of California-grown, conventional long-grain
rice, breeding efforts have been directed toward developing adapted long grains that
cook firmer and less sticky. As part of this approach, L-205 was developed with the
Newrex quality that is characterized by having 2–3% higher amylose content and a
stronger viscogram profile than conventional long grains. Because of these charac-
teristics, Newrex types cook dry, exhibit minimal solids loss during cooking, and
are regarded as a superior type for canned soups, parboiling, and noodle making.
Considerable improvement in whole kernel milling yields have also been achieved
in the more recently released California long grains. Table 6 contains quality char-
acteristics for California long grains.
TABLE 6
Characteristics of California Long-Grain Varietiesa
KOH
% Proteinc Scored Brown Rice Kernelsf
ACb GTe
L-204 Long 22.7 7.8 7.4 4.7 3.9 73.8 7.9 2.3 3.4 24.6
L-205 Newrex 24.5 8.6 7.9 5.3 3.9 73.7 7.3 2.3 3.2 21.8
A-201 Aromatic 24.2 8.1 7.7 5.8 4.1 73.2 7.8 2.2 3.6 22.2
Calmati- Aromatic 23.3 10.0 9.5 6.3 6.0 73.8 7.3 2.1 3.5 20.5
201
a Average of rice experiment station samples from 2000 and 2001.
Specialty types include the waxy short grain Calmochi-101, the long grain aro-
matic A-201, and the aromatic basmati type Calmati-201. These special-purpose
varieties are usually grown under contract; some of their physicochemical charac-
teristics can be found in Tables 5 and 6. Interest in these and other specialty types
has increased significantly in recent years in both the public and private sectors.
Varieties of interest include jasmine, basmati, Mediterranean varieties such as
Arborio, and colored-bran types. Some common features of these types are that they
are generally ethnic foods, have low agronomic productivity, may present milling or
handling challenges, and lack established quality evaluation criteria, making them a
particularly challenging target for rice breeding or marketing.
One new quality type under investigation is “low-amylose” rice. Japanese rice
breeders have been working for many years to develop new rice cultivars with low
amylose content (5–10% amylase content) to improve eating quality and for new or
improved rice products (Okuno et al, 1993). Rice endosperm mutants have been
selected by several researchers (Okuno et al, 1983; Ise et al, 2000), using induced
mutation. The cultivars Milky Queen, Snow Pearl, and Silky Pearl have been
developed in Japan (Higashi et al, 1999; Ise et al, 2001; Takita et al, 2002) and are
in commercial production. The low-amylose rice is used in chilled or frozen rice
products because it is more resistant to retrogradation and to staling. Low-amylose
lines have been developed and are under agronomic and quality evaluation to
determine commercial interest and potential.
REFERENCES
Adair, C. R. 1956. Changes in rice varieties in Adair, C. R., Bollich, C. N., Bowman, D. H.,
the United States from 1931 to 1955. Pages 5- Jodon, N. E., Johnston, T. H., Webb, B. D.,
6 in: Int. Rice Comm. Newsl. 17. and Atkins, J. G. 1973. Rice breeding and test-
Adair, C. R., Beachell, H. M., Jodon, N. E., ing methods in the United States. Pages 22-75
Johnston, T. H., Thysell, J. R., Green, V. E., in: Rice in the United States: Varieties and
Jr., Webb, B. D., and Atkins, J. G. 1966. Rice Production. Agric. Handb. 289, U.S. Dep.
breeding and testing methods in the United Agric., Agric. Res. Serv., Washington, DC.
States. Pages 19-64 in: Rice in the United Ayres, N. M., McClung, A. M., Larkin, P. D.,
States: Varieties and Production. Agric. Bligh, H. F. J., Jones, C. A., and Park, W. D.
Handb. 289. U.S. Dep. Agric., Agric. Res. 1997. Microsatellites and a single-nucleodite
Serv., Washington, DC. polymorphism differentiate apparent amylose
Varieties / 73
classes in an extended pedigree of U.S. rice and sensory properties: How well do our
germ plasm. Theor. Appl. Genet. 94:773-781. measurements correlate with these properties?
Bautista, R. C., and Siebenmorgen, T. J. 2000. Pages 72-82 in: Rice Quality: Foundation for
Rice kernel properties affecting milling quality Value. E. T. Champagne, Ed. U.S. Dep.
at harvest. Pages 238-241 in: B. R. Wells Rice Agric., Agric. Res. Serv., Washington, DC,
Research Studies 1999. R. J. Norman and J. F. and USA Rice Federation, Houston, TX.
Meullenet, Eds. Ark. Agric. Exp. Stn., Fayette- Cnossen, A. G., and Siebenmorgen, T. J. 2000.
ville, AR. Determining minimum tempering durations
Bautista, R. C., Siebenmorgen, T. J., and Counce, while maintaining head rice yield. Pages 331-
P. A. 2000a. Characterization of individual 339 in: B. R. Wells Rice Research Studies
rice kernel moisture content and size distribu- 1999. R. J. Norman and C. A. Beyrouty, Eds.
tions at harvest and during drying. Pages 318- Ark. Agric. Exp. Stn., Fayetteville, AR.
324 in: B. R. Wells Rice Research Studies Dethloff, H. C. 1988. A history of the American
1999. R. J. Norman and C. A. Beyrouty, Eds. rice industry, 1685-1985. Texas A&M Univer-
Ark. Agric. Exp. Stn., Fayetteville, AR. sity Press, College Station.
Bautista, R. C., Siebenmorgen, T. J., and Dilday, R. H. 1990. Contribution of ancestral
Fendley, J. W. 2000b. Fissure formation char- lines in the development of new cultivars of
acterization in rice kernels during drying using rice. Crop Sci. 30:905-911.
video microscopy. Pages 325-330 in: B. R. FGIS. 1989. Federal Grain Inspection Service
Wells Rice Research Studies 1999. R. J. Rice Inspection Handbook. U.S. Dep. Agric.,
Norman and C. A. Beyrouty, Eds. Ark. Agric. Washington, DC.
Exp. Stn., Fayetteville, AR. Gravois, K. A., Moldenhauer, K. A. K., and
Beachell, H. M., and Halick, J. V. 1957. Breed- Rohman, P. C. 1991. Genetic and genotype X
ing for improved milling, processing and environment effects for rough rice and head
cooking characteristics of rice. Int. Rice rice yields. Crop Sci. 31:907-911.
Comm. Newsl. 6(2):1-7. Halick, J. V., and Kelly, V. J. 1959. Gelatiniza-
Beachell, H. M., and Stansel, J. W. 1963. tion and pasting characteristics of rice varieties
Selecting rice for specific cooking characteris- as related to cooking behavior. Cereal Chem.
tics in a breeding program. Pages 25-42 in: Int. 36:91-97
Rice Comm. Newsl. (Special Issue). FAO, Halick, J. V., and Keneaster, K. K. 1956. The use
Bangkok. of a starch-iodine-blue test as a quality indi-
Berrio, L. E., and Cuevas-Perez, F. E. 1989. cator of white milled rice. Cereal Chem.
Cultivar differences in milling yields under 5:315-319
delayed harvesting of rice. Crop Sci. 29:1510- Higashi, T., Saito, S., Takita, T., Yamaguchi, M.,
1512. Sunohara, Y., Yokogami, N., Ikeda, R.,
Berrio, L. E., Jennings, P. R., and Torres, E. A. Tamura, Y., Oyamada, Z., Kowata, H., Inoue,
2002. Breeding rice in Colombia for tolerance M., Matsumoto, S., and Kataoka, T. 1999.
to delayed harvest. Page 55 in: Proc. 29th Rice Breeding of a new rice cultivar with low
Technical Working Group, Little Rock, AR. amylose content “Snow Pearl.” Bull. Tohoku
Calderwood, D. L., Bollich, C. N., and Scott, J. Natl. Agric. Exp. Stn. 95, 1-12. (In Japanese,
E. 1980. Field drying of rough rice: Effect on English abstract)
grain yield. Milling quality, and energy saved. Ise, K., Tomita, K., JiShin, L., You Quan, S.,
Agron. J. 72:649-653. Sunohara, Y., and Kudo, S. 2000. Genetics
Carnahan, H. L., Johnson, C. W., Tseng, S. T., and characterization of mutants with low
and Mastenbrock, J. J. 1978. Registration of amylose content in endosperm induced from
‘M9’ rice. Crop Sci. 18:357-358. the rice variety Hexi 4 of Yunnan, China.
Chambers, B., and Childs, N. 2000. Milled rice Japanese J. Trop. Agric. 44:276-283. (In Japa-
distribution survey shows steady growth in nese, English abstract)
U.S. rice consumption. Pages 35-41 in: Rice Ise, K., Akama, Y., Horisue, N., Nakand, A.,
Situation and Outlook Yearbook. RCS-2000. Yokoo, M., Ando, I., Hata, T., Suto, M.,
Commodity Econ. Div., U.S. Dep. Agric., Namaguchi, K., Nemoto, H., Furutachi, H.,
Agric. Res. Serv., Washington, DC. and Imbe, T. 2001. “Milky Queen,” a new
Chambliss, C. E. 1920. Pages 1-26 in: Prairie rice high-quality rice cultivar with low amylose
culture in the United States. Farmers Bull. content in endosperm. Bull. Natl. Inst. Crop
1092. U.S Dep. Agric., Washington, DC. Sci. 2(12):39-61. (In Japanese, English
Chambliss, C. E., and Jenkins, J. M. 1923. Some abstract)
new varieties of rice. Bull. 1127. U.S. Dep. Iwate, I. 2001. Breeding, production, physiology
Agric., Washington, DC. and quality of the famous Japanese rice variety
Champagne, E. 1999. Traditional and emerging Koshihikari. Pages 243-284 in: Specialty
tools for assessing and predicting processing Rices of the World. R. C. Chaudhary, D. V.
Tran, and R. Duffy, Eds. FAO Science Pub- yields in rough rice. Cereal Chem. 70:650-
lishers, Enfield, NH. 655.
Jennings, P. R., Coffman, W. R., and Kauffman, Linscombe, S. L., Jodari, F., McKenzie, K. S.,
H. E. 1979. Rice Improvement. International Bollich, P. K., White, L. W., Groth, D. E., and
Rice Research Institute, Los Baños, Laguna, Dunand, R. T. 1993. Registration of ‘Bengal’
Philippines. rice. Crop Sci. 33:645-646.
Jodari, F., and Linscombe, S. D. 1996. Grain Linscombe, S. L., Jodari, F., Bollich, P. K.,
fissuring and milling yields of rice cultivars as Groth, D. E., White, L. M., Chu, Q. R.,
influenced by environmental conditions. Crop Dunand, R. T., and Sanders, D. 2000. Regis-
Sci. 36:1496-1502. tration of ‘Cocodrie’ rice. Crop Sci. 40:294.
Johnson, C. W. 2002a. Rice cultivar ‘M-205.’ Lisle, A. J., Martin, M., and Fitzgerald, M. A.
U.S. patent 6,492,582. 2000. Chalky and translucent rice grains differ
Johnson, C. W. 2002b. Rice cultivar ‘M-104’. in starch composition and structure and
U.S. patent 6,492,582. cooking properties. Cereal Chem. 77:627-632.
Johnson, C. W. 2003. Calrose Medium Grains. Little, R. R., Hilder, G. B., and Dawson, E. H.
Pages 29-36 in: 2003 Rice Breeding Progress 1958. Differential effect of dilute alkali on 25
Report and 2003 Research Proposal. Califor- varieties of milled white rice. Cereal Chem.
nia Cooperative Rice Research Foundation, 35:111-126.
Biggs, CA. Lu, R., and Siebenmorgen, T. J. 1995. Cor-
Johnson, C. W., Carnahan, H. L., Tseng, S. T., relation of head rice yield to selected physical
Oster, J. J., and Hill, J. E. 1986. Registration and mechanical properties of rice kernels.
of ‘M-202’ rice. Crop Sci. 26:198. Trans. ASAE 38:889-894.
Johnston, T. H. 1958. Registration of rice varie- Mackill, D. J. 1995. Classifying japonica rice
ties. Agron. J. 50:694-700. cultivars with RPD markers. Crop Sci. 35:889-
Johnston, T. H., Jodon, N. E., Bollich, C. N., and 894.
Rutger, J. N. 1972. The development of early Mackill, D. J., and McKenzie, K. S. 2003. Origin
maturing and nitrogen-responsive rice varie- and characteristics of U.S. rice cultivars.
ties in the United States. Pages 61-76 in: Rice Chapter 1.4 in: Rice: Origin, History, Tech-
Breeding. Int. Rice Res. Inst., Manila, Philip- nology, and Production. C. W. Smith and R.
pines. H. Dilday, Eds. John Wiley & Sons, Hoboken,
Jones, J. W. 1936. Improvement in Rice. Pages NJ.
415-454 in: Yearbook of Agriculture, USDA. Matsuo, T., Futsuhara Y., Kikuchi, F., and
U.S. Government Printing Office, Washington, Yamaguchi, H. 1997. Genetics. Page 1003
DC. in: Science of the Rice Plant. Japanese Min-
Jones, J. W., Jenkins, J. M., Nelson, M., Carter, istry of Agriculture, Forestry and Fisheries,
L. C., Adair, C. R., Wyche, R. H., Beachell, H. Tokyo.
M., Davis, L. L., and King, B. M. 1941. Rice McClung, A., Bergman, C., McKenzie, K.,
varieties and their comparative yields in the Moldenhauer, K., Rister, M. E., and
United States. Circ. 612. U.S. Dep. Agric., Holocombe, R. 1999. Breeding U.S. rice for
Washington, DC. conventional and high value markets. Pages
Jones, J. W., Dockins, J. O., Walker, R. K., and G1-G7 in: Proc. U.S. Japan Natural Resources
Davis, W. C. 1952. Pages 1-36 in: Rice pro- Panel, 27th. A. E. Pavlath and J. P. Cherry,
duction in the southern states. Farmers Bull. Eds. USDA-ARS, Eastern Reg. Res. Center,
2043. U.S. Dep. Agric., Washington, DC. Wyndmoor, PA.
Jones, J. W., Adair, C. R., Beachell, H. M., McKenzie, K. S., and Rutger, J. N. 1983. Genetic
Jodon, N. E., and Williams, A. H. 1953. Rice analysis of amylose content, alkali spreading
varieties and their yields in the United States. score, and grain dimensions in rice. Crop Sci.
Circular 915. U.S. Dep. Agric., Washington, 23:306-313.
DC. McKenzie, K. S., Johnson, C. W., Tseng, S. T.,
Juliano, B. O. 1990. Rice grain quality: Problems Oster, J. J., and Brandon, D. M. 1994. Breed-
and challenges. Cereal Foods World 35:245- ing improved rice cultivars for temperate
253. regions: A case study. Aust. J. Exp. Agric.
Juliano, B. O., and Perez, C. M. 1993. Critical 34:897-905.
moisture content for fissures in rough rice. Meullenet, J. F., Loire, S., and Suwansri, S. 2001.
Cereal Chem. 70:613-655. Prediction of consumer acceptance of cooked
Juliano, B. O., and Villareal, C. P. 1993. Grain jasmine-rice texture by texture profile analysis
Quality Evaluation of World Rices. Int. Rice and stress relaxation tests. Pages 296-303 in:
Res. Inst., Manila, Philippines. B. R. Wells Rice Research Studies 2000. R. J.
Juliano, B. O., Perez, C. M., and Cuevas-Perez, Norman and J. F. Meullenet, Eds. Ark. Agric.
F. 1993. Screening for stable high head rice Exp. Stn., Fayetteville, AR.
Varieties / 75
Meyers, R. 1995. Current domestic markets for Slaton, N., Moldenhauer, K. A. K, Gibbons, J.,
rice. Pages 41-56 in: Beyond the Tradition. E. Blocker, M., Bernhardt, J., Cartwright, R.,
T. Champagne, Ed. U.S. Dep. Agric., Agric. Lee, F., Dilday, R., Robinson, J., Koen, B.,
Res. Serv., Washington, DC, and USA Rice and Black, H. 1999. Arkansas Rice Perfor-
Federation, Houston, TX. mance Tirals, 1997-1999. Univ. Ark. Coop.
Moldenhauer, K. A. K., Gravois, K. A., Lee, F. Ext. Serv., Litle Rock, AR.
N., Norman, R. J., Dilday, R. H., Bernhardt, J. Suzuki, H., and Murayama, N. 1967. Effect of
L., Blocker, M. M., Rohman, P. C., and temperature on the rice grains and rice
McMinn, T. A. 1999. ‘Wells,’ a new high- starches. Pages 82-92 in: Int. Rice Comm.
yielding, long-grain rice cultivar. Pages 463- Newsl. (Special Issue). FAO, Bangkok.
468 in: Arkansas Rice Research Studies 1998. Takita, T., Higashi, T., Yokogami, N., Kataoka, T.,
R. J. Norman and T. H. Johnston, Eds. Univ. Kato, H., Yamaguchi, M., Tamura, Y., Kowata,
Ark. Agric. Exp. Stn., Fayetteville, AR. H., Oyamada, Z., and Sunohara, Y. 2002.
Moldenhauer, K. A. K., Gibbons, J. W., Lee, F. Breeding of a new rice cultivar with low amyl-
N., Bernhardt, J. L., Norman, R. J., Slaton, N. ose content “Silky Pearl.” Pages 1-12 in: Bull.
A., Wilson, C. E., Jr., Cartwright, R., Anders, Tohoku Natl. Agric. Res. Center for Tohoku
M., Blocker, M. M., Tolbert, A. C. 2001. Region 100. (In Japanese, English abstract)
‘Ahrent,’ a blast resistant long-grain rice Torres, E., Jennings, P. R., Duque, M. C., Kuri,
variety. Pages 27-32 in: Arkansas Rice V. E., Corredor, E., and Sierra, J. D. 2002.
Research Studies 2000. R. J. Norman and J. F. Analysis of stability for white belly in rice
Meullenet, Eds. Univ. Ark. Agric. Exp. Stn., (Oriza sativa). Foro Arrocero Latinoam.
Fayetteville, AR. 8(1[15]):9-13.
Moldenhauer, K. A. K., Gibbons, J. W., Lee, F. Tseng, S. T., Carnahan, H. L, Johnson, C. W.,
N., Norman, R. J., Bernhardt, J. L., Anders, M. Oster, J. J., Hill, J. E., and Scardaci, S. C.
M., Wilson, C. E., Cartwright, R., Anders, M. 1984. Registration of ‘L-202’ rice. Crop Sci.
M., Slaton N. A., Blocker, M. M., and Tolbert, 24:1213-1214.
A. C. 2002. ‘Francis,’ a high-yielding, long- USDA, 2000. Rice Situation and Outlook
grain rice variety. Pages 39-44 in: Arkansas Yearbook. RCS-2000. Commod. Econ. Div.,
Rice Research Studies 2001. R. J. Norman and U.S. Dep. Agric., Agric. Res. Serv., Wash-
J. F. Meullenet, Eds. Univ. Ark. Agric. Exp. ington, DC.
Stn., Fayetteville, AR. Webb, B. D. 1985. Criteria of rice quality in the
Okuno, K., Fuwa, H., and Yano, M. 1983. A new United States. Pages 403-442 in: Rice: Chem-
mutant gene for lowering amylose content in istry and Technology, 2nd ed. B. O. Juliano,
endosperm starch of rice. Oryza sativa L. Jpn. Ed. Am. Assoc. Cereal Chem., St. Paul, MN.
J. Breed. 33:387-395. Webb, B. D., and Bollich, C. N. 1980. Cooking
Okuno, K., Nagamine, T., Oka, M., Kawase, and processing quality characteristics of
Katsuta, M., Egawa, U., and Nakagahra, M. Newrex rice. Pages 61-62 in: Proc. Rice
1993. New lines harboring du genes for low Technical Working Group, 18th. Texas Agric.
amylose content in endosperm starch of rice. Exp. Stn., College Station.
Jpn. Agric. Res. Q. 27:102-105. Webb, B. D., Bollich, C. N., Jodon, N. E.,
Resurreccion, A. P., Hara, T., Juliano, B. O., and Johnston, T. H., and Bowman, D. H. 1972.
Yoshida, S. 1977. Effect of temperature during Evaluating the milling, cooking, and process-
ripening on grain quality of rice. Soil Sci. ing characteristic required of rice varieties in
Plant Nutr. 23:109-112. the United States. Pub. ARS-S-1. U.S. Dep.
Rutger, J. N., Peterson, M. L., and Hu, C. H. Agric., Agric. Res. Serv., Southern Region,
1977. Registration of ‘Calrose 76’ rice. Crop New Orleans, LA.
Sci. 17:987. Webb, B. D., Bollich, C. N., Carnahan, H. J.,
Rutger, J. N., McKenzie, K. S., Moldenhauer, K. Kuenzel, K. A., and McKenzie K. S. 1985.
A. K., McClung, A. M., Gravois, K. A., Utilization characteristics and qualities of
Linscombe, S. D., and Kanter, D. G. 2000. United States rice. Pages 26-35 in: Rice Grain
Rice quality breeding efforts in the U.S. Pages Quality and Marketing. Int. Rice Res. Inst.,
311-316 in: Specialty Rices of the World. R. Los Baños, Laguna, Philippines.
C. Chaudhury, D. V. Tran, and R. Duffy, Eds. Wilson, J. H., Ed. 1979. Rice in California. Butte
FAO Science Publishers, Enfield, NH. County Rice Growers, Richvale, CA.
Slaton, N., Moldenhauer, K. A. K., Blocker, M., Yoshida, S., and Hara, T. 1977. Influence of air
Bernhardt, J., Cartwright, R., Lee, F., Dilday, temperature and light on grain filling of an
R., Robinson, J., Black, H. 1998. Arkansas indica and a japonica rice (Oryza sativa L.) in
Rice Performance Trials, 1996-1998. Univ. a controlled environment. Soil Sci. Plant Nutr.
Ark. Coop. Ext. Serv., Litle Rock, AR. 23:93-107.
CHAPTER 4
THE RICE GRAIN AND ITS GROSS COMPOSITION
Elaine T. Champagne
Southern Regional Research Center
New Orleans, Louisiana
Delilah F. Wood
Western Regional Research Center
Albany, California
Bienvenido O. Juliano
Philippine Rice Research Institute Los Baños
Laguna, Philippines
Donald B. Bechtel
Grain Marketing Research Laboratory
Manhattan, Kansas
GROSS STRUCTURE OF THE RICE GRAIN
Botanical nomenclature, particularly of the reproductive plant parts, is based

upon the origin of each component during development and maturation of the seed
or that portion of the plant that carries genetic material of both parents into subse-
quent generations (Esau, 1977). However, the point of this chapter is to acquaint the
reader with rice structure and composition as it relates to processing and consump-
tion; therefore, only the mature grain is discussed.
Overall Structure
Rice (Oryza sativa L.) is a member of the family Poaceae (formerly Gramineae
or grass), and its structure has been described by numerous investigators (Juliano
and Aldama, 1937; Juliano, 1972, 1980; Bechtel and Pomeranz, 1980; Chung and
Wu, 1983; Juliano and Bechtel, 1985; Hoshikawa, 1993; Horigane et al, 2001;
Evers and Millar, 2002; Ogawa et al, 2002). Rice is harvested as paddy rice, or, in
botanical terms, “spikelets.” Spikelets from different cultivars may vary in
proportion but have the same basic structure (Takeoka et al, 1989) (Fig. 1). Each
77
Fig. 1. Scanning electron micrographs of mature, intact paddy rice (spikelets). A, Calmochi-101,
waxy; B, S-102, short grain; C, M-202, medium grain; D, Bengal, medium grain; and E, Cypress, long
grain. Note the multitude of hairlike structures (trichromes) on the outer surfaces of the palea and
lemma, especially toward the top of the micrographs in A and B, and the lack of trichromes on the
surfaces of C–E (which are therefore described as “glabrous”). L = lemma, lg = lower glume, P =
palea, R = rachilla, T = trichomes, ug = upper glume. Scale bar = 2 mm.
Fig. 2. Scanning electron micrographs of mature, dissected spikelets, showing the components. A,
Calmochi-101, waxy; B, Cypress, long grain. The palea and lemma of Calmochi have trichomes on
their surfaces, whereas those of Cypress are glabrous. C = caryopsis, E = embryo, L = lemma, lg =
lower glume, P = palea, R = rachilla, ug = upper glume. Scale bar = 2 mm.
Grain Composition / 79
spikelet consists of a caryopsis (the single, mature brown rice kernel), four bracts
(leaf-like structures subtending an inflorescence), and a rachilla (the main axis of
the grass flower) (Fig. 2). The two major bracts, the palea and lemma, completely
encase the caryopsis. Some rice cultivars have an awn at the tip of the lemma
(Niitsuma, 1993). The palea and lemma are subtended by two smaller bracts, the
upper and lower glumes (Figs. 1 and 2). The bracts and caryopsis are held together
and supported by a rachilla (Figs. 1 and 2) and pedicel (not shown) at the base of the
spikelet. The bracts, rachilla, and pedicel make up the “hull,” “chaff,” or “husk” of
the rice harvest (Hoshikawa, 1993).
Hull
The major structural features of the hull are the lemma and palea. The lemma
and palea may have trichomes on their outer epidermal surfaces in some cultivars
(Figs. 1A,B and 3), while others are glabrous (have a surface devoid of hairs or
other projections) (Fig. 1C–E). The lemma and palea differ in their sizes, shapes,
and numbers of vascular bundles. The lemma has five vascular bundles and the
smaller palea has three (Fig. 3A,B) (Hoshikawa, 1993). The lemma is larger than
the palea, boat-shaped, and nearly encircles the caryopsis (Figs. 3A,B and 4A). The
lemma overlaps and interlocks with the palea at the edges of both structures (Figs.
3A,B and 4A,B). The overlapping arrangement is a structural mechanism that pro-
tects the caryopsis from pests before dehulling (Juliano, 1981; Sauphanor, 1988).
Both the lemma and palea consist of four structural layers, including an outer
epidermis, sclerenchyma, and spongy parenchyma and an inner epidermis (Figs. 3D
and 4C). The cells of the outer epidermis run parallel to the longitudinal axis of the
spikelet. Outer epidermal cells are rigid, with thickened cell walls due to the pres-
ence of silicic acid or silica (Fig. 4C). Trichomes, if they exist, project from be-
tween the longitudinal rows of silica-containing epidermal cells (Fig. 3D) (Juliano
and Bechtel, 1985; Hoshikawa, 1993). The second structural layer beneath the outer
epidermis is a fibrous layer of cells two to three cells in thickness. These cells are
fibrous because they are elongated and have highly lignified cell walls; this tissue is
termed sclerenchyma (Fig. 4C) (Esau, 1977). The third layer consists of crushed,
spongy parenchyma, or undifferentiated cells, where vascular bundles are located
when present (Fig. 4C). Finally, the inner epidermis consists mostly of isodiametric
cells (Figs. 3C,D and 4C) (Juliano and Aldama, 1937; Kaneko et al, 1991).
Caryopsis
The brown rice grain, an indehiscent, single-seeded fruit botanically known as a

“caryopsis” (Figs. 2A,B and 5), consists of a seed and a tightly adherent fruit coat or
pericarp (Esau, 1977).
Pericarp
The pericarp comprises a thin cuticle covering the epidermis; several layers of
partly crushed, “spongy” parenchyma; cross cells (with lignified cell walls), which
are elongated in the transverse axis of the grain; and tube cells, remnants of the
inner epidermis, which are lignified and elongated in the longitudinal axis of the
grain (Figs. 5 and 6) (Esau, 1977; Hoshikawa, 1993). The cross and tube cells give
the pericarp its fibrous structure. The spongy parenchyma layers of the pericarp
contain four vascular bundles, which run along the ventral (embryo), dorsal (back),
and the two lateral sides of the grain. The dorsal vascular bundle is the largest in
diameter and serves to transport nutrients to the seed during development
(Hoshikawa, 1993). The top of the pericarp contains the vestiges of the stigma and
pistil (female flower parts), and the pericarp bottom has a scar where the caryopsis
was connected to the rachilla, just behind the bottom of the embryo.
Fig. 3. Illustrations showing five vascular bundles in the lemma and three in the palea of the hull about
the center (A) and near the base (B). C, the overlapping and interlocking arrangement of the lemma
and palea where they join. D, a cross-section of either a palea or lemma where a trichrome projects
from between silica-containing cells on the epidermis. (Modified and reprinted, with permission, from
Hoshikawa, 1993)
Seed
There is no clear delineation between the seed and the caryopsis due to the fusion
between the outer tissues of the seed and those of the pericarp. However, the seed
includes the testa or seed coat, nucellus (Hoshikawa [1993] terms this the “exo-
sperm”), embryo, and endosperm. The tissues of the seed differ in structure, com-
position, and function. The cells making up the bulk of the seed are storehouses for
Fig. 4. Scanning electron micrographs of a fractured Cypress spikelet at about the midsection of the
embryo. A, transverse fracture of hull; B, the overlapping and “interlocking” of the lemma and palea; C,
cross-sectional view of the lemma, which is also representative of the palea. Em = embryo, En = endo-
sperm, L = lemma, P = palea, Al = aleurone, circle = vascular bundle, oe = outer epidermis, scl = scleren-
chyma, sp = spongy parenchyma, ie = inner epidermis. Scale bars = 0.5 mm (A); 100 µm (B); 25 µm (C).
nutrients, which are important not only as a means of ensuring an adequate energy
supply for the germinating seed, but also as an exploited food supply for humans
and animals.
TESTA AND NUCELLUS

Just beneath the pericarp is the testa (also known as the seed coat), consisting of
a single crushed layer of cells, the testa cuticle, the nucellar cuticle, and the remains
of the nucellar cells (Figs. 5 and 6) (Bechtel and Pomeranz, 1977; Ellis and
Chaffey, 1987). The testa and nucellus are fused to the surrounding tissues. How-
ever, the bond between the cuticles of the testa and nucellus tends to be weak, and
the tissues may separate at the junction between the two (Bechtel and Pomeranz,
1977). The pigments in colored rice are located in the testa or in the pericarp.
EMBRYO
The embryo (germ) is roughly 1–3% by weight of the total grain (Juliano and
Bechtel, 1985) and is located at the basal end on the ventral (abaxial) side of the grain.
The embryo consists of the scutellum, embryonic axis (plumule [leaf], hypocotyl
[stem], and radicle [root]), and various structures that sheath the embryo (Figs. 5 and
7). The scutellum is the largest tissue of the embryo and runs between the starchy
Fig. 5. Illustrations of a longitudinal section of a rice spikelet.

endosperm and the embryonic axis. The scutellum has three appendages, the ventral
scale and two lateral scales, that partially sheath the coleoptile. The coleoptile sheaths
the plumule. The plumule is connected to the hypocotyl, or short stem (consisting of
parenchyma and provascular tissue), and then the radicle. The radicle, made up of the
root cap, root apex, epidermis, hypodermis, cortex, endodermis, pericycle, and provas-
cular tissue, is sheathed by the coleorhiza, a structure at the base of the scutellum.
Upward and ventral to the coleorhiza is the epiblast, which sheaths the lower half of
the embryonic axis (Fig. 7) (Bechtel and Pomeranz, 1978a; Jones and Rost, 1989).
Embryonic cells vary in their contents, which are largely dependent upon their
functions, although virtually all cells contain lipid droplets. Meristematic and
provascular tissues contain lipid droplets and cellular machinery for growth. Tissues
whose function is primarily or partly storage (including the scutellum, coleoptile,
and coleorhiza) also contain protein bodies (Fig. 8A), which contain globoid inclu-
sions (Fig. 8B). Globoids are rich in phytin, a cationic salt of inositol hexaphos-
phoric acid. They are also rich in phosphorus and contain magnesium, potassium,
and (relatively little) calcium (Dikeman et al, 1981). Epidermal tissue contains lipid
droplets, as well as protein bodies that may or may not contain globoid inclusions
(Tanaka et al, 1976; Bechtel and Pomeranz, 1978a; Fulcher and Wong, 1980, 1982;
Fulcher et al, 1981, 1989; Wada and Lott, 1997).
ENDOSPERM
Aleurone. The aleurone layer, the outermost layer of the endosperm, is bound
tightly to the starchy endosperm and may vary from one (Fig. 8C) to several cells in
thickness in the same grain (Fig. 8D). The aleurone layer has more cell thicknesses
on the dorsal (back) than along the lateral (side) and ventral (front, embryo side)
Fig. 6. Illustrations from Hoshikawa (1993) showing the microstructure of outer layers of the caryopsis
and the location of vascular bundles. (Modified and reprinted, with permission, from Hoshikawa,
1993)
surfaces of the grain. Aleurone cells are cuboidal and have densely packed cyto-
plasm with cellular organelles (including nuclei, microbodies, mitochondria) and
endoplasmic reticulum and storage structures (including lipid droplets and protein
bodies). The protein bodies contain globoid inclusions (Fig. 8E) (Juliano and
Bechtel, 1985).
Fig. 7. Montage of a sagittal section of a 45-day-old rice embryo. The embryo has a plumule sheathed
by the coleoptile, a radicle sheathed by the coleorhiza, an epiblast, and a scutellum. A ventral and two
lateral scales (appendages of the scutellum) sheath the apex of the coleoptile. The depleted layer,
consisting of the remnants of digested endosperm, lies between the scutellum and the endosperm. The
modified aleurone layer (AL) is continuous around the embryo except where the embryo was attached
to the nucellus (*, between the small arrows) during the early stages of development. The aleurone
cells are attached to the top and bottom of the embryo and loosely cover the remainder of the embryo
between the ventral scale and the region that joins the scutellum to the coleorhiza, roughly between the
arrowheads. Cr = coleorhiza, Ct = coleoptile, DL = depleted layer, Ep = epiblast, LS = lateral scale, Pl
= plumule, R = radicle, Sc = scutellum, VS = ventral scale. Scale bar = 0.25 mm. (Reprinted, with
permission, from Jones and Rost, 1989)
Fig. 8. Light micrograph of rice caryopsis sections. A, fluorescence micrograph of the scutellum
portion of the embryo of Calrose, showing protein bodies. The section was stained with acriflavine
HCl and observed at 390–420 nm. (Note that proteins are autofluorescent, and the acriflavine HCl
increases the fluorescence intensity.) B, fluorescence micrograph of Calrose scutellum, showing
globoids (G), inclusions of phytic acid located within protein bodies, stained with acriflavine HCl,
excitation 550–590 nm (Fulcher and Wong 1980, 1982). C, Brightfield micrograph showing the
pericarp (P), aleurone cells (Al), compound starch granules (Sg), the subaleurone (sAl), and protein
bodies (Pr) at the periphery of a Calrose caryopsis section; the section was stained with acid fuchsin,
fast green, and iodine potassium iodide. D, fluorescence micrograph of Calrose, showing the pericarp,
a multiple aleurone cell layer, compound starch granules, protein bodies (shown by arrows), the
subaleurone of the starchy endosperm, and the starchy endosperm (En); the section was stained with
Safranin O and viewed at 450–490 nm excitation. E, fluorescence micrograph showing globoids in the
aleurone cells of Calrose. F, Brightfield micrograph of M-202 (medium grain, nonwaxy) stained with
iodine potassium iodide. G, Brightfield micrograph of Calmochi-101 (medium grain, waxy) stained
with iodine potassium iodide. Scale bars = 50 µm (A,B); 25 µm (C–E); 50 µm (F,G).
A modified aleurone cell layer is continuous around the embryo, except where
the embryo was attached to the nucellus (Jones and Rost, 1989) (Fig. 7). This layer
differs from the typical aleurone cells in that the modified cells are rectangular in
shape, have less densely packed cytoplasm, have numerous vesicles and filament
bundles, have fewer and smaller lipid bodies, and lack protein bodies (Bechtel and
Pomeranz, 1977). The modified aleurone cells are attached at the top and bottom of
the embryo and loosely cover the remainder of the embryo between the ventral
scale and the region that joins the scutellum to the coleorhiza (Figs. 5 and 7).
Starchy endosperm. The starchy endosperm constitutes the largest portion of the
rice grain and serves as a huge storage reserve for the germinating and growing
seedling (Krishnan et al, 2001) and as a major food source for humans and animals.
It is composed of nonmetabolic tissue in the mature grain because the cytoplasm
and organelles are digested during the final maturation stage (Esau, 1977; Inouchi et
al, 2000). Mature tissue consists of thin-walled parenchyma cells, usually radially
elongated, filled with compound starch granules and protein bodies. Lateral cells,
those toward the sides of the grain, are polygonal or slightly elongated in shape;
dorsal-ventral cells, those that extend from the ventral to the dorsal side of the grain,
are greatly elongated in the dorsal or ventral direction (Little and Dawson, 1960;
Juliano and Bechtel, 1985). Generally, the radial dimensions of the dorsal-ventral
cells are shorter in long-grain than in medium- or short-grain cultivars (Juliano and
Bechtel, 1985). The starchy endosperm exhibits several types of gradients from its
periphery to its center. At the periphery of the starchy endosperm, also called the
subaleurone region or layer, the cells are relatively small, protein bodies more
abundant, and starch less abundant (Fig. 8C,D) compared with those at the center of
the grain (Fig. 8D,E). Moving toward the center of the grain, the cells become
larger, protein decreases, starch increases, and the compound starch granules be-
come larger (Fig. 8D,E). However, at the center of the grain, a core of small cells
having isodiametric shapes is apparent (Hoshikawa, 1993).
The bulk of the starchy endosperm consists of starch in the form of large,
polygonal, compound starch granules (3–9 µm in diameter) (Bechtel and Pomeranz,
1978b; Asaoka et al, 1985a). Starch granules are composed of amylose, the long-
chain starch molecule, and amylopectin, the branched starch molecule (Whistler et
al, 1984; Baldwin, 2001). Starch of normal, or nonwaxy, cultivars stains blue with
iodine potassium iodide (Fig. 8F). The starch in waxy cultivars (i.e., those
containing an abundance of amylopectin and little or no amylose) stains purple to
brown with iodine potassium iodide (Fig. 8G). Additionally, the starch granules of
waxy cultivars contain pores within the individual starch, and the compound
granules contain spaces between the clusters of individual granules (Fig. 9A),
whereas the nonwaxy cultivars show no evidence of such pores or spaces, in
agreement with Utsunomiya et al (1975). In Figure 9B, the grains were soaked in
water before preparation for scanning electron microscopy to ensure fracturing of
starch granules, whereas, for results shown in Figure 9C and D, the grains were
fractured dry.
In addition to starch, the starchy endosperm contains storage protein (Ellis et al,
1987) in the form of spherical bodies (Figs. 8C,D and 9E,F), of which three types
have been reported (Juliano and Bechtel, 1985). The subaleurone region contains 1)
large spherical protein bodies (1–2 µm in diameter, lamellar, electron-dense, com-
pletely pepsin-digestible, and partially pronase-digestible), 2) small spherical pro-
tein bodies (0.5–0.75 µm in diameter, lamellar, completely pepsin- and pronase-
Fig. 9. Scanning electron micrographs (SEMs). A, Calmochi-101, waxy grain, with pores (Po) or
cavities inside the fractured starch granules and spaces (Sp) between the starch granules in a
compound starch granule. B, M-202, nonwaxy medium grain. The grain was soaked in water before
SEM preparation to fracture the starch granules to show lack of pores or cavities. The spaces between
individual starch granules within a compound granule are due to swelling and subsequent shrinkage, a
preparation artifact. C, dry fracture of M-202 from the periphery (right) to the interior of the starchy
endosperm; cells and compound starch granules are larger toward the center of the grain. D, dry
fracture of M-202 showing unfractured compound starch granules in the central endosperm and very
few protein bodies (Pr). Note that a thin endosperm call wall (cw) is visible. E, Calmochi-101
approaching the central endosperm showing compound starch granules and protein bodies. F, isolated
protein bodies, which have fused during the isolation process. Scale bars = 10 µm (A); 5 µm (B); 25
µm (C); 10 µm (D); 5 µm (E,F).
digestible), and 3) crystalline protein bodies (2–3.5 µm in diameter, crystalline,

completely pepsin-digestible, and partially pronase-digestible). The central starchy
endosperm has only the large spherical protein bodies. When the protein bodies are
isolated, they have a tendency to fuse, supporting the idea that they contain a single
membrane (Fig. 9F) (Juliano and Bechtel, 1985).
COMPOSITION OF THE RICE GRAIN AND ITS MILLING FRACTIONS
Proximate Analysis of Rough Rice and Its Fractions
The composition and properties of rough rice and its fractions are subject to
varietal, environmental, and processing variability. A wide range of values is evi-
dent for all milling fractions (Tables 1–3).
Hull
The rice hull represents 20% of the rough rice grain and is composed of
approximately 20% silica. This high concentration of silica is located in the outer
layers and, with the lignin (9–20%), provides physical protection of the grain from
attack by insects and fungi. The silica is amorphous and has been recovered by various
methods (e.g., acid leaching and gasification) (Luan and Chou, 1990; Chakraverty and
Kaleemullah, 1991). Cutin, a polymer of long-chain hydroxymonocarboxylic acid, a
water-repellent material covering the outer layers of rice hulls, makes up 2–6% of the
hull (Luh, 1991). The major carbohydrates are cellulose, crude fiber, and hemicellu-
lose (Table 1). Unlike the bran and endosperm fractions, the hulls are void of starch.
Protein and lipid contents are low. Hull lipids from a Brazilian variety were shown to
have a content of unsaponifiable matter four times higher than those from the bran
TABLE 1
Range of Mean Proximate Analysis and Content (%) of Organic Fractions of Rough Rice
and Its Milling Fractions at 14% Moisturea
Nutrient Rough Brown Milled Hull Bran Embryo Polish
Protein (N × 5.95) 5.8–7.7 4.3–18.2 4.5–10.5 2.0–2.8 11.3–14.9 14.1–20.6 11.2–12.4
Crude fat 1.5–2.3 1.6–2.8 0.3–0.5 0.3–0.8 15.0–19.7 16.6–20.5 10.1–12.4
Crude fiber 7.2–10.4 0.6–1.0 0.2–0.5 34.5–45.9 7.0–11.4 2.4–3.5 2.3–3.2
Crude ash 2.9–5.2 1.0–1.5 0.3–0.8 13.2–21.0 6.6–9.9 4.8–8.7 5.2–7.3
Available
carbohydrates 64–73 73–87 77–89 22–34 34–62 34–41 51–55
Starch 53.4 66.4 77.6 1.5 13.8 2.1 41.5–47.6
Neutral detergent
fiber 16.4–19.2 2.9–3.9 0.7–2.3 65.5–74.0 23.7–28.6 13.1 –
Pentosans 3.7–5.3 1.2–2.1 0.5–1.4 17.7;18.4 7.0;8.3 4.9;6.4 3.6–4.7
Hemicelluloses – – 0.1 2.9;11.8 9.5–16.9 9.7 –
Cellulose – – 31.4–36.3 5.9–9.0 2.7 –
1,3:1,4 b-glucans – 0.11 0.11 – – – –
Polyuronic acid 0.6 – – – 1.2 0.4 –
Free sugars 0.5–1.2 0.7–1.3 0.22–0.45 0.6 5.5–6.9 8.0–12 –
Lignin 3.4 – 0.1 9.5–18.4 2.8–3.9 0.7–4.1 2.8
Energy (kJ/g) 15.8 15.2–16.1 14.6–15.6 11.1–13.9 16.7–19.9 17.9
a Adapted from Table 1 in Juliano and Bechtel (1985); Gomez (1979); Eggum et al (1982); Pedersen
and Eggum (1983); and Bett-Garber et al (2001).
and caryopsis (Hartman and Lago, 1976). The unsaponifiable matter consisted of
hydrocarbons, alcohols, and sterols, with the sterols consisting of 50% b-sitosterol,
20–40% campesterol, 10–20% stigmasterol, and 2–3% cholesterol. Differences in
fatty acid composition were observed relative to the caryopsis, with the presence of
2–3% saturated C-22 and C-24 acids and a lower proportion of unsaturated acids.
TABLE 2
Range of Mean Content of Elements of Rough Rice and Its Milling Fractionsa
Element Rough Brown Milled Hull Bran Embryo Polish
Macroelementsb
Calcium 0.1–0.8 0.1–0.5 0.1–0.3 0.6–1.3 0.3–1.2 0.2–1.0 0.5–0.7
Magnesium 0.6–1.5 0.2–1.5 0.2–0.5 0.3 5–13 4–13 6–7
Phosphorus 1.7–3.9 1.7–4.3 0.8–1.5 0.3–0.7 11–25 10–21 10–22
Phytin phosphorus 1.8–2.1 1.3–2.7 0.3–0.7 0 9–22 7–16 12–17
Potassium 1.5–3.7 0.6–2.8 0.7–1.3 1.5–7.5 10–20 11–15 7;11
Silicon 10.8 0.6–1.4 0.1–0.4 64–95 3–5 0.4–0.9 1.1;1.6
Sulfur 0.4–1.6 0.3–1.9 0.8 0.4 1.7 – 1.6
Microelementsc
Aluminum 26–540 0.3;26 0.1;2.2 52 200 – –
Cadmium – 0.02–0.16 0.025 – – – –
Chlorine 500–800 210–560 200–300 860 66 1200 –
Cobalt – 0.03–0.04 0.017 – – – 0.05
Copper 2–11 1–6 2–3 30–39 9–34 9–34 5–26
Iodine – 0.03 0.02 – – – –
Iron 14–60 2–52 2–28 39–95 86–430 60–180 43–155
Manganese 17–94 2–36 6–17 100–290 95–230 91–120 –
Molybdenum – 0.3–1.0 1.4 – – – –
Nickel – 0.2–0.5 0.14 – – – –
Selenium – 0.3 0.3 – – – –
Sodium 53–810 17–340 5–86 67–826 71–335 139–636 trace–138
Tin – – <1.1 0 21 – –
Zinc 1.7–31 6–28 6–23 9–40 43–258 57–258 17;60
a Adapted from Table 3 in Juliano and Bechtel (1985).
b mg/g at 14% moisture.
c µg/g at 14% moisture.
TABLE 3
Range of Mean Vitamin Content (µg/g at 14% moisture) of Rough Rice
and Its Milling Fractionsa
Vitamin Rough Brown Milled Hull Bran Embryo Polish
Retinol (A) 0–0.08 0–0.11 0–trace 0 0–3.6 0–1.0 0–0.9
Thiamin (B1) 2.6–3.3 2.9–6.1 0.2–1.1 0.9–2.1 12–24 17–59 3–19
Riboflavin (B2) 0.6–1.1 0.4–1.4 0.2–0.6 0.5–0.7 1.8–4.3 1.7–4.3 1.7–2.4
Niacin (nicotinic acid) 29–56 35–53 13–24 16–42 267–499 28–83 224–389
Pyridoxine (B6) 4–7 5–9 0.4–1.2 – 9–28 13–15 9–27
Panthothenic acid 7–12 9–15 3–7 – 20–61 11–28 26–56
Biotin 0.04–0.08 0.04–0.10 0.01–0.06 – 0.2–0.5 0.3–0.5 0.1–0.6
Inositol, total 800 1000 90–110 – 4000;8000 3200;5500 3700;3900
Choline, total 760–980 950 390–880 – 920–1460 1700;2600 860–1250
p-Aminobenzoic acid 0.3 0.3 0.12–0.14 – 0.65 0.9 0.6
Folic acid 0.2–0.4 0.1–0.5 0.03–0.14 – 0.4–1.4 0.8–4.1 0.9–1.8
Cyanocobalamin (B12) 0–0.003 0–0.004 0–0.0014 – 0–0.004 0–0.01 0–0.003
a-Tocopherol (E) 9–20 9–25 trace–3 – 26–130 76 54–86
a Reprinted from Table 4 in Juliano and Bechtel (1985).
In addition to the protective role, rice hulls exhibit antioxidative defense systems
that protect the ability of the mature rice seed to germinate during storage (Osawa et
al, 1992). Isovitexin, an antioxidant, has been isolated from indica rice hulls (Osawa
et al, 1992). Phenolic-containing fractions isolated from hulls exhibited antioxidant
activity stronger than that of a-tocopherol (Ramarathnam et al, 1986, 1988).
Ferulic, vanillic, p-hydroxybenzoic, p-coumaric, and indolacetic acids and p-
hydroxybenzaldehyde were identified in rice hull extracts that inhibited germination
(Mikkelsen and Sinah, 1961). Hull extracts from some of 91 cultivars were demon-
strated to have significant allelopathic potential in inhibiting barnyard grass germi-
nation, seedling growth, weight, and caloric content (Ahn et al, 2000).
Compounds that regulate plant growth also have been isolated from rice hulls.
One such compound was nicotinamide, a plant-growth regulator (Takeuchi et al,
1975). A chitinase specific to rice hulls also has been isolated and characterized
(Baek et al, 2001). Chitinases play roles in defending plants (Boller, 1988), regu-
lating plant development (De Jong et al, 1992), and signal transduction (Roche et al,
1991). Table 2 lists the mineral content of hulls. These data are from independent
studies in southeast Asia and the United States and were reported in the previous
edition (1985) of this monograph. Marr et al (1995) examined the uptake of ele-
ments by brown rice, hulls, and straw (Table 4). Hulls generally removed smaller
amounts of minerals; the concentrations of phosphorus, potassium, magnesium,
sulfur, zinc, copper, and molybdenum were considerably lower in them than in the
TABLE 4
Removal of Elements by an Average Amaroo Crop Yielding 10 t ha-1 Paddya,b
Nutrient
In 8.4 t ha-1 Total Harvest Indexe
Nutrients Brown In 1.6 t ha-1 In 11 t ha-1 Nutrients (at 10 t ha-1)
Removed Grainc Hullsc Strawd Removed (%)
kg ha–1
Nitrogen 99 3.7 70 172.7 57
Phosphorus 24 0.3 5.5 29.8 81
Potassim 20 9 243 272 7
Magnesium 9 0.6 17 26.6 34
Sulfur 8 0.5 6.6 15.1 53
Calcium 0.8 1.4 26 28.2 3
Manganese 0.3 0.6 6.7 7.6 4
Sodium 0.2 0.3 30.8 31.3 1
Aluminum 0.1 0.1 7.1 7.3 1
Iron 0.1 0.1 2.5 2.7 4
g ha–1
Zinc 134 25 297 456 29
Copper 34 2 40 76 45
Boron 29 20 94 143 20
Molybdenum 5.9 1.4 29 36.3 16
Nickel 3.4 0.6 0.6 4.6 74
a Reprinted, with permission, from Marr et al (1995).
b A crop that yields 10 t ha–1 of paddy will, on average, have 8.4 t ha–1 brown rice, 1.6 t ha–1 hulls, and
produce 11 t ha–1 straw.
c Expressed at 14% moisture.
d Expressed at 0% moisture.
e Nutrient harvest index = amount in grain (kg ha–1)/amount in grain, hulls, and straw (kg ha–1) ×
100/1.
brown rice or straw. Hulls removed higher amounts of calcium, manganese, and
sodium than brown rice.
Vitamin contents are listed in Table 3. Values have been reported for the B vita-
mins (thiamin, riboflavin, and niacin) (Hsu and Luh, 1980).
Bran
Commercial bran makes up 10–15% of rough rice and may contain varying pro-
portions (up to 20%) of polish. The germ usually is included in the bran fraction
unless removed by sieving.
Rice bran is an excellent source of protein (12–15%) and lipids (15–20%).
Chapters 6 and 7 discuss these components in detail. The major proteins in bran are
albumin and globulin, with the albumin-globulin-prolamin-glutelin ratio being
37:36:5:22 for a set of six samples (Cagampang et al, 1966). Rice proteins are con-
sidered to be hypoallergenic (Helm and Burks, 1996). However, proteins with
molecular masses of 14–16, 26, 33, and 56 kDa found in the bran and endosperm
have been shown to be potentially allergenic, based on immunoglobulin E binding
(Urisu et al, 1991). The homologous group of 14- to16-kDa protein recognized by
90–95% of patients with rice allergy are a-amylase/trypsin inhibitors (Izumi et al,
1992, Adachi et al, 1993, Nakase et al, 1996). The 26-kDa protein is an a-globulin
(Limas et al, 1990). The 33-kDa protein is a novel type of plant glyoxalase I (Usui
et al, 2001). The 56-kDa protein has not been identified. More than 150 rice
varieties in Japan were screened for the 16-kDa protein and all varieties tested
contained nearly the same amount (Adachi et al, 1995). In contrast, some varieties
in other Asian countries contained little or none of the allergen.
A thiamin-binding protein in rice bran and germ has been isolated and charac-
terized (Nishino et al, 1980; Nishimura et al, 1984; Shimizu et al, 1996). A
fibronectin-binding protein with cell adhesion activity for animal tumor cells also
has been identified (Shoji et al, 2001) The protein has an amino-terminal acid
sequence identical to that of a putative mature form of hydroxyproline-rich glyco-
protein. A distinct feature of the amino acid composition of the protein was the
high content of hydroxyproline and proline, representing ~45% of the total amino
acids.
Nonstarch lipids are the most abundant form of lipid in bran and are found in the
aleurone, subaleurone, and germ. They are composed primarily of neutral lipid with
lesser amounts of glycolipids and phospholipids. Starch lipids are a much lesser
component and are primarily in the endosperm (Choudhury and Juliano, 1980a).
Minor lipid components are sterols, tocol, tocotrienols, and waxes.
The typical composition of extracted crude rice bran oil is 68–71% triglycerides,
2–3% diglycerides, 5–6% monoglycerides, 2–3% free fatty acids, 2–3% waxes, 5–
7% glycolipids, 3–4% phospholipids, and 4% unsaponifiables (McCaskill and
Zhang, 1999). The phospholipids predominantly include phosphatidylcholine, phos-
phatidylethanolamine, and phosphatidylinositol. Table 5 compares the fatty acid
composition of rice bran oil extracted from U.S. varieties with oil from selected
oilseeds (McCaskill and Zhang, 1999).
The sterol composition of rice bran was found to consist of three 4,4-dimethyl-
sterols, seven 4-monomethylsterols, 14 4-desmethylsterols, and 10 minor sterols
(Narumi and Takatsuto, 2000). Oryzanol, a phytosterol esterified to ferulic acid, is
found at the 2% level or higher in crude rice bran oil. Its health-promoting proper-
ties include plasma cholesterol reduction (Lichtenstein et al, 1994), reduced

cholesterol absorption (Rong et al, 1997), decreased aortic fatty streaks (Rong et al,
1997), increased fecal bile acid excretion (Seetharamaiah and Chandrasekhara,
1990), and inhibition of platelet aggregation (Seetharamaiah et al, 1990).
Crude rice bran oil contains tocotrienols, at ~1,000 ppm, which are antioxidants
with health-protective benefits in prevention of cancer and cardiovascular disease
(Tomeo et al, 1995; Nesaretnam et al, 1998). The wax component of bran contains
policosanols, a collection of C-24 to C-34 primary alcohols. Policosanols from
sugar cane wax, which are composed of 70% octacosanol (C-28), have been dem-
onstrated to effectively lower low-density lipoprotein (LDL) cholesterol and increase
serum high-density lipoprotein (HDL) cholesterol, thus improving the HDL-LDL ratio
(Menendez et al, 1994; Rodriguez-Echenique et al, 1994; Batista et al, 1996; Zardoya
et al, 1996). Rice bran wax contains less octacosanol (20–44%); its effectiveness in
lowering LDL cholesterol has not been demonstrated clinically to date.
Rice bran lipases are a major cause of deterioration of oil in the bran. Lipases are
localized in the testa layer and, to a smaller extent, in the pericarp layers (Sastry et al,
1977). The primary enzyme has a pH optimum of 7.5 and temperature optimum of
30°C (Rajeshwara and Prakash, 1995). It has a molecular mass of 30 kDa and is com-
posed of at least two subunits. Rice bran lipase II has a molecular mass of 33 kDa, pH
optimum at 7.5–8.0, and temperature optimum of 27°C (Aizono et al, 1976). A ther-
mally stable lipase was recently identified and found to be a glycoprotein of 9.4 kDa
(Bhardwaj et al, 2001). The lipase showed a maximum activity at 80°C and pH 11.0.
Lipoxygenase, which is localized in the bran, catalyzes the oxidation of polyun-
saturated fatty acids containing a 1,4-pentadiene structure, such as linoleic and
linolenic acids, into conjugated hydroperoxy fatty acids, which, in turn, are con-
verted into various volatile compounds. Three isozymes have been characterized
(Yamamoto et al, 1980; Ida et al, 1983; Ohta et al, 1986; Zhang et al, 1996). Of 108
varieties from China’s Yunnan province, 22 varieties lacked lipoxygenase-3 (Ise et
al, 2002). Most of these were japonica upland rice varieties that displayed early
heading, round grains, lower amylose content in the endosperm, and glabrous leaves
and hulls. During storage, the increase in off-flavor volatile compounds was found
to be one-third to one-fifth as much in rice without lipoxygenase 3 compared with
the increase in rice containing lipoxygenase 3 (Suzuki et al, 1999).
Starch is found in the aleurone layers of bran in the developing grain, but it dis-
appears with grain maturity (Del Rosario et al, 1968). It is absent in the pericarp and
TABLE 5
Comparison of the Fatty Acid Compositiona of Rice Bran Oil Extracted
from U.S. Varieties with Oil from Selected Oilseedsb
Fatty Acid Rice Bran Peanut Soybean Cottonseed
Myristic (14:0) 0.2 – 0.2 0.8
Palmitic (16:0) 15.0 8.1 10.7 27.3
Stearic (18:0) 1.9 1.5 3.9 2.0
Oleic (18:1) 42.5 49.9 22.8 18.3
Linoleic (18:2) 39.1 35.4 50.8 50.5
Linolenic (18:3) 1.1 – 6.8 –
Arachidic (20:0) 0.5 1.1 0.2 0.3
Behenic (22:0) 0.2 2.1 0.1 –
a Percentage.
b Reprinted, with permission, from McCaskill and Zhang (1999).
seed coat (Luh et al, 1991). Commercial bran, however, also contains starch con-
tributed by endosperm and germ, with values ranging from 10 to 55%. Bran is high
in the nonstarch carbohydrates, cellulose and hemicelluloses. Mod et al (1978,
1979) determined that the water- and alkali-soluble hemicelluloses in the bran con-
tain the sugars rhamnose, arabinose, xylose, mannose, galactose, and glucose, as
well as some protein, hexuronic acid, and ferulic acid. Arabinose was the predomi-
nant sugar, with the arabinose-to-xylose ratio ranging from 2.8:1 to 5.4:1. The ara-
binose-xylose ratios for alkali-soluble hemicelluloses ranged from 0.9:1 to 1.1:1.
The mineral content of bran is listed in Table 2. With the exception of silicon,
the levels of the elements are much higher in the bran and germ than in the other
milling fractions. Distribution in the bran and embryo are similar. Approximately
90% of the phosphorus in the bran is phytic phosphorus. Phytic acid, myo-inositol-
1,2,3,4,5,6-hexakis (dihydrogen phosphate), strongly chelates minerals (e.g., iron
and zinc) and complexes with proteins, with and without mineral mediation, to
reduce mineral bioavailability and alter protein functionality (Cheryan, 1980;
Morris, 1986; Clydesdale, 1988; Erdman and Poneros-Schneier, 1989; Lopez et al,
2002). Besides having a role as an antinutritional factor, phytic acid also has benefi-
cial effects on health, acting as an antioxidant (Minihane and Rimbach, 2002) and
anticancer agent (Shamsuddin, 2002; Steer and Gibson, 2002).
Potassium and magnesium phytates are the principal phytate salts in rice bran
(Tanaka et al, 1973). Ogawa et al (1975) found aleurone particles embedded with
these salts and covered with a protein and carbohydrate coat, which may have a role
in the dephosphorylation of phytic acid during germination of the grain.
Recent research has been directed at reducing phytate content in rice through
genetic and molecular technologies. A low-phytic-acid mutant (Ipa-1-1) in rice,
which reduces the phytic acid portion of the grain phosphorus from 71 to 39% and
increases the inorganic portion of the grain phosphorus from 5 to 32%, has been
isolated (Larson et al, 2000; Raboy et al, 2002). The Ipa allele is nonlethal and is
being used to breed low-phytate rice.
As with the distribution of minerals, the vitamins are present at much higher lev-
els in the bran and germ than in the other milling fractions (Table 3). Rice contains
little or no vitamin A, C (ascorbic acid), or D. Some of the vitamins, such as ribo-
flavin and niacin, are not in a completely free form. The values for inositol and
choline are total values, because inositol is mainly in the form of phytin, and cho-
line is mainly in the phospholipids lecithin and lysolecithin.
Commercial rice bran has a sweet, cereal-like aroma when fresh. During storage,
objectionable odors develop as a result of lipid degradation through lipolytic
hydrolysis and oxidation. Fujimaki et al (1977) and Tsugita et al (1978) identified
170 compounds contributing to the aroma. 4-Vinylphenol, identified as the main
component, has a characteristic unpleasant odor. Sweet and mild odor was attrib-
uted to lactones. 2-Acetylthiazole had a cereal-like odor. It was concluded that lac-
tones and 2-acetylthiazole are major contributors to fresh rice bran aroma.
Milled Rice
Removal of the bran layers (pericarp, tegmen, nucellus, and aleurone) along with
polish (subaleurone), germ (embryo), and a small part of the endosperm results in
milled (white) rice composed entirely of endosperm, with a starch content of
approximately 78% (14% moisture) or 90% (dry weight) (Table 1). Waxy rice
starch has an apparent amylose content of 0.8–1.3%, whereas the amylose content
of nonwaxy rice starch ranges from 8 to 37%. Starch concentration increases from
the surface to the core of milled rice. Protein is the second most abundant
constituent of milled rice, with content ranging from 4 to 11%. Protein concen-
tration is highest on the surface of milled rice and decreases toward the center of the
kernel. The albumin-globulin-prolamin-glutelin ratio for a milled U.S. long-grain
rice was 8:10:12:70 (Padhye and Salunke, 1979). Ratios for japonica rice cultivated
in Korea were 1.8–2.4:10.8–15.1:1.8-5.2:79.1–85.5 (Kim et al, 1997). A basmati
rice had a ratio of 5.9:13.8:5.8:74.5 (Steenson and Sathe, 1995). Six varieties of
Indonesian rice showed ratios of 8:12–14:1–2:77–79 (Damardjati et al, 1985). The
ratios obtained for rice are unique among cereal proteins in that the glutelin fraction
is predominant, whereas other cereals, except oats, have prolamins as the major
fraction (Payne and Rhodes, 1982). This results in the lysine content of rice being
higher than in other cereals because prolamins are lysine-poor.
Crude fiber and lipid content in milled rice are low, ranging from 0.2 to 0.5%
and 0.3 to 0.5%, respectively (Table 1). Phosphorus and potassium are the most
abundant minerals in milled rice (Table 2). Niacin, inositol, and choline are the
three predominant vitamins (Table 3). The distribution of fiber, lipid, and vitamins
in the milled grain follows that of the protein.
A large number of volatile compounds have been observed in uncooked and
cooked milled rice. In all, 100 volatile constituents, including 13 hydrocarbons, 13
alcohols, 16 aldehydes, 14 ketones, 14 acids, 8 esters, 5 phenols, 3 pyridines, and 6
pyrazines, were identified by analysis of steam distillate of cooked rice by glass
capillary gas chromatography (GC) and mass spectrometry (MS) (Yajima et al,
1978). The majority of the compounds were lipid oxidation products and, based on
their high odor thresholds, would contribute minimally to the aroma of fresh rice.
With deterioration during storage, they would contribute to what is referred to as
“old” or “stale” rice aroma. Grimm et al (2001) reported 138 compounds observed
in the headspace of a worldwide collection of uncooked milled rice varieties. The
headspace was collected by selective phase microextraction and analyzed by
GC/MS. The GC profile was dominated by lipid oxidation products, including
straight-chain aldehydes from pentanal to dodecanal, the alcohols from pentanol to
dodecanol, and short-chain fatty acids from acetic to decanoic acid. There was also
a series of unsaturated aldehydes ranging from (E)-2-heptenal to (E)-2-decenal.
Volatile sulfur compounds evolved from cooked rice were hydrogen sulfide,
methyl mercaptan, dimethyl sulfide, n-butyl mercaptan (butane-1-thiol), and
dimethyl disulfide (Sato et al, 1976; Tsuzuki et al, 1978).
Buttery et al (1988) identified 64 major volatile compounds in rice and con-
cluded from odor threshold data that the probable major contributors to rice aroma
in the grain examined were 2-acetyl-1-pyrroline, (E,E)-2,4-decadienal, nonanal,
hexanal, (E)-2-nonenal, octanal, decanal, 4-vinyl-guaiacol, and 4-vinylphenol. 2-
Acetyl-1-pyrroline, which has a characteristic popcorn aroma, is found at concen-
trations ranging from 1 to 10 ppb in nonaromatic rice and in excess of 2 ppm in
scented rice (Buttery et al, 1988). It is largely responsible for the aroma of the aro-
matic basmati- and jasmine-type varieties. In a comparative study of nonfragrant
and fragrant rice, Widjaja et al (1996) identified (E)-2-decenal, (E,E)-2,4-nonadi-
enal, and (E,E)-2,4-decadienal as having a “waxy” aroma. These three lipid oxida-
tion products also are found in waxy rice and contribute to its aroma (Grimm et al,
2000).
Brown Rice
Brown rice, the grain with just the hull removed, has a higher content of all con-
stituents, except starch, compared with milled rice (Tables 1–3). The concentration
of the vitamins is typically 2–10 times higher in brown rice than in milled rice.
Mineral and fiber contents are two to three times higher in the brown rice, and lipid
content is approximately five times higher.
The brown rice protein content of 17,587 cultivars in the world collection at the
International Rice Research Institute ranged from 4.3 to 18.2%, with a mean of
9.5% (Gomez, 1979). Nonprotein nitrogen, mainly free amino acids, accounts for
~3% of the total nitrogen of brown rice, with the highest level in the germ (Juliano,
1972). The major free amino acids are asparagine + aspartic acid, glutamic acid +
glutamine, alanine, serine, proline, g-aminobutyric acid, glycine, arginine, threonine
+ valine, and tryptophan (Tamura and Kenmochi, 1963; Fujimaki et al, 1975).
Calculation of nutrient distribution in the milling fractions of brown rice con-
firms the uneven distribution of nutrients in the various tissues of the rice caryopsis
(Resurreccion et al, 1979). More nonstarch constituents are removed during milling,
with fiber showing the most drastic drop (Table 6), followed by the other nutrients,
except protein and zinc. A less clear-cut effect of milling is noted for ash and its
component phosphorus, potassium, and iron. In 12%-milled brown rice, protein
distribution is 22% in bran-polish and 78% in milled rice (Table 6). Protein distri-
bution becomes more even as protein content increases in the grain (Cagampang et
al, 1966; Juliano et al, 1973; Resurreccion et al, 1979). The highest protein fraction
in the high-protein grain is the polish, rather than the bran (18.8 versus 14.0% pro-
tein) (Resurreccion et al, 1979). Free amino acid distribution is calculated as 30% in
the bran-polish, 53% in the embryo, and 17% in milled rice (Tamura and
Kenmochi, 1963).
TABLE 6
Nutrient Distribution in Milling Fractions of IR32 Rice Caryopsisa
Distribution (% of nutrient in brown rice)
Content in Endosperm (milled rice)b
Brown Rice
(% at 14% Sub-
Nutrient moisture) Bran Polish aleurone Middle Core
Weight (%) of brown rice 100 0–6 6–12 12–20 20–30 30–100
Protein (N × 5.95), % 7.2 11 11 13 14 51
Starch, % 66.4 1 4 7 11 77
Amylose, % 26.5 1 3 8 10 78
Crude fiber, % 0.7 70 19 1 1 9
Neutral detergent fiber, % 2.1 71 19 1 2 7
Crude fat, % 2.9 51 32 10 1 6
Crude ash, % 0.78 42 26 10 5 17
Total P, % 0.15 38 25 11 4 22
Phytin, P, % 0.1 47 33 16 1 3
Total K, % 0.19 37 22 9 5 27
Total Mg, % 0.08 50 34 11 1 4
Total Fe, µg/g 12 47 24 10 3 16
Total Zn, µg/g 27 31 14 7 6 41
a Reprinted, with permission, from Resurreccion et al (1979).
b Distribution (% of nutrient in brown rice).
The distribution of brown rice pentosans by calculation is 43% in bran, 8% in

germ, 7% in polish, and 42% in milled rice (Leonzio, 1967). The corresponding
cellulose distribution is 62% in bran, 4% in germ, 7% in polish, and 27% in milled
rice (Leonzio, 1967). Lignin distribution in brown rice is calculated as 63% in bran,
8% in polish, 8% in germ, and 21% in milled rice (Leonzio, 1966).
About 80% of the nonstarch lipids of brown rice is in the bran and polish, and
about one-third of this is in the embryo (Juliano, 1972). Choudhury and Juliano
(1980a) calculated nonstarch lipids distribution in IR42 brown rice as 41% in bran,
18% in embryo, 15% in polish, and 26% in milled rice (12% in the subaleurone
layer). Starch lipids are mainly in the endosperm, because they are associated with
amylose (Choudhury and Juliano, 1980b).
The ash distribution in brown rice is calculated as 51% in bran, 10% in germ,
11% in polish, and 28% in milled rice (Leonzio, 1967). Iron, phosphorus, and potas-
sium show a distribution similar to that of total ash (Resurreccion et al, 1979)
(Table 6). However, some minerals show a relatively more even distribution in the
grain. For example, calculation has shown that milled rice retains 63% of the
sodium and 74% of the calcium content of brown rice.
By calculation, 65% of the thiamin of brown rice is in the bran (58% in the
embryo), 13% in polish, and the rest (22%) in milled rice. Corresponding values for
riboflavin are 39% in bran (24% in embryo), 8% in polish, and 53% in milled rice.
The distribution of niacin is 54% in bran (18% in embryo), 13% in polish, and 33%
in milled rice. Milling results in the loss to human consumption of ~76% of the
thiamin, 57% of the riboflavin, and 64% of the niacin of brown rice (Kik and
Williams, 1945).
GC/olfactometry of cooked brown rice samples revealed 41 odor-active com-
pounds, of which the following 12 compounds had high flavor dilution factors
(low odor thresholds): 2-acetyl-1-pyrroline (popcorn-like), 2-isobutyl-3-meth-
oxypyrazine (earthy, green bell pepper), (E,E)-nona-2,4-dienal (fatty), 4,5-epoxy-
(E)-dec-2-enal (metallic), unknown (spicy), bis-(2-methyl-3-furyl)-disulfide
(meaty), 3-hydroxy-4,5-dimethyl-3(2H)-furanone (seasoning-like), 4-vinyl-2-
methoxyphenol (spicy, clove-like), 2-amino acetophenone (medicinal, phenolic),
4-vinylphenol (phenolic), phenyacetic acid (honey-like), and vanillin (vanilla-
like) (Jezussek et al, 2002). 2-Amino acetophenone exhibited the highest flavor
dilution factor. Differences in flavor dilution factors for some compounds were
observed among varieties.
FACTORS AFFECTING COMPOSITION
Management and Cultural Practices
Management and cultural practices have a major influence on the protein content
of the rice grain. Protein content was observed to be higher when wider plant spac-
ing was used, and this was attributed to more nitrogen being available per plant (De
Datta et al, 1972). However, a later study showed that spacings had no significant
effect on protein content (Nair, 1975). Protein content was generally higher in hills
bordering an unplanted alley or in hills adjacent to missing hills (de Datta et al,
1972). Protein content was shown to decrease with increasing number of seedlings
per hill, irrespective of spacing (Chandra et al, 1990).
Nitrogen, phosphorus, and potassium fertilizer treatments are reported to

increase concentrations of nitrogen (protein), phosphorus, sulfur, manganese, mag-
nesium, calcium, and copper (Sturgis and Reed, 1937; Patrick et al, 1974; Tsuzuki
et al, 1979; Chen et al, 1997; Marr et al, 1999; Haefele et al, 2002). Nitrogen top-
dressing during the period from the reproductive to the early ripening stage of plant
development plays a major role in determining the protein content of the rice grain.
Top-dressing with nitrogen at flowering resulted in a 30–60% increase in head rice
protein content (Perez et al, 1996). Islam et al (1996) also observed increased crude
protein with nitrogen top-dressing, which was considerably greater in the indica
than in the japonica variety examined. The top-dressed 15N partitioned most
effectively to the glutelin fraction when it was applied at the heading stage. On the
other hand, 15N applied at heading plus 20 days contributed more to increased
prolamin. Souza et al (1993) found that glutelin increased and prolamin decreased
with increases in protein content resulting from supplementary nitrogen. However,
in contrast, Takebe et al (1996) found that the proportions of individual proteins
(albumin + globulin, glutelin, and prolamin) in the total protein remained almost
constant as protein increased with increasing amounts of applied nitrogen. The total
amount of free amino acids has been observed to decrease (Araki et al, 1999) and
increase (Souza et al, 1999) with nitrogen application.
A decrease in amylose content has been observed concurrent with an increase in
protein content with nitrogen application or uptake (Prakash et al, 2002). Zinc
(Chen et al, 1997) and potassium (Vil’gel’, 1986) applications, conversely, in-
creased amylose content of rice. Applied magnesium also increased grain amylose
and protein content, with the response to applied magnesium being negatively
correlated with soil magnesium levels and positively correlated with soil potassium-
magnesium ratio (Lin et al, 1990).
For brown rice, nitrogen application decreased the riboflavin content (from 1.41–
1.43 µg/g to 1.16–1.27 µg/g) and increased the nicotinic acid content (from 54.7–
55.4 µg/g to 58.3–62.1 µg/g), but it had no effect on milled rice content (Taira et al,
1976). Thiamin and folic acid were not affected by nitrogen fertilization.
The composition of organically and conventionally grown rice was compared
over three years at 13 different locations (Nakagawa et al, 2000). At each location,
organically and conventionally managed paddy were grown in adjacent fields. The
organically grown rice had a higher magnesium-(potassium + nitrogen) ratio and
zinc content and lower nitrogen, potassium, and calcium content. All except cal-
cium content could be explained by lower nitrogen application in organic farming
systems. Tamaki et al (1995) determined the mineral content of rice grown organ-
ically over 16 years. The mineral content of rice one year after the start of organic
farming practices was similar to that of rice grown conventionally. Nitrogen,
phosphorus, and potassium content decreased with the duration of organic farming,
while magnesium content gradually increased. These changes were greatest in the
first five years of organic farming. Of the free amino acids, aspartic acid, glutamic
acid, glutamine, and asparagine were significantly higher in milled rice grains from
organically grown rice than in those from conventionally grown rice (Wang et al,
1998). In contrast, the concentrations of hydrolyzed amino acids tended to be higher
in milled rice obtained from conventional than from organic practices.
Protein and amylose content were observed to decrease with delay in harvest
time irrespective of cultivation method (organic or traditional) (Asano et al, 2000).
The composition of the protein, however, was unaffected by harvest time.
Soil
Mineral analysis was undertaken on grain samples of varieties tolerant to adverse

soil conditions that were harvested from plants grown in Maahas clay and in simu-
lated problem soil in the 1978 wet season (IRRI, 1980). Crude ash was reduced 0.5
percentage point and phosphorus content 0.1 percentage point in brown rice of two
pairs of samples grown in phosphorus-deficient soils. Zinc deficiency caused a
greater reduction of potassium content (0.1–0.2 percentage point) than of zinc con-
tent (3 µg/g in brown rice). Acid sulfate soils increased the iron in brown rice by 5
µg/g. Soil salinity resulted in a large increase in potassium content of the brown rice
(0.2–0.4 percentage point) and no change in ash or sodium content).
Climate
Apparent amylose content has been shown to be higher in japonica cultivars

when the grain develops at lower temperatures (15–20°C) than at higher tempera-
tures (Asaoka et al, 1984, 1985b; Sano et al, 1985; He et al, 1990; Hirano and Sano,
1998; Ueda et al, 1998). The amylose content of indica cultivars and isogenic lines
of japonica rice that contain the Wxa allele typical of indica cultivars have not been
observed to correlate with temperature during development (Asaoka et al, 1985b;
Wang and Wessler, 1998). Larkin and Park (1999) showed that the critical
difference is not whether the granule-bound starch synthase (GBSS) allele is
characteristic of indica or japonica cultivars but, rather, whether the 5 leader intron
splice site is AGGTATA or AGTTATA. In the cultivars they examined, amylose
content increased by 8–9.5% in AGTTATA cultivars when the grain was allowed to
develop at 18°C compared with 32°C. Amylose content was as much as 3.2% lower
in some AGGTATA cultivars and up to 1.8% higher in others at the lower
temperature, compared with their values at the higher temperature.
Protein has been reported to be higher in rice grown at cooler temperatures (Lin et
al, 1995; Matsue, 1995; Hirano and Sano, 1998; Ueda et al, 1998) and has been
observed to decrease with increasing soil temperature (Subhani et al, 2001). Lee et al
(1996) reported protein, magnesium, and potassium levels to be highest in a japonica
variety grown at a constant day and night temperature of 22°C, in comparison with
levels when day and night temperatures were 30 and 22, 26 and 18, or 22 and 16°C.
Mineral content (nitrogen, phosphorus, potassium, magnesium, calcium, zinc,
manganese, sodium, iron, and copper) in milled rice was determined for three culti-
vation years for crops of the same cultivar grown in the same region (Koga et al,
2001). In two of the years, the climate was similar. The rice grown in the year in
which the climate differed from the other two years had a mineral content that dif-
fered from that of the rice from the two years with similar climates. The results
indicated that climate affected the content of minerals, especially nitrogen, phos-
phorus, and potassium.
Location on Panicle
Grains on the upper rachis branches of the panicle had a lower protein content
than those on lower rachis branches, with a gradual change from the base to the top
of both primary and secondary rachis branches (Matsue et al, 1994). Grains on pri-
mary rachis branches also had lower protein content than those on secondary rachis
branches at any position within the panicle. In contrast, amylose content from grains
on upper rachis branches was higher than that from grains on lower rachis branches,
and that from grains on primary rachis branches was higher than that from grains on
secondary rachis branches. The maximum ranges of variation of protein and amy-
lose content within the panicle were 1.2 and 3.3%, respectively. The protein of
milled rice showed little difference in prolamin content and albumin + globulin
content between grains on primary and secondary rachis branches (Matsue et al,
1995). Glutelin content of grains on primary rachis branches was noticeably lower
than that of grains on secondary rachis branches (Matsue et al, 1995).
Significant differences in mineral content among panicles have been observed
(Shindoh et al, 1998). Phosphorus and magnesium content was high in heavy grains
on the primary rachis.
Genotypic Differences
Amylose and protein contents have been shown to be controlled by main genetic
effects as well as by genotype–environment interaction effects (Senanayake and
Wijeratne, 1990; Shi et al, 1997; Xu and Zhang, 1998; Shi et al, 1999; Bao et al,
2000). Lysine content was affected mainly by genotype–environment interaction
effects (Shi et al, 2000).
Genotypic differences in concentrations of iron, manganese, copper, and zinc in
milled rice were determined for 285 varieties (Yang et al, 1998). To minimize the
effect of environmental factors on mineral concentrations in the grain, the rice
genotypes were planted at the same location. Copper and zinc concentrations of
indica rice were two times higher than those of japonica rice, while iron concentra-
tions of japonica rice were slightly higher than those of indica rice. In another study,
genotypes differed significantly in response to applied fertilizers and the resultant
nitrogen, phosphorus, and potassium content of the grains (Fageria et al, 1995).
A comparison of the fatty acid composition of rice bran oil from 10 varieties
each of indica and japonica types revealed that the indica type (compared with the
japonica type) had significantly higher proportions of palmitic (19.3 vs. 15.8%),
stearic (2.1 vs. 1.9%), linolenic (1.4 vs. 1.1%), and arachidic (0.9 vs. 0.8%) acids
and lower proportions of oleic (43.2 vs. 45.3%), linoleic (31.1 vs. 32.7%), and
eicosenoic (0.6 vs. 0.7%) acids (Taira, 1989).
Processing
Parboiling is a hydrothermal process accomplished by soaking, steaming, drying,

and milling the rice. The process produces physical and chemical changes in the
kernel and affects the content of some constituents. Parboiling disrupts oil bodies in
the aleurone and germ, which results in a band of oil migrating outward in the bran
layers (Mahadevappa and Desikachar, 1968). Therefore, milled parboiled rice has a
lower oil content than milled raw rice (Subrahmanyan et al, 1938). Conversely,
milled parboiled rice contains higher amounts of the water-soluble vitamins thiamin
and nicotinic acid; sugars; and the minerals phosphorus, calcium, iron, manganese
molybdenum, and chromium (Williams and Bevenue, 1953; Padua and Juliano,
1974; Bhattacharya and Ali, 1985). Possible mechanisms for the reduced loss of
these constituents during milling are discussed by Bhattacharya in Chapter 13. The
magnesium, zinc, and copper content of milled rice is unaffected by parboiling
(Bhattacharya and Ali, 1985). The protein content of brown and milled rice, as well
as the amino acid composition of the protein, are unchanged by parboiling
(Bhattacharya and Ali, 1985).
SUMMARY
Considerable progress has been made in understanding the structure of the rice
grain and its composition. Identification of genotypic differences in structure and
nutrient composition and knowledge of the impact of pre- and postharvest factors on
them are essential. Such knowledge will enable the development of production,
breeding, and processing strategies and technologies to improve not only the proc-
essing and sensory qualities of rice but the nutrition- and health-beneficial charac-
teristics of the grain.
REFERENCES
Adachi, T., Izumi, H., Yamada, T., Tanaka, K., milky stage on amylose content and fine
Takeuchi, S., Nakamura, R., and Matsuda, T. structure of amylopectin of waxy and
1993. Gene structure and expression of rice nonwaxy endosperm starches of rice (Oryza
seed allergenic proteins belonging to the sativa L.) Agric. Biol. Chem. 49:373-379.
alpha-amylase/trypsin inhibitor family. Plant Baek, J.-H., Han, B.-K., and Jo, D.-H. 2001.
Mol. Biol. 21:239-248. Distribution of chitinases in rice (Oryza sativa
Adachi, T., Alvarez, A. M., Aoki, N., Nakamura, L.) seed and characterization of a hull-specific
R., Garcia, V. V., and Matsuda, T. 1995. chitinase. J. Biochem. Mol. Biol. 34:310-315.
Screening of rice strains deficient in 16-kDa Baldwin, P. M. 2001. Starch granule-associated
allergenic protein. Biosci. Biotechnol. proteins and polypeptides: A review.
Biochem. 59:1377-1378. Starch/Staerke 53:475-503.
Ahn, J. K., Chung, I. M., Ahn, J. K., and Chung, Bao, J. S., Xia, Y. W., Bao, J. S., and Xia, Y. W.
I. M. 2000. Allelopathic potential of rice hulls 2000. Effects of genotype × environment
on germination and seedling growth of interaction on eating and cooking quality of
barnyardgrass. Agron. J. 92:1162-1167. indica rice. J. Zhejiang Univ. Agric. Life Sci.
Aizono, Y., Funatsu, M., Fujiki, Y., and 26:144-150.
Watanabe, M. 1976. Purification and charac- Batista, J., Stusser, R., Saez, F., and Perez, B.
terization of rice bran lipase II. Agric. Biol. 1996. Effect of policosanol on hyperlipidemia
Chem. 40:317-324. and coronary heart disease in middle-aged
Araki, M., Matsue, Y., and Kaneko, A. 1999. patients. A 14-month pilot study. Int. J. Clin.
Relation between nitrogen fertilizer and free Pharm. Ther. 34:134-137.
amino acid composition of brown rice. Jpn. J. Bechtel, D. B., and Pomeranz, Y. 1977. Ultra-
Soil Sci. Plant Nutr. 70:19-24. structure of the mature ungerminated rice
Asano, H., Hirano, F., Isobe, K., and Sakurai, H. (Oryza sativa) caryopsis. The caryopsis coat
2000. Effect of harvest time on the protein and aleurone cells. Am. J. Bot. 64:966-973.
composition (glutelin, prolamin, albumin) and Bechtel, D. B., and Pomeranz, Y. 1978a. Ultra-
amylose content in paddy rice cultivated by structure of the mature ungerminated rice
Aigamo duck farming system. Jpn. J. Crop (Oryza sativa) caryopsis. The germ. Am. J.
Sci. 69:320-323. Bot. 65:78-85.
Asaoka, M., Okuno, K., Sugimoto, Y., Bechtel, D. B., and Pomeranz, Y. 1978b. Ultra-
Kawakami, J., and Fuwa, H. 1984. Effect of structure of the mature ungerminated rice
environmental temperature during develop- (Oryza sativa) caryopsis. The starchy
ment of rice plants on some properties of endosperm. Am. J. Bot. 65:684-691.
endosperm starch. Starch/Staerke 36:189-193. Bechtel, D. B., and Pomeranz, Y. 1980. The rice
Asaoka, M., Okuno, K., Sugimoto, Y., and Fuwa kernel. Pages 73-113 in: Advances in Cereal
H. 1985a. Developmental changes in the Science and Technology, Vol. 3. Y. Pomeranz,
structure of endosperm starch of rice (Oryza Ed. Am. Assoc. Cereal Chem., St. Paul, MN.
sativa L.). Agric. Biol. Chem. 49:1973-1978. Bett-Garber, K. L., Champagne, E. T., McClung,
Asaoka, M., Kazutoshi, O., and Fuwa, H. 1985b. A. M., Moldenhauer, K. A., Linscombe, S. D.,
Effect of environmental temperature at the and McKenzie, K. S. 2001. Categorizing rice
cultivars based on cluster analysis of amylose de Datta, S. K., Obcemea, W. M., and Jana, R. K.
content, protein content and sensory attributes. 1972. Protein content of rice grain as affected
Cereal Chem. 78:551-558. by nitrogen fertilizer and some triazines and
Bhardwaj, K., Raju, A., and Rajesekharan, R. substituted ureas. Agron. J. 64:785-788.
2001. Identification, purification, and charac- De Jong, A., Cordewener, J., Lo Schiavo, F.,
terization of a thermally stable lipase from rice Terzi, M., Vandekerckhove, J., Van Kammen,
bran. A new member of the (phospho) lipase A., and De Vries, S. C. 1992. A carrot somatic
family. Plant Physiol. 127:1728-1738. embryo mutant is rescued by chitinase. Plant
Bhattacharya, K. R., and Ali, S. Z. 1985. Cell. 4:425-433.
Changes in rice during parboiling and Del Rosario, A. R., Briones, V. P., Vidal, A. J.,
properties of parboiled rice. Pages 105-167 in: and Juliano, B. O. 1968. Composition and
Advances in Cereal Science and Technology, endosperm structure of developing and mature
Vol. 7. Y. Pomeranz, Ed. Am. Assoc. Cereal rice kernel. Cereal Chem. 45:225-235.
Chem., St. Paul, MN. Dikeman, E., Bechtel, D. B., and Pomeranz, Y.
Boller, T. 1988. Ethylene and the regulation of 1981. Distribution of elements in the rice ker-
antifungal hydrolases in plants. Pages 145-174 nel determined by x-ray analysis and atomic
in: Surveys of Plant Molecular and Cell absorption spectroscopy. Cereal Chem.
Biology, Vol. 5. B. J. Miflin, Ed. Oxford 58:148-152.
University Press, Oxford. Eggum, B. O., Juliano, B. O., and Maniñgat, C.
Buttery, R. G., Turnbaugh, J. G., and Ling, L. C. C. 1982. Protein and energy utilization of rice
1988. Contribution of volatiles to rice aroma. milling fractions by rats. Qual. Plant. Plant
J. Agric. Food Chem. 36:1006-1009. Foods Hum. Nutr. 31:371-376.
Cagampang, G. B., Cruz, L. J., Espiritu, S. G., Ellis, J. R., and Chaffey, N. J. 1987. Structural
Santiago, R. G., and Juliano, B. O. 1966. differentiation of the nucellar epidermis in the
Studies on the extraction and composition of caryopsis of rice (Oryza sativa). Ann. Bot.
rice proteins. Cereal Chem. 43:145-155. 60:671-675.
Chakraverty, A., and Kaleemullah, S. 1991. Ellis, J. R., Gates, P. J., and Boulter, D. 1987.
Conversion of rice husk into amorphous silica Storage-protein deposition in the developing
and combustible gas. Energy Convers. rice caryopsis in relation to the transport tis-
Manage. 32:565-570. sues. Ann. Bot. 60:663-670.
Chandra, D., Mittra, B. N., and Mukherjee, R. K. Erdman, J. W., and Poneros-Schneier, A. 1989.
1990. Effect of nitrogen levels under various Phytic acid interactions with divalent cations
plant populations on wet season rice. J. in foods and in the gastrointestinal tract. Pages
Andaman Sci. Assoc. 6:115-121. 161-172 in: Mineral Absorption in the
Chen, L., Fan, X., Chen, L. J., and Fan, X. M. Monogastric GI Tract. Advances in Exper-
1997. Grain quality response to phosphorus imental Medicine and Biology, Vol. 249. F. R.
and zinc fertilizer in rice genotype. Chinese Dintzis and J. A. Laszlo, Eds. Plenum Press,
Rice Res. Newsl. 5:9-10. New York.
Cheryan, M. 1980. Phytic acid interactions in Esau, K. 1977. Anatomy of Seed Plants, 2nd ed.
food systems. Crit. Rev. Food Sci. Nutr. John Wiley & Sons, Inc., New York.
13:297-335. Evers, T., and Millar, S. 2002. Cereal grain
Choudhury, N. H., and Juliano, B. O. 1980a. structure and development: Some implications
Lipids in developing and mature rice grain. for quality J. Cereal Sci. 36:261-284.
Phytochemistry 19:1063-1069. Fageria, N. K., Santana, E. P., de Castro, E. M.,
Choudhury, N. H., and Juliano, B. O. 1980b. Moraes, O. P., and De-Castro, E. M. 1995.
Effect of amylose content on the lipids of Differential response to soil fertility of upland
mature rice grain. Phytochemistry 19:1385- rice genotypes. Rev. Bras. Cienc. Solo 19:261-
1389. 267.
Chung, M. C., and Wu, H. K. 1983. The devel- Fujimaki, M., Kato, H., Arai, S., and Abe, M.
opment of rice caryopsis. Natl. Sci. Counc. 1975. Changes in free amino acids in rice
Mon. 11:959-976. endosperm during drying. J. Food Sci.
Clydesdale, F. M. 1988. Mineral interactions in Technol. Jpn. 22:83-87.
foods. Pages 73-113 in: Nutrient Interactions. Fujimaki, M., Tsugita, T., and Kurata, T. 1977.
IFT Basic Symposium Series. C. E. Bodwell Fractionation and identification of volatile
and J. W. Erdman, Eds. Marcel Dekker, New acids and phenols in the steam distillate of rice
York. bran. Agric. Biol. Chem. 41:1721-1725.
Damardjati, D. S., Soekarto, S. T., Nur, A., and Fulcher, R. G., and Wong, S. I. 1980. Inside
Siwi, B. H. 1985. Evaluation of protein quality cereals—A fluorescence microchemical view.
and properties in six varieties of Indonesian Pages 1-26 in: Cereals for Food and Bever-
rice. Indones. J. Crop Sci. 1:1-20. ages: Recent Progress in Cereal Chemistry and
Technology. G. E. Inglett and L. Munck, Eds. Hoshikawa, K. 1993. Rice seed, germination and
Academic Press, New York. seedlings, Chapter 1, Part II: Morphology and
Fulcher, R. G., and Wong, S. I. 1982. Fluores- development of vegetative organs,. Pages 91-
cence microscopy of cereal grains. Can. J. Bot. 132 in: Science of the Rice Plant, Vol. 1, Mor-
60:325-329. phology. T. Matsuo and K. Hoshikawa, Eds.
Fulcher, R. G., O’Brien, T. P., and Wong, S. I. Food and Agriculture Policy Research Center,
1981. Microchemical detection of niacin, aro- Tokyo.
matic amine and phytin reserves in cereal Hsu, W. H., and Luh, B. S. 1980. Rice hulls.
bran. Cereal Chem. 58:130-135. Pages 736-763. in: Rice: Production and
Fulcher, R. G., Irving, D. W., and deFrancisco, Utilization. B. S. Luh, Ed. AVI, Westport, CT.
A. 1989. Fluorescence microscopy in food Ida, S., Masaki, Y., and Morita, Y. 1983. The
analysis. Pages 59-108 in: Handbook of Fluo- isolation of multiple forms and product
rescence Analysis in Foods. L. Munck, Ed. J. specificity of rice lipoxygenase. Agric. Biol.
Wiley & Sons, New York. Chem. 47:637-641.
Gomez, K. A. 1979. Effect of environment on Inouchi, N., Ando, H., Asaoka, M., Okuno, K.,
protein and amylose content of rice. Pages 59- and Fuwa, H. 2000. The effect of environ-
68 in: Proc. Workshop Chem. Aspects Rice mental temperature on distribution of unit
Grain Quality. International Rice Research chains of rice amylopectin. Starch/Staerke
Institute, Los Baños, Laguna, Philippines. 52:8-12
Grimm, C. C., Champagne, E. T., Bett-Garber, K. IRRI. 1980. Annual Report for 1979. Inter-
L., and Ohtsubo, K. 2000. The volatile national Rice Research Institute, Los Banos,
composition of waxy rice. Proc. U. S.-Jpn. Laguna, Philippines.
Coop. Program Nat. Resour. (UJNR) Protein Ise, K., Suzuki, Y., Xiong, J., Dai, L., Ye, C.,
Resour. Panel Meet., 29th. U.S. Dep. Agric., Kudo, S., Maekawa, M., Xiong, J. H., Dai, L.
Agric. Res. Serv., Wyndmoor, PA. Y., and Ye, C. R. 2002. Germplasm lacking
Grimm, C. C., Champagne, E. T., and Ohtsubo, lipoxygenase-3 in a wide diversity of rice
K. 2001. Analysis of volatile compounds in genetic resources in Yunnan, China. Jpn. J.
the headspace of rice using SPME/GC/MS. Trop. Agric. 46:33-38.
Pages 229-248 in: Techniques for Analyzing Islam, N., Inanaga, S., Chishaki, N., and
Flavors, Fragrances, and Aromas. R. Marsili, Horiguchi, T.1996. Effect of N top-dressing on
Ed. Marcel Dekker, New York. protein content in japonica and indica rice
Haefele, S. M., Wopereis, M. C. S., and grains. Cereal Chem. 73:571-573.
Wiechmann, H. 2002. Long-term fertility Izumi, H., Adachi, T., Fujii, N., Matsuda, T.,
experiments for irrigated rice in the West Nakamura, R., Tanaka, K., Urisu, A., and
African Sahel: Agronomic results. Field Crops Kurosawa, Y. 1992. Nucleotide sequence of a
Res. 78:119-131. cDNA clone encoding a mjor allergenic
Hartman, L., and Lago, R. C. A. 1976. The protein in rice seeds: Homology of the
composition of lipids from rice hulls and from deduced amino acid sequence with members
the surface of rice caryopsis. J. Sci. Food of alpha-amylase/trypsin inhibitor family.
Agric. 27:939-942. FEBS Lett. 302:213-216.
He, G. C., Kogure, K., and Suzuki, H. 1990. Jezussek, M., Juliano, B. O., and Schieberle, P.
Development of endosperm and synthesis of 2002. Comparison of key aroma compounds in
starch in rice grain. III. Starch property as cooked brown rice varieties based on aroma
affected by the temperature during grain extract dilution analyses. J. Agric. Food Chem.
development. Jpn. J. Crop Sci. 59:340-345. 50:1101-1105.
Helm, R. M., and Burks, A. W. 1996. Hypo- Jones, T. J., and Rost, T. L. 1989. Histochemistry
allergenicity of rice protein. Cereal Foods and ultrastructure of rice (Oryza sativa) zygotic
World 41:839-843. embryogenesis. Am. J. Bot. 76:504-520.
Hirano, H.-Y., and Sano, Y. 1998. Enhancement Juliano, B. O. 1972. The rice caryopsis and its
of Wx gene expression and the accumulation composition. Pages 16-74 in: Rice: Chemistry
of amylose in response to cool temperature and Technology, 1st ed. D. F. Houston, Ed.
during seed development in rice. Plant Cell Am. Assoc. Cereal Chem., St. Paul, MN.
Physiol. 39:807-812. Juliano, B. O. 1980. Properties of the rice
Horigane, A. K. Engelaar, W. M. H. G, caryopsis. Pages 403-438 in: Rice: Production
Maruyama, S. Yoshida, M., Okubo, A., and and Utilization. B.S. Luh, Ed. AVI Publishing
Nagata, T. 2001. Visualisation of moisture Co., Inc., Westport, CT.
distribution during development of rice cary- Juliano, B. O. 1981. Rice grain properties and
opses (Oryza sativa L.) by nuclear magnetic resistance to storage insects: A review. Res.
resonance microimaging. J. Cereal Sci. Paper Ser. 56. International Rice Research
33:105-114. Institute, Los Banos, Laguna, Philippines.
Juliano, B. O., Antonio, A. A., and Esmama, B. hypercholesterolemic humans. Arterioscler.

V. 1973. Effects of protein content on the Thromb. 14:549-556.
distribution and properties of rice protein. J. Limas, G. G., Salinas, M., Moneo, I., Fischer, S.,
Sci. Food Agric. 24:295-306. Wittmann-Liebold, B., and Mendez, E. 1990.
Juliano, B. O., and Bechtel, D. B. 1985. The rice Purification and characterization of ten new
grain and its gross composition. Pages 17-57 rice NaCl-soluble proteins: Identification of
in: Rice: Chemistry and Technology, 2nd ed. four protein-synthesis inhibitors and two
B. O. Juliano, Ed. Am. Assoc. Cereal Chem. immunoglobulin-binding proteins. Planta
St. Paul, MN 181:1-9.
Juliano, J. B., and Aldama, M. J. 1937. Morphol- Lin, Q. M., Lu, B., and Chen, Y. L. 1990. Effect
ogy of Oryza sativa Linnaeus. Philipp. Agric. of magnesium fertilizer on rice and the index
26:1-134. of magnesium fraction in paddy soil. J. Fujian
Kaneko, M., Kataoka, K., and Takahashi, M. Agric. Coll. 4:450-456.
1991. Development of internal structure of rice Lin, M. H., Chang, S. K., Lin, M. H., and Chang,
hull. Bull. Fac. Agric. Tamagawa Univ. S. K. 1995. The effects of the north-east mon-
31:179-187. soon on the quality of early and late maturing
Kik, M. C., and Williams, R. R. 1945. The nutri- japonica rice. Chin. J. Agrometeorol. 2:23-30.
tional improvement of white rice. Bull. 112. Little, R. R., and Dawson, E. H. 1960. Histology
Natl. Res. Counc. Natl. Acad. Sci., Washing- and histochemistry of raw and cooked rice
ton, DC. kernels. Food Res. 25:611-622.
Kim, H.-J., Hong, S.-P., Park, K.-H., Kim, I.-H., Lopez, H. W., Leenhardt, F., Coudray, C., and
and Lee, S.-H. 1997. Comparative study of Remesy, C. 2002. Minerals and phytic acid
japonica rice protein with different eating interactions: Is it a real problem for human
quality. Foods Biotechnol. 6:149-153. nutrition? Int. J. Food Sci. Technol. 37:727-
Koga, H., Muramoto, N., Katayama, O., 739.
Shinohara, A., and Chiba, M. 2001. Compari- Luan, T.-C., and Chou, T.-C. 1990. Recovery of
son of mineral contents in polished rice in dif- silica from the gasification of rice husks/coal
ferent cultivation years. J. Jpn. Soc. Food Sci. in the presence of a pilot flame in a modified
Technol. 48:598-605. fluidized bed. Ind. Eng. Chem. Res. 29:1922-
Krishnan, S., Ebenezer G. A. I., and Dayanandan, 1927.
P. 2001. Histochemical localization of storage Luh , B. S. 1991. Rice hulls. Pages 269-294 in:
components in caryopsis of rice (Oryza sativa Rice, Vol. 2. Utilization. B. S Luh, Ed. Van
L.). Curr. Sci. 80:567-571. Nostrand Reinhold, New York.
Larkin, P. D., and Park, W. D. 1999. Transcript Luh, B. S., Barber, S., and Benedito de Barber,
accumulation and utilization of alternate and C. 1991. Rice bran: Chemistry and technol-
non-consensus splice sites in rice granule- ogy. Pages 313-362 in: Rice, Vol. 2. Utiliza-
bound starch synthase are temperature-sensi- tion. B. S Luh, Ed. Van Hostrand Reinhold,
tive and controlled by a single-nucleotide New York.
polymorphism. Plant Mol. Biol. 40:719-727. Mahadevappa, M., and Desikachar, H. S. R.
Larson, S. R., Rutger, J. N., Young, K. A., and 1968. Some observations on the histology of
Raboy, V. 2000. Isolation and genetic map- raw and parboiled rice. J. Food Sci. Technol.
ping of a non-lethal rice (Oryza sativa L.) Low 5:72-73.
phytic acid 1 mutation. Crop Sci. 40:1397- Marr, K. M., Batten, G. D., and Blakeney, A. B.
1405. 1995. Relationships between minerals in Aus-
Lee, J. I., Kim, J. K., Shin, J. C., Kim, E. H., Lee, tralian brown rice. J. Sci. Food Agric. 68:285-
M. H., and Oh, Y. J. 1996. Effects of ripening 291.
temperature on quality appearance and chemi- Marr, K. M., Batten, G. D., and Lewin, L. G.
cal quality characteristics of rice grain. RDA J. 1999. The effect of nitrogen fertiliser on yield,
Agric. Sci. Rice 38:1-9. nitrogen and mineral elements in Australian
Leonzio, M. 1966. II contenuto di lignina nei brown rice. Aust. J. Exp. Agric. 39:873-880.
sottoprodotti della lavorazione del riso. Riso Matsue, Y. 1995. Studies on palatability of rice
15:219-227. in northern Kyushu: V. Influence of abnormal
Leonzio, M. 1967. Il contenuto di pentosani nel weather in 1993 on the palatability and phys-
riso e nei suoi principals sottoprodotti. Riso icochemical characteristics of rice. Jpn. J.
16:313-320. Crop Sci. 64:709-713.
Lichtenstein, A. H., Ausman, L. M., Carrasco, Matsue, Y., Odahara, K., and Hiramatsu, M.
W., Gualtieri, L. J., Jenner, J. L., Ordovas, J. 1994. Differences in protein content, amylose
M., Nicolosi, R. J., Goldin, B. R., and content and palatability in relation to location
Schaefer, E. J. 1994. Rice bran oil consump- of grains within rice panicle. Jpn. J. Crop Sci.
tion and plasma lipid levels in moderately 63:271-277.
Matsue, Y., Odahara, K., and Hiramatsu, M. and K. Hoshikawa, Eds. Food Agric. Policy
1995. Difference in amylose content, amylo- Res. Center, Tokyo.
graphic characteristics and storage proteins of Nishimura, H., Uehara, Y., Sempuku, and
grains on primary and secondary rachis Iwashima, A. 1984. Purification and some
branches in rice. Jpn. J. Crop Sci. 64:601-606. properties of thiamine-binding protein from
McCaskill, D. R., and Zhang, F. 1999. Use of rice bran. J. Nutr. Sci. Vitaminol. 30:1-10.
rice bran oil in foods. Food Technol. 53:50-51. Nishino, A., Nishino, H., and Iwashima, A. 1980.
Menendez, R., Fernandez, S. I., del Rio, A., Presence of a thiamin-binding protein in rice
Gonzalez, R. M., Fraga, V., Amor, A. M., and bran. J. Nutr. Sci. Vitaminol. 26:415-418.
Mas, R. M. 1994. Policosanol inhibits choles- Ogawa, M., Tanaka, K., and Kasai, Z. 1975.
terol biosynthesis and enhances low density Isolation of high phytin containing particles
lipoprotein processing in cultured human from rice grains using an aqueous polymer
fibroblasts. Biol. Res. 27:199-203. two-phase system. Agric. Biol. Chem. 39:695-
Mikkelsen, D. S., and Sinah, M. N. 1961. Germi- 700.
nation inhibition in Oryza sativa and control Ogawa, Y., Kuensting, H., Sugiyama, J., Ohtani,
by preplanting soaking treatment. Crop Sci. T., Liu, X. Q., Kokubo, M., Kudoh, K., and
1:332-335. Higuchi, T. 2002. Structure of a rice grain rep-
Minihane, A. M., and Rimbach, G. 2002. Iron resented by a new three-dimensional visuali-
absorption and the iron binding and anti-oxi- zation technique. J. Cereal Sci. 36:1-7.
dant properties of phytic acid. Int. J. Food Sci. Ohta, H. Ida, S., Mikami, B., and Morita, Y.
Technol. 37:741-748. 1986. Purification and characterization of rice
Mod, R. R., Conkerton, E. J., Ory, R. L., and lipoxygenase component 3 from embryos.
Normand, F. L. 1978. Hemicellulose compo- Agric. Biol. Chem. 50:3165-3171.
sition of dietary fiber and milled rice and rice Osawa, T., Ramarathnam, N., Kawakishi, S., and
bran. J. Agric. Food Chem. 26:1031-1035. Namiki, M.1992. Antioxidative defense sys-
Mod, R. R., Conkerton, E. J., Ory, R. L., and tems generated by phenolic plant constituents.
Normand, F. L.1979. Composition of water- Pages 122-134. in: Phenolic Compounds in
soluble hemicelluloses in rice bran from four Food and Their Effects on Health, II. (ACS
growing areas of Louisiana, Texas, Arkansas, Symp. Ser. 507.) American Chemical Society,
and California. Cereal Chem. 56:356-358. Washington, DC.
Morris, E. R. 1986. Phytate and dietary mineral Padhye, V. W., and Salunke, D. K. 1979. Extrac-
bioavailability. In: Phytic Acid: Chemistry and tion and characterization of rice proteins.
Applications. E. Graf, Ed. Pilatus Press, Cereal Chem. 56:389-393.
Minneapolis, MN. Padua, A. B., and Juliano, B. O. 1974. Effect of
Nair, R. V. 1975. Note on the influence of spac- parboiling on thiamin, protein and fat of rice.
ing and time of nitrogen application on the J. Sci. Food Agric. 25:697-701.
grain protein in rice. Indian J. Agric. Sci. Patrick, R. M., Hoskins, F. H., Wilson, E., and
45:175-176. Peterson, F. J. 1974. Protein and amino acid
Nakagawa, S. Tamura, Y., and Ogata, Y. 2000. content of rice as affected by application of
Comparison of rice grain qualities as influ- nitrogen fertilizer. Cereal Chem. 51:84-95.
enced by organic and conventional farming Payne, P. I., and Rhodes, A. P. 1982. Cereal
systems. Jpn. J. Crop Sci. 69:31-37. storage proteins: Structure and role in agri-
Nakase , M., Adachi, T., Urisu, A., Miyashita, T., culture and food technology. Pages 346-369
Alvarez, A. M., Nagasaka, S., Aoki, N., in: Encyclopedia of Plant Physiology, Vol.
Nakamura, R., and Matsuda, T. 1996. Rice 14A. D. Boutler and B. Parthier, Eds.
(Oryza sativa L.) Alpha-amylase inhibitors of Springer-Verlag, Berlin.
14-16 kDa are potential allergens and products Pederson, B., and Eggum, B. O. 1983. The influ-
of a multigene family. J. Agric. Food Chem. ence of milling on the nutritive value of flour
44:2624-2628. from cereal grains. 4. Rice Qual. Plant. Plant
Narumi, Y., and Takatsuto, S. 2000. Sterols in Foods Hum. Nutr. 33:267-278.
rice bran. J. Oleo Sci. 49:719-722. Perez, C. M., Juliano, B. O., Liboon, S. P.,
Nesaretnam, K., Stephen, R., Dils, R., and Alcantara, J. M., and Cassman, K. G. 1996.
Darbre, P. 1998. Tocotrienols inhibit the Effects of late nitrogen fertilizer application on
growth of human breast cancer cells irrespec- head rice yield, protein content, and grain
tive of estrogen receptor status. Lipids 33:461- quality of rice. Cereal Chem. 73:556-560.
469. Prakash, Y. S., Bhadoria, P. B. S., Amitava, R.,
Niitsuma, Y. 1993. Upland rice. Chapter 2, 3, and Rakshit, A. 2002. Relative efficacy of
Part I: Morphological Characters of Plants in organic manure in improving milling and
Genus Oryza. Pages 70-88 in: Science of the cooking quality of rice. Int. Rice Res. Notes
Rice Plant, Vol. 1, Morphology. T. Matsuo 27:43-44.
Raboy, V., Young, K. A., Dorsch, J. A., and Senanayake, S. G. J. N., and Wijeratne, V. 1990.
Cook, A. 2002. Genetics and breeding of seed Effect of rice genotype x environment interac-
phosphorus and phytic acid. J. Plant Physiol. tions on quality characters. Int. Rice Res.
158:489-497. Newsl. 15:2-12.
Rajeshwara, A. N., and Prakash, V. 1995. Purifi- Shamsuddin , A. M. 2002. Anti-cancer function
cation and characterization of lipase from rice of phytic acid. Int. J. Food Sci. Technol.
(Oryza sativa L.) bran. Nahrung 39:406-418. 37:769-782.
Ramarathnam, N., Osawa, T., Namiki, M., and Shi, C. H., Zhu, J., Zang, R. C., and Chen, G. L.
Tashiro, T. 1986. Studies on the relationship 1997. Genetic and heterosis analysis for
between antioxidative activity of rice hull and cooking quality traits of indica rice in different
germination ability of rice seeds. J. Sci. Food environments. Theor. Appl. Gen. 95:294-300.
Agric. 37:719-726. Shi, C. H., Zhu, J., Wu, J. G., Yang, X. E., and
Ramarathnam, N., Osawa, T., Namiki, M., and Yu, Y. G. 1999. Analysis of embryo,
Kawakishi, S. 1988. Chemical studies of novel endosperm, cytoplasmic and maternal effects
rice hull antioxidants. I. Isolation, fractiona- for heterosis of protein and lysine content in
tion, and partial characterization. J. Agric. indica hybrid rice. Plant Breed. 118:574-576.
Food Chem. 36:732-737. Shi, C. H., Zhu, J., and Yu, Y. G. 2000. Genotype
Resurreccion, A. P., Juliano, B. O., and Tanaka, × environment interaction effect and genotypic
Y. 1979. Nutrient content and distribution in correlation for nutrient quality traits of indica
milling fractions of rice grain. J. Sci. Food rice (Oryza sativa). Indian J. Agric. Sci. 70:85-
Agric. 30:475-481. 89.
Roche, P., Lerouge, P., Ponthus, C., and Prom, J. Shimizu, M., Yoshida, T., Toda, T., Iwashima,
C. 1991. Structural determination of bacterial A., and Mitsunaga, T. 1996. Isolation of a
nodulation factors involved in the Rhizobium thiamine-binding protein from rice germ and
meliloti-alfalfa symbiosis. J. Biol. Chem. distribution of similar proteins. Biosci. Bio-
266:10933-10940. technol. Biochem. 60:453-457.
Rodriguez-Echenique, C., Mesu, R., Mas, R., Shindoh, K., Toyoshima, H., and Yasui, A. 1998.
Noa, M., Menondey, R., Gonzalez, R, M., Differences of mineral contents in a single
Amor, A. M., Fraga, V., Sotolongo, V., and grain of brown rice on the location of panicle.
Laguna, A. 1994. Effects of policosanol Jpn. J. Crop Sci. 67:478-484.
chronically administered in male monkeys Shoji, Y., Mita, T., Isemura, M., Mega, T., Hase,
(Macaca arctoides). Food Chem. Toxicol. S., Isemura, S., and Aoyagi, Y. 2001. A
32:565-575. fibronectin-binding protein from rice bran with
Rong, N., Ausman, L. M., and Nicolosi, R. J. cell adhesion activity for animal tumor cells.
1997. Oryzanol decreases cholesterol absorp- Biosci. Biotechnol. Biochem. 65:1181-1186.
tion and aortic fatty streaks in hamsters. Lipids Souza, S. R., Stark, E. M. L. M., and Fernandes,
32:303-309. M. S. 1993. Effect of supplemental nitrogen
Sano, Y., Maekawa, M., and Kikuchi, H.1985. on the quality of rice proteins. J. Plant Nutr.
Temperature effects on the Wx protein level 16:1739-1751.
and amylose content in the endosperm of rice. Souza, S. R., Starck, E. M. L. M., Fernandes, M.
J. Hered. 76:221-222. S., and Magalhaes, J. R. 1999. Effects of sup-
Sastry, B. S., Ramakrishna, M., and plemental nitrogen on nitrogen-assimilation
Raghavendra-Rao, M. R. 1977. Histochemical enzymes, free amino nitrogen, soluble sugars,
localization of lipase in the rice grain. J. Food and crude protein of rice. Commun. Soil Sci.
Sci. Technol. 14:317-324. Plant Anal. 30:711-724.
Sato, S., Tadenuma, M. Oba, T., Takahashi, K., Steenson, D. F., and Sathe, S. K. 1995. Charac-
and Koike, K. 1976. Changes of volatile sulfur terization and digestibility of basmati rice
compounds of white rice during sake making. (Oryza sativa L. var. Dehraduni) storage pro-
Nippon Jozo Kyokai Zasshi 71:387-390. teins. Cereal Chem. 72:275-280.
Sauphanor, B. 1988. Influence of some chaff Steer, T. E., and Gibson, G. R. 2002. The micro-
characteristics on varietal resistance of rice to biology of phytic acid metabolism by gut
storage insects. Entomol. Exp. Appl. 47:55-67. bacteria and relevance for bowel cancer. Int. J.
Seetharamaiah, G. S., and Chandrasekhara, N. Food Sci. Technol. 37:783-790.
1990. Effect of oryzanol on cholesterol Sturgis, M. B., and Reed, J. F. 1937. The relation
absorption and biliary and fecal bile acids in of organic matter and fertilizer to the growth
rats. Indian J. Med. Res. 92:471-475. and composition of rice. J. Am. Soc. Agron.
Seetharamaiah, G. S., Krishnakantha, T. P., and 29:360-366.
Chandrasekhara, N. 1990. Influence of Subhani, A., Huang, C. Y., Xie, Z. M., Liao, M.,
oryzanol on platelet aggregation in rats. J. El-ghamry, A. M., Huang, C. Y., Xie, Z. M.,
Nutr. Sci. Vitaminol. 36:291-297. and Liao, M. 2001. Impact of some agronomic
practices on paddy field soil health under var- Tsuzuki, E., Matsuki, C., Morinaga, K., and
ied ecological conditions: II. Influence of soil Shida, S. 1978. Studies on the characteristics
temperature. Pedosphere 11:143-150. of scented rice. IV. Volatile sulfur compounds
Subrahmanyan, V., Sreenvasan, A., and das evolved from cooked rice. Nippon Sakumotsu
Gupta, H. P. 1938. Studies on quality of rice. Gakkai Kiji 47:375-380.
I. Effect of milling on the chemical composi- Tsuzuki, E., Machii, H., Shida, S., and Miura, M.
tion and commercial qualities of raw and par- 1979. Studies on the characteristics of scented
boiled rice. Indian J. Agric. Sci. 8:459-486. rice. V. Protein and amino acid contents in
Suzuki, Y., Ise, K., Li, C., Honda, I., Iwai, Y., brown rice. Bull. Fac. Agric. Miyazaki Univ.
and Matsurkura, U. 1999. Volatile components 26:433-441. (In Japanese)
in stored rice (Oryza sativa L.) of varieties Ueda, K. Kusutani, A., Asanuma, K., and Ichii,
with and without lipoxygenase-3 in seeds. J. M. 1998. Effect of transplanting time on
Agric. Food Chem. 47:1119-1124. growth of rice cultivar “Kinuhikari” in
Taira, H. 1989. Fatty acid composition of indica Kagawa prefecture—Meteorological factors
and Japonica-types of rice bran and milled affecting grain yield and palatability of rice.
rice. J. Am. Oil. Chem. Soc. 66:1326-1329. Jpn. J. Crop Sci. 67:289-296.
Taira, H., Tair, H., Matsuzaki, A., and Urisu, A., Yamada, K., Masuda, S., Komada, H.,
Matsushima, S. 1976. Effect of nitrogen top- Wada, E., Kondo, Y., Horiba, F., Tsuruta, M.,
dressing on B-vitamin content of rice grain. Yasaki, T., et al. 1991. 16-Kilodalton rice
Nippon Sakumotsu Gakkai Kiji 45:69-74. protein is one of the major allergens in rice grain
Takebe, M., Oikawa, T., Matsuno, K., Shimizu, E., extract and responsible for cross-allergenicity
and Yoneyama, T. 1996. Influence of nitrogen between cereal grains in the Poaceae family. Int.
application on the contents of glutelin and Arch. Allergy Appl. Immunol. 96:244-252.
prolamin of polished rice (Oryza sativa L.) Usui, Y., Nakase, M., Hotta, H., Urisu, A., Aoki,
grains. Jpn. J. Soil Sci. Plant Nutr. 67:139-146. N., Kitajima, K., and Matsuda, T. 2001. A 33-
Takeoka, Y., Ogawa, K., Kawai, T., and Wada, kDa allergen from rice (Oryza sativa L.
T. 1989. Scanning electron microscopic Japonica). CDNA cloning, expression, and
observations on morphogenesis of the panicle identification as a novel glyoxalase I. J. Biol.
and spikelet in rice plants. Jpn. J. Crop Sci. Chem. 276:11376-11381.
58:119-125. Utsunomiya, H., Yamagata, M., and Doi, Y.
Takeuchi, S. Kono, Y., Kawarada, A., Ota, Y., 1975. Scanning electron microscopy of the
and Nakayama, M. 1975. Nicotinamide as a endosperm of cereal crops. IV. Starch cell
plant growth regulator isolated from rice hulls. layer of imperfect grain of rice (nonglutinous)
Agric. Biol. Chem. 39:859-861. and glutinous rice. Yamaguti Daigaku
Tamaki, M., Yoshimatsu, K., and Horino, T. Nogakubu Gakujutsu Hokoku 26:19-44.
1995. Relationship between the duration of Vil’gel’, M. A. 1986. On the effects of potassium
organic farming culture and amylographic fertilizer on the protein and carbohydrate
characteristics and mineral contents of rice. complex formation in rice grain.
Jpn. J. Crop Sci. 64:677-681. Sel’skokhozyaistvennaya-Biologiya 7:29-31.
Tamura, S., and Kenmochi, K. 1963. Studies on Wada, T. and Lott, J. N. A. 1997. Light and
amino acid content of rice. III. Distribution of electron microscopic and energy dispersive X-
amino acid in rice grain. Nippon Nogei ray microanalysis studies of globoids in pro-
Kagaku Kaishi 37:753-756. tein bodies of embryo tissues and the aleurone
Tanaka, K., Ogawa, M., and Kasai, Z. 1976. The layer of rice (Oryza sativa L.) grains. Can. J.
rice scutellum: Studies by scanning electron Bot. 75:1137-1147.
microscopy and electron microphone X-ray Wang, G., Abe, T., Sasahara, T., and Wang, G.
analysis. Cereal Chem. 53:643-649. Y. 1998. Concentrations of Kjeldahl-digested
Tanaka, K., Yoshida, T., Asada, K., and Kasai, Z. nitrogen, amylose, and amino acids in milled
1973. Subcellular particles isolated from aleu- grains of rice (Oryza sativa L.) cultivated
rone layer of rice seeds. Arch. Biochem. Bio- under organic and customary farming prac-
phys. 155:136-143. tices. Jpn. J. Crop Sci. 67:307-311.
Tomeo, A. C., Geller, M., Watkins, T. R., Gapor, Wang, L., and Wessler, S. 1998. Inefficient
A., and Bierenbaum, M. L. 1995. Antioxidant reinitiation is responsible for upstream open
effects of tocotrienols in patients with hyper- reading frame-mediated translational repression
lipidemia and carotid stenosis. Lipids of the maize R gene. Plant Cell 10:1733-1745.
30:1179-1183. Whistler, R. L., BeMiller, J. N., and Paschall, E.
Tsugita, T., Kurata, T., and Fujimaki, M. 1978. F., Eds. 1984. Starch Chemistry and Technol-
Volatile components in the steam distillate of ogy. Academic Press, New York.
rice bran; Identification of neutral and basic Widjaja, R., Craske, J. D., and Wootton, M.
compounds. Agric. Biol. Chem. 42:643-651. 1996. Comparative studies on volatile compo-
nents on non-fragrant and fragrant rices. J. Sci. Yang, X., Ye, Z. Q., Shi, C. H., Zhu, M. L., and
Food Agric. 70:151-161. Graham, R. D. 1998. Genotypic differences in
Williams, K. T., and Bevenue, A. 1953. A note concentrations of iron, manganese, copper,
on sugars in rice. Cereal Chem. 30:267-269. and zinc in polished rice grains. J. Plant Nutr.
Xu, C. W., and Zhang, A. H. 1998. Identification 21:1453-1462.
for genetic expression of rice quality traits in Zardoya, R., Tula, L., Castano, G., Mas, R.,
several crosses between indica and japonica Illnait, J., Fernandez, J. C., Diaz, E., and
cultivars. Chin. J. Rice Sci. 12:51-54. Fernandez, L. 1996. Effects of policosanol on
Yajima, L, Yanai, T., Nakamura, M., Sakakibara, hypercholesterolemic patients with abnormal
H., and Habu, T. 1978. Volatile flavor compo- serum biochemical indicators of hepatic func-
nents of cooked rice. Agric. Biol. Chem. tion. Curr. Ther. Res. 57:568-577.
42:1229-1233. Zhang, L., Hamberg, M., and Zhang, L. Y. 1996.
Yamamoto, A., Fujii, Y., Yasumoto, K., and Specificity of two lipoxygenases from rice:
Mitsuda, H. 1980. Product specificity of rice Unusual regiospecificity of a lipoxygenase
germ lipoxygenase. Lipids 15:1-5. isoenzyme. Lipids 31:8003-809.
CHAPTER 5
STARCH
Melissa Fitzgerald
NSW Agriculture
Yanco, NSW, Australia
Starch is one of the most important agricultural products for the world’s popula-
tion. Currently, 2,050 million tonnes of starch each year are produced from cereals
(Khush, 1997), and rice contributes significantly to that figure (Burrell, 2003). Food
use accounts for most of the starch, insofar as it can represent up to 80% of daily
calorific intake in some cultures (Burrell, 2003). Starch is also used in many non-
food applications, including adhesives, coatings, pharmaceuticals, fillers, viscosity
modifiers, and a host of others. Each application requires particular properties of the
starch, and the properties of starch are largely defined by the macro-, micro-, and
ultrastructure of the starch.
Starch granules constitute approximately 90% of the dry weight of a milled rice
grain. Starch is polyglucose, and the two types of glucose polymers that constitute
starch are amylose and amylopectin. Starch determines the physical and cooking
properties of rice grains, or at least contributes to them through interactions with
other components of the rice endosperm (proteins, lipids, water) or through interac-
tions with other ingredients used to process the rice. The ratio of amylose and amy-
lopectin in the starch, the solubility of each, and the structure of each fraction all
contribute to the performance of the rice grain. The synthesis and structure of starch
has long been a mystery, but because of the evolution of technology over the past
decade, knowledge of starch synthesis and structure has grown exponentially. This
deeper understanding makes it a real possibility that starch structure can one day be
manipulated to target starches for particular markets and applications and to assist
in identifying novel and potential applications.
THE STARCH GRANULE
Starch granules are made from two polymers of glucose: amylose and amy-
lopectin. Amylopectin molecules are highly branched, are of high molecular weight,
and constitute the skeleton of the starch granule (Kossmann and Lloyd, 2000).
Amylose molecules are essentially linear, but have some branching, and exist in
starch granules along with amylopectin. Amylopectin is essential for the synthesis
of a starch granule, but amylose is not. Amylose has received far more attention in
rice research than has amylopectin, even though it accounts for much less of the
starch, since it is considered to be a key indicator of cooking quality.
109
Starch granules that lack amylose are called either waxy, because of their muta-
tion at the waxy locus (Mizuno et al, 1993), or glutinous, because of their opaque
appearance. The structure of the starch granules differs among the cereals, and
starch granules in cereal endosperms differ from those in tubers and roots (Martin
and Smith, 1995). Starch granules in rice differ from starch granules in wheat, corn,
and barley. Those in corn range in shape from round to angular (Whistler and
BeMiller, 1997). Wheat grains contain two populations of granules, one about 40
µm in diameter, and the other ranging from 5 to 16 µm (Whistler and BeMiller,
1997). Starch granules of rice are small, about 3-5 µm each, and each granule is
angular. The granules are compound (Fig. 1). Each compound starch granule is
polyhedral and contains many (at least 16) individual granules (Fig. 1). The only
other cereal with compound starch granules in the endosperm is oats. The formation
of compound granules has received little attention, perhaps because the compound
structure itself is not a recognized source of variation affecting the end use of the
grain.
Starch granules contain some proteins and some lipids. The proteins are mainly
enzymes remaining from the synthesis of starch (Martin and Smith, 1995) and are
mostly accounted for by granule-bound starch synthase (GBSS) (Baldwin, 2001).
By contrast, in the endosperm, the protein is storage protein in bodies outside the
starch granules (Chrastil, 1990).
Lipids complexed with starch molecules occur in starch granules. During
synthesis of the starch molecules, complexes form between amylose and lipids as a
result of the lipid molecule lodging in the cavity of the amylose helix (Morrison et
Fig. I. Scanning electron micrograph of a fractured plane of a rice grain showing two compound
amyloplasts broken open, clearly revealing at least 16 single granules (see arrows) inside the
compound structure.
Starch / 111
al, 1993). There is also some suggestion of a complex forming between lipids and
long chains on the amylopectin molecule. The presence of the lipids can
substantially affect the cooking, processing, and sensory properties of the starch,
and the structure of the starch can affect the types of lipids present (Ozcan and
Jackson, 2002).
Macrostructure of the Starch Granule
CRYSTALLINE REGIONS
Starch granules contain crystalline and amorphous areas and are therefore
described as semicrystalline. The radial and ordered arrangement of the starch
molecules confers the semicrystallinity. Starch granules are birefringent when
viewed in polarized light, showing a cross, commonly known as the maltese cross
(Whistler and BeMiller, 1997). The birefringence is due to the high degree of
molecular order of the starch granules but does not actually define crystallinity. The
technique of X-ray diffraction provides information about crystallinity and is used
to show the structure of the crystallites, changes to crystallite structure after differ-
ent treatments, and relative amounts of crystalline and amorphous regions. Figure 2
(top) shows a diagram of the molecular packing in the A and B types of poly-
morphs. X-ray diffraction of intact granules of cereal starch shows that starch gran-
ules from cereals give the A pattern of crystallinity (Whistler and BeMiller, 1997).
The A pattern indicates that the crystalline regions contain parallel double helices
Fig. 2. Top, simple diagram of the A and B polymorphs of starch. The A polymorph is common in
cereals and contains less interhelical water than the B polymorph, which is common in tubers. The
hexagons represent the double helices depicted in the bottom diagram, each of which has six
residues/turns Bottom, three-dimensional diagram of a double helix. Courtesy Alfred French and
Glenn Johnson, USDA-ARS Southern Regional Research Center)
separated by interhelical water (Gidley, 1987). The A polymorph contains four

molecules of water per unit cell, and the B polymorph, commonly found in starch
from tubers, holds 36 molecules of interhelical water per unit cell (Baker et al,
2001). Figure 2 (bottom) depicts a double helix.
Microscopic studies suggest that the crystalline and amorphous regions are
arranged in blocklets that appear as protrusions on the surface of the granule when
viewed with scanning electron microscopy and atomic force microscopy (AFM)
(Gallant et al, 1997; Dang and Copeland, 2003). Gallant et al (1997) propose that
the different layers of the starch granule contain different-sized blocklets; larger
blocklets occur in the crystalline shells, and smaller blocklets occur in the
semicrystalline shells. The studies so far suggest that the size of the blocklets varies
also with the botanical origin of the starch (Gallant et al, 1997). The size of block-
lets in rice starch from several varieties has been estimated by AFM to be about 100
nm (Dang and Copeland, 2003). Blocklets in rice starch are proposed to contain
about 280 amylopectin side-chain clusters alternating with amorphous regions esti-
mated to be about 4 nm long (Dang and Copeland, 2003).
Amylopectin clusters arise from the crystallization of side chains into mature
amylopectin, removing those chains from further synthetic processes. Crystalliza-
tion generally occurs by the formation of double helices between adjacent A chains
of amylopectin. A single turn of a single chain in a double helix requires a maltooli-
gosaccharide to be six units long. Studies show that chains as short as six glucose
units can cocrystallize, but optimally chains need to be about 10–12 glucose units to
crystallize (Gidley and Bulpin, 1987; Imberty et al, 1991). Double helices are stabi-
lized by hydrogen bonding and van der Waals forces (Imberty et al, 1991). A double
helix can form between the outer chain of one amylopectin molecule and the outer
adjacent chain of a neighboring molecule (Hizukuri, 1986). Groups of A chains are
clustered in the same plane on different molecules, and the planar clustering allows
double helices to form within and between molecules on that plane, leading to
arrays of crystallinity. Studies show that, within an amylopectin molecule, clusters
are approximately 9 nm long, 120–400 nm wide, and separated by amorphous areas
about 4 nm long (Watanabe and French, 1980; Gallant et al, 1997; Knight et al,
1998).
AMORPHOUS REGIONS
Two types of amorphous areas occur in starch granules: channels extending from
the outer surface to the center of the granule; and regions that alternate with crystal-
line regions (Gallant et al, 1997). The amorphous regions that alternate with crys-
talline regions probably account for most of the amorphous area of the starch gran-
ule. These areas are rich in amylopectin branch points (Robin et al, 1974) and are
also likely to contain the amylose (Jacobs and Delcour, 1998). A study using cross-
linking agents showed amylose cross-linked with amylopectin, which led the
authors to suggest that amylose chains extend from the amorphous regions into the
crystalline regions of the amylopectin (Jane et al, 1992). More recently, gold-
labeled enzymes were used to label amylose and amylopectin; that work indicated
that discrete bands of amylose alternate with discrete bands of amylopectin, thereby
placing amylose in the amorphous region (Atkin et al, 1999). Another study, on the
synthesis of amylose, also suggests that, while amylose is being extended, the
growing polymer is extruded into the spaces in the amorphous areas (Vandewal et
al, 1998).
Starch / 113
Surface pores have been observed in starch granules from rice, wheat, and pota-
toes (Baldwin et al, 1994), and both pores and channels have been observed in
starch from wheat, rye, barley, millet, corn, and sorghum (Gallant et al, 1997). The
channels are thought to extend from the surface pores, creating a serpentine path-
way into the granule along the junction zones of the blocklets (Gallant et al, 1997).
Channels are thought to facilitate the exchange of water to and from the starch
granule (Baldwin et al, 1994) and to provide exit points for amylose as it leaches
during gelatinization of the starch (Gallant et al, 1997).
The ratio of amylose and amylopectin in the starch granule has been shown to
alter the relative sizes of the crystalline and amorphous areas (Jenkins and Donald,
1995). The length of the glucan chains, the structure of the individual starch mole-
cules, and the packing of the individual molecules all contribute to the radial and
ordered arrangement of both the crystalline and amorphous regions (Gidley, 1987;
Gidley and Bulpin, 1987).
Microstructure of the Components
AMYLOPECTIN
The structure of starch differs with botanical origin (Smith et al, 1997). Mutants
carrying a lesion in GBSSI—i.e., waxy mutants—accumulate starch granules
morphologically similar to those in wild-type endosperms, except that they lack
amylose (Yun and Matheson, 1993). Amylopectin molecules therefore are required
for the skeleton of the starch granules, and amylose perhaps fills the spaces.
The structure of an amylopectin molecule has never been visualized microscopi-
cally, but interpretations have been drawn from studies using enzymatic digestion of
the molecule, gel permeation and size-exclusion chromatography, electrophoresis,
absorption spectroscopy, or reducing-end chemistry. Two classes of enzymes have
been useful in investigations of the structure of amylopectin. One of these classes
contains the debranching enzymes isoamylase and pullulanase. These enzymes
digest a-(1-6) linkages only, which essentially decomposes the amylopectin mole-
cule into its constituent chains, allowing the chain-length distribution of amy-
lopectin to be determined. The second class of enzymes, which includes b-amylase,
digests a-(1-4) linkages only, liberating maltose. Sequential digestion with such
enzymes, coupled with measurements of the glucose released as maltose and of the
reducing power of the product, has led to interpretations of the length of the differ-
ent chains and a proposal of amylopectin structure, the cluster model, which is the
currently accepted one (Kainuma and French, 1972; Robin et al, 1974). The central
chain, carrying the reducing end, is termed the C chain; B chains subtend the C
chain, and A chains and other B chains subtend the B chains. The A chains are con-
nected to other chains through a-(1-6) linkages but are not branched themselves.
There are many more A chains than B chains, and the A chains are arranged in
ordered clusters and define the areas of crystallinity. The B chains can be divided
further into four populations of chains. B4 chains span four clusters of A chains; B3
chains span three clusters; B2 span two clusters; and B1 chains span only one clus-
ter (Hizukuri, 1986). The C chains range in size from 10 to 130 glucose units, with
most being around 40 glucose units (Hanashiro et al, 2002).
The architecture of the amylopectin molecule is determined by the length, fre-
quency, and placement of its constituent chains. Amylopectin molecules in rice fall
into three broad groups, with average degrees of polymerization (DP) of 700–2,100
(small), 4,400–8,400 (medium), and 13,400–26,500 (large) (Takeda et al, 2003).

The chain-length distribution of amylopectin from B-type starches is longer than
that from A-type starches (Hizukuri et al, 1983), and within the polymorph classes,
the chain-length distribution differs among species (Hizukuri et al, 1983; Reddy et
al, 1993; Ong et al, 1994; Hanashiro et al, 1996, 2002; Hanashiro and Takeda, 1998;
Takeda et al, 2003), and between indica and japonica rices (Hizukuri et al, 1989;
Umemoto et al, 2002). The chain-length distribution of starches has become much
simpler to assay since the development of capillary electrophoresis techniques
(Oshea et al, 1998). Chain-length distribution by capillary electrophoresis has been
applied to rice starch and used to compare varieties, describing more fully the
difference between the chain-length distributions of indica and japonica rices
(Nakamura et al, 2002), and to determine the function of different enzymes of
starch synthesis (Nakamura et al, 2002; Umemoto et al, 2002).
AMYLOSE
Amylose is the other polymer of glucose in starch granules. Amylose chains, like
amylopectin, are glucose polymers, but unlike amylopectin, are linked mainly
through a-(1-4) linkages. Amylose chains are extremely long relative to those of
amylopectin, about 1,000 DP (Takeda and Hizukuri, 1987). Experimental evidence
confirms that some branching occurs on the amylose chains. Incomplete hydrolysis
of potato amylose, using b-amylase, which completely digests linear a-(1-4)-linked
chains, indicated that amylose carries branches (Hizukuri et al, 1981). The
researchers then coupled a-amylase with the enzyme pullulanase, which hydrolyzes
a-(1-6) linkages, and the two enzymes completely hydrolyzed the amylose
(Hizukuri et al, 1981), again suggesting branching. Branching has recently been
essentially confirmed microscopically in amylose from pea (Gunning et al, 2002).
In the rice variety Sasanishiki, Takeda and Hizukuri (1987) found that each amylose
molecule was subtended by 8–11 branches and that the average length of those
chains was 115 DP. Furthermore, the branching structure of amylose differs
between and within species (Takeda and Hizukuri, 1987).
It is not surprising that the branching structure of amylose differs among varieties
within a species. Different varieties of rice carry different alleles of GBSSI (Ayres et
al, 1997; Bergman et al, 2001), possibly different alleles of starch branching enzyme
(SBE) (Bao et al, 2002), and possibly different factors affecting the expression of
GBSS or SBE (Larkin and Park, 1999; Cai et al, 2002). Therefore, different varieties
of rice are likely to differ in the frequency, length, and placement of branches on amy-
lose backbones, and therefore the structure of amylose is likely to differ among varieties.
When starch is heated in water, amylose leaches from the granules and consti-
tutes the hot-water-soluble component (Tsai et al, 1997). Presuming that the amy-
lose is free to leach, the structure of the amylose could cause it to become entangled
in the gelatinizing and swelling starch granule, thus causing it to be unable to leach.
In different varieties of rice, amylose differs in the amount that is soluble in hot
water (Bhattacharya et al, 1978; Ramesh et al, 1999; Mizukami et al, 1999). Figure
3 shows the distribution of hot-water-soluble amylose by gel permeation chroma-
tography from three different varieties of rice, all of 25% amylose. Molecules with
weights between 5.103 and 5.105 (~30–3,000 DP) elute in the separating phase of the
column. Clearly, the total amylose in a rice grain does not reflect the amount of hot-
water-soluble amylose. In that case, the structure of amylose in each individual
starch granule could differ.
Starch / 115
Fig. 3. Chromatogram from a gel-permeation column of solubilized starch from three rice varieties,
each with 25% amylose. Polymers greater than about 3,000 degrees of polymerization (DP) elute in
the void volume, and those less than 3,000 DP elute in the separating phase of the column. Amylose
elutes in the separating phase. Each variety shows a different amount of soluble amylose and a
different molecular weight of the amylose. DRI = differential refractive index.
There may be little variation in the molecular weight distribution of amylopectin

molecules for any given rice (Takeda et al, 2003), but there certainly is a broad
molecular weight distribution of amylose chains. Figure 3 shows the molecular
weight distribution of hot-water-soluble amylose from three varieties of rice by gel
permeation chromatography. It is clear from Figure 3 that a significant amount of
material with molecular weight greater than 5.105 (~3,000 DP) elutes in the void
volume, and a significant amount of material elutes in the separating phase of the
column, which ranges in molecular weight distribution from about 30 to 3,000 DP.
Such a range in the molecular weight of amylose from any given rice suggests that
the length of the amylose chains is not controlled by GBSSI. Consistent with this is
the idea that amylose is synthesized in accordance with the space available in the
starch granule, and as space becomes progressively limiting, the size of amylose
progressively decreases (Vandewal et al, 1998).
Starch granules have evolved to store a large amount of glucose in a condensed,
relatively dehydrated form, to provide energy to nourish the next generation during
germination and during the early stages of growth. However, the starch in different
genera, species within a genus, and cultivars within a species differ in macro-,
micro-, and ultrastructure, and consequently these variations cause different func-
tional properties. Because starch accounts for at least 90% of the dry weight of the
rice grain, the functional properties of starch, alone and interactively, are of extreme
interest to any customer or consumer of rice.
FUNCTIONAL PROPERTIES OF STARCH
During cooking, several stages occur to transform a raw rice grain into a cooked
grain of pleasing textural attributes. These include glass transition, gelatinization,
swelling, pasting and leaching of amylose, and retrogradation. Glass transition and
gelatinization are characteristics of starch as a semicrystalline and glassy polymer

system (Biliaderis et al, 1985; Slade and Levine, 1988, 1989; Waigh et al, 2000).
Glass transition occurs before gelatinization, and therefore determines the
gelatinization temperature (Slade and Levine, 1988). Water is a plasticizer that
depresses the temperature at which both glass transition and gelatinization occur,
and temperature affects the plasticizing capacity of water (Slade and Levine, 1988).
The addition of water and heat begins the process of cooking, and thus glass
transition and gelatinization affect many of the rheological properties of the rice as
whole grains and as an ingredient in other applications.
Following gelatinization, the granule can swell to many times its original vol-
ume, and in the absence of shear, it maintains its structure (Parker and Ring,
2001). During swelling, the amylose hydrates and leaches from the granule. The
leached amylose is then able to interact with other amylose molecules and form a
gel embedding the swollen amylopectin molecules and hydrated proteins, thus
forming a cooked rice grain or gel with a particular collection of textural attri-
butes.
Glass Transition
Before gelatinization of starch can occur, the glassy (amorphous) regions of the
starch must become soft and rubbery (Biliaderis et al, 1986; Slade and Levine,
1987). This process is called glass transition (Biliaderis et al, 1985, 1986; Slade and
Levine, 1988). Water depresses the glass transition temperature (Tg), and so at
higher moisture contents, glass transition occurs at lower temperatures (Biliaderis et
al, 1986; Slade and Levine, 1988, 1989; Matveev et al, 2000). In the conditions
required to cook rice, the rice is generally in excess water, so glass transition is
always depressed. When rice is used as an ingredient in a process, the moisture
content can be used to manipulate the glass transition and thereby the processing
properties of foods (Slade and Levine, 1988, 1989). For a synthetic polymer,
glass transition occurs when a glassy region is at a certain moisture content and
temperature that allows it to become rubbery, and when those conditions are
removed, the effects of glass transition are reversed. In terms of starch, though,
the glassy, or amorphous, areas contain the branch points of the amylopectin,
many of which subtend chains that traverse crystalline clusters on either side of
the amorphous regions (Biliaderis et al, 1986; Slade and Levine, 1987), so
reversing glass transition may not necessarily reverse the secondary effect of
rubbery amorphous regions.
Gelatinization
Gelatinization is the non-equilibrium melting of the crystalline regions (Slade

and Levine, 1988), and requires presoftening of the amorphous regions (glass tran-
sition). Gelatinization temperature is an important property of rice, or of any grain,
because it correlates strongly with the cooking time and the texture of the cooked
product (Maningat and Juliano, 1978). Gelatinization temperature is influenced by
the extent of crystallinity (Slade and Levine, 1988), the proportion of short chains in
amylopectin (Qi et al, 2003)—which is probably another way of expressing the
extent of crystallinity, and the proportion of long chains in amylopectin (Umemoto
et al, 2002).
Starch / 117
Swelling and Pasting
Once starch gelatinizes, the starch granules begin to swell and, in the absence of
shear, can swell to many times their initial volume while maintaining their integrity
(Parker and Ring, 2001). The swelling is accompanied by leaching of amylose
molecules into the continuous phase (Bhattacharya et al, 1972, 1978; Chinnaswamy
and Bhattacharya, 1986; Jacobs and Delcour, 1998; Nguyen et al, 1998; Rani and
Bhattacharya, 1989; Mizukami et al, 1999; Ramesh et al, 1999; Fitzgerald et al,
2003). In many studies, the amount of amylose that becomes soluble in hot water
has been reported to indicate and influence cooked texture more than the effect of
the total amount of amylose (Chinnaswamy and Bhattacharya, 1986; Sowbhagya et
al, 1987; Ramesh et al, 1999). However, in an earlier study, the importance of the
insoluble amylose was emphasized (Bhattacharya et al, 1978). The confusion
probably arose because several things influence the solubility of amylose. Solubility
of the amylose differs among varieties, and the extent of solubility could be due to
the structure of the amylose, entanglement with gelatinized amylopectin (and there-
fore the structure of amylopectin), or complexes between amylose and lipids.
Retrogradation
Gelatinized starch contains no crystalline regions, but under certain conditions of

storage and temperature, the molecules in a starch gel can reassociate into an
ordered structure (Baik et al, 1997). This process is called retrogradation, and in
terms of food technology, refers to the hardening of cooked rice or a starch gel upon
storage. The deterioration of the textural characteristics of cooked rice is due,
largely, to retrogradation of the starch (Yao and Ding, 2002).
Retrogradation describes the rapid recrystallization of amylose and the slow
recrystallization of amylopectin (Slade and Levine, 1987). The degree of retrogra-
dation and the nature of the newly formed crystals can depend on the time and tem-
perature of storage, the source of the starch, and the presence of other molecules in
the system (Slade and Levine, 1987; Baik et al, 1997).
Recrystallization of amylose is essentially the rapid formation of double helices
in parts of the amylose chains followed by aggregation of those helices (Gidley,
1989). The first stage of retrogradation depends on the amylose content of the rice
(Baik et al, 1997) and the amount of amylose that is free, rather than complexed
with lipids (Yao et al, 2002). Further, hot-water-soluble components of rice starch
with high molecular weight promote retrogradation more than lower-molecular-
weight polymers (Tsai and Lii, 2000), suggesting that the molecular weight distri-
bution of the amylose contributes significantly to the first phase of retrogradation.
Retrogradation due to amylose is not reversible at temperatures less than 100°C
(Miles et al, 1985) because amylose crystals melt only above 100°C.
Recrystallization of short amylopectin chains constitutes the second process of
retrogradation (Baik et al, 1997). Several studies suggest that the fine structure or
the chain-length distribution of amylopectin contributes to differences in the degree
of retrogradation by amylopectin, in particular, the proportion of short A chains in
the amylopectin (Lu et al, 1997; Fredriksson et al, 1998; Silverio et al, 2000; Tsai
and Lii, 2000; Yao et al, 2002). Further, Yao et al (2002) suggest that the interaction
of amylose with amylopectin increases the rate of amylopectin retrogradation. Ret-
rogradation due to amylopectin is reversible if the retrograded gel is exposed to a
temperature greater than the gelatinization temperature of crystalline amylopectin

(Lu et al, 1997).
Clearly, the functional properties of rice differ greatly among different varieties
of rice. These functional properties include the extent of crystallinity, glass transi-
tion, gelatinization temperature, the ability of granules to swell following gelatini-
zation, the solubility of the amylose, and the rate and degree of retrogradation.
These variations must be due, at least in part, to different macro-, micro-, and
molecular structures of the starch and must arise from different synthetic processes
or from different regulatory mechanisms affecting similar processes. Understanding
the processes of starch synthesis in different species, genera, and cultivars provides
opportunities to manipulate starch synthesis to generate novel starch structures.
THE ENZYMES OF STARCH SYNTHESIS
During the day, the vegetative parts of the plant produce sugars, mainly sucrose
and other nonreducing sugars, from photosynthesis. These sugars are transported
from their source, the bundle sheath cells of the leaves, to the various sink organs
within the plant. During the generative phase of plant growth (i.e., during grain
development), the developing grains exert the strongest sink demand in the plant.
Sucrose produced from photosynthesis is then transported in the phloem to the
developing grains.
Starch accumulates in the rice grain during the grain-filling period. Its deposition
involves a suite of enzymes and several processes. The first process is the
acquisition of the precursors, which are small sugars, including sucrose and small
polymers of glucose. That process relates more to plant physiology than to starch
chemistry and is not dealt with here. The second process is the synthesis of the
amylose and the amylopectin. A rich body of literature, which is constantly being
updated, discusses the synthetic processes (Martin and Smith, 1995; Sinclair et al,
1995; Ball et al, 1996, 1998; Morell et al, 1997; Smith et al, 1997; Isshiki et al,
1998; Vandewal et al, 1998; Denyer et al, 1999a,b, 2001; Kossmann and Lloyd,
2000; Myers et al, 2000). The third process is the organization of the starch poly-
mers into crystalline, semicrystalline, and amorphous regions and the organization
of amylose into its location within the starch granules. The fourth process, unique to
rice and oats, is the organization of individual starch granules into compound starch
granules. The third and fourth processes are described above.
Adenosine 5-Diphosphatase Glucose Pyrophosphorylase
Starch synthesis in rice becomes most active about 10 days after pollination;
after the endosperm sac has expanded and cellularized (Evers and Millar, 2002). In
those initial 10 days, the endosperms acquire the nutrients and precursors they
require for the synthesis and deposition of starch (and proteins and lipids) (Jenner et
al, 1993; Jenner, 1994; Bhullar and Jenner, 1986). Sucrose is the main form of car-
bon delivered to the developing endosperm, but the sucrose must be metabolized to
glucose-1-phosphate (G1P), most likely by the agency of sucrose synthase and
uridine 5-diphosphatase (UDP) glucose pyrophosphylase (Kossmann and Lloyd,
2000; Emes et al, 2003) to allow the glucose to participate in the pathway of starch
synthesis. Four classes of enzymes contribute to the synthesis of starch from adeno-
sine 5-diphosphatase (ADP) glucose: ADP glucose pyrophosphorylase (AGPase),
Starch / 119
the starch synthases, SBEs, and starch-debranching enzymes (Kossmann and Lloyd,
2000). Figure 4 shows a diagram of the pathways of starch synthesis.
The first committed step in the synthesis of starch is the conversion of G1P and
adenosine 5-triphosphatase (ATP) to ADP-glucose (ADPG) by the action of
AGPase (EC 2.7.7.23), producing pyrophosphate (Martin and Smith, 1995). It was
once believed that the main activity of AGPase occurred in the amyloplast and that
AGPase was the major regulatory step in the pathway of starch synthesis (Preiss
and Sivak, 1996). However, an increasing body of evidence shows, at least for bar-
ley, wheat, and maize, that an isoform of AGPase occurs in the cytosol, which
accounts for most of the synthesis of ADPG and is less sensitive to regulation
(Denyer et al, 1996b; Emes et al, 2003). Synteny among cereals could allow the
assumption that, in rice also, ADPG is synthesized mainly in the cytosol and partly
in the starch granule. The dual location of AGPase has at least two consequences:
first, it raises interesting possibilities for the regulation of starch synthesis, and sec-
ond, ADPG synthesized in the cytosol must be transported from the cytosol, across
the amyloplast membrane into the starch granule, where it can be incorporated into
starch.
Recent studies have identified a transporter for ADPG in the amyloplast enve-
lope of wheat (Emes et al, 2003). The transporter catalyzes the exchange of ADPG
with ADP, adenosine 5-triphosphatase, and ATP, but it does not bind UDP glucose.
Also, it is expressed in wheat endosperm from about 10 days after anthesis and
thereafter is highly expressed during the rest of grain filling in wheat (Emes et al,
2003). Emes et al (2003) suggest that this protein could be the major route for the
Fig. 4. Starch synthesis pathway. Adenosine 5-diphosphatase (ADP) glucose pyrophosphorylase

(AGPase) produces ADP-glucose (ADPG) in the cell but not inside the developing plastid. ADPG is
transported into the plastid and then participates in the synthesis of amylose (AM) by granule-bound
starch synthase (GBSS) or of amylopectin (AP) by the isoforms of starch synthase (SS) and branching
enzyme (BE), and by the debranching enzyme (DBE) and the disproportionating enzyme (D).
supply of substrate for starch synthesis. ADPG is the substrate for the starch
synthases, and Clarke et al (1999) have shown that GBSSI has a 10-fold lower
affinity for ADPG than the soluble starch synthases have. Therefore, if the activity
of the transporter regulates the amount of ADPG reaching the starch granules, it
could have some impact on the balance of amylose and amylopectin accumulated in
the starch granules.
The Starch Synthases
The family of starch synthases consists of the granule-bound and the other forms.
They catalyze the addition of glucose, from ADPG, to the nonreducing end of a
glucose polymer through the formation of an a-(14) linkage (Leloir et al, 1961;
Martin and Smith, 1995; Rahman et al, 1998; Smith, 1999, 2001; Kossmann and
Lloyd, 2000; Myers et al, 2000; Denyer et al, 2001; Nakamura, 2002). Recently,
though, it has been suggested that the starch synthases elongate chains from the
reducing end of a starch molecule (Mukerjea et al, 2002). It is possible that both
pathways operate in the granule, as is discussed later, but the evidence is not yet
strong enough to determine conclusively the relationship between the two types of
pathways in the process of starch synthesis.
The starch synthases can be classified based on the amino acid sequence of each.
Aside from the granule-bound form, GBSSI, there are four other forms: SSI, SSII,
SSIII, and SSIV (Kossmann and Lloyd, 2000; Denyer et al, 2001). GBSSI is exclu-
sively granule bound, but several reports suggest that the other starch synthases are
not exclusively soluble, in that a proportion of some are strongly associated with the
starch granule (Knight et al, 1998; Li et al, 1999; Kossmann and Lloyd, 2000). In all
plants analyzed so far, amylose has been shown in vitro and through the use of
waxy mutants to be synthesized by GBSSI (Sano, 1984; Villareal and Juliano, 1986;
Visser et al, 1991; Maddelein et al, 1994; Kossmann and Lloyd, 2000). In rice, the
waxy locus on chromosome 6 controls the synthesis of amylose (Sano, 1984). For
the synthesis of amylopectin, other isoforms of starch synthase identified are SSI,
SSIIa, SSIIb, and SSIII (Nakamura, 2002).
GRANULE-BOUND STARCH SYNTHASE—AMYLOSE CHAIN

ELONGATION
GBSSI is the product of the waxy (wx) gene and in rice is expressed in the
endosperm and in the pollen (Hirano and Sano, 2000). Some rices, commonly
known as waxy or glutinous rices, carry the recessive wx gene and contain no amy-
lose. In rice, there are at least three functional alleles at the waxy (Wx) locus—Wxa
and Wxb (Sano, 1984)—and a third recently identified, Wxop (Mikami et al, 1999).
Indica rices generally carry the Wxa allele, and the Wxb allele is generally found in
the japonica rices (Sano, 1984; Sano et al, 1986; Hirano and Sano, 2000; Mikami et
al, 2000; Umemoto and Terashima, 2002). The Wxop allele is expressed in the
opaque rices (Mikami et al, 1999). Opaque endosperms, first reported in Nepal,
look similar to waxy rices, but (unlike waxy rices, which contain no amylose), in
opaque endosperms, amylose accounts for 10% of the starch (Mikami et al, 1999).
The Wxb allele was derived from the Wxa allele during domestication of rice (Sano,
1984; Mikami et al, 1999; Olsen and Purugganan, 2002); the Wxop allele is thought
to have evolved from the Wxa allele independently of the Wxb gene (Mikami et al,
1999).
Starch / 121
Expression of the Wxa allele is about 10 times higher than that of the Wxb allele,
resulting in much more GBSSI protein and hence more amylose in the indica than
in the japonica varieties (Sano et al, 1986; Isshiki et al, 1998). The genetic basis for
the difference in GBSSI and amylose content was discovered recently. The Wxa
allele differs from the Wxb by a single base at the 5 end of the splice site on intron 1
of the Wx gene (Wang et al, 1995; Isshiki et al, 1998; Mikami et al, 1999; Olsen and
Purugganan, 2002). The mutation in Wxb is a single-base substitution from G to T at
the 5 splice site of the first intron of GBSSI, changing the sequence at the splice
site from AGGTATA to AGTTATA. The mutation decreases the efficiency of
splicing at that site to 17.5% of that in rices carrying the Wxa allele (Isshiki et al,
1998) and causes the accumulation of large amounts of unspliced transcript, which
still contains the first intron (Wang et al, 1995; Ayres et al, 1997; Isshiki et al,
1998). The unspliced transcript is not translated to produce GBSS protein or
amylose (Liu et al, 2003). The inefficient splicing at the proper site activates two
other, weaker splice sites in exon 1, one being a single base upstream and the other
about 100 bases upstream of the mutated site (Bligh et al, 1998; Cai et al, 1998;
Hirano et al, 1998; Isshiki et al, 1998). The overall decrease in GBSS protein in the
Wxb varieties, both from the decreased splicing efficiency at the proper site, and
from the weaker action of the alternative sites, explains the lower amylose content
of those rices carrying the Wxb allele.
Recent advances in technology have simplified efforts to understand the relation-
ship of the Wxa and Wxb alleles with amylose content. Isshiki et al (1998) introduced
a single-point mutation in both Wx alleles to convert the Wxa to Wxb in one rice and
the Wxb to Wxa in another rice. They showed in vitro that substituting the T in the
Wxb allele for a G significantly increased the amount of correctly spliced RNA so
that it equaled that of the Wxa allele. Conversely, substituting the G in the Wxa allele
for a T decreased the amount of spliced RNA (Isshiki et al, 1998). Another study
developed near-isogenic lines of IR36 (indica) and Taichung 65 (T65) (japonica) by
introducing the Wxb into IR36 and the Wxa into T65 by back-crossing (Mikami et al,
2000). This study showed in vivo that the near-isogenic line T65Wxa accumulated a
large amount of transcript correctly spliced and contained the same amount of
amylose as the wild-type IR36, but the near-isogenic line IR36Wxb contained 5%
less amylose than the wild-type T65 (Mikami et al, 2000); the researchers suggest
that modifiers in the indica background contribute to the lower amylose content of
the near-isogenic line IR36Wxb. Recently, it was reported that two of the dull
mutants, du-1 and du-2, which contain very low amylose (Satoh and Omura, 1981),
contain mutations of the genes required to activate the two weaker splice sites in
exon 1 of the Wxb gene (Isshiki et al, 2000). Perhaps the indica background of the
IR36Wxb created by Mikami et al (2000) was less able to activate the alternative
splice sites and thus resulted in less amylose in that modified line.
If we accept the distinction described above, that the Wxa allele carries the
sequence AGGTATA and the Wxb allele carries the sequence AGTTATA at the
splice site of intron 1, the distribution among indica and japonica does not hold true.
Several studies investigating the G/T polymorphism show that varieties such as
Labelle (Isshiki et al, 1998), Belle Patna, Dawn, Bluebonnet, Katy, L202, Lemont,
and many others (Ayres et al, 1997; Larkin and Park, 1999) have the correct splic-
ing sequence. Therefore, by definition, those varieties carry the Wxa allele. Those
varieties are all of intermediate amylose and are clearly not indica types. Further-
more, some of those varieties have been shown, by restriction fragment length
polymorphism mapping, to be japonica types (Larkin and Park, 1999). However, the
pedigree of some of these indicates that the Wx gene was introgressed from a tropi-
cal japonica variety during the selection process. Tropical japonica types have
intermediate amylose (Juliano, 1979; Sano et al, 1991) and also have an intermedi-
ate amount, relative to indica and temperate japonica, of GBSS protein (Sano et al,
1991).
The evolution of tropical japonica types from indica types has been postulated
but not proven (Sano et al, 1991). The evidence presented above suggests that the
tropical japonica types carry the Wxa allele but do not accumulate as much amylose
as the indica varieties, which also carry the Wxa allele. Given the sophistication of
technology this century, if the tropical japonicas carried a Wx allele that differed
from that found in the indica types, most likely it would have been discovered by
now. Either a fourth allele for intermediate amylose at the Wx locus is extremely
elusive, or some other factors modulate the expression of the Wxa allele in the tropi-
cal japonica background.
Much of the variability in the amylose content of different rices, including the
problem of the tropical japonica types, correlates with a microsatellite that is a
simple sequence repeat 59 bp upstream of exon 1 of the waxy gene (Ayres et al,
1997). The simple sequence repeat of 8–20 cytosine and thymine (CT) dinucleotide
repeats in the 5 untranslated region of exon 1 of the GBSSI gene (Bligh et al, 1995;
Ayres et al, 1997; Olsen and Purugganan, 2002) explains 83% of the variation in
amylose content of 92 varieties in the U.S. rice program (Ayres et al, 1997), and 74
in China (Shu et al, 1999). Trends are emerging to describe a relationship between
amylose content and the number of CT repeats. However, only a fraction of the
world’s rice germ plasm has been screened, so the relationships outlined below are
still quite preliminary. A summary of the results of several analyses (Bligh et al,
1995; Ayres et al, 1997; Bergman et al, 2000; Bao et al, 2002; Olsen and
Purugganan, 2002) and data from varieties and breeding lines analyzed in the
Australian Rice Improvement Program, shows the following relationships: Rices
with 18 or 19 CT repeats are, so far exclusively, low-amylose, temperate japonica
types carrying the Wxb allele; those with 14 or 20 CT repeats are, so far exclusively,
intermediate-amylose types descended from tropical japonica lines and carrying the
Wxa allele; those with 8, 10, or 11 CT repeats are, so far exclusively, indica types
carrying the Wxa allele. These last types generally contain high amylose, but a small
proportion of the varieties so far tested with 10 or 11 CT repeats classify as
intermediate amylose. Those with 17 CT repeats have evolved in several
phylogenetic categories, including waxy, and carry either the Wxa or the Wxb allele
(Bao et al, 2002; Olsen and Purugganan, 2002).
The microsatellite seems to discriminate between rices of different amylose
contents or those of the same amylose content but with different processing or
cooking characters. For example, the properties of progeny of a cross between a
standard, Calrose-type medium grain and a variety from Japan with superior
characteristics in the Japanese market, such as Koshihikari, could not be
distinguished using wet chemistry to determine amylose content. However, by
knowing the microsatellite alleles (CT19 and CT17), the progeny can be
categorized. Another example was given by Bergman et al (2001). In that study,
progeny of a cross between parents of CT20 and CT11 could not be distinguished
Starch / 123
by wet chemistry. Those with 11 CT repeats were desired, because experience has
shown that they suit a particular application (Bergman et al, 2001). In all analyses
performed so far, across divergent genetic backgrounds, a strong relationship exists
between the number of CT repeats and the amylose class (Bergman et al, 2001).
This strong relationship suggests that the microsatellite discriminates between
different types of amylose and thus between different pathways or processes of
amylose synthesis.
The mechanism of amylose synthesis has not been measured in vivo in any
organism, and rice starch has not been used in any in vitro studies of amylose syn-
thesis. However, knowledge gained from in vitro studies of amylose synthesis in
other plants can probably be applied to rice since the starch synthases have been
strongly conserved throughout evolution (Li et al, 1999; Myers et al, 2000).
Amylose synthesis occurs in starch granules by the addition of the glucose moi-
ety from ADPG to the nonreducing end of a glucan chain in a reaction catalyzed by
GBSSI (Leloir et al, 1961). GBSSI acts processively, in that it adds more than one
glucose molecule before dissociating from the polymer (Denyer et al, 1999a). The
synthesis of amylose lags behind the synthesis of amylopectin (Denyer et al, 2001),
but the two processes overlap. In rice, the activity of GBSSI in the endosperm is
highest from about 15 days after flowering until about 30 days after flowering
(Baun et al, 1970; Nakamura and Yuki, 1992; Takeda et al, 1993a; Nakamura et al,
1996b). ADPG is the substrate for amylose (and amylopectin) synthesis, and when
the supply of ADPG is limited, through a mutation either in AGPase or
phosphoglucomutase, the resultant starch contains no amylose (Vandenkoornhuyse
et al, 1996), suggesting that the starch synthases compete for ADPG (Kossmann and
Lloyd, 2000). GBSSI has the lowest affinity for ADPG among the starch synthases
(Frydman and Cardini, 1967; Clarke et al, 1999), so any decrease in the
concentration of ADPG would affect amylose synthesis before affecting
amylopectin synthesis (Kossmann and Lloyd, 2000; Denyer et al, 2001). Hence,
conditions that supply a sufficiently high concentration of ADPG to sustain the
synthesis of amylose, as well as amylopectin, are needed during the time when both
the granule-bound and other starch synthases are active.
ADPG is necessary for the synthesis of amylose, but it has long been known
from experiments in vitro that maltooligosaccharides (MOS) stimulate the synthesis
of amylose and provide primers for GBSSI (Leloir et al, 1961). The stimulation of
amylose synthesis by MOS has been observed in experiments in vitro using potato,
maize, and Arabidopsis (Denyer et al, 2001) and in the unicellular green alga Chla-
mydomonas rheinhardtii (Vandewal et al, 1998). Both roles for MOS, stimulating
and priming, were later confirmed by (Denyer et al, 1996a,b). In those later
experiments, washed starch granules were supplied with MOS and radiolabeled
ADPG. In the absence of MOS, the label was detected in amylopectin, but in the
presence of MOS, the label was detected in amylose (Denyer et al, 1996a). In
another experiment, radiolabeled ADPG and MOS with modified reducing ends
were supplied to washed starch granules, and the MOS were elongated by more
than one glucose moiety (Denyer et al, 1999a). The findings from those in vitro
experiments led Denyer and co-workers to conclude that MOS play a significant
role in the priming and synthesis of amylose in vivo and that GBSSI elongates MOS
progressively (Denyer et al, 1996a,b, 2001).
A role for MOS in the synthesis of amylose in vivo requires that a supply of
MOS be available for GBSSI activity. There is increasing evidence that amylose
synthesis occurs within the matrix of the granule (Tatge et al, 1999), so a sufficient
concentration of MOS must be able to penetrate the granule. French (1984) sug-
gested that molecules with molecular mass less than about 1 kDa could permeate
starch granules, so maltoheptaose would probably exceed the limit for entry. Even if
MOS can permeate the starch granule, questions have been raised about their role in
amylose synthesis. The concentrations of MOS used by Denyer and co-workers
were high, from 10 to 100 mM, and the concentration of MOS in plant tissues is
low, possibly too low to contribute to amylose synthesis (Kossmann and Lloyd,
2000). Furthermore, when expression of the gene was down-regulated, by silencing
it using antisense, the mechanisms that could provide MOS in vivo did not lead to a
decrease in amylose content (Kossmann and Lloyd, 2000).
An alternative theory to the MOS model described above has been proposed
using the green alga C. rheinhardtii as a model for the synthesis of starch in higher
plants (Vandewal et al, 1998). In that study, starch granules were incubated in
radiolabeled ADPG and then in unlabeled ADPG. The label was first detected in
external chains on amylopectin molecules and then later in polymers corresponding
to amylose in size (Vandewal et al, 1998). As time progressed, more label was
detected in the amylose fraction, prompting the conclusion that GBSSI extends an
amylopectin chain, which is then cleaved by an unidentified hydrolase reaction
(Vandewal et al, 1998). This model would not require a MOS primer, which,
according to Kossmann and Lloyd (2000), would be more consistent with the con-
ditions likely to occur inside the starch granule.
The experiment described above was repeated using pea embryos (Denyer et al,
1999b). In that study, labeled glucose was detected in the amylopectin fraction, but
during the pulse period, label could not be recovered in the amylose fraction (Denyer
et al, 1999b). The results prompted those authors to suggest that the mechanism of
amylose synthesis by the extension and subsequent cleavage of an amylopectin chain
was a process unique to the green alga and not a process common to higher plants
(Denyer et al, 1999a, 2001). However, in the experiments with higher plants, both the
pulse and chase were shorter than the experiments with Chlamydomonas. Before
discounting this model of amylose synthesis in higher plants, the possibility of
experimental parameters affecting the result must be ruled out.
The synthesis of amylose by the extension and subsequent cleavage of an exter-
nal amylopectin chain is more consistent with the in vivo conditions of the starch
granule than the MOS primer model (Ball et al, 1998; Kossmann and Lloyd, 2000).
The main difficulty with that model, though, is that a cleavage process has not been
identified. It is perhaps the time and the place to review the synthesis of amylose
and to consider experimental data in the literature in the context of the synthesis of
starch granules during grain filling in rice. The synthesis of amylose occurs in the
starch granules by GBSSI (Sano, 1984). Amylose synthesis in rice occurs during
grain filling but reaches its maximum activity some days after the peak activity of
the soluble starch synthases (Baun et al, 1970; Asaoka et al, 1985; Nakamura and
Yuki, 1992; Takeda et al, 1993a; Nakamura et al, 1996b). The substrate for the
starch synthases is ADPG (Recondo and Leloir, 1961; Vandenkoornhuyse et al,
1996; Tetlow et al, 1998; Emes et al, 2003). The soluble starch synthases have a
much higher affinity for ADPG than GBSS does (Frydman and Cardini, 1967), so if
ADPG is limiting, amylose synthesis will decline (Clarke et al, 1999). At the
growing regions of the starch granule, active amylopectin synthesis could
physiologically decrease the availability of ADPG for amylose synthesis. As amylo-
Starch / 125
pectin synthesis progresses and the activity of the soluble starch synthases decreases
at a previous growth region, the availability of ADPG for GBSSI activity could
increase at that previous growth region. GBSSI is presumed to be bound to the
amylopectin matrix of the granule, with no freedom to move (Vandewal et al,
1998). If GBSSI were to extend an amylopectin chain from the nonreducing end
(Vandewal et al, 1998), the active site of the enzyme needs to be close to the
nonreducing end of the amylopectin chain. Alternatively, as suggested by Mukerjea
et al (2002), GBSS could act by inserting the glucose moiety from ADPG to the
reducing end of a chain. If GBSS does act by adding a glucose unit to the reducing
end of a chain by a two-stage insertion mechanism (Mukerjea et al, 2002), the need
for GBSS to be near a nonreducing end of a chain becomes irrelevant, since GBSS
could open an amylopectin chain anywhere and begin inserting glucose units.
Extension of the chain from the reducing end closest to the active site of GBSS
would continue until the chain is removed from the active site of the enzyme by
hydrolysis (Mukerjea et al, 2002). Since the GBSS protein is immobile, the
hydrolysis reaction would detach the extended chain from the GBSS protein and
thus from the amylopectin primer.
THE OTHER STARCH SYNTHASES—AMYLOPECTIN CHAIN

ELONGATION
The other isoforms of starch synthase elongate the chains of amylopectin. In rice,
the isoforms active in the endosperm are SSI, SSII, and SSIII (Nakamura, 2002), and
by evolution, SSII has diverged into SSIIa and SSIIb (Myers et al, 2000), and
apparently SSIII has also diverged into SSIIIa and SSIIIb (Nakamura, 2002). Until
recently, these starch synthases, other than granule-bound one, have routinely been
called “soluble starch synthases.” Recent work suggests that some of the “soluble”
isoforms are strongly associated with the starch granule (Knight et al, 1998) and that
SSII, for example, enters both the soluble and insoluble phase at different stages
during development in pea and potato (Kossmann and Lloyd, 2000) and in wheat (Li
et al, 1999). Therefore, in this work, the term soluble is not used to describe the starch
synthases involved in amylopectin synthesis, but all discussion in this section about
starch synthases excludes any reference to the granule-bound isoform.
Currently no tool exists to measure or characterize the synthetic processes of
chain elongation in vivo. The tools most commonly used to determine the role of
each of the starch synthases consist primarily of mutational or antisense analyses.
Mutants, lacking one or several of the starch genes, produce starch with character-
istics that differ from the wild-type. Antisense plants, in which particular genes are
silenced, also produce starch that is structurally different from starch produced by
wild-type plants. With these techniques, individual enzymes cannot be conclusively
assigned a particular role in starch synthesis; each enzyme participates in a path-
way, so pleiotropic effects cannot be ruled out. For example, a mutant or antisense
plant lacking activity of a particular isoform might produce a particular type of
amylopectin that differs from that produced in control plants. Structural differences
between starch in control and treatment plants do not automatically assign a role for
that isoform, since the absence of that isoform could modify the precursor or
template for branching enzymes and debranching enzymes, which could modulate
their actions relative to their modes of action in control plants. Nevertheless,
mutants and antisense are currently the best tools in the box, and the following
discussion makes use of the knowledge generated by them.
SSI was first cloned from rice (Baba et al, 1993), and a similar cDNA was then
cloned from maize (Imparlradosevich et al, 1998), potato (Kossmann and Lloyd,
2000), and wheat (Li et al, 1999). SSI was the major form of the enzyme purified
from rice endosperm, which suggests, but does not confirm, that SSI accounts for
most starch synthase activity in rice (Baba et al, 1993).
It is clear from many studies that SSI, II, and III act on specific chain lengths. No
mutants for SSI have been discovered, and possibly this is because SSI activity is integral
to the successful growth of the plant, so a mutation in that enzyme could be lethal. The
activity of each isoform depends greatly on the botanical species of the starch (Smith et
al, 1997), so findings from plants other than rice cannot be readily transferred to rice.
Efforts to elucidate the role of SSI thus concentrate on antisense techniques and genetic-
suppression techniques. Recently, Nakamura (2002) described some work in which a
retrotransposon was inserted into the gene encoding SSI in rice. The mutant amylopectin
was enriched in chains of 6 and 7 DP and was correspondingly depleted in chains of 8–12
DP. These findings led Nakamura to conclude that SSI was responsible for creating short
A chains by specifically elongating chains of 6–7 DP (cited in Nakamura 2002). No other
work reports the action of SSI.
The activity of SSII has been characterized further than the activity of SSI, pre-
sumably because interfering with SSII is not as potentially devastating as interfering
with SSI. Antisense techniques to suppress SSII activity in potato (Lloyd et al, 1999)
have shown that the resultant amylopectin was enriched in chains of 8–12 DP, but the
alteration was not drastic. A mutation eliminating the activity of SSII in pea plants
(Rug5 locus) showed less starch and an overall decrease in chains 15–45 DP,
suggesting that SSII, at least in pea, is involved with the elongation of chains of 15–45
DP (Craig et al, 1998). In monocots, SSII has diverged further into the classes SSIIa
and SSIIb. Studies in wheat (Yamamori et al, 2000) and rice (Nakamura, 2002;
Nakamura et al, 2002) suggest functions for SSIIa but not for SSIIb.
In wheat, a line was generated with a deficiency in starch granule protein 1
(SGP1) at each of the three genomes, generating a line completely lacking SGP1
(Yamamori et al, 2000). Analysis of SGP1 indicated that it was, in fact, SSII
(Yamamori et al, 2000). The resulting amylopectin lacked intermediate chains of
DP 11–25, showed less crystallinity, and had no detectable gelatinization endotherm
(Yamamori et al, 2000), suggesting a key role for SSII in the synthesis of interme-
diate chains in amylopectin in wheat.
The activity of SSII in rice has been the subject of at least two recent studies
(Nakamura et al, 2002; Umemoto et al, 2002), and both suggest that SSII is a very
important enzyme for determining the structure and consequently the properties of
the starch. In the first study, mapping showed that SSIIa, alk, gel(t), and acl(t) are
most likely to be identical, all mapping to the same locus on chromosome 6
(Umemoto et al, 2002). Analyzing the progeny of a cross between Nipponbare, a
japonica, and Kasalath, an indica, Umemoto et al (2002) determined that the amy-
lopectin of the indica lines analyzed contained fewer short and more intermediate
chains than the japonica lines, and they identified a polymorphism at the alk locus
(SSIIa) which was responsible for the loss of SSIIa activity in japonica lines
(Umemoto et al, 2002). A further study, using many more indica and japonica lines,
showed that the chain-length distribution of indica lines was, in general, character-
ized by fewer short chains and a greater proportion of intermediate chains relative
to japonica lines (Nakamura et al, 2002). Furthermore, Nakamura et al (2002) sug-
gested that SSIIa activity defined these chain-length ratios; those with the higher
Starch / 127
average chain length (L-type) were indicas, and japonicas had a lower average
chain length (S-type). More specifically, Nakamura et al (2002) suggested that the
absence of SSIIa activity in japonica lines led to proportionally more short chains
(<11 DP) and less intermediate chains (12–25 DP), whereas the presence of SSIIa
activity in indica lines led to fewer short and more intermediate chains.
Little is known about the activity of SSIII, at least in rice. In other cereals and
tubers, mutations and antisense techniques have suggested that SSIII does contrib-
ute to amylopectin synthesis. The du-1 mutation in maize carries a lesion in SSIII
(Gao et al, 1998). The resultant amylopectin contains a fraction with a different
chain-length distribution, more like phytoglycogen, but the response was very small
(Singletary et al, 1997). Interpretation of those results is difficult, since pleiotropic
effects have been reported in the du-1 mutation, which decrease the activity of a
branching enzyme and increase the activity of another starch synthase (Seo et al,
2002). Suppression of the activity of SSII and SSIII simultaneously in potato affects
amylopectin structure, but the result is not an identifiable mixture of the individual
effects of each isoform (Edwards et al, 1999; Lloyd et al, 1999), suggesting that the
isoforms are interdependent (Smith, 2001).
The activity of the starch synthases, other than GBSS, remains a field of conjec-
ture and is based largely on indirect measurements or in vitro assays. The combined
results so far, though, suggest that the absence of one of the isoforms of starch syn-
thase cannot be compensated for by the others, indicating a specific individual or
interactive role for each isoform.
Recently, the combined efforts of starch and polymer scientists have provided a
way to assign the activity of the different isoforms of the starch synthases (J. Castro,
H. Chiou, M. Fitzgerald, M. Morell, and R. Gilbert, unpublished data). The
technique is in its infancy and utilizes complex mathematics (Clay and Gilbert,
Fig. 5. Plot of the molecular weight distribution of debranched amylopectin, expressed as the natural
log of the proportion (lnP) of chains of a particular molecular weight. Straight-line regions of these
plots indicate random-chain elongation or chain-termination events, suggesting that chains of about
degree of polymerization (DP) 10–35 and 45–70 are synthesized by random processes, and chains
shorter than 10 DP and between 35 and 45 DP are synthesised by nonrandom processes.
1995) to transform the absolute molecular weight distribution of the starch and to
plot these as functions of the natural log (J. Castro, H. Chiou, M. Fitzgerald, M.
Morell, and R. Gilbert, unpublished data) (Figure 5). Straight-line regions in these
lnP plots imply that, for each particular DP, the growth and stoppage events are
random (Clay and Gilbert, 1995) and therefore proportional to the amount of
glucose. The slope of the linear regions indicates the rate coefficients of the
processes, and regions of the plot where the rate is upward (for example, <10 DP
and between 30 and 40 DP) indicate nonrandom events (Clay and Gilbert, 1995).
The new technique (J. Castro, H. Chiou, M. Fitzgerald, M. Morell, and R. Gilbert,
unpublished data) relies on determining the absolute molecular weight of the starch
correctly and represents the first tool for analyzing starch synthetic processes in
vivo. It can be used to analyze the starch structure of mutants. Then the individual
rate constants for each enzyme and isoform can be defined, and thus the
contribution of each to the synthesis of a particular amylopectin structure can be
determined.
The Branching Enzymes
Branching enzymes create branches through dual action. First, they catalyze the
cleavage of a-(14) linkage, and then they catalyze the transfer of the product,
through its reducing end, to an existing glucose polymer by creating a new a-(16)
linkage. Multiple isoforms of branching enzymes exist and are distributed differ-
ently among species and tissues. Two nomenclatures exist for branching enzymes;
the one that has evolved in maize research will be used here. Branching enzymes
form two classes, BEI and BEII. In monocots, BEII has diverged, perhaps by a gene
duplication event (Baker et al, 2000), into BEIIa and BEIIb. In maize, the three iso-
forms, BEI, BEIIa, and BEIIb have been defined clearly, by both biochemical frac-
tionation and primary cDNA sequence. BEIIa and BEIIb seem closely related, but
encoded by different genes (Fisher et al, 1996), whereas BEI is distinctly different
from BEII (Seo et al, 2002). All three isoforms are present in rice endosperms.
BEIIb is unique to the endosperm (Nishi et al, 2001), and its activity is greater than
the activity of BEIIa (Yamanouchi and Nakamura, 1992). Studies in vitro suggest
that each isoform possesses distinct branching enzyme activity, and mutation and
antisense studies suggest that the activity of each isoform depends, to some degree,
on the activity of the other isoforms.
BEI has been isolated from rice (Mizuno et al, 1992), maize (Guan and Preiss,
1993), and wheat (Morell et al, 1997), and some of its properties have been charac-
terized in experiments in vitro. BEI expression in wheat (Morell et al, 1997) and
maize (Gao et al, 1996) is not readily detectable throughout the first half of grain
development but is expressed much more strongly throughout the second half in
both species. This differential expression suggests that the two isoforms of BEII
create the precursor for the action of BEI (Seo et al, 2002). BEI of both maize and
wheat branches amylose 10-fold more efficiently than it branches amylopectin, and
it transfers longer chains than either of the other two isoforms (Guan and Preiss,
1993; Takeda et al, 1993a).
Mutational and antisense analyses have been used to determine the function of
BEI. Rahman et al (2001) report that the absence of BEI activity, through either
mutation or down-regulation, has little impact on starch structure. More
specifically, the absence of BEI in a maize mutant carrying a mutator transposon in
Starch / 129
BEI did not alter the chain-length distribution of the amylopectin in the endosperm
(Blauth et al, 2002). Seo et al (2002) expressed the branching enzymes both
individually and in combination in yeast; the yeast lines lacked the endogenous
glycogen branching enzyme but retained the endogenous glycogen synthase. BEI
was unable to act when the other isoforms were missing (Seo et al, 2002). Taken
together, the data described above suggest that BEI either does not participate in
amylopectin synthesis, or, in its absence, the actions of BEIIa and BEIIb can
compensate for the loss of BEI. By contrast with data for maize, the chain-length
distribution of the amylopectin in the endosperm of a rice line mutated at BEI
showed fewer intermediate chains between 16 and 23 DP, fewer chains longer than
37 DP and more chains shorter than 12 DP (Nakamura, 2002). The difference in the
chain-length distribution between the BEI mutant of rice and the wild-type rice was
very small in the report by Nakamura (2002), and perhaps further work is needed to
determine conclusively whether the differences were significant.
The studies described above suggest that BEI, when absent, does not affect
amylopectin chain length, and when present in vitro, transfers longer chains and is
more active toward amylose. Amylose molecules carry branches (Hizukuri et al,
1981; Takeda et al, 1992, 1993b), which presumably result from the action of a
branching enzyme, and BEI, at least in vitro, is the only isoform that actively trans-
fers branches onto amylose (Guan and Preiss, 1993; Takeda et al, 1993a). A recent
study suggests grounds to explore further a link between amylose synthesis and
BEI. Cai et al (2002) have found that the bZIP family of transcriptional factors can
bind to regions in the promoter sequence of both the BEI and GBSS genes in rice
endosperm. They suggest, therefore, that the expression of GBSS and BEI in rice
endosperm is coordinately regulated at the genetic level (Cai et al, 2002).
The specific function of BEI in rice could depend on the variety and the allele of
BEI present. In rice, several microsatellite allele classes of BEI have been found at
the SBE locus in a region of the BEI gene (Bao et al, 2002). The microsatellite
classes are of either 1) 8 or 10 CT dinucleotide repeats with an insertion sequence
or 2) eight CT dinucleotide repeats alone (Bao et al, 2002). Bao et al (2002)
classified these starch branching enzyme (SBE) alleles as SBEA, SBEB, and SBEC.
SBEA was 10 CT repeats with the insertion sequence, SBEB was eight CT repeats
with the insertion sequence, and SBEC was eight CT repeats only. It is not known
whether these alleles describe different functions or activities of the BEI gene, but it
is interesting to note their distribution in a study with 66 rices (Bao et al, 2002). Bao
et al (2002) used only 66 rices, predominantly waxy, so their finding, that SBEA is
unique to indica rices and SBEB and C alleles are unique to japonica rices, remains
to be tested more rigorously.
The activity of BEIIa accounts for about 15% of total branching enzyme activity
in rice (Yamanouchi and Nakamura, 1992), and it is expressed throughout grain
development in rice (Yamanouchi and Nakamura, 1992), wheat (Morell et al,
1997), and maize (Seo et al, 2002). Mutants lacking BEIIa in rice (Nakamura, 2002)
and maize (Blauth et al, 2001) show no differences in the chain-length distribution
of the amylopectin.
Only BEI and BEIIa isoforms carry out branching enzyme activity in the synthe-
sis of transitory starch in leaves. A mutation in BEIIa leads to a large decrease in
the number of short chains in the transitory starch of rice (Nakamura, 2002) and
maize (Blauth et al, 2001) but no change to the endosperm starch. These findings
led Nakamura (2002) to suggest that, in endosperms, BEIIa might act in a way that
supports the functions of the other isoforms in amylopectin synthesis.
The function of BEIIb has been well characterized in several species with the aid
of a mutant with a lesion in the BEIIb gene. This mutant was found first in maize
(Boyer and Preiss, 1978) and was later identified in rice (Mizuno et al, 1993). This
mutant was identified early because it could be selected on the basis of the pheno-
type of the starch. Mutations in other branching enzyme genes could not be so eas-
ily selected, since the absence of BEI or BEIIa does not affect the structure of the
amylopectin (Seo et al, 2002). The absence of BEIIb was determined after the
mutant was found (Mizuno et al, 1993). The mutant, the amylose extender (ae)
mutant, results in high amylose, measured by wet chemistry, and a difference in the
structure of the amylopectin (Boyer and Preiss, 1981; Colleoni et al, 1999a;
Yoshimoto et al, 2000).
The differences in the structure of amylopectin in the ae mutant were once the
subject of debate (see references in Baba and Arai, 1984). Now, the differences in
the structure of the starch in the mutant have been well characterized for maize
(Baba and Arai, 1984; Yun and Matheson, 1993) and rice (Nishi et al, 2001).
The structure of the starch in the ae mutant was shown first for maize (Baba and
Arai, 1984). The amylopectin contains unusually long chains and is less branched
than amylopectin from the wild-type, and an intermediate fraction is also present
that is branched and carries chains of 40–50 DP (Baba and Arai, 1984; Yun and
Matheson, 1993; Nakamura, 2002). The lmaxof the complex between iodine and the
intermediate material is higher than that for amylopectin and lower than that for
amylose, leading to the conclusion that the intermediate material is responsible for
the higher amylose content of the ae mutant (Baba and Arai, 1984). The ae muta-
tion has the same effect on the structure of the starch in rice (Mizuno et al, 1993;
Nishi et al, 2001; Nakamura, 2002).
A recent study of the ae mutant in rice used several mutants of one variety to
measure the chain-length distribution of the amylopectin. The mutants used were
wx, ae + wx, and ae + Wx, and the wild-type was Kinmaze (Wx, non-ae) (Nishi et
al, 2001). Nishi et al (2001) showed that the ae mutant, in both the waxy and
nonwaxy backgrounds, led to a marked decrease in the number of glucan chains
from 6 to 17 DP and a slight increase in the number of chains from about 40 to 60
DP. The activity of all the enzymes of starch synthesis was measured, and in the ae
mutants, the activity of SSI, but not SSII, was significantly decreased (Nishi et al,
2001). Nishi et al (2001) ruled out a pleiotropic effect and suggested that the
decrease in SSI activity was caused by a decrease in the substrate for SSI. In
particular, they suggested that BEIIb transfers short chains for subsequent elongation
by a starch synthase; the absence of BEIIb activity in the ae mutants thus decreased
the amount of substrate for starch synthase activity, in particular the SSI isoform
(Nishi et al, 2001). This theory would explain the observed decrease in chains less than
17 DP but not the increase in the number of chains between 40 and 60 DP.
In their work, Nishi et al (2001) did not separate fractions of the starch, as did
others (Baba and Arai, 1984), so the chain-length analysis of the debranched starch
would include chains of 6–60 DP from amylose, the intermediate material, and
amylopectin. The length of the chains in the intermediate material is reported to be
between 40 and 50 DP (Baba and Arai, 1984). Nishi et al (2001) observed an
increase in the number of chains of 40–60 DP, which is consistent with the idea that
the intermediate material accounts for the increase they noted in that region.
Starch / 131
The studies described above, all indicate an important role for BEIIb. In its
absence, other isoforms cannot compensate for its activity, and quite a different
starch is produced. BEIIb activity is either directly responsible for short branches on
the amylopectin, or indirectly responsible, as suggested by Nishi et al (2001), by
providing the substrate in which SSI can elongate. The altered starch in the ae
mutants has proved suitable for some applications, and the ae mutant of maize is
now grown commercially (Brown et al, 1995).
The Debranching Enzymes
Debranching enzymes catalyze the removal of branches by hydrolyzing a-(16)

linkages. There are two types of debranching enzymes: isoamylases and
pullulanases. Isoamylase-type debranching enzymes seem to be able to hydrolyze
a-(16) linkages if the branches are widely spaced, as in glycogen and
amylopectin, but not when they are close together, as in pullulan (Nakamura et al,
1996a). By contrast, pullulanases can hydrolyze the closely spaced a-(16)
linkages present in pullulan and limit dextrins, but they do not readily digest
amylopectin (Nakamura et al, 1996a).
Historically, debranching enzymes were not considered to contribute to starch
synthesis, since theoretically, the synthesis of semicrystalline starch requires only
starch synthases and SBEs (Pan and Nelson, 1984; Smith, 2001). A unique property
of amylopectin—crystallization—is thought to be facilitated by the frequency and
position of the branches in the amylopectin molecule. Frequency and position of
branches would be governed by either addition or removal of branches. Several
studies suggest that addition of branches does not determine either frequency or
position. Guan et al (1995) replaced the glycogen branching enzyme of E. coli with
branching enzymes from maize, and a glycogenlike polymer was generated (Guan
et al, 1995). In a similar study, with yeast as the host, the branching frequency of
the glucan produced from glycogen synthase and maize branching isoforms was
intermediate between those of native amylopectin and native glycogen (Seo et al,
2002). The failure to generate crystalline amylopectin by expressing maize
branching enzymes in a host that usually accumulates glycogen suggests that
branching enzymes do not determine the frequency and position of branches in an
arrangement that favors crystallization. The corollary of those studies is that an
additional process, removal of branches, contributes to the synthesis of semicrystal-
line amylopectin.
Debranching enzymes were once thought to contribute to the degradation of
starch, but the body of evidence is increasingly convincing that debranching
enzymes are critical to the synthesis of amylopectin. Central to this argument was
identification of the factor that makes sweet corn sweet. Pan and Nelson (1984) first
determined that the sugary-1 (su-1) mutant of maize (i.e., sweet corn) carried a bio-
chemical lesion in the debranching enzyme. Previously, it had been shown that the
glucan produced by su-1 mutants of maize was highly branched and water-soluble;
it was thereafter termed phytoglycogen (Sumner and Somers, 1944). Phytoglycogen
was proposed as an intermediate molecule of starch synthesis, the substrate for a
debranching enzyme, which allowed the branches to crystallize and the removed
branches to splice to form amylose chains (Erlander, 1958). Pan and Nelson (1984)
suggested that the absence of debranching activity in su-1 mutants of maize was
consistent with the model proposed by Erlander (1958). James et al (1995) showed
that the su-1 mutants of maize carried a lesion in the gene of a debranching enzyme
that shared significant homology with a bacterial isoamylase. Similar sugary
mutants were also identified in rice, and they also contained phytoglycogen and a
lesion in the debranching enzyme gene (Nakamura et al, 1996a,b; Watanabe et al,
1997). Mutants lacking debranching enzyme in Chlamydomonas also contained
phytoglycogen (Mouille et al, 1996). Analysis of the expression of a debranching
enzyme from barley (iso1) also suggested that a debranching enzyme is involved in
starch synthesis (Sun et al, 1999). Studies across a range of taxa thus established
strong grounds for the involvement of a debranching enzyme in the synthesis of
amylopectin.
Both the isoamylase and pullulanase types of debranching enzymes occur in
potato tubers (Ishizaki et al, 1983), maize (Beatty et al, 1999), rice (Wong et al,
2003), and Chlamydomonas (Dauvillee et al, 2000). The exact role of each in the
synthesis of starch has not yet been defined, although pullulanase in maize, ZPU1,
possibly removes branches of only about three glucose units (Wu et al, 2002).
Although the su-1 gene encodes isoamylase in maize (James et al, 1995) and rice
(Kubo et al, 1999) many of the sugary mutants show a decrease in the activity of
either or both isoamylase and pullulanase. A difficulty in conclusively ascribing
function to debranching enzymes is that the endosperms and leaf tissues of the
mutants do not lack starch entirely; they accumulate phytoglycogen, but they also
accumulate a proportion of crystallized amylopectin, and they still retain some
debranching activity (Nakamura et al, 1996b, 1997; Zeeman et al, 1998; Burton et
al, 2002). Two models, glucan trimming (Ball et al, 1996) and water-soluble poly-
mer clearing (Zeeman et al, 1998) have been proposed to explain the action of the
debranching enzymes, but both are yet to be proven.
The glucan trimming hypothesis was first proposed by Ball et al (1996).
According to that model, preamylopectin is the precursor of mature amylopectin.
Preamylopectin occurs at the surface of the granule. In a particular potential cluster,
starch synthases elongate chains until specific lengths are reached that enable
branching enzymes to act. Branching enzymes then create branches randomly,
giving a structure analogous to phytoglycogen on the developing cluster (Ball et al,
1996). By the action of debranching enzymes, the randomly branched structure is
trimmed to facilitate crystallization (Ball et al, 1996). Furthermore, the specificity
of branching enzymes determines the size of each cluster, and the action of
debranching enzymes prevents the formation of phytoglycogen (Ball et al, 1996).
The model has been refined and expanded recently (Myers et al, 2000). Central to
the model is the theory that crystallization represents a powerful sink for carbon,
and, by this model, an essential step in crystallization is the transition from an
unordered state (preamylopectin) to an ordered state (mature amylopectin) (Myers
et al, 2000). In the absence of debranching activity, the preamylopectin would
continue to be an available substrate for starch synthases and branching enzymes,
eventually producing phytoglycogen. Mutants that retain some debranching activity,
such as those in which the mutation lies in either isoamylase or pullulanase,
accumulate some amylopectin and some phytoglycogen. Myers et al (2000) suggest
that the partitioning between amylopectin and phytoglycogen could happen very
early in the synthesis of the molecule, and if a crystallization event failed to occur,
because of low debranching activity, then further disorder would occur rapidly. If
debranching activity was present and crystallization did occur, then that event could
lead to conditions conducive to crystallization across that particular zone in the
Starch / 133
cluster (Myers et al, 2000). Low levels of debranching activity would lead to fewer
crystallization events and more potential for disorder; thus, some amylopectin and
some phytoglycogen would occur in mutants retaining some debranching activity
(Myers et al, 2000).
An alternative model has been proposed to explain the role of debranching
enzymes in amylopectin synthesis (Zeeman et al, 1998; Smith, 2001). In this model,
semicrystalline amylopectin is synthesized exclusively by the action of the different
isoforms of starch synthase and branching enzymes (Zeeman et al, 1998), and
debranching enzyme acts indirectly to prevent the accumulation of water-soluble
polymers. By this model, starch synthases and branching enzymes in the stroma
synthesize water-soluble glucans from small MOS, providing alternative, erroneous
substrates for the enzymes and decreasing the potential activity of the enzymes at
the surface of the granule (Zeeman et al, 1998). Debranching enzymes in the stroma
scavenge and prevent such soluble material from accumulating, thus directing the
enzymes of starch synthesis to the surface of the granule (Zeeman et al, 1998; Smith,
2001). Furthermore, debranching activity recovers the sugars from the water-soluble
polymers, recycling them for amylopectin synthesis (Zeeman et al, 1998).
The two models described above differ significantly in the action they ascribe to
the debranching enzymes. Smith (2001) suggests that resolving the issue is currently
beyond experimental capability, but the issue continues to generate significant sci-
entific interest.
The Disproportionating Enzyme
Genetic experiments implicate the disproportionating (D) enzymes in amy-

lopectin synthesis, suggesting that they act in concert with starch synthases and
branching enzymes to affect chain length (Myers et al, 2000). D enzyme transfers a
segment from one chain to another. The absence of D enzyme in mutants of Chla-
mydomonas led to less amylopectin, an increase in the proportion of short chains in
the amylopectin synthesized, and an increase in the size of the pool of MOS
(Colleoni et al, 1999b). So far, D enzyme has been found in the leaves and tubers of
some species (Colleoni et al, 1999b), but its activity has not been characterized. D
enzyme has been implicated in the metabolism of MOS during the degradation of
starch (Critchley et al, 2001).
The processes of starch synthesis are clearly not yet resolved. Analysis of
mutants and antisense plants are providing the most information about the activity
of the enzymes. Linking the activity of individual enzymes with the structure of the
starch presents scientists with a great challenge, which is increasingly surmountable
as technology advances.
CONCLUSION
Starch is one of the most important agricultural products in the world, and in this
millennium, research is rapidly unraveling the genetic basis for different starch
structures and the molecular and physical mechanisms that underlie the functional
properties of the starch. The growth in knowledge and information about starch will
significantly improve the ability of food technologists to design particular starch
properties, cereal chemists to define the starch properties, and rice-breeding pro-
grams to deliver varieties having those properties.
REFERENCES
Asaoka, M., Okuno, K., and Fuwa, H. 1985. Bao, J. S., Corke, H., and Sun, M. 2002. Micro-
Effect of environmental temperature at the satellites in starch-synthesizing genes in
milky stage on amylose content and fine relation to starch physicochemical properties
structure of amylopectin of waxy and non- in waxy rice (Oryza sativa L.). Theor. Appl.
waxy endosperm starches of rice (Oryza sativa Genet. 105:898-905.
L.). Agric. Biol. Chem. 49:373-379. Baun, L. C., Palmiano, E. P., Perez, C. M., and
Atkin, N. J., Cheng, S. L., Abeysekera, R. M., Juliano, B. O. 1970. Enzymes of starch
and Robards, A. W. 1999. Localisation of metabolism in the developing rice grain. Plant
amylose and amylopectin in starch granules Physiol. 46:429-434.
using enzyme-gold labeling. Starch/Staerke Beatty, M. K., Rahman, A., Cao, H. P.,
51:163-172. Woodman, W., Lee, M., Myers, A. M., and
Ayres, N. M., McClung, A. M., Larkin, P. D., James, M. G. 1999. Purification and molecular
Bligh, H. F. J., Jones, C. A., and Park, W. D. genetic characterization of ZPU1, a pullu-
1997. Microsatellites and a single-nucleotide lanase-type starch-debranching enzyme from
polymorphism differentiate apparent amy- maize. Plant Physiol. 119:255-266.
lose classes in an extended pedigree of US Bergman, C. J., Delgado, J. T., Bryant, R.,
rice germ plasm. Theor. Appl. Genet. Grimm, C., Cadwallader, K. R., and Webb, B.
94:773-781. D. 2000. Rapid gas chromatographic tech-
Baba, T., and Arai, Y. 1984. Structural charac- nique for quantifying 2-acetyl-1-pyrroline and
terisation of amylopectin and intermediate hexanal in rice (Oryza sativa, L.). Cereal
material in amylomaize starch granule. Agric. Chem. 77:454-458.
Biol. Chem. 48:1763-1775. Bergman, C. J., Delgado, J. T., McClung, A. M.,
Baba, T., Nishihara, M., Mizuno, K., Kawasaki, and Fjellstrom, R. G. 2001. An improved
T., Shimada, H., Kobayashi, E., Ohnishi, S., method for using a microsatellite in the rice
Tanaka, K., and Arai, Y. 1993. Identification, waxy gene to determine amylose class. Cereal
cDNA cloning, and gene expression of soluble Chem. 78:257-260.
starch synthase in rice (Oryza sativa L.) Bhattacharya, K. R., Sowbhagya, C. M., and
immature seeds. Plant Physiol. 103:565-573. Indudhara Swamy, Y. M. 1972. Interrelation-
Baik, M. Y., Kim, K. J., Cheon, K. C., Ha, Y. C., ship between certain physicochemical proper-
and Kim, W. S. 1997. Recrystallization kinet- ties of rice. J. Food Sci. 37:733-735.
ics and glass transition of rice starch gel Bhattacharya, K. R., Sowbhagya, C. M., and
system. J. Agric. Food Chem. 45:4242-4248. Indudhara Swamy, Y. M. 1978. Importance of
Baker, A. A., Helbert, W., Sugiyama, J., and insoluble amylose as a determinant of rice
Miles, M. J. 2000. New insight into cellulose quality. J. Sci. Food Agric. 29:359-364.
structure by atomic force microscopy shows Bhullar, S. S., and Jenner, C. F. 1986. Effects of
the I-alpha crystal phase at near-atomic reso- temperature on the conversion of sucrose to
lution. Biophys. J. 79:1139-1145. starch in the developing wheat endosperm.
Baker, A. A., Miles, M. J., and Helbert, W. 2001. Aust. J. Plant Physiol. 13:605-615.
Internal structure of the starch granule Biliaderis, C. G., Page, C. M., Slade, L., and
revealed by AFM. Carbohydr. Res. 330:249- Sirett, R. R. 1985. Thermal behavior of amy-
256. lose-lipid complexes. Carbohydr. Polym.
Baldwin, P. M. 2001. Starch granule-associated 5:367-389.
proteins and polypeptides. Starch/Staerke Biliaderis, C. G., Page, C. M., Maurice, T. J., and
53:475-503. Juliano B. O. 1986. Thermal characterization
Baldwin, P. M., Adler, J., Davies, M. C., and of rice starches: A polymer approach to phase
Melia, C. D. 1994. Holes in starch granules— transition of granular starch. J. Agric. Food
Confocal, SEM and light microscopy studies Chem. 34:6-14.
of starch granule structure. Starch/Staerke Blauth, S. L., Yao, Y., Klucinec, J. D., Shannon,
46:341-346. J. C., Thompson, D. B., and Guilitinan, M. J.
Ball, S., Guan, H. P., James, M., Myers, A., 2001. Identification of Mutator insertional
Keeling, P., Mouille, G., Buleon, A., Colonna, mutants of starch-branching enzyme 2a in
P., and Preiss, J. 1996. From glycogen to corn. Plant Physiol. 125:1396-1405.
amylopectin—A model for the biogenesis of Blauth, S. L., Kim, K. N., Klucinec, J.,
the plant starch granule. Cell 86:349-352. Shannon, J. C., Thompson, D., and
Ball, S. G., van de Wal, M. H. B. J., and Visser, Guiltinan, M. 2002. Identification of Mutator
R. G. F. 1998. Progress in understanding the insertional mutants of starch-branching
biosynthesis of amylase. Trends Plant Sci. enzyme 1 (sbe1) in Zea mays L. Plant Mol.
3:462-467. Biol. 48:287-297.
Starch / 135
Bligh, H. F. J., Till, R. I., and Jones, C. A. 1995. zations. 1. Model development and implica-
A microsatellite sequence closely linked to the tions for data interpretation. Macromolecules
waxy gene of Oryza sativa. Euphytica 86:83- 28:552-569.
85. Colleoni, C., Dauvillee, D., Mouille, G., Buleon,
Bligh, H. F. J., Larkin, P. D., Roach, P. S., Jones, A., Gallant, D., Bouchet, B., Morell, M.,
C. A., Fu, H. Y., and Park, W. D. 1998. Use of Samuel, M., Delrue, B., d’Hulst, C., Bliard, C.,
alternate splice sites in granule-bound starch Nuzillard, J. M., and Ball, S. 1999a. Genetic
synthase mRNA from low-amylose rice varie- and biochemical evidence for the involvement
ties. Plant Mol. Biol. 38:407-415. of alpha-1,4 glucanotransferases in amy-
Boyer, C. D., and Preiss, J. 1978. Evidence for lopectin synthesis. Plant Physiol. 120:993-
independent genetic control of starch branch- 1003.
ing enzymes from developing maize kernels. Colleoni, C., Dauvillee, D., Mouille, G., Morell,
Plant Physiol. 61:39. M., Samuel, M., Slomiany, M. C., Lienard, L.,
Boyer, C. D., and Preiss, J. 1981. Evidence for Wattebled, F., d’Hulst, C., and Ball, S. 1999b.
independent genetic control of the multiple Biochemical characterization of the Chlamy-
forms of maize endosperm branching enzymes domonas reinhardtii alpha-1,4 glucanotrans-
and starch synthases. Plant Physiol. 67:1141- ferase supports a direct function in amy-
1145. lopectin biosynthesis. Plant Physiol.
Brown, I. L., Mcnaught, K. J., and Moloney, E. 120:1005-1013.
1995. Hi-Maize(TM)—New directions in starch Craig, J., Lloyd, J. R., Tomlinson, K., Barber, L.,
technology and nutrition. Food Austr. 47:272- Edwards, A., Wang, T. L., Martin, C., Hedley,
275. C. L., and Smith, A. M. 1998. Mutations in the
Burrell, M. M. 2003. Starch: The need for gene encoding starch synthase II profoundly
improved quality or quantity—An overview. J. alter amylopectin structure in pea embryos.
Exp. Bot. 54:451-456. Plant Cell 10:413-426.
Burton, R. A., Jenner, H., Carrangis, L., Fahy, B., Critchley, J. H., Zeeman, S. C., Takaha, T., Smith,
Fincher, G. B., Hylton, C., Laurie, D. A., A. M., and Smith, S. M. 2001. A critical role for
Parker, M., Waite, D., van Wegen, S., disproportionating enzyme in starch breakdown
Verhoeven, T., and Denyer, K. 2002. Starch is revealed by a knock-out mutation in
granule initiation and growth are altered in Arabidopsis. Plant J. 26(1):89-100.
barley mutants that lack isoamylase activity. Dang, J. M. C., and Copeland, L. 2003. Imaging
Plant J. 31(1):97-112. rice grains using atomic force microscopy. J.
Cai, X. L., Wang, Z. Y., Xing, Y. Y., Zhang, J. Cereal Sci. 37:165-170.
L., and Hong, M. M. 1998. Aberrant splicing Dauvillee, D., Mestre, V., Colleoni, C.,
of intron 1 leads to the heterogeneous 5 UTR Slomianny, M. C., Mouille, G., Delrue, B.,
and decreased expression of waxy gene in rice d’Hulst, C., Bliard, C., Nuzillard, J. M., and
cultivars of intermediate amylose content. Ball, S. 2000. The debranching enzyme com-
Plant J. 14:459-465. plex missing in glycogen accumulating
Cai, Y., Xie, D. L., Wang, Z. Y., and Hong, M. mutants of Chlamydomonas reinhardtii dis-
M. 2002. Interaction of rice bZIP protein REB plays an isoamylase-type specificity. Plant Sci.
with the 5-upstream region of both rice sbe1 157:145-156.
gene and waxy gene. Chinese Sci. Bull. Denyer, K., Clarke, B., Hylton, C., Tatge, H., and
47:310-314. Smith, A. M. 1996a, The elongation of amy-
Chinnaswamy, R., and Bhattacharya, K. R. 1986. lose and amylopectin chains in isolated starch
Characteristics of gel-chromatographic frac- granules. Plant J. 10:1135-1143.
tions of starch in relation to rice and expanded Denyer, K., Dunlap, F., Thorbjornsen, T.,
rice-product qualities. Starch/Staerke 38:51- Keeling, P., and Smith, A. M. 1996b. The
57. major form of ADP-glucose pyrophosphory-
Chrastil, J. 1990. Protein-starch interactions in lase in maize endosperm is extra-plastidial.
rice grains. Influence of storage on oryzenin Plant Physiol. 112:779-785.
and starch. J. Agric. Food Chem. 38:1804- Denyer, K., Waite, D., Edwards, A., Martin, C.,
1809. and Smith, A. M. 1999a. Interaction with
Clarke, B. R., Denyer, K., Jenner, C. F., and amylopectin influences the ability of granule-
Smith, A. M. 1999. The relationship between bound starch synthase I to elongate malto-
the rate of starch synthesis, the adenosine 5- oligosaccharides. Biochem. J. 342:647-653.
diphosphoglucose concentration and the Denyer, K., Waite, D., Motawia, S., Moller, B.
amylose content of starch in developing pea L., and Smith, A. M. 1999b. Granule-bound
embryos. Planta 209:324-329. starch synthase I in isolated starch granules
Clay, P. A., and Gilbert, R. G. 1995. Molecular elongates malto-oligosaccharides processively.
weight distributions in free-radical polymeri- Biochem. J. 340:183-191.
Denyer, K., Johnson, P., Zeeman, S., and Smith, Gidley, M. J. 1987. Factors affecting the crystal-
A. M. 2001. The control of amylose synthesis. line type (A-C) of native starches and model
J. Plant Physiol. 158:479-487. compounds: A rationalisation of observed
Edwards, A., Fulton, D. C., Hylton, C. M., effects in terms of polymorphic structures.
Jobling, S. A., Gidley, M., Rossner, U., Carbohydr. Res. 161:301-304.
Martin, C., and Smith, A. M. 1999. A com- Gidley, M. J. 1989. Molecular mechanisms
bined reduction in activity of starch synthases underlying amylose aggregation and gelation.
II and III of potato has novel effects on the Macromolecules 22:351-358.
starch of tubers. Plant J. 17:251-261. Gidley, M. J., and Bulpin, P. V. 1987. Crystalli-
Emes, M. J., Bowsher, C. G., Hedley, C., Burrell, sation of malto-oligosaccharides as models of
M. M., Scrase-Field, E. S. F., and Tetlow, I. J. the crystalline forms of starch: Minimum
2003. Starch synthesis and carbon partitioning chain-length requirement for the formation of
in developing endosperm. J. Exp. Bot. 54:569- double-helices. Carbohydr. Res. 161:291-300.
575. Guan, H. P., and Preiss, J. 1993. Differentiation
Erlander, S. 1958. Proposed mechanism for the of the properties of the branching isozymes
synthesis of starch by glycogen. Enzymologia from maize (Zea mays). Plant Physiol.
19:278-283. 102:1269-1273.
Evers, T. and Millar, S. 2002. Cereal grain Guan, H. P., Kuriki, T., Sivak, M., and Preiss, J.
structure and development: Some implications 1995. Maize branching enzyme catalyzes
for quality. J. Cereal Sci. 36:261-284. synthesis of glycogen-like polysaccharide in
Fisher, D. K., Gao, M., Kim, K. N., Boyer, C. D., GLGB-deficient Escherichia coli. Proc. Natl.
and Guiltinan, M. J. 1996. Allelic analysis of Acad. Sci. U.S.A. 92:964-967.
the maize amylose-extender locus suggests Gunning, A. P., Giardina, T. P., Faulds, C. B.,
that independent genes encode starch-branch- Juge, N., Ring, S. G., Williamson, G., and
ing enzymes IIa and IIb. Plant Physiol. Morris, V. J. 2002. Surfactant-mediated solu-
110:611-619. bilisation of amylose and visualisation by
Fitzgerald, M. A., Martin, M., Ward, R. M., Park, atomic force microscopy. Carbohydr. Polym.
W. D., and Shead, H. J. 2003. Viscosity of rice 51:177-182.
flour: A rheological and biological study. J. Hanashiro, I., and Takeda, Y. 1998. Examination
Agric. Food Chem. 51:2295-2299. of number-average degree of polymerization
Fredriksson, H., Silverio, J., Andersson, R., and molar-based distribution of amylose by
Eliasson, A. C., and Aman, P. 1998. The influ- fluorescent labeling with 2-aminopyridine.
ence of amylose and amylopectin characteris- Carbohydr. Res. 306:421-426.
tics on gelatinization and retrogradation prop- Hanashiro, I., Abe, J., and Hizukuri, S. 1996. A
erties of different starches. Carbohydr. Polym. periodic distribution of the chain length of
35:119-134. amylopectin as revealed by high performance
French, D. 1984. Organization of the starch gran- anion exchange chromatography. Carbohydr.
ules: Pages 183-247 in: Starch: Chemistry and Res. 283:151-159.
Technology, 2nd ed. R. L. Whistler, J. N. Hanashiro, I., Tagawa, M., Shibahara, S., Iwata,
BeMiller, and E. F. Paschall, Eds. Academic K., and Takeda, Y. 2002. Examination of
Press, Orlando, FL. molar-based distribution of A, B and C chains
Frydman, R. B., and Cardini, C. E. 1967. Studies of amylopectin by fluorescent labeling with 2-
on the biosynthesis of starch. II. Some proper- aminopyridine. Carbohydr. Res. 337:1211-
ties of the adenosine diphosphate glucose: 1215.
Starch glucosyltransferase bound to the starch Hirano, H. Y., and Sano, Y. 2000. Comparison of
granule. J. Biol. Chem. 242:312-317. waxy gene regulation in the endosperm and
Gallant, D. J., Bouchet, B., and Baldwin, P. M. pollen in Oryza sativa L. Genes Genet. Sys.
1997. Microscopy of starch—Evidence of a 75:245-249.
new level of granule organization. Carbohydr. Hirano, H. Y., Eiguchi, M., and Sano, Y. 1998. A
Polym. 32:177-191. single base change altered the regulation of the
Gao, M., Fisher, D. K., Kim, K. N., Shannon, J. waxy gene at the posttranscriptional level
C., and Guiltinan, M. J. 1996. Evolutionary during the domestication of rice. Mol. Biol.
conservation and expression patterns of maize Evol. 15:978-987.
starch branching enzyme I and IIB genes sug- Hizukuri, S. 1986. Polymodal distribution of the
gests isoform specialization. Plant Mol. Biol. chain lengths of amylopectins, and its signifi-
30:1223-1232. cance. Carbohydr. Res. 147:342-347.
Gao, M., Wanat, J., Stinard, P. S., James, M. G., Hizukuri, S., Takeda, Y., and Yasuda, M. 1981.
and Myers, A. M. 1998. Characterization of Multi-branched nature of amylose and the
DULLI, a maize gene coding for a novel action of debranching enzymes. Carbohydr.
starch synthase. Plant Cell 10:399-412. Res. 94:205-213.
Starch / 137
Hizukuri, S., Kaneko, T., and Takeda, Y. 1983. Kainuma, K., and French, D. 1972. Naegeli
Measurement of the chain length of amy- amylodextrin and its relationship to starch
lopectin and its relevance to the origin of granule structure. II. Role of water in crystalli-
crystalline polymorphism of starch granules. zation of B-starch. Biopolymers 11:2241-
Biochim. Biophys. Acta 760:188-191. 2250.
Hizukuri, S., Takeda, Y., Maruta, N., and Juliano, Khush, G. S. 1997. Origin, dispersal, cultivation
B. O. 1989. Molecular structures of rice starch. and variation of rice. Plant Mol. Biol. 35(1-
Carbohydr. Res. 189:227-235. 2):25-34.
Imberty, A., Buleon, A., Vinh, T., and Perez, S. Knight, M. E., Harn, C., Lilley, C. E. R., Guan,
1991. Recent advances in knowledge on starch H. P., Singletary, G. W., Muforster, C. M.,
structure. Starch/Staerke 43:375-384. Wasserman, B. P., and Keeling, P. L. 1998.
Imparlradosevich, J. M., Li, P., Zhang, L., Molecular cloning of starch synthase I from
Mckean, A. L., Keeling, P. L., and Guan, H. P. maize (w64) endosperm and expression in
1998. Purification and characterization of Escherichia coli. Plant J. 14:613-622.
maize starch synthase I and its truncated Kossmann, J., and Lloyd, J. 2000. Understanding
forms. Arch. Biochem. Biophys. 353(1):64-72. and influencing starch biochemistry. Crit. Rev.
Ishizaki, Y., Taniguchi, H., Maruyama, Y., and Plant Sci. 19:171-226.
Nakamura, M. 1983. Debranching enzymes of Kubo, A., Fujita, N., Harada, K., Matsuda, T.,
potato tuber (Solanum tuberosum L.). I. Purifi- Satoh, H., and Nakamura, Y. 1999. The starch-
cation and some properties of potato isoamy- debranching enzymes isoamylase and pullu-
lase. Agric. Biol. Chem. 47:771-779. lanase are both involved in amylopectin bio-
Isshiki, M., Morino, K., Nakajima, M., Okagaki, synthesis in rice endosperm. Plant Physiol.
R. J., Wessler, S. R., Izawa, T., and 121:399-409.
Shimamoto, K. 1998. A naturally occurring Larkin, P. D. and Park, W. D. 1999. Transcript
functional allele of the rice waxy locus has a accumulation and utilization of alternate and
GT to TT mutation at the 5 splice site of the non-consensus splice sites in rice granule-
first intron. Plant J. 15(1):133-138. bound starch synthase are temperature-sensi-
Isshiki, M., Nakajima, M., Satoh, H., and tive and controlled by a single-nucleotide
Shimamoto, K. 2000. Dull: Rice mutants with polymorphism. Plant Mol. Biol. 40:719-727.
tissue-specific effects on the splicing of the Leloir, L. F., de Fekete, M. A. R., and Cardini, C.
waxy pre-mRNA. Plant J. 23:451-460. E. 1961. Starch and oligosaccharide synthesis
Jacobs, H., and Delcour, J. A. 1998. Hydrother- from uridine diphosphate glucose. J. Biol.
mal modifications of granular starch, with Chem. 236:636-641.
retention of the granular structure. A review. J. Li, Z., Rahman, S., Kosar-Hashemi, B., Mouille,
Agric. Food Chem. 46:2895-2905. G., Appels, R., and Morell, M. K. 1999.
James, M. G., Robertson, D. S., and Myers, A. Cloning and characterization of a gene
M. 1995. Characterization of the maize gene encoding wheat starch synthase. I. Theor.
sugary-1, a determinant of starch composition Appl. Genet. 98:1208-1216.
in kernels. Plant Cell 7:417-429. Liu, Q. Q., Wang, Z. Y., Chen, X. H., Cai, X. L.,
Jane, J., Xu, A., Radosavljevic, M., and Seib, P. Tang, S. Z., Yu, H. X., Zhang, J. L., Hong, M.
A. 1992. Location of amylose in normal starch M., and Gu, M. H. 2003. Stable inheritance of
granules. I. Susceptibility of amylose and the antisense waxy gene in transgenic rice
amylopectin to cross-linking reagents. Cereal with reduced amylose level and improved
Chem. 69:405-409. quality. Transgenic Res. 12(1):71-82.
Jenkins, P. J., and Donald, A. M. 1995. The Lloyd, J. R., Landschutze, V., and Kossmann, J.
influence of amylose on starch granule struc- 1999. Simultaneous antisense inhibition of
ture. Int. J. Biol. Macromol. 17:315-321. two starch-synthase isoforms in potato tubers
Jenner, C. F. 1994. Starch synthesis in the kernel leads to accumulation of grossly modified
of wheat under high temperature conditions. amylopectin. Biochem. J. London 338:515-
Aust. J. Plant Physiol. 21:791-806. 521.
Jenner, C. F., Siwek, K., and Hawker, J. S. 1993. Lu, S., Chen, L. N., and Lii, C. Y. 1997. Correla-
The synthesis of [C-14] starch from [C-14] tions between the fine structure, physiochemi-
sucrose in isolated wheat grains is dependent cal properties, and retrogradation of amy-
upon the activity of soluble starch synthase. lopectins from Taiwan rice varieties. Cereal
Aust. J. Plant Physiol. 20:329-335. Chem. 74:34-39.
Juliano, B .O. 1979. Amylose analysis in rice—A Maddelein, M. L., Libessart, N., Bellanger, F.,
review. Pages 251-260 in: Proc. Workshop on Delrue, B., Dhulst, C., Vandenkoornhuyse, N.,
Chemical Aspects of Rice Grain Quality. Fontaine, T., Wieruszeski, J. M., Decq, A., and
International Rice Research Institute, Los Ball, S. 1994. Toward an understanding of the
Baños, Laguna, Philippines. biogenesis of the starch granule—Determina-
tion of granule-bound and soluble starch syn- Myers, A. M., Morell, M. K., James, M. G., and
thase functions in amylopectin synthesis. J. Ball, S. G. 2000. Recent progress toward
Biol. Chem. 269:25150-25157. understanding biosynthesis of the amylopectin
Maningat, C. C., and Juliano, B. O. 1978. Prop- crystal. Plant Physiol. 122:989-997.
erties of lintnerized starch granules from rices Nakamura, Y. 2002. Towards a better under-
of different amylose content and gelatinization standing of the metabolic system for amy-
temperature. Int. Rice Res. Newsl. 3:7-8. lopectin biosynthesis in plants: Rice
Martin, C., and Smith, A. M. 1995. Starch bio- endosperm as a model tissue. Plant Cell
synthesis. Plant Cell 7:971-985. Physiol. 43:718-725.
Matveev, Y. I., Grinberg, V. Y., and Nakamura, Y., and Yuki, K. 1992. Changes in
Tolstoguzov, V. B. 2000. The plasticizing enzyme activities associated with carbohydrate
effect of water on proteins, polysaccharides metabolism during the development of rice
and their mixtures. Glassy state of bio- endosperm. Plant Sci. 82(1):15-20.
polymers, food and seeds. Food Hydrocolloids Nakamura, Y., Umemoto, T., Ogata, N., Kuboki,
14:425-437. Y., Yano, M., and Sasaki, T. 1996a. Starch
Mikami, I., Aikawa, M., Hirano, H. Y., and Sano, debranching enzyme (R-enzyme or pullu-
Y. 1999. Altered tissue-specific expression at lanase) from developing rice endosperm: Puri-
the Wx gene of the opaque mutants in rice. fication, cDNA and chromosomal localization
Euphytica 105:91-97. of the gene. Planta 199:209-218.
Mikami, I., Dung, L. V., Hirano, H. Y., and Sano, Nakamura, Y., Umemoto, T., Takahata, Y.,
Y. 2000. Effects of the two most common Wx Komae, K., Amano, E., and Satoh, H. 1996b.
alleles on different genetic backgrounds in Changes in structure of starch and enzyme
rice. Plant Breed. 119:505-508. activities affected by sugary mutations in
Miles, M. J., Morris, V. J., Orfore, P. D., and developing rice endosperm. Possible role of
Ring, S. G. 1985. The roles of amylose and starch debranching enzyme (R-enzyme) in
amylopectin in the gelation and retrogradation amylopectin biosynthesis. Physiol. Plant.
of starch. Carbohydr. Res. 135:271-281. 97:491-498.
Mizukami, H., Takeda, Y., and Hizukuri, S. Nakamura, Y., Kubo, A., Shimamune, T.,
1999. The structure of the hot-water soluble Matsuda, T., Harada, K., and Satoh, H. 1997.
components in the starch granules of new Correlation between activities of starch
Japanese rice cultivars. Carbohydr. Polym. debranching enzyme and alpha-polyglucan
38:329-335. structure in endosperms of sugary-1 mutants of
Mizuno, K., Kimura, K., Arai, Y., Kawasaki, T., rice. Plant J. 12(1):143-153.
Shimada, H., and Baba, T. 1992. Starch Nakamura, Y., Sakurai, A., Inaba, Y., Kimura,
branching enzymes from immature rice seeds. K., Iwasawa, N., and Nagamine, T. 2002. The
J. Biochem. 112:643-651. fine structure of amylopectin in endosperm
Mizuno, K., Kawasaki, T., Shimada, H., Satoh, from Asian cultivated rice can be largely clas-
H., Kobayashi, E., Okumura, S., Arai, Y., and sified into two classes. Starch/Staerke 54:117-
Baba, T. 1993. Alteration of the structure of 131.
starch components by the lack of an isoform of Nguyen, Q. D., Jensen, C. T. B., and Kristensen,
starch branching enzyme in rice seeds. J. Biol. P. G. 1998. Experimental and modelling stud-
Chem. 268:19084-19091. ies of the flow properties of maize and waxy
Morell, M. K., Blennow, A., Kosarhashemi, B., maize starch pastes. Chem. Eng. J. 70:165-
and Samuel, M. S. 1997. Differential expres- 171.
sion and properties of starch branching Nishi, A., Nakamura, Y., Tanaka, N., and Satoh,
enzyme isoforms in developing wheat H. 2001. Biochemical and genetic analysis of
endosperm. Plant Physiol. 113(1):201-208. the effects of amylose-extender mutation in
Morrison, W. R., Law, R. V., and Snape, C. E. rice endosperm. Plant Physiol. 127:459-472.
1993. Evidence for inclusion complexes of Olsen, K. M., and Purugganan, M. D. 2002.
lipids with v-amylose in maize, rice and oat Molecular evidence on the origin and evolu-
starches. J. Cereal Sci. 18:107-109. tion of glutinous rice. Genet. 162:941-950.
Mouille, G., Maddelein, M. L., Libessart, N., Ong, M. H., Jumel, K., Tokarczuk, P. F.,
Talaga, P., Decq, A., Delrue, B., and Ball, S. Blanshard, J. M. V., and Harding, S. E. 1994.
1996. Preamylopectin processing—A manda- Simultaneous determinations of the molecular
tory step for starch biosynthesis in plants. weight distributions of amyloses and the fine
Plant Cell 8:1353-1366. structures of amylopectins of native starches.
Mukerjea, R., Yu, L. L., and Robyt, J. F. 2002. Carbohydr. Res. 260(1):99-117.
Starch biosynthesis: Mechanism for the elon- Oshea, M. G., Samuel, M. S., Konik, C. M., and
gation of starch chains. Carbohydr. Res. Morell, M. K. 1998. Fluorophore-assisted car-
337:1015-1022. bohydrate electrophorsis (face) of oligosaccha-
Starch / 139
rides—Efficiency of labelling and high-reso- 5. Inter- and intraspecific differentiation in the

lution separation. Carbohydr. Res. 307:1-12. waxy gene expression of rice. Euphytica
Ozcan, S., and Jackson, D. S. 2002. The impact 35(1):1-9.
of thermal events on amylose-fatty acid com- Sano, Y., Hirano, H. Y., and Nishimura, M. 1991.
plexes. Starch/Staerke 54:593-602. Evolutionary significance of differential regu-
Pan, D., and Nelson, O. E., 1984. A debranching lation at the wx locus of rice. Pages 11-20 in:
enzyme deficiency in endosperms of the sug- Rice Genetics II. International Rice Research
ary-1 mutants of maize. Plant Physiol. 74:324- Institute, Manila.
328. Satoh, H., and Omura, T. 1981. New endosperm
Parker, R., and Ring, S. G. 2001. Aspects of the mutations induced by chemical mutagenes in
physical chemistry of starch. J. Cereal Sci. rice, Oryza sativa L. Jpn. J. Breed. 31:316-326.
34:1-17. Seo, B. S., Kim, S., Scott, M. P., Singletary, G.
Preiss, J., and Sivak, M. N. 1996. Starch synthe- W., Wong, K. S., James, M. G., and Myers, A.
sis in sinks and sources. Pages 63-95 in: M. 2002. Functional interactions between het-
Photoassimilate Distribution in Plants and erologously expressed starch-branching en-
Crops: Source-Sink Relationships. Marcel zymes of maize and the glycogen synthases of
Dekker, Inc., New York. brewer’s yeast. Plant Physiol. 128:1189-1199.
Qi, X., Tester, R. F., Snape, C. E., and Ansell, R. Shu, Q. Y., Wu, D. X., Gao, M. W., Ayres, N.
2003. Molecular basis of the gelatinisation and M., Larkin, P. D., and Park, W. D. 1999.
swelling characteristics of waxy rice starches Microsatellites polymorphism on the waxy
grown in the same location during the same gene locus and their relationship to amylose
season. J. Cereal Sci. 37:363-376. content in indica and japonica rice. Acta
Rahman, S., Regina, A., Li, Z. Y., Mukai, Y., Genet. Sinica 26:350-358.
Yamamoto, M., Kosar-Hashemi, B., Abrahams, Silverio, J., Fredriksson, H., Andersson, R.,
S., and Morell, M. K. 2001. Comparison of Eliasson, A. C., and Aman, P. 2000. The effect
starch-branching enzyme genes reveals evo- of temperature cycling on the amylopectin ret-
lutionary relationships among isoforms. rogradation of starches with different amy-
Characterisation of a gene for starch-branching lopectin unit-chain length distribution. Carbo-
enzyme IIa from the wheat D genome donor hydr. Polym. 42:175-184.
Aegilops tauschii. Plant Physiol. 125:1314- Sinclair, P. J., Neeson, R. J., and Barlow, E. W.
1324. R. 1995. Osmotic potential and soluble solids
Rahman, A., Wong, K. S., Jane, J. L., Myers, A. concentration in onion (Allium cepa) bulbs. J.
M., and James, M. G. 1998. Characterisation Sci. Food Agric. 69:211-214.
of su1 isoamylase, a detrement of storage Singletary, G. W., Banisadr, R., and Keeling, P.
starch structure in maize. Plant Physiol. L. 1997. Influence of gene dosage on carbo-
117:425-435. hydrate synthesis and enzymatic activities in
Ramesh, M., Ali, S. Z., and Bhattacharya, K. R. endosperm of starch-deficient mutants of
1999. Structure of rice starch and its relation maize. Plant Physiol. 113:293-304.
to cooked-rice texture. Carbohydr. Polym. Slade, L., and Levine, H. 1987. Recent advances
38:337-347. in starch retrogradation. Industrial polysaccha-
Rani, M. R., and Bhattacharya, K. R. 1989. rides. Pages 387-430 in: The Impact of Bio-
Slurry viscosity as a possible indicator of rice technology and Advanced Methodologies.
quality. J. Texture Stud. 20:139-149. Gordon and Breach, New York.
Recondo, E., and Leloir, L. F. 1961. Adenosine Slade, L., and Levine, H. 1988. Non-equilibrium
diphosphate glucose and starch synthesis. Bio- melting of native granular starch. Part I. Tem-
chem. Biophys. Res. Comm. 6:85-88. perature location of the glass transition associ-
Reddy, K. R., Ali, S. Z., and Bhattacharya, K. R. ated with gelatinisation of A-type cereal
1993. The fine structure of rice-starch amy- starches. Carbohydr. Polym. 8:183-208.
lopectine and its relation to the texture of Slade, L., and Levine, H. 1989. A food polymer
cooked rice. Carbohydr. Polym. 22:267-275. science approach to selected aspects of starch
Robin, J. P., Mercier, C., Charbonniere, R., and gelatinization and retrogradation. Pages 215-
Guilbot, A. 1974. Lintnerized starches. Gel 270 in: Frontiers in Carbohydrate Research.
filtration and enzymatic studies of insoluble Aspen Publishers Inc., New York.
residues from prolonged acid treatment of Smith, A. M. 1999. Making starch. Curr. Opin.
potato starch. Cereal Chem. 51:389-406. Plant Biol. 2:223-229.
Sano, Y. 1984. Differential regulation of waxy Smith, A. M. 2001. The biosynthesis of starch
gene expression in rice endosperm. Theor. granules. Biomacromolecules 2:335-341.
Appl. Genet. 68:467-473. Smith, A. M., Denyer, K., and Martin, C., 1997.
Sano, Y., Katsumata, M., and Okuno, K. 1986. The synthesis of the starch granule. Annu.
Genetic studies of speciation in cultivated rice. Rev. Plant Physiol. Plant Mol. Biol. 48:65-87.
Sowbhagya, C. M., Ramesh, B. S., and Vandenkoornhuyse, N., Libessart, N., Delrue, B.,
Bhattacharya, K. R. 1987. The relationship Zabawinski, C., Decq, A., Iglesias, A., Carton,
between cooked-rice texture and the physico- A., Preiss, J., and Ball, S. 1996. Control of
chemical characteristics of rice. J. Cereal Sci. starch composition and structure through sub-
5:287-297. strate supply in the monocellular alga Chlamy-
Sumner, J. B., and Somers, G. F. 1944. The domonas reinhardtii. J. Biol. Chem.
water-soluble polysaccharides of sweet corn. 271:16281-16287.
Arch. Biochem. 4:7-9. Vandewal, M., Dhulst, C., Vincken, J. P.,
Sun, C. X., Sathish, P., Ahlandsberg, S., and Buleon, A., Visser, R., and Ball, S. 1998.
Jansson, C. 1999. Analyses of isoamylase gene Amylose is synthesised in vitro by extention
activity in wild-type barley indicate its of and cleavage from amylopectin. J. Biol.
involvement in starch synthesis. Plant Mol. Chem. 273:22232-22240.
Biol. 40:431-443. Villareal, C. P., and Juliano, B. O. 1986. Waxy
Takeda, Y., and Hizukuri, S. 1987. Structures of gene factor and residual protein of rice starch
branched molecules of amyloses of various granules. Starch/Staerke 38:118-119.
origins, and molar fractions of branched and Visser, R. G. F., Ponstein, A. S., Jacobsen, E.,
unbranched molecules. Carbohydr. Res. Feenstra, W. J., and Van der Leij, F. R. 1991.
165:139-145. Sequence of the structural gene for granule-
Takeda, Y., Maruta, N., and Hizukuri, S. 1992. bound starch synthase of potato (Solanum
Structure of amylose subfractions with differ- tuberosum L.) and evidence for a single point
ent molecular sizes. Carbohydr. Res. 226:279- deletion in the amf allele. Mol. Gen. Genet.
285. 228:240-248.
Takeda, Y., Guan, H. P., and Preiss, J. 1993a. Waigh, T. A., Gidley, M. J., Komanshek, B. U.,
Branching of amylose by the branching isoen- and Donald, A. M. 2000. The phase transfor-
zymes of maize endosperm. Carbohydr. Res. mations in starch during gelatinisation: A liq-
240:253-263. uid crystalline approach. Carbohydr. Res.
Takeda, Y., Tomooka, S., and Hizukuri, S. 328:165-176.
1993b. Structures of branched and linear Wang, Z. Y., Zheng, F. Q., Shen, G. Z., Gao, J.
molecules of rice amylase. Carbohydr. Res. P., Snustad, D. P., Li, M. G., Zhang, J. L., and
246:267-272. Hong, M. M. 1995. The amylose content in
Takeda, Y., Shibahara, S., and Hanashiro, I. rice endosperm is related to post-transcrip-
2003. Examination of the structure of amy- tional regulation of the waxy gene. Plant J.
lopectin molecules by fluorescent labeling. 7:613-622.
Carbohydr. Res. 338:471-475. Watanabe, T., and French, D. 1980, Structural
Tatge, H., Marshall, J., Martin, C., Edwards, E. features of Naegeli amylodextrin as indicated
A., and Smith, A. M. 1999. Evidence that by enzymic degradation. Carbohydr. Res.
amylose synthesis occurs within the matrix of 84:115-123.
the starch granule in potato tubers. Plant Cell Watanabe, Y., Nakamura, Y., and Ishii, R. 1997.
Environ. 22:543-550. Relationship between starch accumulation and
Tetlow, I. J., Blissett, K. J., and Emes, M. J. activities of the related enzymes in the leaf
1998. Metabolite pools during starch synthesis sheath as a temporary sink organ in rice
and carbohydrate oxidation in amyloplasts (Oryza sativa). Aust. J. Plant Physiol. 24:563-
isolated from wheat endosperm. Planta 569.
204(1):100-108. Whistler, R. L., and BeMiller, J. N. 1997. Carbo-
Tsai, M. L., Li, C. F., and Lii, C. Y. 1997. Effects hydrate Chemistry for Food Scientists. Am.
of granular structures on the pasting behaviors Assoc. Cereal Chem., St. Paul, MN.
of starches. Cereal Chem. 74:750-757. Wong, K. S., Kubo, A., Jane, J. L., Harada,
Tsai, M. L., and Lii, C. Y., 2000. Effect of hot-water- K., Satoh, H., and Nakamura, Y. 2003.
soluble components on the rheological properties Structures and properties of amylopectin
of rice starch. Starch/Staerke 52:44-53. and phytoglycogen in the endosperm of
Umemoto, T., and Terashima, K. 2002. Activity sugary-1 mutants of rice. J. Cereal Sci.
of granule-bound starch synthase is an impor- 37:139-149.
tant determinant of amylose content in rice Wu, C. Y., Colleoni, C., Myers, A. M., and
endosperm. Funct. Plant Biol. 29:1121-1124. James, M. G. 2002. Enzymatic properties and
Umemoto, T., Yano, M., Satoh, H., Shomura, A., regulation of ZPU1, the maize pullulanase-
and Nakamura, Y. 2002. Mapping of a gene type starch debranching enzyme. Arch. Bio-
responsible for the difference in amylopectin chem. Biophys. 406:21-32.
structure between japonica-type and indica- Yamamori, M., Fujita, S., Hayakawa, K.,
type rice varieties. Theor. Appl. Genet. Matsuki, J., and Yasui, T. 2000. Genetic
104(1):1-8. elimination of a starch granule protein, SGP-1,
Starch / 141
of wheat generates an altered starch with Yoshimoto, Y., Tashiro, J., Takenouchi, T., and
apparent high amylase. Theor. Appl. Genet. Takeda, Y. 2000. Molecular structure and
101:21-29. some physicochemical properties of high-
Yamanouchi, H., and Nakamura, Y. 1992. Organ amylose barley starches. Cereal Chem.
specificity of isoforms of starch branching 77:279-285.
enzyme (Q-enzyme) in rice. Plant Cell Yun, S. H., and Matheson, N. K. 1993. Structures
Physiol. 33:985-991. of the amylopectins of waxy, normal, amy-
Yao, Y., and Ding, X. L. 2002. Pulsed nuclear lose-extender, and wx-ae genotypes and of the
magnetic resonance (PNMR) study of rice phytoglycogen of maize. Carbohydr. Res.
starch retrogradation. Cereal Chem. 7:751- 243:307-321.
756. Zeeman, S. C., Umemoto, T., Lue, W. L.,
Yao, Y., Zhang, J. M., and Ding, X. L. 2002. Auyeung, P., Martin, C., Smith, A. M., and
Structure-retrogradation relationship of rice Chen, J. 1998. A mutant of Arabidopsis lack-
starch in purified starches and cooked rice ing a chloroplastic isoamylase accumulates
grains: A statistical investigation. J. Agric. both starch and phytoglycogen. Plant Cell
Food Chem. 50:7420-7425. 10:1699-1711.
CHAPTER 6
RICE PROTEINS
Frederick F. Shih
Rice protein is an important source of nutrition and energy for 50% of the
world’s population, for whom rice has long been a staple diet. The protein content
of rice, at approximately 7%, is relatively low compared with that of other cereal
grains. However, because of the huge quantity of rice produced worldwide
(approximately 400 million metric tons annually), the amount of rice protein poten-
tially available is considerable. On the other hand, rice protein has a significant
influence on the structural, functional, and nutritional properties of rice. It is a major
factor in determining the texture (e.g., stickiness), pasting capacity, and sensory
characteristics of rice. In recent years, rice protein has been recognized to be
uniquely nutritious and hypoallergenic, which makes rice increasingly popular for
use in foods all over the world.
Extensive research has been conducted on rice proteins because scientists recog-
nize the importance of protein for the understanding and utilization of rice. How-
ever, only limited efforts have been made to keep up with and summarize the
information on rice proteins in the literature. Earlier reviews on the subject include
those of Houston (1972), Lasztity (1984), and Hamaker (1994). Of particular sig-
nificance was the review by Juliano (1985) in the second edition of this book, a
revision of his work (Juliano, 1972) in the first edition. A lot more research has
been done since, particularly on the characterization of rice proteins, processing of
rice protein products, and development of better-quality rice proteins. This chapter
is an update of information in the literature on the chemistry and technology of rice
proteins. It is an overview covering materials from sources old and new, but the
emphasis is on studies reported since Juliano’s review (Juliano, 1985).
PROTEIN DISTRIBUTION AND COMPOSITION
As in other cereal grains, the protein content of the embryo and aleurone layer in
rice is as much as 20 percentage points higher than that of the endosperm. Never-
theless, the greatest part of the total proteins is located in the endosperm. Protein
content usually is calculated from Kjeldahl nitrogen multiplied by the factor 5.95.
This factor is based on the nitrogen content (16.8%) of the major rice protein,
glutelin. Approximate protein contents of rough rice and its milling fractions are listed
143
TABLE 1
Proximate Protein Content of Rough Rice
and Its Milling Fractionsa
Crude Protein
Rice Fraction (g of N × 5.95)
Rough rice 5.6–7.7
Brown rice 7.1–8.3
Milled rice 6.3–7.1
Rice bran 11.3–14.9
Rice hull 2.0–2.8
a Source: Juliano (1993); reprinted with permission.
in Table 1. Values reported in the literature vary widely, depending on the variety,
climate, and other agrotechnical conditions (Juliano, 1985). Invariably, milled rice
contains lower quantities of protein than rough rice or brown rice because, during
milling, a part of the protein-rich outer layers (aleurone cells) is removed. For the
same reason, the protein content of the co-products of milling, such as the bran and the
polish of the outer layers, normally is higher than that of the milled rice.
Protein Bodies
Up to 95% of the endosperm rice protein is in the form of discrete particles

called protein bodies (PBs). They range in size from 1 to 4 µm, and mostly are con-
centrated in the peripheral-lateral and peripheral-dorsal cells (Fig. 1). There are at
least two types of PBs in the rice endosperm, designated PB-I and PB-II. As
observed under an electron microscope, PB-I displays a spherical shape and PB-II
exhibits an irregular crystalline morphology (Bechel and Juliano, 1980; Tanaka et
al, 1980; Ogawa et al, 1987, 1989). PB-I is highly enriched with prolamins and con-
stitutes approximately 20% of milled rice protein. PB-II contains predominantly
glutelins and constitutes 60–65% of milled rice protein. However, Barber et al
(1998a), using polyclonal antibodies to purify protein fractions in immunocyto-
chemical analyses, reported that the prolamins were not exclusively confined to PB-
I or glutelins to PB-II. In studies on fecal particles of monogastric animals fed with
rice proteins, the origin of the fecal protein particles remained unclear (Barber et al,
1998b). Intact PB-I and the fecal protein particle showed no significant difference
in sizes, indicating that they might be the same protein particles. However, the
residual particles contained immunocytochemically identified unique proteins that
were also found in both PB-I and PB-II, raising questions about linking fecal protein
particles exclusively with PB-I.
Protein Fractions
Osborne’s classification of proteins based on solubility has been widely used in

plant protein chemistry (Osborne, 1924). Essentially, proteins are separated into
albumin (water-soluble), globulin (salt-soluble), prolamin (alcohol-soluble), and
glutelin (alkaline-soluble). New approaches and techniques have shown that the
distinction among the four Osborne fractions of plant proteins is not always straight-
forward and that the fractions are actually composed of mixtures of proteins
(Juliano, 1985; Wilson, 1987). Nevertheless, the Osborne classification has pro-
Proteins / 145
Fig. 1. Schematic diagram of various protein bodies and the compound starch granule in the
endosperm subaleurone layer. (Reprinted, with permission, from Coffman and Juliano, 1987)
vided pertinent solubility information and a solid basis for the studies of plant pro-
teins. Typically, in traditional Osborne sequential extractions, ratios of albumin,
globulin, prolamin, and glutelin have been reported to be in the range of 3.0–18.7:
0–17:1.6–20.6:55.0–88.1 for brown rice and 0.9–9.9:1.4–19.9:0.4–10.3:61.8–91 for
milled rice (Lasztity, 1984). These ratios and the range for each fraction vary
widely, depending on the variety of rice and the extraction conditions. For instance,
Huebner et al (1990) reported that, while the mean ratio of the salt-soluble (albumin
plus globulin) prolamin to glutelin for 33 short, medium, long, and extra-long rice
cultivars was 7:9:84, the salt-soluble fraction varied twofold (from 4 to 10%) and
the prolamin content varied by more than threefold (from 5 to 18%). The ratios also
may change depending on the maturity of the grain. Lower molecular weight
subunits were found to have increased and higher molecular weight subunits
decreased at 14, 30, and 15 days after flowering; the effect was particularly
pronounced for the glutelin subunit, resulting in a significantly lower average for
this protein fraction (Chrastil and Zarins, 1994).
The approximate distribution of albumin in brown rice is bran, 31% (including
germ, 13%); polish, 9%; and milled rice, 40% (Juliano, 1972). The corresponding
distribution for globulin is bran, 40% (including germ, 10%); polish, 4%; and milled
rice, 56%. The prolamin distribution is bran, 21% (including germ, 10%); polish, 4%;
and milled rice, 75%. Approximately 5% of the glutelin is in the bran (principally in
the germ), 2% in polish, and 93% in milled rice. The proportion of albumin in protein
is highest in the outer layers of milled rice and decreases toward the center, whereas
the proportion of glutelin has an inverse distribution (Houston et al, 1968).
A summary of the amino acid composition for milled rice protein fractions is
shown in Table 2. Albumin has the highest lysine content, followed by glutelin, then
globulin and prolamin. Globulin is richest in the sulfur amino acids cysteine and
methionine; prolamin is the poorest. When brown rice protein content increases
from planting or climatic conditions, it is due mainly to the increase in glutelin
content (Cagampang et al, 1966; Souza et al, 1999). The slight increase in low-
lysine prolamin may explain the decrease in the lysine content of protein as protein
content increases.
TABLE 2
Amino Acid Composition (g/16.8 g of N) of Milled Rice and Its Protein Solubility Fractionsa
Amino acid Whole Albumin Globulin Prolamin Glutelin
Alanine 5.9–6.2 7.1–8.5 5.6–6.3 6.7–7.6 5.6–5.9
Arginine 8.5–9.3 7.9–10.0 7.2–13.9 6.1–6.9 9.0–10.8
Aspartic acid 9.8–10.2 10.2–11.2 7.1–13.5 8.3–8.7 10.2–11.2
Cystine 1.5–1.9 1.9–2.3 3.3–4.0 0–0.8 1.2–1.8
Glutamic acid 19.1–20.8 12.7–17.8 17.4–18.8 22.2–33.4 19.2–20.8
Glycine 4.6–5.4 6.3–8.4 1.7–2.7 3.0–3.7 4.3–5.3
Histidine 2.4–2.8 2.9–3.4 2.4–4.1 1.3–2.1 2.6–2.7
Isoleucine 4.2–4.9 3.5–3.8 6.6–6.8 4.6–5.2 4.3–4.7
Leucine 6.7–8.9 6.6–8.0 6.6–6.8 12.8–146 7.3–9.3
Lysine 3.5–4.1 5.1–6.4 1.9–3.7 0.3–1.2 2.7–4.3
Methionine 2.5–3.3 1.9–2.1 3.0–5.4 0.5–0.9 2.0–3.1
Phenylalanine 5.3–5.8 3.7–4.6 3.3–4.8 5.8–6.7 5.4–6.0
Proline 4.6–5.0 4.5–7.1 3.8–7.5 5.0–6.7 4.9–6.2
Serine 5.8–6.1 4.2–5.4 5.5–6.5 4.2–6.1 4.5–6.2
Threonine 3.4–3.9 4.2–5.2 2.5–2.7 2.5–2.8 2.8–5.1
Tryptophan 1.3–1.8 1.5–1.8 1.4–1.5 0.5–2.6 1.0–1.6
Tyrosine 4.9–5.8 4.4–5.1 5.5–6.3 9.2–9.9 5.3–5.5
Valine 6.1–6.3 5.9–7.8 5.4–6.5 6.5–7.1 6.3–6.9
Ammonia 2.2–2.5 1.6–1.7 1.6–1.8 3.3–3.5 2.2–2.4
a Source: Juliano (1985).
ALBUMIN
Separation of the water-soluble albumin and salt-soluble globulin cannot be
achieved by simple sequential extraction because of minerals present in the rice
grain that dissolve in the water solvent during the extraction (Villareal and Juliano,
1981). Albumins also readily denature from solution within a few days, even at
0–4°C. Traditionally, albumins are obtained by precipitating off the globulins from
0.7–0.9M NaCl extracts of milled rice, either by dialysis against water or by addi-
tion of ammonium sulfate to a 1.3M concentration (30% saturation) (Houston and
Mohammad, 1970). Gel filtration of the 0.1–1.8M ammonium salt cut showed
major albumins with molecular mass of 18–20 kDa. Analytical disc gel elec-
trophoresis at pH 8.3 and 4.5 showed two bands for basic albumins at pH 4.5 and no
migration at pH 8.3. One slow-migrating band for peak 2 (albumin I), one faster-
migrating band for peak 4 (albumin II), and two bands for peak 3 (a mixture of
albumin I and albumin II) at pH 8.3 also were found. Albumins in the 0.1–1.8M
ammonium sulfate cut had a lower lysine content, but both ammonium sulfate frac-
tions had a high cystine content, which could explain their sensitivity to denatura-
tion during storage of milled rice (Cagampang et al, 1976) or their aqueous extrac-
tions (Villareal and Juliano, 1981).
GLOBULIN
A major globulin was obtained by ammonium sulfate precipitation from a 2.5–5%
NaCl solution of milled rice (Houston and Mohammad, 1970). It was classified as a-
globulin with a molecular mass of 25.5 kDa by gel filtration. Ultracentrifugation
and gel filtration studies also showed the presence in the rice caryopsis of a-, b-, g-,
and d-globulins with increasing molecular mass (Morita and Yoshida, 1968).
a-Globulin was present in the low molecular mass gel filtration fraction of 0.5M
NaCl extracts of the embryo, bran, outer endosperm, and inner endosperm. The 15-
Proteins / 147
kDa fraction of inner endosperm corresponded to b-globulin and also to g-globulin

in rice embryo and bran. The void volume fraction (>200 kDa) corresponded to
d-globulin. Further chromatographic studies of purified g-globulin showed three
components, designated g1, g2, and g3 in order of increasing basicity and decreasing
migration rate on electrophoresis at pH 9.8 (Sawai and Morita, 1970).
PROLAMIN
Prolamin usually is extracted with 70% ethanol after the removal of albumin and
globulin from rice flour. However, the prolamin thus obtained often was contami-
nated with lipids, phenols, and carbohydrates (Mandac and Juliano, 1978). Repeated
precipitation from the 70% ethanol extract by adding three volumes of acetone was
required to purify the protein. Generally, rice prolamin is low in lysine and sulfur-
containing amino acids (Table 2). Prolamin is located in PB-I, and consists of three
polypeptide subunits with molecular masses of 10, 13, and 16 kDa (Ogawa et al,
1987). Hibino et al (1989) analyzed the amino acid composition of these subunits
and reported that while, as expected, sulfur-containing amino acids were low in the
dominating 13-kDa prolamin, they were unusually high in the 10- and 16-kDa
subunits. The high methionine content of the 10-kDa subunit, at up to 22.5 g/16 g of
N, was particularly noteworthy (Table 3).
GLUTELIN
Glutelin is the major storage protein in rice. Studies on native glutelin are diffi-
cult, because native glutelin is extremely insoluble, has high molecular mass, and is
heterogeneous. Glutelin is soluble below pH 3 and above pH 10, but it is not very
soluble in 3M urea or aluminum lactate buffer or in 0.1N acetic acid (Cagampang et
al, 1966; Palmiano et al, 1968; Tecson et al, 1971). The traditional alkaline extrac-
tion could cause glutelin aggregation and the degradation of sulfur amino acids.
Typical effort to fractionate glutelin by successive extraction of milled rice with
water, NaCl in phosphate buffer, and 70% ethanol with acetic acid extracted less
than 23% of the glutelins (Wieser et al, 1980).
TABLE 3
Aminogram (g/16 g of N) of Subunits of Rice Glutelin and Prolamina
Glutelin Subunitsb
30–39 kDa 19–25 kDa Prolamin Subunits
Amino Acid (acidic) (basic) 13 kDa 10 kDa 16 kDa
Histidine 2.2–2.5 2.6–2.7 2.0–2.4 1.7 4.2
Isoleucine 3.2–3.3 4.1–4.9 3.8–5.4 1.6 3.6
Leucine 6.4–7.5 7.0–8.5 17.9–26.4 4.7 8.1
Lysine 2.2–3.0 3.0–4.1 0.4–5.5 1.0 3.3
Methoinine + cystinec 0.2–1.9 0.1–2.4 0.7–1.2 22.5 5.3
Phenylalanine + tyrosine 10.0–10.5 10.1–10.8 12.7–21.6 4.3 7.6
Threonine 2.8–3.7 2.5–3.7 1.8–2.8 6.8 2.7
Valine 5.1–5.7 5.7–7.0 2.7–3.9 4.4 3.9
Amino acid scored (%) 38–52 52–71 7–8 18 57
a Source: Juliano (1993); reprinted with permission.
b S-Cyanoethyl glutelin subunits.
c Only fraction 3 from isoelectric focusing of the 13-, 10- and 16-kDa prolamin subunits had cystine.
All glutelins had substituted cystine residues.

d Based on 5.8% lysine as 100% (WHO, 1985).
Present evidence indicates that rice glutelin consists of three acidic or a subunits
of 30–39 kDa and two basic or b subunits of 19–25 kDa (Kagawa et al, 1988).
These two groups of polypeptides, formed by the cleavage of a 57-kDa polypeptide
precursor (Sarker et al, 1986), are believed to cross-link by disulfide bonding,
resulting in glutelin molecules with molecular masses ranging from 64 to 500 kDa
(Sugimoto et al, 1986).
The essential amino acid contents of the glutelin and prolamin subunits are
shown in Table 3. Glutelin subunits have an even distribution of lysine and the
sulfur-containing methionine and cystine, whereas prolamin subunits have a much
wider distribution. Overall, lysine is limiting in these polypeptides except in fraction
3 from isoelectric focusing of the 13-kDa prolamin subunit, which has 5.5% lysine
and is limiting in methionine plus cysteine. Thus, glutelin has a better amino acid
score than prolamin except for the 16-kDa prolamin subunit. Also, the 10-kDa
prolamin subunit has a exceptionally high (6.8%) cysteine content.
Identification of Rice Varieties
Rice varietal identification is important for breeding new varieties and for mar-
keting varieties for specific uses. Investigators have characterized rice proteins for
that purpose by gel electrophoresis (Mandac and Juliano, 1978; Villareal and
Juliano, 1978; Chauhan and Manda, 1984; Kusama et al, 1984; Sengupta and
Chattopadhyay, 2000), two-dimensional electrophoresis and electrofocusing (Du
Cros et al, 1979; Guo et al, 1986), gel filtration chromatography (Iwasaki et al,
1982), and reversed-phase high-performance liquid chromatography (RP-HPLC)
(Lookhart et al, 1987; Hussain et al, 1989; Huebner et al, 1990). Huebner et al
(1990) reported that, using improved extraction and RP-HPLC conditions, glutelin
as the major rice protein was most useful in characterizing rice varieties, rather than
prolamin, as in traditional analysis for cereal proteins.
With advances in gene technology, DNA fingerprinting studies have provided
valuable information on the genetic diversity in rice (Harp and Correll, 1998; Sonti,
1998; Don et al, 1999). DNA fingerprints were generated, utilizing rice-derived
minisatellite sequences, for use as genetic markers mapped to a rice recombinant
inbred line population (Zhou et al, 1997; Gustafson and Yano, 2000). The technique
provides molecular probes to detect genetic variation in rice cultivars.
Protein Mutants
Rice protein is unique among the cereal proteins in being richest in glutelin and
lowest in prolamin (Table 4). Lysine is the limiting essential amino acid of rice
protein; therefore, efforts to improve the protein quality of rice include increasing
the level of the lysine-rich glutelin (or the glutelin-rich PB-II) at the expense of the
lysine-poor prolamin (or the prolamin-rich PB-I) (Kumamaru et al, 1988, 1997;
Schaeffer and Sharpe, 1997). Some mutants that met these criteria were identified,
and their protein bodies were isolated and characterized (Ogawa et al, 1989). Also
for that purpose, starch mutants have been developed and transferred to IR36 rice
(Juliano et al, 1990). The amylose extender IR36 mutants had higher lysine than
IR36 by 0.8% in brown rice and 0.5% in milled rice. Efforts have been made to
investigate hybrids of rice and sorghum and other cereal grains. Because rice pro-
tein is richer in lysine (3.5–4%) than sorghum (1–2%), amino acid composition for
Proteins / 149
TABLE 4
Osborne Protein Fractions and Lysine Content of Cereal Proteinsa
Osborn Protein Fractions (% of total) Lysine
Cereal Albumin Globulin Prolamin Glutelin (g/16 g of N)
Wheat 5–15 5–10 40–50 30–45 2.3
Rice 2–5 2–10 1–5 75–90 3.8
Corn 2–10 2–10 50–55 30–45 2.5
Barley 3–10 10–20 35–50 25–45 3.2
Oat 5–10 50–60 10–16 5–20 4.0
Sorghum 5–10 5–10 55–77 30–40 2.7
Millet 5–10 0–5 50–60 30–40 2.7
Rye 20–30 5–10 20–30 30–40 3.7
rice/sorghum and rice/wheat hybrids were of particular interest. The lysine content
of four milled rice/sorghum hybrids of 3.1–3.6 g/1.68 g of N was found to be closer
to that of rice than to sorghum (IRRI, 1980). One milled rice/wheat hybrid had 4.1 g
of lysine per 16.8 g of N at 10.8% protein, with a sodium dodecyl sulfate-
polyacrylamide gel electrophoresis profile similar to that of milled rice glutelin
(IRRI, 1983).
PROCESSING OF PROTEIN PRODUCTS
High-protein rice products have become commercially available in recent years

in countries such as the United States. The reasons for the earlier lack of interest in
these products are twofold; namely, the low content of protein in rice and the lack
of knowledge of rice proteins by consumers. The industry used to process broken
rice kernels or rice flour for its starch and consider the residual protein a waste co-
product. On the other hand, rice bran from the milling of rice, in spite of being rela-
tively rich in protein, has mostly been ignored as a valuable resource because it
contains little of the starch. Only in recent years, when plant proteins in general and
rice proteins in particular were recognized as being equal to or healthier than animal
proteins for human consumption, did the processing of rice proteins from sources
such as rice bran and rice flour begin to get attention from industry and rice
research facilities.
Bran Proteins
Rice bran is a particularly attractive source of protein because it is protein-rich,

plentiful, and low-cost. Rice bran makes up approximately 10–15% of rough rice,
with an annual yield of approximately 50 million metric tons worldwide as a co-
product from the milling of rice. It contains 12–15% protein but is used primarily as
animal feed. Efforts to add value to this co-product have not been successful
because of difficulties in the processing of the protein component. Rice proteins are
bonded strongly by disulfide linkages, forming high molecular mass complexes,
which are extremely insoluble in water (Cagampang et al, 1966; Sawai and Morita,
1968; Tecson et al, 1971; Sugimoto at al, 1986). Furthermore, because of the pres-
ence of lipid, which is easily degraded by lipase and causes rancidity in rice, rice
bran must be stabilized, for instance, by heat or acid treatments to deactivate the
lipase or by defatting to remove the lipid (Sayre et al, 1982; Prabhakar, 1987).
These treatments can cause protein denaturation or aggregation, which makes the
processing of rice bran proteins even harder to achieve.
CHEMICAL METHODS
High-protein rice products with protein contents from 20 to 80% can be pro-
duced by alkaline extraction of the rice bran followed by precipitation at the
isoelectric pH of approximately 4.5 (Connor at al, 1976; Betschart et al, 1977; Bera
and Mukherjee, 1989; Prakash and Ramanatham, 1994; Gnanasambandam and
Hettiarachchy, 1995). For example, commercially available unstabilized and heat-
stabilized rice bran were extracted in a slurry at pH 9.5 for 30 min, resulting in
protein concentrates with protein contents of 71.5 and 50.9%, respectively
(Gnanasambandam and Hettiarachchy, 1995). Invariably, heat stabilization not only
impaired the extractability of the protein in bran but also changed the amino acid
composition and electrophoretic profile of the extracted protein products (Prakash
and Ramanatham, 1994; Gnanasambandam and Hettiarachchy, 1995). Generally,
alkaline conditions are effective in solubilizing and extracting protein from rice
bran, but alkaline extraction has its shortcomings. High-pH alkaline conditions can
lead to undesirable protein modifications, including molecular cross-linking and
rearrangements, resulting in the formation of toxic compounds such as lysinoalanine
(Cheftel et al, 1985; Otterburn, 1989). Molecular denaturation and degradation often
occur, which reduce the food-use functional property and nutritional quality of the
protein (Prakash and Ramanatham, 1994, 1995a,b).
ENZYMATIC METHODS
For food-use purposes, high-protein products from rice bran are often and pref-
erably produced by more specific enzymatic methods. Hamada (1999) reported a
procedure for the preparation of a rice protein isolate using alkaline protease. How-
ever, at low degrees of hydrolysis (<2%), when protein degradation was insignifi-
cant, the improvement in protein extractability remained low. On the other hand, at
high degrees of hydrolysis, when >90% protein was extractable, the functional and
nutritional properties of the extracted protein may be adversely affected. More gen-
eral and effective processes involve the use of carbohydrate-hydrolyzing enzymes
such as cellulase, pectinase, hemicellulase, and xylanase to remove the cell wall
tissues and thus enhance the extractability of the entrapped protein components
(Ansharullah and Chesterman, 1997; Shih et al, 1999; Wang et al, 1999). In a par-
ticularly successful extraction (Fig. 2), rice bran protein isolate containing approxi-
mately 92% protein was prepared from unstabilized and defatted rice bran using
phytase and xylanase (Wang et al, 1999). However, before the recovery of the pro-
tein, the enzymes were deactivated by adjusting the reaction mixture to pH 10.
Therefore, in addition to the enzyme function, the treatment under alkaline condi-
tions also may contribute to the enhancement of protein extractability.
Endosperm Proteins
Rice flour from milled rice or broken kernels contains mostly endosperm storage
proteins, approximately 7% by weight and readily available for extraction. As dis-
cussed earlier, alkaline solvent is effective in facilitating the extraction of proteins
from rice flour (Cagampang et al, 1966; Tecson et al, 1971). Physical methods also
Proteins / 151
have been reported for the protein separation. When a slurry of rice flour was emul-
sified in the presence of sodium stearoyl lactylate, the emulsion that was formed
separated into two-layer fractions on centrifugation (Kung et al, 1987). The upper
fraction contained 48.5% protein, and the lower fraction contained only 1.6% pro-
tein.
The major component in rice flour is starch; therefore, starch-hydrolyzing
enzymes such as a-amylase, glucoamylase, and pullulanase often are used to sepa-
rate the protein in rice flour by solubilizing the starch. Depending on the processing
conditions in the production of the rice flour and the conditions of the subsequent
enzymatic treatment, the protein content of the resulting products ranges from
slightly enriched (25% protein) to highly enriched (>90% protein) (Resurreccion et
al, 1978; Hansen et al, 1981; Chen and Chang, 1984; Griffin and Brooks, 1989;
Morita and Kiriyama, 1993; Shih et al, 1999). Typically, Morita and Kiriyama
(1993) achieved the processing of protein isolate by treating rice flour with heat-
stable a-amylase at 97°C for 2 hr. As starting material, the flour was obtained at
70% milling. The product, after washing off the solubles with boiling water, was
recovered with a protein content of >90%.
Because milled rice and rice flour are sold at premium prices and contain only a
small amount of protein, it is possible, but not practical, to produce rice isolate with
>90% protein from milled rice or regular rice flour. An alternative is to use, as
starting material, protein-rich co-products from the processing of rice ingredients,
such as in syrup manufacturing. After removing the hydrolyzed starch as syrup, the
insoluble residue contains up to 50% protein and is a low-cost industrial co-product
Fig. 2. Procedure for the preparation of protein isolate from unstabilized defatted rice bran. (Reprinted,
with permission, from Wang et al, 1999; ©American Chemical Society)
(Chen and Chang, 1984; Shaw and Sheu, 1992). Shih and Daigle (2000) reported
that treatment of this co-product with a-amylase and glucoamylase resulted in a
product containing 85% protein. Follow-up treatment with a mixture of cellulase
and xylanase raised the protein content to 92%. Inorganic impurities, particularly
the metal manganese that was present at an unusually high level of 47 mg/kg in the
starting rice flour, also were removed from the protein product. The removal of
manganese is desirable because the consumption of high levels of manganese is
known to cause liver problems and nerve system disorders in humans (Butterworth
et al, 1995; Fell et al, 1996).
Functional Properties
Compared with other plant proteins, rice protein has relatively poor food-use
functional properties (Barber and Barber, 1974). Rice proteins are extremely
insoluble because of the intermolecular disulfide linkages and high molecular
weights of the major protein glutelin (Cagampang et al, 1966; Sawai and Morita,
1968; Tecson et al, 1971). The solubility and, with it, other food-use functional
properties are further reduced during processing, particularly by modifications such
as heat treatment for processing or stabilization (Knorr, 1982; Knorr and Betschart,
1983; Prakash and Ramanatham, 1995a; Devi et al, 1997). Functional properties
also can be influenced by the drying technique employed. Freeze-dried protein
concentrates from both heat-stabilized and parboiled rice bran had low water-
absorption capacity and high fat-absorption capacity. Cabinet-dried samples did not
exhibit any foaming or emulsification capacity, indicating significant molecular
denaturation (Prakash and Ramanatham, 1995a).
The protein in rice flour, though only about 7%, plays a significant role in deter-
mining the functional properties of the starch, which makes up approximately 80%
of the rice kernel (Cheng, 1987; Marshall et al, 1990). The removal of protein by
the treatment of buffer or protease (Fig. 3) tends to bring about a decrease in the
gelatinization temperature of the flour (Marshall et al, 1990), indicating that rice
protein has an inhibitory effect on the swelling of rice starch granules. Measuring
the interaction of protein and starch in a model system, Chrastil (1990) found
stickiness of whole cooked rice to be positively correlated with the binding ratios of
the rice protein oryzenin to starch, amylose, and/or amylopectin. Cooked rice
stickiness was reported to increase with the disruption of the disulfide linkages in
the protein component (Hamaker and Griffin, 1990, 1993; Hamaker et al, 1991).
Proteins from the outer layers of the rice kernel, tightly bound to the starch, were
found to be responsible for reducing the pasting and crystallizing capacities of the
starch (Yang and Chang, 1999). Protein content also has been reported to correlate
negatively with the expansion ratio for gun-puffed milled rice or oil-puffed par-
boiled rice (Villareal and Juliano, 1987). It is negatively correlated with the peak
viscosity and positively correlated with the pasting temperature of the isolated rice
starch (Lim et al, 1999).
Utilization
In spite of their limited food-use functional properties, rice proteins have been
successfully utilized in food. For years, with rice proteins as a key ingredient, rice
flour or rice bran has been incorporated into foodstuffs such as bread, beverages,
Proteins / 153
Fig. 3. Differential scanning calorimetry thermal curves of different whole grain milled rice varieties:
untreated (curve 1), buffer-treated (curve 2), and Pronase-treated (curve 3). The water content was 70%
(w/w), and the heating rate of the calorimeter was 1.0°C/min. (Reprinted, with permission, from
Marshall et al, 1990)
pasta, and confections (Lynn, 1969; Carroll, 1990; Saunders, 1990). Recently, as
high-protein rice products have become more available, the use of rice proteins in
food has surged. They have been used in infant foods (Hanson et al, 1981),
breakfast cereal (Bakar and Hin, 1985), snack foods (Capansana et al, 1984), and
edible films (Shih, 1996). To meet the needs of hypoallergenic protein products, a
rice product developed by enzymatically removing the allergic factor in the protein
component has been produced and made commercially available in Japan
(Watanabe, 1993).
NUTRITIONAL PROPERTIES
Percent protein in rice generally is calculated based on the Kjeldahl conversion

factor of 5.95. However, in nutritional studies, the factor 6.25 is used to maintain a
common nitrogen-based calculation for all proteins. Rice protein is nutritionally
important to many people. In tropical Asia, in particular, rice is the principal source
of the dietary protein (35–40%) and energy (50%) of the population (Juliano, 1993).
Protein-intake studies showed that intakes of 0.94–1.23 g/kg per day, at energy lev-
els of 37–63 kcal/kg per day, were adequate for Chilean, Chinese, Filipino, Korean,
and Thai subjects (Intengan et al, 1984; Rand et al, 1984). Based on the safe level of
protein requirement for milk of 0.89 g/kg of body weight (Huang et al, 1980), the
high-protein rice (IR58) had 62% of the protein quality of milk. In general, rice
protein, like most cereal proteins, is deficient in the essential amino acid lysine, but
it has an excess of the essential sulfur-containing amino acids cysteine and me-
thionine. By contrast, legume proteins, such as soy protein, are deficient in sulfur
amino acids but have excess lysine. Therefore, a diet of rice and legumes provides a
better amino acid balance or score than either a rice or a legume diet alone (Rand et
al, 1984; Eggum et al, 1987).
Effect of Processing
MILLING
Milling reduces protein in the rice and lysine in the protein (Resurreccion et al,
1979; Villareal et al, 1991), resulting in a loss in nutrient content for the milled rice.
Brown rice is also more nutritious by having a higher vitamin and mineral content,
but it is affected negatively by having the higher fiber, phytate, and antinutrition
factors of rice bran. The mean nutritional properties of various raw and cooked,
freeze-dried, milled rices are shown in Table 5. The true digestibility (TD) of
cooked rice protein in humans is approximately 88% and, based on 5.8% lysine as
100%, the amino acid score is approximately 65% (WHO, 1985). In general, the
values of net protein utilization (NPU) decreased and those of biological value (BV)
increased during cooking. Milling also removes pigment and phytate in pigmented
rice; black- and red-pigmented brown rice had lower TD and higher BV than non-
pigmented brown rices, but only red rice had NPU comparable to that of
nonpigmented rices in growing rats (Eggum et al, 1981). Milling removes these dif-
ferences in protein TD, BV, and NPU.
HEATING
Most thermal processes cause decomposition of lysine or cysteine to various
extents. Extrusion cooking of milled rice batter at 15% moisture and 120–150°C
reduced the total content of lysine 11–13% and cysteine 14–29% but had no effect
on tryptophan or cystine (Eggum et al, 1986). Gun-puffing of milled rice lowered
the cysteine content from 3.1 to 1.6 g/16.8 g of N without significant lysine degra-
dation (Villareal and Juliano, 1987); however, lysine decomposition occurred in
commercial oven-puffed rice, most likely due to the toasting step in the puffing
process (Khan and Eggum, 1979; Chopra and Hira, 1986). Yellow or stack-burned
rice, which occurs when wet grain, particularly unthreshed grain, is piled without
TABLE 5
Mean Nutritional Properties of Various Raw and Processeda Ricesb
Balance Data in Five Growing Rats
Lysine Cysteine
Crude True Biological Net Protein Digest- Digest-
Protein Lysine Digestibility Value Utilization ibility ibility
(% N × (g/16 g (% of N (% of (% of (% of (% of
Rice Type 6.25) of N) intake) digested N) intake) intake) intake)
Rawc 8.9 3.6 99.7 67.7 96.8 99.9 99.5
Processedc 9.0 3.5 88.6 78.2 95.4 99.4 82.0
Rawd 11.8 3.5 99.1 68.8 97.0 … …
Processedd 12.7 3.5 85.8 73.7 92.5 … …
a Cooked and freeze-dried after being milled.
b Source: Juliano (1993); reprinted with permission.
c Low-protein rices (IR29, IR32, IR450-5-9).
d High-protein rice (IR58).
Proteins / 155
provision for aeration, has a lower lysine content and lower NPU in rats than nor-
mal rice (Eggum et al, 1984). Boiling for 30 min reduces the TD of rice protein by
10–15% but has no effect on other cereal proteins (Eggum et al, 1977). Parboiling
also reduces the TD of rice protein, but both parboiling and boiling improve BV
without any adverse effect on NPU (Eggum et al, 1984).
Digestibility
The undigested protein, which passes out of the alimentary system as fecal protein
particles, is believed to be the lipid-rich core containing spherical protein bodies (PB-
I) (Tanaka et al, 1978). The core protein is low in lysine because lysine is readily
digested and its digestibility is unaffected by cooking. It is rich in sulfur-containing
cysteine because cysteine has the lowest digestibility among amino acids in rice
protein. Prolamin subunits of 10 and 16 kDa are rich in sulfur-containing amino acids;
therefore, mutants with reduced or increased levels of prolamin and these prolamin
subunits are expected to affect the digestibility of the protein in cooked rice. Tanaka
and Ogawa (1988) found greater amounts of large spherical protein bodies (PB-I) in
indica rice (30%) than in japonica rice (20%) and suggested that the protein of cooked
indica rice may be less digestible than that of cooked japonica rice. However, Tanaka
et al (1978) reported similar in vitro digestibilities (10–15% of total protein) for
protein bodies from japonica and indica rices. Also, with the development of a low-
glutelin (high-prolamin) mutant, Iida et al (1993) predicted that the mutant would have
only 72% of the TD of the parent cultivars, but a comparable mutant was found to
have a TD value similar to that of the parent (Eggum et al, 1994; Eggum and Juliano,
1997). In other studies, a comparison of the protein quality of cooked, IR36-based,
mutant and IR36 milled rices confirmed the higher lysine (by 0.5%) and better protein
quality of the mutant, although the NPU was comparable (Juliano et al, 1990). The
mutant had the highest content (0.9%) of waxy gene protein (Villareal and Juliano,
1989); therefore, the poor digestibility of its protein may be due, in part, to its tight
amylose-waxy gene protein complex. Recently, studies on indica and japonica milled
rices with low amylose content and low starch gelatinization temperature showed that
the high TD of cooked rices was not significantly correlated with low prolamin
content in raw rice or low waxy-gene product but was significantly correlated with
low cysteine content in protein and with low denatured-prolamin content in cooked
rice (Boisen et al, 2001). On the other hand, as discussed earlier, questions recently
have been raised about the assumption that the fecal particles are a digested product of
PB-I. Immunocytochemical analysis conducted by Barber et al (1998a,b) revealed that
intact PB-I and fecal protein particles had no significant difference in size and that all
protein bodies and the fecal particles contained the same unique protein component,
indicating that prolamins were not exclusively confined to PB-I or glutelins to PB-II.
More studies are needed to resolve questions on the digestibility of rice proteins.
Protein Quality
The FAO/WHO method of protein quality evaluation is based on the amino acid
score multiplied by the TD value in rats (FAO, 1990). Application of this method to
the cooked composite rice diets of preschool and adult Filipinos and to their cooked
rice components gave protein quality values lower than those based on lysine
digestibility (Juliano, 1992) (Table 6). TD was 88–90% and lysine digestibility was
TABLE 6
Nutritional Properties of Protein in Typical Cooked-Rice Diets
and Their Cooked Milled Rice Componentsa
Dietb Cooked Milled
Property Preschool Adult Rice
Protein, % N × 6.25 14.2 12.7 8.8
Lysine, g/16 g of N 5.8 0.2 5.3 0.2 3.6 0.2
Amino acid score,c % 100 3 92 3 62 3
Digestible energy, % 91.3 0.8 93.0 0.8 95.3 0.2
True digestibility, % 88.8 0.8 87.8 0.9 90.0 0.8
Biological value, % 90.0 0.8 86.6 1.1 82.5 1.1
Net protein utilization, % 79.9 1.2 75.5 1.4 74.3 1.2
Lysine digestibility, % 95.4 0.5 95.7 0.9 100 0.3
Protein quality,d % 88.8 3.1 80.4 3.3 56.0 3.0
Protein quality,e % 95.4 3.0 88.1 3.5 62.2 3.0
Difference in protein quality, % 6.6 7.7 6.2
b Filipino cooked-rice diets for preschoolers and adults.
c Based on 5.8 g lysine/16 g of N as 100%.
d Amino acid score × true digestibility/100.
e Amino acid score × lysine digestibility/100.
95–96% for cooked-rice diets and 100% for cooked milled rice. Milled rice had
higher digestible energy values but lower protein, BV, and NPU than the rice diets.
The amino acid scores and protein quality of the rice diets were as high or higher
than their NPU, but the NPU of milled rice was higher than its amino acid score and
protein quality. Thus, the new FAO/WHO method underestimates the protein
quality of cooked rice but not that of raw rice with 100% protein and lysine digesti-
bilities in growing rats.
Hypoallergenicity
Rice has been a staple diet in 50% of the world’s population, and yet a relatively
small proportion of adverse food reactions has been attributed to rice proteins
(Helm and Burks, 1996). Generally perceived as hypoallergenic, rice is the only
grain allowed on an extensive elimination diet for allergy testing (Van Hooser and
Crawford, 1989). Therefore, rice is a viable protein source for the development of
hypoallergenic formulas. To achieve that goal, it is necessary to identify and char-
acterize pertinent rice protein allergens. Matsuda et al (1988) reported the isolation
of an allergenic protein from rice grain, which positively reacted with serum class E
immunoglobulins (IgE) from three rice-allergic individuals. The protein had a
molecular mass of 16 kDa and was present mainly in the endosperm portion of the
rice seed. When measured by single immunodiffusion, the allergenic protein con-
stituted 5 and 0.2% of total extractable protein from endosperm and bran, respec-
tively. In later work, Matsuda et al (1991) extracted three immunoreactive proteins
from rice endosperm with molecular masses of 16, 15.5, and 14 kDa. Landers and
Hamaker (1994) found a small amount of highly antigenic protein (approximately
16 kDa) in the alkali-extracted rice bran protein concentrates but not in the albumin
and globulin protein fractions. A substance is antigenic when it induces the forma-
tion of an antibody of any class, and a substance is allergenic when it elicits a Type
I hypersensitivity reaction that specifically involves IgE. Matsuda et al (1988, 1991)
Proteins / 157
demonstrated that, in the case of rice proteins, antigenicity was an indicator of

allergenicity. In studies of cross-allergenicity among cereal grain extracts, the 16-
kDa rice protein was shown to be one of the major allergens by immunoblotting
analysis, histamine release assay from human leukocytes, and RAST inhibition
(Tsai et al, 1990; Urisu et al, 1991; Matsuda, 1991). Recently, Nakase et al (1998)
reported another major allergen in rice, a protein subunit at 33 kDa, which exhibited
structural similarity to wheat and barley allergens.
Watanabe et al (1990) treated milled rice with a surfactant and enzyme mixture
to prepare a hypoallergenic milled rice grain. In clinical tests, this hypoallergenic
rice can dramatically improve rice-associated atopic dermatitis (Wantanabe, 1993).
However, the enzyme treatment is expensive and not suitable for mass production.
Alternative approaches have been reported, including simple alkaline extraction to
remove the antigens (Ikezawa et al, 1999) and genetic modification to achieve the
hypoallergenization (Nakamura and Matsuda, 1996).
CONCLUSION
A lot has been learned and a lot more needs to be studied to fully understand and
utilize rice proteins. Tremendous amounts of data are available on the distribution,
fractionation, composition, processing, and utilization of rice proteins. We have a
good understanding of the basics of rice proteins, are capable of developing high-
protein rice products for practical uses, and have established useful standards for
effective evaluation of rice protein quality. However, more studies are needed to
clarify the complicated relationship between protein bodies and the polypeptide
subunits and, in general, to understand the interrelationships among the structural,
nutritional, mutational, and hypoallergenic properties of the rice proteins. With
advances in technology, more tools are now available to probe into the fine struc-
ture of rice and monitor the functional mechanisms of the rice proteins. Using these
tools, scientists are expected to be even more successful in modifying and develop-
ing higher-quality rice proteins to promote the use of rice and to meet the future
needs of the populace.
REFERENCES
Ansharullah, J., Hourigan, J. A. and Chesterman, Barber, D. L., Lott, J. N. A., and Yang, H. 1998b.
C. F. 1997. Application of carbohydrases in Properties of rice (Oryza sativa L.) Fecal
extracting protein from rice bran. J. Sci. Food protein particles: Light and electron micro-
Agric. 74:141-146. scopic observations. J. Cereal Sci. 27:83-93.
Bakar, J., and Hin, Y. S. 1985. High-protein Bechtel, D. B., and Juliano, B. O. 1980. For-
rice-soya breakfast cereal. J. Food Process mation of protein bodies in the starchy
Preserv. 8(3/4):163-174. endosperm of rice (Oryza sativa L.): A rein-
Barber, S., and Barber, C. B. 1974. Basic and vestigation. Ann. Bot. 45:503-509.
applied research needs for optimizing rice Bera, M. B., and Mukherjee, R. K. 1989. Solu-
bran utilization as food and feed. Pages 1-94 bility, emulsifying and foaming properties of
in: Proc. Rice By-Products Utilization. S. rice bran protein concentrates. J. Food Sci.
Barber and E. Tortosa, Eds. Instituto de 54:142-143.
Agroquimica y Techologia de Alimentos, Betschart, A. A., Fong, R. Y., and Sounders, R.
Valencia. M. 1977. Rice byproducts: Comparative
Barber, D. L., Lott, J. N. A., and Yang, H. 1998a. extraction and precipitation of nitrogen from
Immunocytochemical reactions to intact pro- U.S. and Spanish bran and germ. J. Food Sci.
tein bodies in rice (Oryza sativa L.): Using 42:1088-1093.
antibodies to purified fractions of some rice Boisen, S., Duldulao, J. B. A., Mendoza, E. M.
polypeptides. J. Cereal Sci. 27:71-81. T., and Juliano, B. O. 2001. Comparative
protein digestibility in growing rats of Society of Agronmomy/Crop Science Soci-

cooked rice and protein properties of indica ety of America, Madison, WI.
and japonica milled rices. J. Cereal Sci. Connor, M. A., Saunders, R. M., and Kohler,
33:183-191. G. O. 1976. Rice bran protein concentrates
Butterworth, R. F., Spahr, L., Fontaine, S., and obtained by wet alkaline extraction. Cereal
Layragues, P. 1995. Manganese toxicity, Chem. 53:488-496.
dopaminergic dysfunction, and hepatic Devi, K., Geervani, P., and Sumathi, S. 1997.
encephalopathy. Metab. Brain Dis. 10:259- Changes in solubility and in vitro digestibil-
267. ity of rice proteins subjected to heat proc-
Cagampang, G. B., Cruz, L. J., Espiritu, S. G., essing. J. Food Sci. Technol. 34:345-347.
Santiago, R. G., and Juliano, B. O. 1966. Don, L. D., Kusaba, M., Urashima, A. S., Tosa,
Studies on the extraction and composition of Y, Nakayashiki, H., and Mayama, S. 1999.
rice proteins. Cereal Chem. 43:145-155. Population structure of rice blast in Japan
Cagampang, G. B., Perdon, A. A., and Juliano, examined by DNA fingerprinting. Ann.
B. O. 1976. Changes in salt-soluble proteins Phytopathol. Soc. Jpn. 65:15-24.
of rice during grain development. Phyto- Du Cros, D. L., Wrigley, C. W., and Blakeney,
chemistry 15:1425-1429. B. 1979. Fractionation of rice-grain proteins
Capansana, M. V., Cilindro, A. G., Martin, A. by gradient gel electrophoresis and gel
M. J., Garcia, E. I., and Guerrero, A. C. isoelectric focusing. Characterization of rice
1984. Feasibility study of high protein snack genotypes. Riso Milano Entenazionale
food production. I. Rice-soy curls. J. NSTA 28:275-284.
Technol. 9:58-72. Eggum, B. O., and Juliano, B. O. 1997. Proper-
Carroll, L. E. 1990. Functional properties and ties and protein quality in growing rats of a
applications of stabilized rice bran in bakery low-glutelin content rice mutant. Cereal
products. Food Technol. 44(4):74-76. Chem. 74:200-201.
Chauhan, J. S., and Manda, J. S. 1984. Varietal Eggum, B. O., Resurreccion, A. P., and Juliano,
identification in rice (Oryza sativa L.) by B. O. 1977. Effect of cooking on nutritional
physico-chemical characters of grain and value of milled rice in rats. Nutr. Rep. Int.
electrophoretic variants of salt soluble seed 15:649-655.
proteins. Seed Res. 12:78-89. Eggum, B. O., Alabata, E. P., and Juliano, B.
Cheftel, J. C., Cuq, J. L., and Lorient, D. 1985. O. 1981. Protein utilization of pigmented
Amino acids, peptides and proteins. Page and nonpigmented brown and milled rice by
245 in: Food Chemistry. O. R. Fennema, Ed. rats. Qual. Plant. Plant Foods Hum. Nutr.
Marcel Dekker, New York. 31:175-179.
Chen, W. P., and Chang, Y. C. 1984. Produc- Eggum, B. O., Juliano, B. O., Villareal, C. P.,
tion of high-fructose rice syrup and high- and Perez, C. M. 1984. Effect of treatment
protein rice flour from broken rice. J. Sci. on composition and protein and energy utili-
Food Agric. 35:1128-1135. zation of rice and mung bean by rats. Qual.
Cheng, H. H. 1987. Physicochemical properties Plant. Plant Foods Hum. Nutr. 14:261-272.
of rice protein fractions, and their effect on Eggum, B. O., Juliano, B. O., Ibabao, M. B.,
the pasting rice flour. Ph.D. dissertation, and Perez, C. M. 1986. Effect of extrusion
Graduate Institute of Food Science and cooking on nutritional value of rice flour.
Technology, National Taiwan University, Food Chem. 19:235-240.
Taipei. Eggum, B. O., Juliano, B. O., Ibabao, M. G. B.,
Chopra, N., and Hira, C. K. 1986. Effect of Perez, C. M., and Carangal, V. R. 1987.
roasting on protein quality of cereals. J. Plant foods. Hum. Nutr. 37:237.
Food Sci. Technol. 23:233-235. Eggum, B. O., Satoh, H., and Juliano, B. O.
Chrastil, J., 1990. Protein-starch interactions in 1994. Protein quality evaluation of cooked
rice grain. Influence of storage on oryzenin rice for protein mutants in growing rats.
and starch. J. Agric. Food Chem. 38:1804- Cereal Chem. 71:199-201.
1809. FAO. 1990. Protein quality evaluation. Report
Chrastil, J., and Zarins, Z. 1994. Changes in of a Joint FAO/WHO Expert Consultation,
peptide subunit composition of albumins, Bethesda, MD, 4–8 Dec. 1989. FAO, Rome.
globulins, prolamins, and oryzenin in Fell, J. M. E., Reynolds, A. P., Meadows, N.,
maturing rice grains. J. Agric. Food Chem. Khan, K., Long, S. G., Quaghebeur, G.,
42:2152-2155. Taylor, W. J., and Milla, J. P. 1996. Manga-
Coffman, W. R., and Juliano, B. O. 1987. Rice. nese toxicity in children receiving long-term
Pages 101-131 in: Nutritional Quality of parenteral nutrition. Lancet 347:1218-1221.
Cereal Grains: Genetic and Agronomic Gnanasambandam, R., and Hettiarachchy, N.
Improvement, R. A. Olson, Ed. American S. 1995. Protein concentrates from unsta-
Proteins / 159
bilized and stabilized rice bran: Preparation major globulin from rice endosperm. Cereal
and properties. J. Food Sci. 60:1066-1069, Chem. 47:5-12.
1074. Houston, D. F., Iwasaki, T., Mohammad, A.,
Griffin, V. K., and Brooks, J. R. 1989. Produc- and Chen, L. 1968. Radial distribution of
tion and size distribution of rice maltodex- protein by solubility classes in the milled
trins hydrolyzed from milled rice flour using rice kernel. J. Agric. Food Chem. 16:720-
heat-stable alpha-amylose. J. Food Sci. 724.
54:190-193. Huang, P. C., Lin, C. P., and Hsu, J. Y. 1980.
Guo, Y. J., Bishop, R., Ferhnstrom, H., Yu, G. Protein requirements of normal infants at the
Z., Lian, Y. N., and Huang, S. D., 1986. age of about one year: Maintenance nitrogen
Classification of Chinese rice varieties by requirements and obligatory nitrogen losses.
electrofocussing. Cereal Chem. 63:1. J. Nutr. 110:1727-1735.
Gustafson, J. P., and Yano, M. A. D. 2000. Huebner, F. R., Bietz, J. A., Webb, B. D., and
Genetic mapping of hypervariable Juliano, B. O. 1990. Rice cultivar identifica-
minisatellite sequences in rice (Oryza sativa tion by high-performance liquid chromatog-
L.). Theor. Appl. Genet. 100:447-475. raphy of endosperm proteins. Cereal Chem.
Hamada, J. S. 1999. Use of proteases to 67:129-135.
enhance solubility of rice bran proteins. J. Hussain, A., Scanlon, M. G., and Juliano, B. O.
Food Biochem. 23:307-321. 1989. Discrimination of rice cultivars by
Hamaker, B. R. 1994. The influence of rice polyacrylamide gel electrophoresis and high-
protein on rice quality. Pages 177-194 in: performance liquid chromatography. Cereal
Rice Science and Technology. W. E. Chem. 66:353.
Marshall and J. I. Wadsworth, Eds. Marcel Iida, S., Amano, E., and Nishio, T. 1993. A rice
Dekker, Inc., New York. (Oryza sativa L.) mutant having a low con-
Hamaker, B. R., and Griffin, V. K. 1990. tent of glutelin and a high content of prola-
Changing the viscoelastic properties of min. Theor. Appl. Genet. 87:374-378.
cooked rice through protein disruption. Ikezawa, Z., Tsubaki, K., Okuna, H., Shimada,
Cereal Chem. 67:261-264. T., Moteki, K., Sugiyama, H., Katsunuma,
Hamaker, B. R., and Griffin, V. K. 1993. Effect K., Anzai, H., and Amano, S. 1999. Use-
of disulfide bond-containing protein on rice fulness of hypoallergenic rice (AFT-RI) and
starch gelatinization and pasting. Cereal analysis of the salt in soluble rice allergen
Chem. 70:377-380. molecule. Areugi 48:40-49.
Hamaker, B. R., Griffin, V. K., and Intengan, C. L., Roxas, B. V., Cabrera-
Moldenhauer, K. A. K. 1991. Potential Santiago, M. I. Z., Loyola, A. S., Alejandro,
influence of a starch granule-associated E. R., Abadilla, J. N., and Yu, G. F. B. 1984.
protein on cooked rice stickiness. J. Food Protein requirements of Filipino children 18-
Sci. 56:1327-1329, 1346. 30 months old concerning local diets. I.
Hansen, L. P., Hosek, R., Callan, M., and Rice-fish and rice-mung bean diets. Philipp.
Jones, F. T. 1981. The development of high- J. Nutr. 37:87-95.
protein rice flour for early childhood feed- IRRI. 1980. Annual report for 1979. Pages 25-
ing. Food Technol. 35(11):38-42. 38. IRRI, Manila.
Harp, T. L., and Correll, J.C. 1998. Recovery IRRI. 1983. Annual report for 1982. Pages 70-
and characterization of spontaneous, sele- 75. IRRI, Manila.
nate-resistant mutants of Magnaporthe Iwasaki, T., Shibuya, N., Suzuki, T., and
grisea, the rice blast pathogen. Mycologia Chikubu, S. 1982. Gel filtration and electro-
90:954-963. phoresis of soluble rice proteins extracted
Helm, R. M., and Burks, A. W. 1996. Hypoal- from long, medium, and short grain varie-
lergenicity of rice protein. Cereal Foods ties. Cereal Chem. 59:192-195.
World 41:839-843. Juliano, B. O. 1972. The rice caryopsis and its
Hibino, T., Kidzu, K., Masumura, T., Ohtsuki, composition. Pages 16-74 in: Rice: Chem-
K., Tanaka, K., Kawabata, M., and Fujii, S. istry and Technology, 1st ed. D. F. Houston,
1989. Amino acid composition of rice Ed. Am. Assoc. Cereal Chem., St. Paul,
prolamin polypeptides. Agric. Biol. Chem. MN.
53:513-518. Juliano, B. O. 1985. Polysaccharides, Proteins, and
Houston, D. F. 1972. Rice bran and polish. Lipids of Rice. Pages 59-160 in: Rice: Chem-
Pages 272-291 in: Rice: Chemistry and istry and Technology, 2nd ed. B. O. Juliano, Ed.
Technology, 1st ed. D. F. Houston, Ed. Am. Am. Assoc. Cereal Chem., St. Paul, MN.
Assoc. Cereal Chem., St. Paul, MN. Juliano, B. O. 1992. Structure, chemistry, and
Houston, D. F., and Mohammad, A. 1970. function of the rice grain and its fractions.
Purification and partial characterization of a Cereal Food World 37:772-779.
Juliano, B. O. 1993. Rice in Human Nutrition. Mandac, B. E., and Juliano, B. O. 1978. Prop-
Food and Agriculture Organization of the erties of prolamin in mature and developing
United Nations, Rome. rice grain. Phytochemistry 17:611-614.
Juliano, B. O., Perez, C. M., Kaushik, R., and Marshall, W. E., Normand, F. L., and Goynes,
Khush, G. S. 1990. Some grain properties of W. R. 1990. Effects of lipid and protein
IR-36-based starch mutants. Starch 42:256- removal on starch gelatinization in whole
260. grain milled rice. Cereal Chem. 67:458-
Kagawa, H., Hirano, H., and Kikuchi, F. 1988. 463.
Variation of glutelin seed storage protein in Matsuda, T., Sugiyama, M., Nakamura, R., and
rice (Oryza sativa L.). Jpn. J. Breed. 38:327- Torii, S. 1988. Purification and properties of
332. an allergenic protein in rice grain. Agric.
Khan, M. A., and Eggum, B. O. 1979. Effect of Biol. Chem. 52:1465-1470.
home and industrial processing on protein Matsuda, T., Nomura, R., Sugiyama, M., and
quality of baby foods and breakfast cereals. Nakamura, R. 1991. Immunochemical stud-
J. Sci. Food Agric. 30:369-376. ies on rice allergenic proteins. Agric. Biol.
Knorr, D. 1982. Effects of recovery methods on Chem. 55:509-513.
the functionality of protein concentrates Morita, T., and Kiriyama, S. 1993. Mass pro-
from food processing wastes. J. Food Proc. duction method for rice protein isolate and
Eng. 5:215-230. nutritional evaluation. J. Food Sci. 58:1393-
Knorr, D., and Betschat, A. A. 1983. Effect of 1396, 1406.
dehydration methods on the functionality Morita, Y., and Yoshida, C. 1968. Studies on
of plant protein concentrates. Starch/Starke g-globulin from rice embryo. Agric. Biol.
35:23-28. Chem. 32:664-670.

Kumamaru, T., Satoh, H., Iwata, N., Omura, T., Nakamura, R., and Matsuda, T. 1998. Rice
Ogawa, M., and Tanaka, K. 1988. Mutants allergenic protein and molecular approach
of rice storage proteins. I. Screening of for hypoallergenic rice. Biosci. Biotechnol.
mutants for rice storage proteins of protein Biochem. 60:1215-1221.
bodies in the starchy endosperm. Theor. Nakase, M., Usui, Y., Alvarez-Nakase, A. M.,
Appl. Genet. 76:11-16. Adachi, T., Urisu, A., Nakamura, R., Aoki,
Kumamaru, T., Sato, H., and Satoh, H. 1997. N., Kitajima, K., and Matsuda, T. 1996.
High-lysine mutants of rice, Oryza sativa L. Cereal allergens: Rice-seed allergens with
Plant Breed. 116:245-249. structural similarity to wheat and barley
Kung, L., Chen, H., and Sung, H. 1987. A new allergens. Allergy 53(Suppl. 46):55-57.
method for separation of rice protein and Ogawa, M., Kumamaru, T., Satoh, H., Iwata,
starch. J. Chin. Agric. Chem. Soc. 25:299-307. N., Omura, T., Kasai, Z., and Tanaka, K.
Kusama, T., Yanagi, S. O., and Iwamato, M. 1987. Purification of protein body-I of rice
1984. Comparison of rice (Oryza sativa L.) seed and its polypeptide composition. Plant
seed proteins from several varieties. Agric. Cell Physiol. 28:1517-1527.
Biol. Chem. 48:1649-1651. Ogawa, M., Kumamaru, T., Satoh, H., Omura, T.,
Landers, P., and Hamaker, B. R. 1994. Anti- Park, T., Shintaku, K., and Baba, K. 1989.
genic properties of albumin, globulin, and Mutants for rice storage proteins. 2. Isolation
protein concentrate fractions from rice bran. and characterization of protein bodies from
Cereal Chem. 71:409-411. rice mutants. Theor. Appl. Genet. 78:305-
Lasztity, R. 1984. Rice proteins. Pages 165-176 310.
in: The Chemistry of Cereal Proteins. R. Osborne, T. B. 1924. The Vegetable Proteins.
Lasztity, Ed. CRC Press, Inc., Boca Raton, Longmans, Green and Co., London.
FL. Otterburn, M. S. 1989. Protein crosslinking.
Lim, S. T., Lee, J. H., Shin, D. H., and Lim, H. Page 247 in: Protein Quality and the Effects
S. 1999. Comparison of protein extraction of Processing. R. D. Phillips and J. W.
solutions for rice starch isolation and effects Finley, Eds. Marcel Dekker Inc., New York.
of residual protein content on starch pasting Palamiano, E. P., Almazan, A. M., and Juliano,
properties. Starch/Staerke 51:120-125. B. O. 1968. Physicochemical properties of
Lookhart, G. L., Albers, L. B., Pomeranz, Y., protein of developing and mature rice grain.
and Webb, B. D. 1987. Identification of U.S. Cereal Chem. 45:1-12.
rice cultivars by high-performance liquid Prabhakar, J. V. 1987. Stabilization of rice
chromatography. Cereal Chem. 64:199-206. bran. Page 330 in: Handbook on Rice Bran.
Lynn, L. 1969. Edible rice bran foods. Page 62 B. M. Sheth and B. V. Mehta, Eds. Solvent
in: Protein Enriched Cereal Foods for World Extractors Association of India, Bombay.
Needs. M. Milner, Ed. Am. Assoc. Cereal Prakash, J., and Ramanatham, G. 1994. Effect
Chem., St. Paul, MN. of stabilization of rice bran on the extracta-
Proteins / 161
bility and recovery of proteins. Nahrung Shih, F. F. 1996. Edible films from rice protein
38:87-95. concentrate and pullulan. Cereal Chem.
Prakash, J., and Ramanatham, G. 1995a. Effect 73:406-509.
of stabilization treatment of rice bran on Shih, F. F., and Daigle, K. W. 2000. Prepara-
functional properties of protein concentrates. tion and characterization of rice protein iso-
J. Sci. Food Agric. 67:181-187. lates. J. Am. Oil Chem. Soc. 77:885-889.
Prakash, J., and Ramanatham, G. 1995b. Effect Shih, F. F., Champagne, E. T., Daigle, K., and
of stabilization of rice bran on nutritional Zarins, Z. 1999. Use of enzymes in the proc-
quality of protein concentrates. Int. J. Food essing of protein products from rice bran and
Sci. Nutr. 46:177-184. rice flour. Nahrung 43:14-18.
Rand, W. M., Uauy, R., and Scrimshaw, N. S., Sonti, R. V. 1998. Bacterial leaf blight of rice:
Eds. 1984. Protein-Energy-Requirement New insights from molecular genetics. Curr.
Studies in Developing Countries: Results of Sci. (Bangalore) 74:206-212.
International Research. Food Nutr. Bull. Souza, S. R., Stark, E. M. L. M., and
Suppl. UN University, Tokyo. Fernandes, M. S. 1999. Foliar spraying of
Resurreccion, A. P., Juliano, B. O., and Eggum, rice with nitrogen: Effect on protein levels,
B. O. 1978. Preparation and properties of protein fractions, and grain weight. J. Plant
destarched milled rice. Nutr. Rep. Int. 18:17- Nutr. 22:579-588.
25. Sugimoto, T., Tanaka, K., and Kasai, Z. 1986.
Resurreccion, A. P., Juliano, B. O., and Tanaka, Molecular species in the protein body II
Y. 1979. Nutrient content and distribution in (PB-II) of developing rice endosperm. Agric.
milling fractions of rice grain. J. Sci. Food Biol. Chem. 50:3031-3035.
Agric. 30:475-481. Tanaka, K., and Ogawa, M. 1988. Storage
Resurreccion, A. P., Li, X., Okita, T. W., and protein genes and their expression control:
Juliano, B. O. 1993. Characterization of Specially focusing on the improvement of
poorly digested protein of cooked rice pro- PB-I digestibility. Abstracts, Rockefeller
tein bodies. Cereal Chem. 70:101-104. Foundation Program on Rice Biotechnology.
Sarker, S. C., Ogawa, M., Takahshi, M., and IRRI, Los Baños, Philippines.
Asada, K. 1986. The processing of a 57-kDa Tanaka, Y., Resurreccion, A. P., Juliano, B. O.,
precursor peptide to subunits of rice glutelin. and Bechtel, D. B. 1978. Properties of whole
Plant Cell Physiol. 27:1579-1986. and undigested fraction of protein bodies of
Saunders, R. M. 1990. The properties of rice milled rice. Agric. Biol. Chem. 42:2015-
bran as a foodstuff. Cereal Chem. 35:632- 2023.
635. Tanaka, K., Sugimoto, T., Ogawa, M., and Kasai,
Sawai, H., and Morita, Y. 1968. Studies on rice Z. 1980. Isolation and characterization of two
glutelin. II. Cross-structure of glutelin from types of protein bodies in the rice endosperm.
rice endosperm. Agric. Biol. Chem. 32:496- Agric. Biol. Chem. 44:1633-1639.
500. Tecson, E. M. S., Esmama, B. V., Lontok, L.
Sawai, H., and Morita, Y. 1970. Studies on P., and Juliano, B. O. 1971. Studies on the
g-globulin of rice embryo. III. Molecular extraction and composition of rice endo-
dimension and chemical composition of sperm glutelin and prolamin. Cereal Chem.
g1 globulin. Agric. Biol. Chem. 34:61-67. 48:168-181.
Sayre, R. N., Saunders, R. M., Enochian, R. V., Tsai, Y.-T., Chen, S.-H., Lin, K.-L., and Hsieh,
Schultz, W. G., and Beagle, E. C. 1982. K.-H. 1990. Rice pollen allergy in Taiwan.
Review of rice bran stabilization systems Ann. Allergy 65:4359.
with emphasis on extrusion cooking. Cereal Urisu, A., Yamada, K., Masuda, S., Komada,
Foods World 27:317-322. H., Wada, E., Kondo, Y., Horiba, F.,
Schaeffer, G. W., and Sharpe, F. T. 1997. Elec- Tsuruta, M., Yasaki, T., Ymada, M., Torii,
trophoretic profiles and amino acid compo- S., and Nakamura, R. 1991. 16-Kilodalton
sition of rice endosperm proteins of a mutant rice protein is one of the major allergens in
with enhanced lysine and total protein after rice grain extract and responsible for cross-
backcrosses from germplasm improvements. allergenicity between cereal grains in the
Theor. Appl. Genet. 95:230-238. Poaceae family. Int. Arch. Allergy Appl.
Sengupta, S., and Chattopadhyay, N. C. 2000. Immunol. 96:244-252.
Rice varietal identification by SDS-PAGE. Van Hooser, B., and Crawford, L. V. 1989.
Seed Sci. Technol. 28:871-873. Allergy diets for infants and children.
Shaw, J. F., and Sheu, J. R. 1992. Production of Compr. Ther. 15:38-47.
high-maltose syrup and high-protein flour Villareal, R. M., and Juliano, B. O. 1978. Prop-
from rice by an enzymatic method. Biosci. erties of glutelin from mature and develop-
Biotechnol. Biochem. 56:1071-1073. ing rice grain. Phytochemistry 17:177-182.
Villareal, R. M., and Juliano, B. O. 1981. Prop- Watanabe, M. 1993. Hypoallergenic rice as a
erties of albumins of milled rice. Phyto- physiological functional food. Trends Food
chemistry 20:1785-1789. Sci. Technol. 3:125-128.
Villareal, C. P., and Juliano, B. O. 1987. WHO. 1985. Energy and Protein Requirements.
Varietal differences in quality character- Tech. Rep. Ser. World Health Organization,
istics of puffed rices. Cereal Chem. 64: Geneva.
337-342. Wieser, H., Seilmeier, W., and Belitz, H. D.
Villareal, C. P., and Juliano, B. O. 1989. Com- 1980. Vergleichende Untersuchungen uber
parative level of waxy gene product of Particlle Aminosauresequenzen von Prola-
endosperm starch granules of different rice minen und Glutelinen verschiedener
ecotypes. Starch 41:369-371. Geteidearten. II. Fractionierung der gluteline.
Villareal, C. P., Maranville, J. W., and Juliano, Z. Lebensm. Unters. Forsch. 171:430-436.
B. O. 1991. Nutrient content and retention Wilson, C. M. 1987. Proteins of the kernel.
during milling of brown rices from the Inter- Pages 273-310 in: Corn: Chemistry and
national Rice Research Institute. Cereal Technology, 1st ed. S. A. Watson and P. E.
Chem. 68:437-439. Ramstad, Eds. Am. Assoc. Cereal Chem., St.
Wang, M., Hettiarachchy, N. S., Qi, M., Paul, MN.
Burks, W., and Seibenmorgen, T. 1999. Yang, C.-H., and Chang, W.-H. 1999. Effects of
Preparation and functional properties of protein and lipid binding to starch on the
rice bran protein isolate. J. Agric. Food physicochemical and pasting properties of rice
Chem. 47:411-416. flour. Food Sci. Agric. Chem. 1:277-285.
Wantanabe, M. M., J., Ikezawa, Z., Suzuki, Y., Zhou, Z., Bebeli, P. J., Somers, D. J., and
Hirao, T., Yoshizawa, T., and Arai, S. 1990. Gustafson, J. P. 1997. Direct amplification
Production of hypoallergenic rice by enzy- of minisatellite-region DNA with VNTR
matic decomposition of constituent proteins. core sequences in the genus Oryza. Theor.
J. Food Sci. 55:781-783. Appl. Genet. 95:942-949.
CHAPTER 7
RICE LIPIDS
J. Samuel Godber
Department of Food Science
Louisiana State University Agricultural Center
Baton Rouge, Louisiana
Bienvenido O. Juliano
Philippine Rice Research Institute Los Baños
Laguna, Philippines
Lipids in rice, although not as abundant as the carbohydrate and protein compo-
nents, are important because they contribute to nutritional, sensory, and functional
qualities. Lipid rancidity in brown rice is well known as a major deterrent to more
widespread food usage of this form of rice. In addition, lipids associated with the
starch component have been found to affect the pasting and other functional char-
acteristics of starch. Recent evidence for the contribution of minor lipid components
of rice to protection against chronic diseases such as heart disease and cancer have
heightened interest in rice lipids from a nutritional standpoint. This chapter charac-
terizes the lipids in rice, with a focus on specific categories and classes of lipids and
emphasis on their functionality.
Several reviews have been published on the chemistry of rice lipids (Juliano,
1977, 1983; Fujino, 1978; Morrison, 1978), although little additional information
has been published on the gross composition of rice lipids since the publication of
the previous edition of this text. This edition includes recent developments in rice
lipid composition, with emphasis on factors that affect that composition. It also in-
cludes recent information on the nutritionally important minor constituents.
Lipids are present in the form of spherosomes, or lipid droplets, with diameters
of <1.5 µm in the aleurone layer, <1 µm in the subaleurone layer, and <0.7 µm in
the embryo of rice grain (Juliano, 1983). In addition, endosperm cell walls stain for
lipid dyes such as Sudan IV (Yoshizawa et al, 1980). Most of the lipids in the
endosperm are associated with protein bodies, but it is believed that starch granules
also have bound lipids.
Lipids are generally classified into nonstarch lipids, principally those in the
spherosomes of the aleurone layer and embryo, and starch lipids (Morrison, 1978).
The former represent the majority of the lipids present in rice, but the latter have
considerable importance relative to starch functionality (Morrison, 1995). Morrison
(1981) proposed the term starch surface lipids for nonstarch lipids that contaminate
internal starch lipids (i.e., lipids inside the granule), but Gailliard and Bowler (1987)
163
contended that lipids found on the surface of the granule should be considered
starch lipids. With current interest in the functionality of starch and the potential
role of lipids in functionality, additional efforts have been made to further clarify
the origin and roles of lipids associated with starch, as discussed later.
EXTRACTION OF LIPIDS
Nonpolar solvents such as diethyl ether, petroleum ether, and chloroform/metha-

nol (2:1, v/v) readily extract nonstarch lipids from dry grain (Morrison, 1978;
Choudhury and Juliano, 1980a). Residual nonstarch lipids may be further extracted
with water-saturated butanol (WSB) for 30 min to 1 hr at 20-30°C. Starch granules
damaged as a result of milling steps, unlike undamaged granules, leach out starch
lipids into WSB at ambient temperature and contaminate the nonstarch lipids.
Lipids of protein bodies of raw milled rice were 84% extracted by 8-hr contact
with chloroform/methanol (2:1, v/v) at 25C; an 11% additional extraction by 5-min
contact with cold WSB produced a total of 95% extraction (Choudhury and Juliano,
1980a). Thus, protein body lipids would mostly be in the nonstarch lipid fraction of
brown or milled rice.
Starch lipids may be extracted with WSB after extraction of nonstarch lipids.
Although WSB gave an appreciable yield of internal wheat starch lipids from five
overnight extractions at 2-4°C, an estimated 2.5 × 105 hr would be needed to extract
all internal rice starch lipids under these conditions (Morrison, 1981). Extraction of
pronase-digested rice flour with WSB twice for 12 hr each at ambient temperature
removed 83-85% of the lipids from waxy rices and 50–86% of the lipids from non-
waxy rices. For a sample of IR42 brown rice from which the nonstarch lipids had
been extracted, further extraction by cold WSB reduced fat-by-hydrolysis from 0.41
to 0.09%, indicating a 78% extraction of starch lipids by cold WSB (Choudhury and
Juliano, 1980a). Refluxing WSB (67% butanol and 33% water at 92°C), which ge-
latinizes the starch granules, is required to completely extract starch lipids. Reflux-
ing 95% ethanol was only as effective as cold WSB in extracting lipids from rice
starch (IRRI, 1983a). Azudin and Morrison (1986) contended that the most appro-
priate method to obtain starch lipid quantification would be to extract all lipids from
finely milled rice with hot propanol-water and extract total lipids from purified
starch in the same manner, with the difference being nonstarch lipid.
MAJOR LIPID CATEGORIES
Starch Lipids
Starch lipids represent a relatively small proportion of the total lipid composition
of rice but may play a role in starch synthesis and appear to make a major contribu-
tion to starch functionality (Morrison, 1995). As mentioned previously, recognition
of the difficulties in separating lipid associated with the starch in the granule from
lipid that occurs incidentally on the surface of the starch granule has been the sub-
ject of considerable debate and study. Lipids can be associated with starch in sev-
eral different ways (Morrison, 1988). Morrison (1988) suggested that the classifica-
tion system used for wheat starch may be applicable for all starches. In this system,
lipids inside the starch granule would be considered true starch lipids. Lipids from
Lipids / 165
the surrounding proteinaceous matrix of the endosperm would be classified as starch

surface lipids. Both of these categories are composed primarily of monoacyl lipids.
All other lipids associated with starch would be called nonstarch lipids (Morrison,
1988). Vasanthan and Hoover (1992) reported levels of lipid in what they termed
“highly purified starch granules,” in which the degree of starch damage was also as-
sessed. They classified starch lipids as total lipid by acid hydrolysis, unbound lipid
by chloroform/methanol (2:1, v/v), and bound lipid by hot n-propanol/water (3:1,
v/v), with the unbound lipid being considered primarily surface lipid. Kitahara et al
(1997) employed glucoamylase as a means to more appropriately characterize lipids
associated with starch, and recent use of more sophisticated spectroscopic tech-
niques has further clarified starch-lipid interactions (Morrison et al, 1993). How-
ever, the exact nature of starch-associated lipid has not been entirely delineated.
Starch lipids extracted with cold WSB from defatted brown rice amounted to
0.6-0.7% for nonwaxy rice and 0.2% for waxy rice (Choudhury and Juliano, 1980b).
Corresponding values for milled rice were 0.5% for nonwaxy and 0.1% for waxy
rice. Since cold WSB extracted only 78% of brown-rice starch lipids relative to hot
WSB, the true starch lipid level would be 0.7-0.8% for nonwaxy rice. Thus, non-
waxy rice has more starch lipid and less nonstarch lipid in brown and milled rice
than does waxy rice. However, the total content of grain lipids seems to be similar.
IR42 brown rice (29% amylose) had 2.5% nonstarch lipids and 0.7% starch lipids;
IR480-5-9 brown rice (24% amylose) had 2.7% nonstarch lipids and 0.8% starch
lipids; and IR4445-63-1 (waxy) had 2.9% nonstarch lipids and 0.2% starch lipids,
for a total lipid content of 3.1-3.5% at 14% moisture. Corresponding values for
over-milled rice were 0.4% nonstarch plus 0.6% starch lipids for IR24 and IR480-
5-9, and 0.7% plus 0.1% for waxy rice, for a total of 0.8-1.0% total lipids.
Kawashima and Kiribuchi (1980) reported 0.6% nonstarch and 0.6% starch lipids in
a nonwaxy milled rice and 0.8% nonstarch and 0.4% starch lipids in a waxy milled
rice (a total of 1.2%).
Morrison and Azudin (1987) reported a wide range of starch lipids, from less than
0.1% in waxy varieties to as high as 1.3% in nonwaxy varieties. These researchers
evaluated the influence of genetics and environmental conditions on starch lipids
and looked for a potential correlation between starch lipids and amylose content.
For some varieties, a clear relationship was seen between amylose and starch lipid
but not in all cases.
Starch lipids consist primarily of lysophospholipids and free fatty acids (Morrison,
1988). Neutral lipids and glycolipids have been reported (Choudhury and Juliano,
1980b), but these may be contaminants from nonstarch lipid (Morrison, 1988).
Major phospholipid species were lysophosphatidylethanolamine and lysophosphati-
dylcholine. Starch lipids of waxy milled rice had more free fatty acids and fewer
lysophospholipids than starch lipids of nonwaxy milled rice (Choudhury and
Juliano, 1980b). Hirayama and Matsuda (1973) reported that lysophosphatidylcho-
line was the major phospholipid of starch lipids, followed by lysophosphatidyletha-
nolamine, and that sterol glycosides were the major glycolipids.
Major fatty acids of nonwaxy milled-rice starch lipids were palmitic and linoleic,
together with a lesser amount of oleic acid (Choudhury and Juliano, 1980a,b).
Starch lipids had less oleic acid and more palmitic acid than nonstarch lipids. Starch
lipids from waxy milled rice had 47% palmitic, 18% oleic, and 30% linoleic acid
(Choudhury and Juliano, 1980b). Kawashima and Kiribuchi (1980) also reported
starch lipids of milled rice to be 35% palmitic, 14% oleic, and 46% linoleic in non-
waxy rice and 38% palmitic, 15% oleic, and 36% linoleic in waxy rice. The neutral
fraction of starch lipids from nonwaxy milled rice had 21-25% palmitic, 9-22%
oleic, and 48-68% linoleic acid; glycolipids had 45-56% palmitic, 5-9% oleic, and
26-45% linoleic acid; and phospholipids had 46-50% palmitic, 5-16% oleic, and
30-42% linoleic acid (Choudhury and Juliano, 1980b; Maniñgat and Juliano, 1980).
None of the three fractions had a fatty-acid composition similar to that of the total
starch lipids.
The quantification of fat-by-hydrolysis probably corresponds to the fatty-acid
fraction of starch lipids (Choudhury and Juliano, 1980a). Both had identical fatty-
acid compositions. Cold-WSB extraction of 0.57% starch lipids from IR42 defatted
brown rice reduced fat-by-hydrolysis from 0.35 to 0.08%, and the fat-by-hydrolysis
corresponded to 48% by weight of starch lipids.
Influence of Starch Purification
The level and composition of the lipids of purified starch granules depends to a
large measure on the method of preparation. Dodecyl benzene sulfonate (DoBS)
extraction of protein during starch preparation of IR480-5-9 milled rice reduced
cold-WSB lipids of starch from 0.69 to 0.38% and changed the neutral lipid-
glycolipid-phospholipid ratio from 26:16:58 (Choudhury and Juliano, 1980b) to
30:36:34 (Maniñgat and Juliano, 1980). For eight milled rices, hot-WSB lipids
ranged from 0.09 to 0.79% for DoBS-prepared starch, as compared to 0.19-1.42%
for starch prepared by alkali protease (Maniñgat and Juliano, 1980; IRRI, 1983b).
The corresponding cold-WSB lipids were 0.03-0.38% and 0.17-0.81% at 14%
moisture. The major neutral lipid species of DoBS-prepared starch was free fatty
acid (88-97%), with minor amounts of monoglycerides, diglycerides, and sterols.
Major glycolipids were diglycosyl monoglycerides and monoglycosyl monoglyc-
erides, together with smaller amounts of sterol glycosides and contaminant DoBS.
The major phospholipid was lysophosphatidylcholine (61-91%), followed by lyso-
phosphatidylethanolamine, with smaller amounts of their parent phospholipids. Ito
et al (1979) reported 0.59% starch lipids for a DoBS-prepared starch consisting of
32% free fatty acids, 30% lysophosphatidylcholine, 11% monoglycosyl monoglyc-
erides, 8% diglycosyl monoglycerides, 7% lysophosphatidylethanolamine, and 3%
monoglycerides, equivalent to a fraction ratio of 37:20:43. Component sugars of
glycolipids were predominantly galactose, plus glucose. DoBS-prepared starch had
no nonstarch lipids extractable with petroleum ether (Maniñgat and Juliano, 1980).
A DoBS-prepared rice starch contained 47% palmitic, 9% oleic, and 40% lino-
leic acid in its lipids (0.59% at 14% moisture) (Ito et al, 1979). Free fatty acids were
30% palmitic, 10% oleic, and 54% linoleic acid; lysophosphatidylcholine had 79%
palmitic, 7% oleic, and 9% linoleic acid; and lysophosphatidylethanolamine had
51% palmitic, 4% oleic, and 42% linoleic acid. Starch lipids from waxy rice pre-
pared by the DoBS method had 58% palmitic, 19% oleic, and 14% linoleic acid
(Maniñgat and Juliano, 1980).
The lipid fractions of two DoBS-prepared starches also showed differences in
fatty-acid composition but were not identical to the corresponding fractions of
starch lipids from milled rice (Maniñgat and Juliano, 1980). Neutral lipids had 24 or
34% palmitic, 11 or 12% oleic, and 62 or 50% linoleic acid; glycolipids had 37 or
41% palmitic, 3% oleic, and 58 or 52% linoleic acid; and phospholipids contained
57 or 63% palmitic, 4% oleic, and 35 or 26% linoleic acid.
Lipids / 167
In contrast to the neutral lipid-glycolipid-phospholipid ratio for starch lipids ex-

tracted with cold WSB from IR480-5-9 brown rice (28:20:52) and from milled rice
(26:15:59) (Choudhury and Juliano, 1980b), starch prepared by alkaline protease
treatment had 0.7% cold-WSB lipids with a ratio of 44:7:49 uncorrected for contami-
nant nonstarch lipids (Maniñgat and Juliano, 1980). Sawada and Mano (1980) re-
ported ratios of 51:15:34 and 33:9:58 for two protease-prepared tropical rice starches.
Mano (1982) reported a ratio of 37:20:43 for DoBS-prepared starch of a Japanese
variety. Azudin and Morrison (1982, unpublished data) reported 0.02% free fatty
acids and no lysophospholipids as true starch lipids in two waxy rice starches and
0.14–0.30% free fatty acids and 0.45–0.64% lysophospholipids as true lipids for six
nonwaxy rice starches, equivalent to a ratio of 27-40:0:60-73 (mean ratio, 33:0:67).
Free fatty acids, lysophosphatidylcholine, and lysophosphatidylethanolamine were
the major hot-WSB starch lipids of two nonwaxy starches; the amounts were 34, 23,
and 10% and 36, 18, and 18%, respectively (Sawada and Mano, 1980).
Contaminant nonstarch lipids in protease-prepared rice starches were 0.75 and
0.98% in two waxy rices at 14% moisture (0.02% starch lipids) and 0.45-1.08% in
six nonwaxy starches with 0.65-0.93% starch lipids (Azudin and Morrison, 1982,
unpublished data). The mean content was 0.68% nonstarch lipids and 0.77% starch
lipids. In waxy rice starch, major nonstarch lipids were triglycerides, diglycerides
plus free fatty acids, lysophospholipids, sterol esters, monoglycerides plus acyl
sterol glycosides, and glycosyl glycerides. Among the six nonwaxy starches, the
major type of nonstarch lipid was triglycerides, followed by lysophospholipids,
diglycerides plus free fatty acids, and variable amounts of monoglycerides plus acyl
sterol glycosides, sterol esters, and glycosyl glycerides.
Alkali treatments to prepare starch probably saponify starch lipids and render
them extractable as soap (sodium salt of fatty acids). Thus, Schoch (1942) extracted
soap from commercial rice starch with 0.59% lipids, and the residual lipid content
after seven 1-hr extractions with three parts of 85% methanol under reflux was
0.03% (95% extraction); it was 0.07% (88% extraction) after 48 hr of Soxhlet
extraction with 80% dioxane in water. Ohashi et al (1980) removed >60% of starch
lipids by 3-hr reflux with five parts of 99% methanol; these lipids consisted of 44
mol% linoleic, 37 mol% palmitic, and 13 mol% oleic acid. A starch prepared with
0.1% NaOH had 0.71% fat-by-hydrolysis at 14% moisture and 0.71% hot-WSB
starch lipids (Ohashi et al, 1980).
The 0.1% NaOH-prepared rice starch had starch lipids consisting of 70% free
fatty acids plus diglycerides, 22% lysophospholipids (mainly lysophosphatidylcho-
line), 4% diglycosyl glycerides, 3% monoglycerides plus acyl sterol glycosides, 2%
each of monoglycosyl glycerides and sterol esters, and <1% triglycerides (Azudin
and Morrison, 1982, unpublished data). The neutral lipid-glycolipid-phospholipid
ratio was approximately 72:6:22. Alkali treatment to remove protein reduced cold-
WSB lipids of IR480-5-9 rice from 0.69 to 0.24% for 0.1% NaOH and to 0.08% for
0.2% NaOH, with corresponding lipid fraction ratios of 80:9:11 and 79:9:12,
respectively (Maniñgat and Juliano, 1980), using a milled rice with a starch lipid
ratio of 26:16:58 (Choudhury and Juliano, 1980a). Rice starch prepared with 0.1%
NaOH had 0.13% nonstarch lipids in addition to 0.47% starch lipids (N. S. Azudin
and W. R. Morrison, 1982, unpublished data). Major nonstarch lipids were 91%
free fatty acids plus diglycerides and 6% triglycerides.
Vasanthan and Hoover (1992) evaluated the lipid composition of a commercially
prepared rice starch (Sigma Chemical Co., St. Louis, MO) using their system of
classifying lipids as total, bound and unbound. They found total lipid, by acid hy-
drolysis, to be 0.76%, of which 0.05% was unbound and 0.71% was bound. Neutral
lipid was the predominate class of lipid in the unbound, or surface lipids, whereas
phospholipids were the predominate class in the bound lipids. Hoover et al (1996)
found 0.68% total lipid, with 0.17% unbound and 0.50% bound in long-grain brown
rice (IR64).
Nonstarch Lipids
Nonstarch lipids are found in the aleurone, subaleurone, and germ of brown rice,
and a small amount is in the hull of rough rice. Obviously, the type and degree of
milling influences the content of nonstarch lipid in rice products. In fact, it has been
suggested that degree of milling could be determined based on grain-surface lipid
quantification (Bennett et al, 1993; Chen et al, 1998).
The hull of IR42 rice had about 0.4% lipid (nonstarch) at 14% moisture and
contributed little to total rough rice lipids (Choudhury and Juliano, 1980a). It con-
tributed 0.4% of total nonstarch lipids (0.3% of total neutral lipids, 17% of total
glycolipids, and 7% of total phospholipids) of rough rice and consisted mainly of
neutral lipids (Table 1). Hexane-extracted hull lipids had similar fatty acids as
brown-rice nonstarch lipids (Hartman and Lago, 1976) (Table 1), but it contained <1%
arachidic (C20:0), <2% behenic (C22:0), and <4% 1ignoceric (C24:0) acids (Morrison,
1978).
Using 13C nuclear magnetic resonance spectroscopy, Bradbury and Collins (1982)
confirmed a lipid content of 3.1% for IR480-5-9 brown rice, of which 0.75% was
for endosperm, 27.3% for germ, and 26.7% for aleurone cells plus grain coat at 14%
moisture, which is equivalent to 22% of brown-rice nonstarch lipids each in endo-
sperm and germ, and 56% in aleurone cells plus grain coat.
In three brown-rice samples, nonstarch lipids at 14% moisture were 2.5% for
29%-amylose rice, 2.7% for 24%-amylose rice, and 2.9% for waxy (1.7%-amylose)
rice (Choudhury and Juliano, 1980b). The difference in fat content may be due in
part to the higher nonstarch lipid content of the inner endosperm, which makes up
85% of brown rice. The nonstarch lipid content was 0.7% of the inner endosperm of
waxy rice, as compared with 0.4% for the two nonwaxy rices. Reported nonstarch
lipids for Japanese brown rice are 2.4% (Hirayama and Matsuda, 1973) and 2.6%
(Fujino and Mano, 1972). The mean lipid content of brown rice from cultivars
grown in the United States was 2.87% (n = 14) for long-grain type, 2.76% (n = 10)
for medium-grain type, 2.49% (n = 2) for short-grain type, and 3.11% (n = 2) for
short-grain (waxy) type (Taira and Itani, 1988).
Nonstarch lipids of brown rice had saponification number, iodine number, and
unsaponifiable matter content mostly similar to those of its fractions (Table 1). Rice
hull lipids differ from brown rice lipids in saponification number and iodine
number, probably because of their higher level of unsaponifiable matter (Hartman
and Lago, 1976) and the lower linoleic acid content (Table 1). Major fatty acids are
36-38% linoleic (C18:2), 35-37% oleic (C18:1) and 23-24% palmitic (C16:0) acids
(Choudhury and Juliano, 1980a,b). The higher level of palmitic acid and lower level
of oleic acid in milled-rice nonstarch lipids may be due in part to contaminant starch
lipids. Minor fatty acids are myristic (C14:0), stearic (C18:0), and linolenic (C18:3), plus
a trace of lauric (C12:0), 0.1-0.5% palmitoleic (C16:1), and 0.3-0.7% arachidic acids
(Lugay and Juliano, 1964). Tanaka et al (1982) found 0.4-0.5% gadoleic (C20:1),
Lipids / 169
0.2-1.4% behenic, and 0.1-0.6% lignoceric acids, whereas Izzo et al (1972) found
0.3-2.5% eicosenoic, a trace to 0.8% behenic, and a trace to 2.4% lignoceric acids.
Silicic acid fractionation of nonstarch lipids of brown rice, bran, polish, and
germ showed 86-91% neutral (nonpolar) lipids (eluted with chloroform), 2-5%
glycolipids (eluted with acetone), and 7-9% phospholipids (eluted with methanol)
(Choudhury and Juliano, 1980b) (Table 1). Again, milled-rice nonstarch lipids had a
lower neutral lipid fraction and a higher glycolipid fraction, probably from con-
tamination with starch lipids. Reported ratios for the neutral lipid, glycolipid, and
phospholipid fractions of Japanese rice nonstarch lipids are 77:10:12 (Fujino and
Mano, 1972) and 87:9:4 (Hirayama and Matsuda, 1973) for brown rice; 89:9:2 for
bran; and 76:12:12 for milled rice (Hirayama and Matsuda, 1973). Miyazawa et al
(1978) reported that bran lipids have 90.6% neutral lipids. Other reported ratios are
90:3:7 for bran lipids (Fujino and Miyazawa, 1979) and 91:4:5 for bran-polish lipids
(Ito et al, 1981a).
The fatty-acid composition of the three fractions of nonstarch lipids of rice grain
(embryo, bran, and polish) showed 22-25% palmitic, 37-41% oleic, and 37-41%
linoleic acid for neutral lipids; 25-30% palmitic, 26-34% oleic, and 35-39% lino-
leic acid for glycolipids; and 20-26% palmitic, 34-43% oleic, and 32-37% linoleic
acid for phospholipids (Choudhury and Juliano, 1980b). Fujino and Mano (1972) re-
ported the major fatty acids of nonstarch lipids of brown rice to be 24% palmitic,
36% oleic, and 34% linoleic for total lipids; 24% palmitic, 41% oleic, and 31% lino-
leic for neutral lipids; 26% palmitic, 18% oleic, and 41% linoleic for glycolipids;
and 28% palmitic, 19% oleic, and 50% linoleic for phospholipids. As expected, the
major fraction, neutral lipids, had the fatty-acid composition closest to that of the
total lipids. Glycolipids had more palmitic and less oleic acid than either neutral or
total lipids; phospholipids may have less or more linoleic acid than neutral lipids.
Taira and Itani (1988) surveyed the brown-rice fatty-acid composition of cultivars
typically grown in the United States. They found very similar levels of all individ-
ual fatty acids when comparing long-grain, medium-grain, and short-grain types.
However, short-grain waxy types had significantly higher levels of palmitic acid.
The distribution of brown-rice nonstarch lipids was 14-18% in germ, 39-41% in
bran, 15-21% in polish, and 25-33% in milled rice (12-14%) in the subaleurone
layer and 12-19% in the inner endosperm (Choudhury and Juliano, 1980b). The
distribution of the neutral lipids of brown rice was 43% in bran, 19% in germ, 15%
in polish, and 21% in milled rice. The distribution for glycolipids was 28% in bran,
7% in germ, 14% in polish, and 51% in milled rice, and, for brown-rice phosphol-
ipids, it was 30% in bran, 14% each in germ and polish, and 42% in milled rice.
Hirayama and Matsuda (1973) reported that 80% of the nonstarch lipids were in
bran-polish and 20% in milled rice; of these, 83% of the neutral lipids, 74% of the
glycolipids, and 40% of the phospholipids were in bran-polish, and the remainder
were in milled rice. Thus, the endosperm lipids were richer in polar lipids than were
the bran lipids. Shin and Godber (1996) determined that the oil extracted from rice
bran contained 20.1% total lipid, 89.2% neutral lipid, 6.8% glycolipid, and 4.0%
phospholipid.
The major neutral lipid fraction of nonstarch lipids by thin-layer chromatography
(TLC) was triglycerides, constituting 83-87% of the neutral lipids and 71-79% of
the total lipids (Choudhury and Juliano, 1980b) (Table 1). The other important neu-
tral lipid species was free fatty acids, at 4-7% of total lipids, together with diglycer-
ides, sterols, and trace monoglycerides. Fujino and Mano (1972) also found triglyc-
Lipids / 171
erides as the major neutral lipid of brown rice, followed by sterol esters, sterols, and
free fatty acids, in addition to hydrocarbon, wax, diglycerides, and monoglycerides.
The major molecular species of rice-bran triglycerides were palmitic-linoleic-oleic,
oleic-linoleic-palmitic, palmitic-linoleic-linoleic, linoleic-linoleic-palmitic, and tri-
olein for the esters of the three hydroxyl groups of glycerol (Fujino, 1982). The
fatty-acid composition of the major fraction of crude-rice-bran oil can be seen in
Table 2 (Shin and Godber, 1996). Oleic and linoleic were the major fatty acids in all
fractions. Glycolipids were higher in palmitic acid and the phospholipids had a
higher unsaturated-to-saturated fat ratio.
The major glycolipids of nonstarch lipids of brown rice and its milling fractions
were acyl sterol glycosides, sterol glycosides, and diglycosyl diglycerides, together
with ceramide glycoside, monoglycosyl diglyceride, monoglycosyl monoglyceride,
and diglycosyl monoglyceride, trace ceramide, and one unknown spot (Choudhury,
1979; Choudhury and Juliano, 1980b). Fujino and Mano (1972) reported the major
glycolipid of brown-rice nonstarch lipids to be acyl sterol glycosides, followed by
sterol glycosides and cerebroside, together with lesser amounts of glyceroglycolip-
ids, such as diglycosyl diglyceride, monoglycosyl glyceride, and sulfoquinovosyl
diglyceride (Fig. 1). Hirayama and Matsuda (1973) obtained acyl sterol glycosides,
sterol glycosides, diglycosyl diglycerides, and monoglycosyl diglycerides as major
glycolipids in nonstarch lipids of brown rice, bran, and milled rice, together with
lesser amounts of diglycosyl monoglycerides, triglycosyl diglycerides, monoglyco-
syl monoglycerides, ceramides, and sulfolipids.
Major phospholipids (Fig. 2) of nonstarch lipids of brown rice and its milling
fractions were 3-4% phosphatidylcholine (lecithin) and 3-4% phosphatidylethanol-
amine (cephalin) plus lysophosphatidylcholine (lysolecithin) and two unknown spots
(Choudhury, 1979; Choudhury and Juliano, 1980b) (Table 1). The high lysophos-
phatidylcholine content in milled rice may be contaminant starch lipids. Fujino and
Mano (1972) reported phosphatidylethanolamine, phosphatidylinositol, and phos-
phatidylcholine as principal phospholipids of brown-rice nonstarch lipids, plus
phosphatidylglycerol, diphosphatidylglycerol, phosphatidylserine, phosphatidic acid,
and lysophosphatidylcholine. Miyazawa et al (1977) also reported that phosphatidyl-
inositol was a major phospholipid of bran of waxy and nonwaxy rices, in addition to
phosphatidylcholine and phosphatidylethanolamine. Hirayama and Matsuda (1973)
found phosphatidylethanolamine and phosphatidylcholine to be the principal phos-
TABLE 2
Major Lipid Classes of Crude-Bran Oil Extracted from Raw Rice Bran
and Their Fatty-Acid Compositiona,b
Fatty-Acid Composition (%)
Lipid Satu- Unsatu-
Classc wt%d 14:0 16:0 18:0 18:1 18:2 18:3 20:0 rated rated
TL 20.1 0.40 22.21 2.21 38.85 34.58 1.14 0.61 25.43 74.57
NL 89.2 0.43 23.41 1.88 37.24 35.29 1.07 0.68 26.40 73.60
GL 6.8 0.09 27.34 0.18 36.45 35.76 0.18 ¡ 27.61 72.39
PL 4.0 0.11 22.13 0.16 38.11 39.32 0.17 ¡ 22.40 77.60
a Adapted from Shin and Godber (1996).
b Values are means of three replicate analyses.
c TL, total lipids; NL, neutral lipids (nonpolar lipid and free fatty acids); GL, glycolipids; PL, phospho-
lipids.
d TL is wt% of bran, and NL, GL, PL are wt% of TL.
Fig. 1. Monoglycosyl and oligoglycosyl diglycerides of rice lipids.
Fig. 2. Rice phospholipids.

Lipids / 173
pholipids of nonstarch lipids of brown rice, bran, and milled rice, plus phosphati-
dylglycerol, phosphatidylserine, lysophosphatidylcholine, lysophosphatidylethanol-
amine, and traces of diphosphatidylglycerol and phosphatidylinositol.
A protein-body preparation made by using macerating and cellulase enzymes on
milled rice contained 80% of the proteins and 87% of the lipids of milled rice
(Hirayama and Matsuda, 1973). The molar ratio was 12 µmol of sugar and 11 µmol
of phosphorus per 100 µeq of fatty acid ester. Major phospholipids were lysophos-
phatidylcholine and lysophosphatidylethanolamine, together with phosphatidic acid,
diphosphatidylglycerol, phosphatidylethanolamine, phosphatidylcholine, and phos-
phatidylinositol. Major glycolipids were acyl sterol glycosides, and sterol glyco-
sides, plus the four glycosyl glycerides.
Protein-body preparations made by destarching cooked milled rice with Asper-
gillus a-amylase and sieving through nylon cloth had 79-84% protein and 6-10%
lipids (Resurreccion and Juliano, 1981). The lipids consisted of 92% neutral lipids,
5% glycolipids, and 3% phospholipids. The ratio of lipid fractions was affected by
the source of a-amylase used for destarching. Mano (1982) reported neutral lipid-
glycolipid-phospholipid ratios of 66:20:14 and 74:12:14 for two protein-body prep-
arations using bacterial a-amylase. Earlier preparations using Rhizopus glucoamy-
lase had ratios of 76-79:10-12:12-13 (Tanaka et al, 1978). Major neutral lipids by
TLC were free fatty acids plus triglycerides, diglycerides, and monoglycerides
(Resurreccion and Juliano, 1981; Mano, 1982). The major phospholipids were lyso-
phosphatidylcholine and lysophosphatidylethanolamine, plus traces of phosphati-
dylcholine and phosphatidylethanolamine. Major glycolipids were sterol glycosides,
acyl sterol glycosides, diglycosyl diglycerides, monoglycosyl diglycerides, and ce-
ramide monohexosides (Mano, 1982).
The fatty-acid composition of protein-body lipids of two samples was 1% myris-
tic, 30-35% palmitic, 2% stearic, 20-26% oleic, 38-40% linoleic, and 2% linolenic
acid (Mano, 1982). The corresponding values were 32% palmitic, 24% oleic, and
42% linoleic acid for neutral lipids; 32% palmitic, 19% oleic, and 43% linoleic acid
for glycolipids; and 41% palmitic, 19% oleic, and 36% linoleic acid for phospho-
lipids (Resurreccion and Juliano, 1981).
Pepsin-treated protein bodies and fecal protein particles isolated from humans
consuming a rice diet, representing the core of the spherical protein bodies, had a
higher lipid content (18-22%) than whole protein bodies, but the lipids had lipid
species and fatty-acid compositions similar to those of lipids of whole protein
bodies (Resurreccion and Juliano, 1981).
SPECIFIC CLASSES OF LIPIDS
Ferulate Esters: “Oryzanol Fraction”
g-Oryzanol, as this class of compounds was originally designated, was thought to

be a single compound when first isolated in the early 1950s from the soap stock of
rice-oil refining (Seetharamaiah and Prabhakar, 1986). It has since been found to be
a mixture of steryl and other triterpenyl esters of ferulic acid (4-hydroxy-3-methoxy
cinnamic acid). The earliest investigations of this class of compounds classified the
major ferulate esters as oryzanol A, B, and C, which were present at 1.5-2.9% of
bran oil and had a melting point (mp) of 138.5°C (Tsuchiya et al, 1957). They con-
sisted of cycloartenyl ferulate (C40H58O4, mp 165.2-166.5°C) (Ohta and Shimizu,
1957), called oryzanol A, and 24-methylenecycloartanyl ferulate (C41H60O4, mp

164-164.5°C), termed oryzanol C (Ohta, 1960) (Fig. 3). Oryzanol B was considered
a mixture of oryzanol A and C (Shimizu and Ohta, 1960). Further work indicated
that an unsaponifiable fraction from ferulates of soap stock of rice-bran oil was
mainly the triterpenoid alcohols, cycloartenol and 29-methylene cycloartenol, and
sterol C28H48O (Kato, 1961). Other constituents of bran oil included b-sitosteryl
ferulate (C39H58O4, mp 131–131.5°C) (Tanaka et al, 1964) and methyl ferulate (mp
64°C) (Tanaka et al, 1971), the latter at 0.006-0.027% of three rice-bran oils.
More recent studies utilizing more sophisticated analytical techniques have
expanded our understanding of the composition of the “oryzanol fraction,” as it is
now designated. Rogers et al (1993) utilized reverse-phase high performance liquid
chromatography (HPLC) to separate individual ferulates in the oryzanol fraction.
They found that cycloartenyl ferulate, 24-methylene cycloartanyl ferulate, and
campesteryl ferulate (Fig. 3) were the predominant ferulates in oryzanol, along with
b-sitosteryl ferulate, stigmasteryl ferulate and cycloartanyl ferulate, which is in
basic agreement with the consensus of earlier reports. Norton (1995) found cyclo-
branyl ferulate, in addition to the ferulates found by Rogers et al (1993), and indi-
cated that the oryzanol fraction from corn was much different from that of rice bran.
Xu and Godber (1999) separated a total of 10 different ferulates from the oryzanol
fraction; they included the major ferulic acid esters of cycloartenol, 24-methylene
cycloartanol, and campesterol, along with previously identified stigmasterol and
sitosterol; as well as five that had not been previously identified in rice bran: D7-
stigmasterol, D7-campesterol, D7-sitostenol, campestanol, and sitostanol. Recently,
an even more thorough evaluation of the nonsaponifiable lipid fraction of a metha-
nol extract of rice bran has been completed (Akihisa et al, 2000). Additional infor-
Fig. 3. The three major ferulates composing the oryzanol fraction of rice oil.
Lipids / 175
mation was obtained regarding cis-trans isomerization at the ferulic acid ester link-
age and previously unidentified ferulates (including ferulic acid esters of 24-methyl-
cholesterol, 24-methylenecholesterol, and cycloeucalenol). These same researchers
(Akihisa et al, 1999) have also found the acyclic triterpene alcohol achilleol A and
the incompletely cyclized triterpene alcohol graminol A in rice-bran oil.
Oryzanol is now considered a promising nutraceutical because of evidence that it
may lower serum cholesterol and possibly protect against certain types of cancer.
Rukmini and Raghuram (1991) summarized the early research evaluating the
potential of rice-bran oil to lower serum cholesterol and suggested that the effect
was possibly due to the oryzanol component. In recent years, considerable research
has been done relative to the potential health benefits of the oryzanols, which is
covered in Chapter 19.
Because of the interest in the potential health benefits of oryzanols, additional
scrutiny is being given to factors that could affect its abundance in rice and rice
bran. That many factors can affect the levels of oryzanol components in both the
bran and oil is now well established. Little is known concerning the effects of pro-
duction variables on oryzanol levels, although a major effort supported by the
USDA (Howard, 2001) is currently underway. Lloyd et al (2000) found that there
was a significant difference between long (6.42 mg/g) and medium (5.17 mg/g)
grain rice varieties in the oryzanol concentration of the bran after the first milling
operation. This would suggest that the oryzanol fraction resides in the outer layers
of the bran fraction. They found that the total oryzanol concentration in composite
bran samples was 4.74 mg/g for long grain and 4.00 mg/g for medium grain. Norton
(1995) determined the oryzanol concentration of a rice-bran sample obtained from a
health-food store to be 3.4 mg/g.
Shin et al (1997) found that the extrusion stabilization process that is typically
used to prevent hydrolytic rancidity of rice bran can also influence the level of
oryzanol in the bran. They found that, as the temperature of extrusion increased, the
concentration of oryzanol decreased from 3.13 mg/g at an extrusion temperature of
110°C to 2.98 mg/g at an extrusion temperature of 140°C. An attempt to stabilize
rice bran using irradiation also revealed that oryzanol concentration was reduced
from 3.1 mg/g with no irradiation to 2.4 mg/g with 15 kGy irradiation (Shin and
Godber, 1996).
Hu et al (1996) found that extraction conditions could significantly affect the
efficiency of oryzanol extraction. They found that the highest extraction rate was
obtained when isopropanol preheated to 60°C was used at a 3:1 (w/w) solvent ratio,
yielding an oryzanol concentration of 2.93 mg/g. Xu and Godber (2000) found that
supercritical CO2 extraction could be used to obtain concentrated fractions of
oryzanol. When extract was collected between 15 and 20 min at 50°C and 680 atm,
the concentration of oryzanol increased to 674.6 mg/g from a level of 11.8 mg/g
with conventional hexane extraction. Kuk and Dowd (1998) found that increasing
the pressure in a supercritical CO2 extraction of rice bran from 483 to 621 atm at
80°C increased the sterol fraction (presumed to be oryzanol) from 1.5 to 3%.
Considerable variability has been noted in the oryzanol concentration of rice oil.
Rogers et al (1993) found that, in five commercially available refined rice-bran oils,
the concentration of oryzanol ranged from 0.115 to 0.787 mg/g, whereas Norton
(1995) found concentrations of 0.4 and 3.0 mg/g in rice oils from two different
commercial sources. Obviously, the refining process can have a great impact on the
oryzanol concentration. Yoon and Kim (1994) found that a crude oil with 1.61%
oryzanol was reduced to 0.77% oryzanol by refining. Xu and Godber (2000) noted
that the concentration of oryzanol in a hexane extract of rice bran was 9.8 mg/g
without saponification during extraction and 4.6 mg/g with saponification, indicat-
ing that the process of alkali refining may be responsible for the reduction of oryza-
nol concentration in refined rice oil products, as was suggested by Orthoefer (1996).
Glycosyl Glycerides
The chemical nature of the glycosyl glycerides of rice bran was established by
Fujino and Miyazawa (1979) (Table 3). Sugar configurations in mono- and diglyco-
syl monoglycerides are galactose b-(1-3) and glucose b-(1-3). The subsequent
galactose moiety is an a-(1-6) configuration (Fig. 1). The galactosyl diglycerides
showed differences in fatty-acid composition (Table 3).
Molecular species of glyceroglycolipids of rice bran were reported by Miyazawa
and Fujino (1978). The major fatty acids of sulfoquinovosyl diglycerides were 63%
palmitic acid, 17% linoleic, and 11% oleic.
Sphingolipids
The ceramides of brown rice constituted about 1% of the glycolipids. About 90%
of ceramide fatty acids were 2-hydroxy acids, and 10% were normal (Kondo et al,
1975). Lignoceric (C24:0) acid accounted for 48% of 2-hydroxy and normal acids.
The hydroxy acids had a higher than normal content of odd-numbered carbon atoms.
Nine sphingosine bases were detected, principally 38% 4-hydroxysphinganine and
25% sphinganine (Fig. 4). Thus, the principal ceramide of brown rice is 2-N-2-
hydroxy-lignoceroyl-4-hydroxysphinganine (Fig. 5).
Ceramides constituted 0.7% of the rice-bran glycolipids, and ceramide fatty
acids showed 28% normal, 70% 2-hydroxy, and 2% 2,3-dihydroxy acids (Fujino
and Ohnishi, 1976). In every class, lignoceric acid was the principal component:
47% of normal, 50% of 2-hydroxy, and 42% of dihydroxy acids. The sphingosine
bases were principally trihydroxy bases, 93% 4-hydroxysphinganine. Thus, the major
rice-bran ceramide is the same as the brown-rice ceramide (Fig. 5). Fatty acids of
ceramide hexosides were 96% 2-hydroxy acids and 4% normal acids (Fujino et
TABLE 3
Fatty-Acid and Sugar Composition (percent of total fatty acid or sugar)
of Rice Bran Mono-, Di-, Tri-, and Tetraglycosyl Glyceridesa
Monoglycosyl Diglycerides Diglycerides
Fatty Acid Monoglycosyl Galac- Glu- Di- Tri- Tetra-
or Sugar Monoglycerides Total tosyl cosyl glycosyl glycosyl glycosyl
Myristic 1.1 2.8 0.3 7.5 1.4 1.7 2.1
Palmitic 13.8 23.8 7.2 52.5 27.0 31.6 33.3
Stearic 0.8 0.4 ... 1.1 2.2 4.5 2.8
Oleic 26.4 20.3 22.5 25.5 18.7 18.0 8.9
Linoleic 34.1 34.8 45.3 13.7 28.4 35.1 47.2
Linolenic 23.8 17.9 24.7 ... 22.3 9.0 5.7
Glucose 12 36 0 100 5 7 14
Galactose 88 64 100 0 95 93 86
a Adapted from Fujino and Miyazawa (1979).
Lipids / 177
al, 1974). They consisted of 59% hydroxyarachidic, 20% hydroxylignoceric (pheno-

sinic), 10% hydroxybehenic, and 7% hydroxystearic, plus 0.2–1.0% of myristic,
palmitic, stearic, oleic, linoleic, arachidic (major), behenic, and lignoceric acids.
Ceramide mono-, di-, and trihexosides were isolated from rice-bran oil, but only the
monohexoside was characterized (Fujino and Ohnishi, 1976). Fatty acids were 11%
normal and 89% 2-hydroxy acids, and the principal acids were 31% cerotoleic
(C26:1) for normal acids and 41% 2-hydroxyarachidic for hydroxy acids. Fatty acids
containing an odd number of carbons, from 15 to 25, were detected for both fatty-
acid types. The long-chain base components were 83% sphinga-4,8-dienine isomers
(65% 4,8-trans,cis and 18% 4,8-trans,trans) and 11% 4-hydroxy-8-sphingenine.
The hexose portion was 95.8% glucose, 3.7% mannose, and 0.5% galactose. Thus,
Fig. 4. The four major chain bases of rice sphingolipids, each 18 carbon atoms long.
Fig. 5. Formulas of the major ceramide (A), monoglycosyl ceramide (cerebroside) (B), and diglycosyl
ceramide (C) found in rice lipids.
the principal ceramide monohexoside of brown rice is b-(1-1)-glucosyl-2-N-2-

hydroxyarachidoylsphinga-4,8-dienine (Fig. 5) (Fujino, 1982).
The ceramide dihexoside and trihexoside had structures similar to those of the
monohexoside and ceramide, with hydroxylignoceroyl-4-hydroxy-8-sphingenine
as the backbone, with diglycosyl as either b-(1-4)-glucosyl-b-glucosyl (Fig. 5) or b-
(1-4)-mannosyl-b-glucosyl, and with triglycosyl as either b-(1-4)-glucosyl-b-(1-4)-
mannosyl-b-D-glucosyl or b-(1-4)-mannosyl-b-(1-4)-mannosyl-b-D-glucosyl (Fujino,
1982). For the tetraglycosyl ceramide, the glycosyl group was either b-(1-4)-
glucosyl-b-(1-4)-mannosyl-b-(1-4)-mannosyl-b-D-glucosyl or b-(1-4)-mannosyl-b-
(l-4)-mannosyl-b-(1-4)-mannosyl-b-D-glucosyl (Fujino, 1982).
Sterol Lipids
The general classification of sterol components in rice has taken on greater im-
portance with the recognition that sterols of plants (phytosterols) may have health
benefits related to alterations in cholesterol metabolism (Weststrate and Meijer,
1998; Piironen et al, 2000). In a comparison of plant sterols in edible oil, Piironen
et al (2000) found that rice-bran oil had the highest concentration of sterols (com-
pared with the oils of corn, cottonseed, olive, palm, rapeseed, soybean, or sun-
flower). Sterol concentration in crude-rice-bran oil (32.25 g/kg) was up to four
times higher than that in corn oil (8.09–15.57), the oil with the second-highest
concentration.
The oryzanol component of rice has already been depicted as having a phytos-
terol component esterified to ferulic acid. Thus, quantification of sterols in rice
TABLE 4
Composition of Triterpene Alcohols and Sterols in Crude-Rice-Bran Oila and Sterol Lipidsb
Crude Sterol Free Acyl Sterol Sterol
Triterpene Alcohols or Sterols Oilc Esters Sterols Glycosides Glycosides
Triterpene alcohols (4,4-dimethylsterols)
Cycloartanol 9 2 4
Cycloartenol 41 27 20
24-Methylenecycloartanol 42 68 67
Cyclobranol <0.5 1 3
Unknown 7 2 6
4-Methyl sterols
Gramisterol 39 55 60
Citrostadienol 40 28 24
Obtusifoliol 7 10 8
Unknown 14 7 8
4-Demethyl sterols
b-Sitosterol 49 52 57 80 77
Campesterol 28 20 16 8 9
Stigmasterol 15 9 19 11 12
-5-Avenasterol 5 11 6 ... 1
-7-Stigmastenol 1 4 Trace <1 1
-7-Avenasterol(isofucosterol) 2 3 ... ... ...
Cholesterol <0.5 1 2 ... ...
a Data from Itoh et al (1973a, 1973b).
b Data from Kuroda et al (1977).
c 1.2% triterpene alcohols, 0.4% 4-methyl sterols, and 1.8% 4-demethyl sterols (Itoh et al, 1973a).
Lipids / 179
Fig. 6. Triterpene alcohols (4,4-dimethyl sterols), 4-methyl sterols, and 4-demethyl sterols of rice lipids.
would likely include sterols in the oryzanol component. Kuroda et al (1977) found
six classes of sterol lipids in rice bran: free triterpene alcohols (4,4-dimethyl ster-
ols), free 4-methyl sterols, free 4-demethyl sterols, sterol esters, acyl sterol gly-
cosides, and sterol glycosides. Quantitative amounts of each class of sterols would
be highly dependent on extraction, sample preparation, and method of quantifica-
tion. The relative composition of the four major classes is given in Table 4.
Itoh et al (1973a,b) reported on the composition of triterpene alcohols, 4-methyl
sterols, and 4-demethyl sterols in crude-rice-bran oil. Of the 4.2% unsaponifiables
in Japanese crude-rice-bran oil, 19% were less-polar compounds (hydrocarbons, ali-
phatic alcohols, etc.), 28% were triterpene alcohols, 10% were 4-methyl sterols, and
43% were sterols (Table 4) (Fig. 6). Reported values for 4-demethyl sterols were
14-33% campesterol, 3-6% stigmasterol, and 55-63% b-sitosterol (Itoh et al,
1973a). Imai et al (1974) reported 20-24% campesterol, 17-20% stigmasterol, and
56-63% b-sitosterol in four refined rice-bran oils. Melchert (1975) reported 23%
campesterol, 14% stigmasterol, 61% b-sitosterol, and 2% D5-avenasterol in rice-
bran oil sterols. Sterols of rice-hull lipids were 52% b-sitosterol, 22% campesterol,
20% stigmasterol, and 3% cholesterol, and caryopsis-surface-lipid sterols were 46%
b-sitosterol, 40% campesterol, 12% stigmasterol, and 2% cholesterol (Hartman and
Lago, 1976).
Itoh et al (1980) also reported the occurrence of 24-ethylidene sterols of the E-
configuration (fucosterol and 28-isocitrostadienol) in rice-bran oil in addition to
those of the Z-configuration (28-isofucosterol and citrostadienol) (Itoh et al, 1973a;
1973b).
Major fatty acids of sterol esters were 58% linoleic, 30% oleic, and 7% palmitic,
whereas those of acyl sterol glycosides were 42% linoleic, 30% palmitic, and 23%
oleic (Kuroda et al, 1977). Earlier, Sakata et al (1973) reported that sterol esters of
brown rice differed from free fatty acids and glyceryl esters in that they contained
0.6% pentadecanoic (C15:0), 0.3% pentadecanoleic (C15:1), 0.2% heptadecanoic
(C17:0), and 0.4% heptadecanoleic acids (C17:1).
Only D-glucose was detected in sterol glycosides and acyl sterol glycosides
(Kuroda et al, 1977). Oligoglycosylsterols of rice bran were shown recently to have
glucose as the only component sugar, with a b-(1-3)-glucopyranosyl configuration
for monoglycosyl sterols and a b-(1-4)- for the subsequent glucose moiety (Fujino
and Ohnishi, 1979). However, cellotetraosylsitosterol contained 97.5% glucose and
2.5% mannose (Ohnishi and Fujino, 1980). Component sterols in mono- to penta-
glucosyl sterols were 64-71% sitosterol, 11-19% campesterol, 10-13% stigmas-
terol, and 1-3% cholesterol, with trace amounts of D5- and D7-avenasterols and D7-
stigmastenol (Fujino and Ohnishi, 1979; Ohnishi and Fujino, 1980).
Tocopherols and Tocotrienols
Earlier studies showed that the rice tocopherols were mainly a-tocopherol and g-
tocopherol (Morrison, 1978; Ito et al, 1981b) together with g-tocotrienol (Slover,
1971) (Fig. 7). More recent evidence indicates that there is a greater abundance of
tocotrienols, especially g-tocotrienol (Rogers et al, 1993; Martin, 1994; Shin and
Godber, 1996; Shin et al, 1997) (Table 5).
Crude-rice-bran oils were found to contain (per 100 g of oil) 19-46 mg of a-
tocopherol, 1-3 mg of b-tocopherol, 1-10 mg of g-tocopherol, and 0.4-0.9 mg of d-
tocopherol, for a total of about 50 mg/100 g (Tanabe et al, 1982; Kanematsu et al,
Lipids / 181
1983), plus 14-33 mg of a-tocotrienol and 9-69 mg of g-tocotrienol per 100 g of oil
(Tanabe et al, 1981, 1982). Refined rice-bran oils contain less tocols, having 13-37
mg of a-tocopherol, 1–2 mg of b-tocopherol, 3–6 mg of g-tocopherol, and 0.1–0.5
mg of d-tocopherol (Kato et al, 1981; Tanabe et al, 1982; Kanematsu et al, 1983),
plus 10-24 mg of a-tocotrienol and 8-20 mg of g-tocotrienol per 100 g of oil (Kato
et al, 1981; Tanabe et al, 1981, 1982). The mean tocol content was 93 mg/100 g for
crude oil and 50 mg/100 g for refined oil (Tanabe et al, 1982). About 7% of the
tocols is in esterified form and the rest is in free form (Kato et al, 1981).
Ito et al (1981b), using TLC, gas liquid chromatography, and HPLC, estimated
the total tocopherol on a fresh-weight basis to be 10.3 mg/kg of brown rice, 48.0
mg/kg of rice bran, and 7.0–9.4 mg/kg of milled rice.
Hu et al (1996) found that the method of extraction affected the concentration of
tocols. Their research found that isopropanol (3:1, w/w) preheated to 60°C yielded
171 mg of total tocols per kilogram of rice bran, whereas preheated hexane (3:1,
w/w) yielded only 157 mg/kg. Shin et al (1997) found that rice-bran stabilization
and subsequent storage affect both total and individual tocols in rice bran. The total
concentration of tocols in bran stabilized by extrusion at 110°C was 304 mg/kg,
Fig. 7. Structure of tocopherols and tocotrienols.
TABLE 5
Tocopherol (T) and Tocotrienol (T3) Concentrations (mg/100 g) in Raw Rice Bran,
Theoretical Levels in Brown Rice and Crude Oil, and Actual Levels
in Commercially Available Refined Oil
Source a-T b-T K-T L-T I-T3 K-T3 L-T3
Rice brana 6.3 0.9 3.2 0.2 3.8 12.0 0.7
Brown riceb 0.63 0.09 0.32 0.02 0.38 1.20 0.07
Crude oilc 31.5 4.5 16.0 1.0 19.0 60.0 3.5
Refined oild 8.2 12.8 1.3 2.1 42.9 3.5
a Adapted from Shin and Godber (1996).
b Calculated based on 10% rice-bran yield, estimated assuming little contribution from milled rice.
c Calculated based on 20% lipid in rice bran as determined by Shin and Godber (1996).
d Data from Rogers et al (1993); average of five brands of refined rice oil.
whereas stabilization at 140°C yielded only 273 mg/kg. Furthermore, g-tocotrienol

was more stable and persisted to a greater extent during one year of storage than the
other homologues. Other factors that have been shown to influence the concentra-
tion of tocols include milling (Lloyd et al, 2000), irradiation stabilization (Shin and
Godber, 1996), and variety (Martin, 1994; Lloyd et al, 2000). Martin (1994) found
that the long-grain varieties of Maybelle and Tebonnet generally had higher levels
of total tocols (236-286 mg/kg) in their raw bran than did the medium-grain va-
rieties Mars and Rico (199-220 mg/kg). The greatest difference was for g-toco-
trienol, for which the long-grain varieties ranged from 128 to 152 mg/kg, while the
medium-grain varieties ranged from 98 to 109 mg/kg.
Lloyd et al (2000) also found that long-grain varieties had higher levels of toco-
trienols than medium-grain varieties, regardless of milling fraction.
Wax
Waxes are esters of long-chain fatty acids with fatty alcohols, methanol, and
ethanol. The nonstarch lipids of rice bran-polish, with a neutral lipid-glycolipid-
phospholipid ratio of 91:4:5, contained wax in the neutral fraction. Sterol esters,
longer alkyl esters, and shorter alkyl esters were recovered in a 9.5% yield in the
ratio 8:1:1 (Ito et al, 1981a). Fatty-acid analysis showed major acids to be linoleic,
oleic, and palmitic for sterol esters; behenic, lignoceric, and palmitic for longer
alkyl esters; and oleic and palmitic for shorter alkyl esters (Table 6). Major alcohols
were tetratriacontanol (C34:0, mainly branched), triacontanol (C30:0), dotriacontanol
(C32:0), octacosanol (C28:0), and tetracosanol (C24:0) for longer alkyl ester (Table 6)
and methanol and ethanol for shorter alkyl esters. The composition of 4-demethyl-
sterol of sterol esters was 64% sitosterol, 22% campesterol, and 9% stigmasterol,
plus smaller amounts of four others. Major sterol esters were linoleoyl sitosterol,
plus palmitoyl sitosterol and linoleoyl campesterol (Ito et al, 1983). Major longer
alkyl esters were 17% behenoyl dotriacontanol, 16% palmitoyl octacosanol, 14%
behenoyl tetratriacontanol, and 12% palmitoyl triacontanol, in decreasing order.
Major shorter alkyl esters were 43% methyl oleate, 30% methyl palmitate, and 17%
ethyl oleate.
Straight-chain alkanes and alkenes and branched-chain alkenes (squalene) were
detected in the hydrocarbon fraction (Ito et al, 1981a). Major alkanes were nonaco-
sane (C29:0) and untriacontane (C31:0), and major alkenes were nonacosene (C29:1),
untriacontene (C31:1), and tritriacontene (C33:1) (Table 6). Squalene content was re-
ported as 120 mg/100 g of rice-bran oil (Melchert, 1975).
Hard and soft waxes recovered from tank settlings of crude-rice-bran oil at 6.9
and 5.9% recovery, respectively, had an mp of 79.5 and 74°C (Yoon and Rhee,
1982). Hard wax was 64.5% fatty alcohols (C24:0, C26:0, C30:0, and C22:0), 33.5% fatty
acids (C26:0, C24:0, and C22:0), and 2% hydrocarbons (C31:0 and C29:0). Soft wax was
51.8% fatty alcohols (C24:0 and C30:0), 46.2% fatty acids (C26:0, C24:0, and C16:0), and
2% hydrocarbons (C21:0 and C29:0).
Belavadi and Bhowmick (1988) found that the composition of tank settlings was
not consistent from batch to batch and was dependent on many variables from proc-
essing to environmental. Thus, these researchers performed a controlled study using
a simulation of typical commercial handling of rice oil. They isolated the waxy com-
ponent as isopropanol-insoluble material and separated it by silicic column chroma-
tography. They found three fractions: one that eluted with petroleum-benzene (1:1)
Lipids / 183
TABLE 6
Chain Lengths (percent of total-fatty-acid chains) of Fatty Acids of Sterol and Alkyl Esters,
Alcohols of Longer Alkyl Esters, and Alkanes and Alkenes of Rice-Bran Waxy Lipidsa
Fatty Acids of
Carbon Longer Shorter Alcohols of
and Double Sterol Alkyl Alkyl Longer Alkyl
Bond Number Esters Esters Esters Esters Alkanes Alkenes
14.0 0.6 1.8 2.2 ... ... ...
16.0 11.1 23.8 35.5 ... ... ...
18.0 1.0 3.8 0.8 ... ... ...
18.1 33.1 2.9 60.2 ... ... ...
18.2 51.6 0.3 1.5 ... ... ...
18.3 2.0 ... ... ... ... ...
20.0 0.7 3.6 ... 0.1 ... ...
22.0 ... 32.6 ... 2.0 ... ...
23.0 ... ... ... ... 1.3 ...
24.0 ... 31.2 ... 11.2 0.2 ...
25.0 ... ... ... ... 2.0 ...
26.0 ... ... ... 6.3 0.8 ...
27.0 ... ... ... ... 9.5 7.9
28.0 ... ... ... 12.5 3.6 1.3
29.0 ... ... ... ... 46.5 38.8
30.0 ... ... ... 19.1 3.0 0.9
31.0 ... ... ... ... 23.7 20.7
32.0 ... ... ... 10.5 (5.1)b 1.4 0.5
33.0 ... ... ... ... 6.5 18.8
34.0 ... ... ... 6.6 (18.3)b 0.7 1.4
35.0 ... ... ... ... 0.8 8.4
36.0 ... ... ... 3.0 (5.4)b ... 0.1
37.0 ... ... ... ... ... 1.6
a Adapted from Ito et al (1981a).
b Branched.
had an mp of 70.74°C and constituted 33.9% of the total wax; one that eluted with
ethyl acetate, had an mp of 62.64°C, and constituted 13.3% of the total wax; and a
fraction that was uneluted, did not melt even at 300°C, and constituted 52.9% of the
total wax. The fatty-alcohol composition was fairly similar among the various frac-
tions. From these results, the authors reasoned that a fairly large percentage of the
wax esters exist as polymers in which an aromatic component (possibly ferulic acid)
links the fatty acids and alcohols together. A similar situation exists with carnauba
wax, to which rice wax has been compared from a functional standpoint (Belavadi
and Bhowmick, 1988).
FACTORS AFFECTING LIPID COMPOSITION
Genetically induced variability of rice lipids could be expected. Some studies

have suggested that the waxy gene had no consistent effect on the content and fatty-
acid composition of the nonstarch lipids of brown rice (Vidal and Juliano, 1967).
However, Taira et al (1981) showed a higher content of crude fat (nonstarch lipids)
for waxy than for nonwaxy brown rice for both lowland and upland varieties. Waxy
rice lipids tended to be richer in palmitic acid but poorer in oleic, and, to a lesser
extent, in linoleic acids. Similar trends were obtained for waxy mutants and non-
waxy mutants of Nihonmasari rice, for both brown and milled rices (Taira and
Hiraiwa, 1982). The higher nonstarch lipid content of waxy milled rice is consistent
with its lower content of starch lipid relative to that of nonwaxy milled rice. Taira et
al (1988) conducted an extensive survey of lipids in the primary rice groups: indica,
sinica, javanica, and japonica (Table 7). The indica and sinica groups tended to have
higher levels of palmitic and lower levels of linoleic than the javanica and japonica
groups. In a similar study, the fatty-acid composition of rice cultivars grown in the
United States appeared to be affected by parental lines of javanica or japonica types
(Taira and Itani, 1988). The sophistication of potential genetic manipulation of lipid
content can be seen in the work of Suzuki et al (1996), in which the elimination of
lipoxygenase-3 from seeds was found to improve storage stability by reducing the
tendency for lipids to become rancid by oxidative reactions.
Cultural practices could also affect the composition of lipids in rice. Because
light is such an important component of plant metabolism, Inanaga et al (1996)
theorized that light might affect the accumulation of lipid in rice seeds. They found
that the bran from rice cultured in a vinyl shelter, which was exposed to lower qual-
ity light than rice grown in a field, had higher lipid content. They suggested that the
increased lipid content was the result of stimulation of the synthesis of triglycerides
from diglycerides via the depression of synthesis of some unidentified compounds
from diglycerides.
In Japanese rice, the earlier cropping season (because of the mean daily ripen-
ing temperature of 27.1-27.8°C) resulted in a crude- fat (nonstarch lipid) content
of brown rice (2.43% at 14% moisture) of six nonwaxy varieties that was higher
than that of normal-season (24.5-26.5°C) culture (2.28% fat) and late-season
(17.0-21.3°C) culture (2.04% fat) (Taira et al, 1979). Also, with decreasing ripen-
ing temperature, oleic acid decreased progressively (41, 40, and 37% in the early-,
normal-, and late-season crops, respectively), and linoleic acid increased (37, 37,
and 40%). Oleic and linoleic acid contents of the oil were negatively correlated (r =
–0.98). Fat-by-hydrolysis (starch lipids) remained constant at 0.83-0.84% in the six
milled rices (Taira et al, 1980).
During grain development, hull lipids decreased in total amount starting at 12–16
days after flowering (DAF), mainly from decreases in glycolipids and phospholipids
per grain (Choudhury and Juliano, 1980a).
TABLE 7
Mean Lipid Content and Fatty-Acid Composition of Brown Rice of 15 Cultivars Each
of Indica, Sinica, Javanica, and Japonica Groups of Ricea
Group Indica Sinica Javanica Japonica
Lipid, % dry wt 2.52 2.65 2.63 2.51
Composition, wt of total fatty acids
Myristic 0.23 0.25 0.24 0.23
Palmitic 19.56 18.33 15.20 16.79
Palmitoleic 0.21 0.21 0.15 0.22
Stearic 1.95 2.00 1.69 1.84
Oleic 43.95 44.48 44.75 40.48
Linoleic 30.44 30.70 34.37 36.26
Linolenic 1.30 1.59 1.26 1.46
Arachidic 0.76 0.81 0.63 0.78
a Adapted from Taira et al (1988).
Lipids / 185
In the developing grain of IR42 brown rice, phospholipids and glycolipids in-
creased only up to 8 DAF, whereas neutral lipids increased up to 12-16 DAF for
nonstarch lipids (Choudhury and Juliano, 1980a). The fatty-acid accumulation in
nonstarch lipids was essentially complete by 12 DAF, with an accompanying de-
crease in linolenic acid (from 26 to 4%) and increase in oleic acid (from 14 to 28%).
Nonstarch free fatty acids did not change in composition (20-22% palmitic, 24-27%
oleic, 45-46% linoleic, and 5% linolenic) and increased fivefold between 4 and 8
DAF.
Accumulation of fat-by-hydrolysis (starch lipids) in a developing IR42 caryopsis
occurred until 20 DAF and followed closely the accumulation of nonwaxy starch
(Choudhury and Juliano, 1980a). The proportion of linolenic acid decreased and that
of linoleic acid increased until 16 DAF. These changes in lipid composition during
grain development more closely followed the earlier development of the aleurone
and embryo (which were richer in nonstarch lipids) than that of the nonwaxy
endosperm, which was richer in starch lipids.
Processing factors and storage can also influence the relative amount of lipid in
rice. With an increase in the degree of milling (from 80 to 60% milled-rice recovery
from brown rice), the changes in fatty-acid composition and content of nonstarch
lipids (extracted with diethyl ether) were greater than those in starch lipids (fat-by-
hydrolysis) (Yoshizawa et al, 1973). Palmitic acid increased and oleic acid de-
creased in the triglyceride fraction, and the ratio of saturated fat to unsaturated fatty
acids increased with increased degree of milling (Table 8). The triglyceride fraction
decreased, whereas the monoglyceride fraction and free fatty acids increased as the
degree of milling increased. Crude-fat content decreased from 0.19% in 80%-milled
rice to 0.07% in 60%-milled rice on a dry-weight basis. By contrast, fat-by-hydrolysis
in the four samples representing 80-, 75-, 70-, and 60%-milled rice contained
44-46% palmitic, 12-13% oleic, and 36-38% linoleic acid, and it remained at
0.53-0.57% of milled rice at 14% moisture.
A recent report (Begum and Bhattacharyya, 2000) indicated the surprising effect
of cooking on the fat content of several India rices, including both glutinous and
nonglutinous varieties. It found that the process of cooking rice reduced the average
fat content from 2.38 to 0.26% on a dry basis. The authors attributed this effect to
the influence of heat and water and noted that the amount of fat in the raw rice
TABLE 8
Effect of Degree of Milling on the Composition (%) of Neutral Nonstarch Lipids of Milled Ricea
Percentage Recovery from Brown Rice
Lipids 80 75 70 60
Crude fatb 0.19 0.08 0.08 0.07
Waxc 1.0 1.0 1.2 0.0
Sterolsc 12.1 13.2 9.3 10.1
Triglycerides (TG)c 49.6 47.6 38.9 30.6
Saturation ratio of TGd 0.23 0.35 0.37 0.41
Diglycerides (DG)c 3.3 1.0 1.8 1.3
Monoglyceridesc 24.7 24.5 25.8 38.7
Free fatty acidsc 9.3 12.8 23.0 15.9
a Adapted from Yoshizawa et al (1973).
b Percent of bran dry weight.
c Weight percent of crude fat.
d Saturated fatty acids divided by unsaturated fatty acids.
affected cooking time. Washing and cooking raw rice in excess water removes 69–
79% of crude fat as an emulsion (Perez et al, 1987).
Shin et al (1986) found that storage of brown rice at 35°C reduced the lipid con-
tent from 1.75 to 1.65% during 12 months of storage. This was presumed to be due
to lipid oxidation because the n-hexanal content (in the headspace of a 2-ml vial)
increased from 0.022 µg to 0.304 µg during that same period. Linoleic acid was
found to be more highly affected than oleic acid.
SUMMARY
Rice lipids£although a minor constituent of edible rice products£have consid-

erable influence on rice quality, and lipid constituents in rice by-products may have
important value as nutraceuticals. The major classifications of rice lipids include
starch and nonstarch lipids, as well as minor components such as sterols, tocols,
tocotrienols, and waxes. Starch lipids are much less abundant than nonstarch lipids,
but they affect the functionality and metabolism of starch molecules. Considerable
debate has focused on the proper classification of starch lipids because of difficul-
ties in extraction, but the consensus is that the predominant types of starch lipids are
monoacyl lipids (free fatty acids and phospholipids), although neutral and glyco-
lipids may also be present, possibly as the result of contamination from nonstarch
lipids during extraction. Nonstarch lipid, on the other hand, is the most abundant
form of lipid in rice and is found in the aleurone, subaleurone and germ of brown
rice, with a small amount in the hull of rough rice. It is composed primarily of neu-
tral lipid, with lesser amounts of glycolipids and phospholipids. Rice is one of the
richest plant sources of sterols. A specific type of phytosterol found in rice is es-
terified to ferulic acid and is referred to as “oryzanol.” It has been the focus of con-
siderable attention due to its potential health-promoting properties. Likewise, rice is
an excellent source of tocopherols and tocotrienols, which have also been touted for
their potential health-promoting properties. Rice waxes have been compared to car-
nauba wax from a functional standpoint and can be readily salvaged from crude oil
tank settlings from the refining of rice oil. Finally, many factors can affect rice lip-
ids, including genetics, production practices, and processing. The importance of the
rice lipids will continue to expand as we develop our understanding of their pres-
ence and abundance in rice and rice products.
ACKNOWLEDGMENT
Approved for publication by the director of the Louisiana Agricultural Experiment Station as manu-
script number 03-32-1388.
REFERENCES
Akihisa, T., Koike, K., Kimura, Y., Sashida, N., from rice bran and their anti-inflammatory ef-
Matsumoto, T., Ukiya, M., and Kikaido, T. fects. J. Agric. Food Chem. 48:2313-2319.
1999. Acyclic and incompletely cyclized tri- Azudin, M. N., and Morrison, N. R. 1986. Non-
terpene alcohols in the seed oils of Thaeceae starch lipid and starch lipids in milled rice. J.
and Gramineae. Lipids 34:1151-1157. Cereal Sci. 4:23-31.
Akihisa, T., Yasukawa, K., Yamaura, M., Ukiya, Begum, M., and Bhattacharyya, R. 2000. Fat
M., Kimura, Y., Shimizu, N., and Arai, K. contents in raw and cooked Assam rice varie-
2000. Triterpene alcohol and sterol ferulates ties. J. Food Sci. Technol. 37:58-60.
Lipids / 187
Belavadi, V. K., and Bhowmick, D. N. 1988. An long grain brown rice. Food Res. Int. 29:99-
investigation of rice bran oil tank settling. J. 107.
Am. Oil Chem. Soc. 65:241-245. Howard, L. 2001. Development of health benefits of
Bennett, K. E., Siebenmorgan, T. J., and functional foods derived from mid-south crops.
Mauromoustakos, A. 1993. Effects of McGill USDA-IFAFS grant, agreement number 00-
No. 2 miller settings on surface fat concentration 52102-9635. U.S. Dep. Agric., Washington, DC.
of head rice. Cereal Chem. 70:734-739. Hu, W., Wells, J. H., Shin, T.-S., and Godber, J.
Bradbury, J. H., and Collins, J. G. 1982. Lipids S. 1996. Comparison of isopropanol and hex-
of the major histological components of rice ane for extraction of vitamin E and oryzanol
studied by 13C nuclear magnetic spectroscopy. from stabilized rice bran. J. Am. Oil Chem.
Cereal Chem. 59:159-162. Soc. 73:1653-1656.
Chen, H., Siebenmorgen, T. J., and Griffin, K. Imai, C., Watanabe, H., Haga, N., and Ii, T. 1974.
1998. Quality characteristics of long grain rice Detection of adulteration of cottonseed oil by
milled in two commercial systems. Cereal gas chromatography. J. Am. Oil Chem. Soc.
Chem. 75:560-565. 51:326-330.
Choudhury, N. H. 1979. Chemical properties of Inanaga, S., Nishi, N., and Kizuka, T. 1996. The
lipids in developing and mature rice grains. effect of light on the accumulation of lipids in
Ph.D. dissertation, Univ. of Philippines, Los rice seed. Soil Sci. Plant Nutr. 42:853-858.
Baños. IRRI. 1983a. Annual Report for 1981. Int. Rice
Choudhury, N. H., and Juliano, B. O. 1980a. Lip- Research Institute, Los Baños, Laguna, Philip-
ids in developing and mature rice grain. Phyto- pines.
chemistry 19:1063-1069. IRRI. 1983b. Annual Report for 1982. Int. Rice
Choudhury, N. H., and Juliano, B. O. 1980b. Ef- Research Institute, Los Baños, Laguna, Philip-
fects of amylose content on lipids of mature pines.
rice grain. Phytochemistry 19:1385-1389. Ito, S., Sato, S., and Fujini, Y. 1979. Internal
Fujino Y. 1978. Rice lipids. Cereal Chem. 55: lipids in rice bran. Starch/Staerke 31:217-221.
559-571. Ito, S., Susuki, T., and Fujino, Y. 1981a. Wax
Fujino, Y. 1982. Fundamental studies on lipids in lipid in rice bran. Nippon Nogeikagaku Kaishi
agricultural and dairy products. Nippon Nogei- 55:247-253.
kagaku Kaishi 56:353-367. Ito, S., Tsukidate, Y., and Fujino, Y. 1981b. Sepa-
Fujino, Y., and Mano, Y. 1972. Classification of ration and determination of tocopherols homo-
lipids and composition of fatty acid and brown logues in cereal grains and bean seeds. Obihiro
rice. Eiyo To Shokuryo 25:472-474. Chikusan Daigaku Gakujutsu Kenkyu Hokoku
Fujino, Y., and Miyazawa, T. 1979. Chemical Dai-1-Bu 12:143-148.
structures of mono-, di-, tri-, and tetraglycosyl Ito, S., Susuki, T., and Fujino, Y. 1983. Wax
glycerides in rice bran. Biochim. Biophys. lipid in rice bran. Cereal Chem. 60:252-253.
Acta 572:442-451. Itoh, T., Tamura, T., and Matsumoto, T. 1973a.
Fujino, Y., and Ohnishi, M. 1976. Constituents of Sterol composition of 19 vegetable oils. J. Am.
ceramide and ceramide monohexoside in rice Oil Chem. Soc. 50:122-125.
bran. Chem. Phys. Lipids 17:275-289. Itoh, T., Tamura, T., and Matsumoto, T. 1973b.
Fujino, Y., and Ohnishi, M. 1979. Isolation and Methylsterol compositions of 19 vegetable
structure of diglycosylsterols and triglycosyl- oils. J. Am. Oil Chem. Soc. 50:300-303.
sterols in rice bran. Biochim. Biophys. Acta Itoh, T., Sakurai, S., Tamura, T., and Matsumoto,
574:94-102. T. 1980. Occurrence of 24(E)-ethylidene ster-
Fujino, Y., Sakata, S., and Nakano, M. 1974. Ce- ols in two Solanaceae seed oils and rice bran
rebroside in rice grain. J. Food Sci. 39:471- oil. Lipids 15:22-25.
473. Izzo, R., Fabbrini, F., and Lotti, G. 1972. Ricer-
Gailliard, J., and Bowler, P. 1987. Morphology che sugli olii separati dalle diverse parti delle
and composition of starch. Pages 55-78 in: cariossiodi del riso. Riso 21:161-171.
Starch: Properties and Potential, Vol. 13. T. Juliano, B. O. 1977. Rice lipids. Riso 26:3-21.
Galliard, Ed. John Wiley and Sons, New York. Juliano, B. O. 1980. Properties of the rice cary-
Hartman, L., and Lago, R. C. A. 1976. The com- opsis. Pages 403-438 in: Rice: Production and
position of lipids from rice hulls and from the Utilization. B. S. Luh, Ed. AVI Publishing
surface of rice caryopsis. J. Sci. Food Agric. Co., Westport, CT.
27:939-942. Juliano, B. O. 1983. Lipids in rice and rice proc-
Hirayama, O., and Matsuda, H. 1973. Lipid com- essing. Pages 305-330 in: Lipids in Cereal
ponents and distribution in brown rice. Nippon Technology. P. J. Barnes, Ed. Academic Press,
Nogeikagaku Kaisha 47:371-377. London.
Hoover, R., Sailaja, Y., and Sosulski, F. W. 1996. Kanematsu, H., Ushigusa, T., Maruyama, T.,
Characterization of starches from wild and Niiya, I., Fumoto, D., Toyoda, T., Kawaguchi,
Y., and Matsumoto, T. 1983. Comparison of Morrison, W. R. 1978. Cereal lipids. Pages 221-
tocopherol contents in crude and refined ed- 348 in: Advances in Cereal Science and Tech-
ible vegetable oils and fats by high perform- nology, Vol. 2. Y. Pomeranz, Ed. Am. Assoc.
ance liquid chromatography. Yukagaku 32: Cereal Chem., St. Paul, MN.
122-126. Morrison, W. R. 1981. Starch lipids: A reap-
Kato, A. 1961. Isolation of ferulates from soap- praisal. Starch/Staerke 33:408-410.
stock of rice bran oil. Yukagaku 10:679-683. Morrison, W. R. 1988. Lipids in cereal starches:
Kato, A., Tanabe, K., and Yamaoka, M. 1981. A review. J. Cereal Sci. 8:1-15.
Esterified tocopherols and tocotrienols in rice Morrison, W. R. 1995. Starch lipids and how they
bran oil, soybean oil, and sesame oil. Yuka- relate to starch granule structure and functional-
gaku 30:515-516. ity. Cereal Foods World 40:437-438, 440-446.
Kawashima, K., and Kiribuchi, T. 1980. Studies Morrison, W. R. and Azudin, M. N. 1987. Varia-
on lipid components and heat dependent past- tion in the amylose and lipid contents and
ing behavior of non-waxy and waxy rice some physical properties of rice starches. J.
starches. Kaseigaku Zasshi 31:625-628. Cereal Sci. 5:35-44.
Kitahara, K., Tanaka, T., Suganuma, T., and Morrison, W. R., Law, R. V, and Snape, C. E.
Nagahama, T. 1997. Release of bound lipids in 1993. Evidence for inclusion complexes of
cereal starches upon hydrolysis by glucoamy- lipids with a-amylose in maize, rice, and oat
lase. Cereal Chem. 74:1-6. starches. J. Cereal Sci. 18:107-109.
Kondo, Y., Nakano, M., and Fujino, Y. 1975. Norton, R. A. 1995. Quantitation of steryl feru-
Ceramide in rice grain. Agric. Biol. Chem. 39: late and p-coumarate esters from corn and rice.
719-721. Lipids 30:269-274.
Kuk, M. S, and Dowd, M. K. 1998. Supercritical Ohashi, K., Goshima, G., Kusuda, H., and Tsuge,
CO2 extraction of rice bran. J. Am. Oil Chem. H. 1980. Effect of embraced lipid on the ge-
Soc. 75:623-628. latinization of rice starch. Starch/Staerke 32:
Kuroda, N., Ohnishi, M., and Fujino, Y. 1977. 54-58.
Sterol lipids in rice bran. Cereal Chem. 54: Ohnishi, M., and Fujino, Y. 1980. Structural study
997-1006. on new sterylglycosides in rice bran: Cello-
Lloyd, B. J., Siebenmorgen, T. J., and Beers, K. tetraosylsitosterol and cellopentaosylsitosterol.
W. 2000. Effects of commercial processing on Agric. Biol. Chem. 44:333-338.
antioxidants in rice bran. Cereal Chem. 77: Ohta, G. 1960. Constituents of rice bran oil. III.
551-555. Structure of oryzanol-C. Chem. Pharm. Bull.
Lugay, J. C., and Juliano, B. O. 1964. Fatty acid 8:5-9. (In: Chem. Abstr. 55:5570, 1961)
composition of rice lipids by gas-liquid chro- Ohta, G., and Shimizu, M. 1957. Constituents of
matography. J. Am. Oil Chem. Soc. 41:273- rice bran oil. II. Structure of Oryzanol-A.
275. Chem. Pharm. Bull. 5:40-44. (In: Chem. Abstr.
Maniñgat, C. C., and Juliano, B. O. 1980. Starch 52:1190, 1958)
lipids and their effect on rice starch properties. Orthoefer, F. T. 1996. Rice bran oil: Healthy
Starch/Staerke 32:76-82. lipid source. Food Technol. 50(12):62-64.
Mano, Y. 1982. Lipids of starch and protein body Perez, C. M., Juliano, B. O., Pascual, C. G., and
in rice grain. Bull. Obihiro Ohtani Junior Col- Novenario, V. G. 1987. Extracted carbohy-
lege 19:37-44. (In Japanese) drates and lipids during washing and boiling
Martin, D. E. 1994. Extrusion stabilization and of milled rice. Starch/Staerke 39:386-390.
near-infrared analysis of rice bran. M.S. thesis, Piironen, V., Lindsay, G. D., Miettinen, T. A.,
Louisiana State Univ., Baton Rouge. Toivo, J, and Lampi, A.-M. 2000. Review—
Melchert, H. U. 1975. Lipophile Gelchroma- Plant sterols: Biosynthesis, biological function
tographie zur Isolierung von Sterinen und and their importance to human nutrition. J.
Squalen aus dem Unverseifbaren von Fetten. Sci. Food Agric. 80:939-966.
Dtsch. Lebensm. Rundsch. 71:400-402. Resurreccion, A. P., and Juliano, B. O. 1981.
Miyazawa, T., and Fujino, Y. 1978. Chemical Properties of poorly digestible fraction of
constituents and molecular species of glycero- protein bodies of cooked milled rice. Qual.
glycolipids in rice bran. Nippon Nogeikagaku Plant. Plant Foods Hum. Nutr. 31:119-128.
Kaishi 52:37-43. Rogers, E. J., Rice, S. M., Nicolosi, R. J., Carpenter,
Miyazawa, T., Yoshino, Y., and Fujino, Y. 1977. D. R., McClelland, C. A, and Romanczyk, L.
Studies on phospholipids in non-glutinous and J. 1993. Identification and quantitation of g-
glutinous rice bran. J. Sci. Food Agric. 28: oryzanol components and simultaneous assess-
889-894. ment of tocols in rice bran oil. J. Am. Oil
Miyazawa, T., Tazawa, H., and Fujino, Y. 1978. Chem. Soc. 70:301-307.
Molecular species of triglycerides in rice bran. Rukmini, C. and Roghuram, T. C. 1991. Nutri-
Cereal Chem. 55:138-145. tional and biochemical aspects of the hypol-
Lipids / 189
ipidemic action of rice bran oil: A review. J. lowland and upland irrigation cultures. Nippon
Am. Coll. Nutr. 10:593-601. Sakumotsu Gakkai Kiji 50:109-114.
Sakata, S., Mano, Y., and Fujino, Y. 1973. Com- Taira, H., Nakagahra, M., and Nagamine, T.
position of fatty acid and sterol in neutral lipid 1988. Fatty acid composition of indica, sinica,
of rice grain. Eiyo To Shokuryo 26:75-77. javanica and japonica groups of nonglutinous
Sawada, T., and Mano, Y. 1980. Combined lipid brown rice. J. Agric. Food Chem. 36:45-47.
in tropical rice starch. Bull. Obihiro Ohtani Tanabe, K., Yamaoka, M., and Kato, A. 1981.
Junior Coll. 17:31-38. (In Japanese) High performance liquid chromatography and
Schoch, T. J. 1942. Non-carbohydrate substances mass spectra of tocotrienols in rice bran oils. J.
in the cereal starches. J. Am. Chem. Soc. 64: Jpn. Oil Chem. Soc. 30:116-118.
2954-2956. Tanabe, K., Yamaoka, M., Tanaka, A., Kato, A.,
Seetharamaiah, G. S, and Prabhakar, J. V. 1986. and Amemiya, J. 1982. Determination of toco-
Oryzanol content of Indian rice bran oil and its pherols in rice bran oils by high performance
extraction from soap stock. J. Food Sci. Tech- liquid chromatography. Yukagaku 31:205-208.
nol. 23:270-273. Tanaka, A., Kato, A., and Tsuchiya, T. 1964.
Shimizu, M., and Ohta, G. 1960. Constituents of Isolation of b-sitosteryl ferulate from rice bran
rice bran oil. V. Re-examination of oryzanol- oil. Yukagaku 13:260-263.
B. Chem. Pharm. Bull. 8:108-11. Tanaka, A., Kato, A., and Tsuchiya, T. 1971.
Shin, T. S., and Godber, J. S. 1996. Changes of Isolation of methyl ferulate from rice bran oil.
endogenous antioxidants and fatty acid com- J. Am. Oil Chem. Soc. 48:95-97.
position in irradiated rice bran during storage. Tanaka, A., Resurreccion, A. P., Juliano, B. O.,
J. Agric. Food Chem. 44:567-573. and Bechtel, D. B. 1978. Properties of whole
Shin, M. G., Yoon, S. H., Rhee, J. S, and Kwon, and undigested fraction of protein bodies of
T. W. 1986. Correlation between oxidative milled rice. Agric. Biol. Chem. 42:2015-2023.
deterioration of unsaturated lipid and n-hexa- Tanaka, A., Tanabe, K., Kato, A., Hayashi, N.,
nal during storage of brown rice. J. Food Sci. and Takizawa, N. 1982. Fatty acids in Japa-
51:460-463. nese rice bran oils. Yukagaku 31:363-366.
Shin, T. S., Godber, J. S., Martin, D. E., and Tsuchiya, T., Kaneko, R., and Okubo, O. 1957.
Wells, J. H. 1997. Hydrolytic stability and Oryzanol content of rice bran oil. Tokyo Kogyo
changes in E vitamers and oryzanol of ex- Shikensho Hokoku 52:1-3.
truded rice bran during storage. J. Food Sci. Vasanthan, T, and Hoover, R. A. 1992. Com-
62:704-728. parative study of the composition of lipids as-
Slover, H. T. 1971. Tocopherols in foods and fat. sociated with starch granules from various bo-
Lipids 6:291-296. tanical sources. Food Chem. 43:19-27.
Suzuki, Y., Yasui, T., Matsukura, U., and Terao, Vidal, A. J., and Juliano, B. O. 1967. Compara-
J. 1996. Oxidative stability of bran lipids from tive composition of waxy and non-waxy rice.
rice variety [Oryza sativa (L.)] lacking lipoxy- Cereal Chem. 44:86-91.
genase-3 in seeds. J. Agric. Food Chem. 44: Westsrate, J. A., and Meijer, G. W. 1998. Plant
3479-3483. sterol-enriched margarines and reduction of
Taira, H., and Hhiraiwa, S. 1982. Lipid content plasma total and LDL-cholesterol concentra-
and fatty acid composition of brown rice and tions in normocholesterolaemic and mildly
its milled rice of glutinous mutant. Nippon hypercholesterolaemic subjects. Eur. J. Clin.
Sakumotsu Gakkai Kiji 51:159-164. Nutr. 52:334-343.
Taira, H, and Itani, T. 1988. Lipid content and Xu, Z., and Godber, J. S. 1999. Purification and
fatty acid composition of brown rice of cul- identification of components of g-oryzanol in
tivars of the United States. J. Agric. Food rice bran oil. J. Agric. Food Chem. 47:2724-
Chem. 36:460-462. 2728.
Taira, H., Taira, H., and Fujii, K. 1979. Influence Xu, Z., and Godber, J. S. 2000. Comparison of
of cropping season on lipid content and fatty supercritical fluid and solvent extraction meth-
acid composition of lowland non-glutinous ods in extracting g-oryzanol from rice bran. J.
brown rice. Nippon Sakumotsu Gakkai Kiji Am. Oil Chem. Soc. 77:547-551.
48:371-377. Yoon, S. H, and Kim, S. K. 1994. Oxidative sta-
Taira, H., Taira, H., and Fujii, K. 1980. In- bility of high-fatty acid rice bran oil at differ-
fluence of cropping season on lipid content ent stages of refining. J. Am. Oil Chem. Soc.
and fatty acid composition of rice bran and 71:227-229.
milled rice. Nippon Sakumotsu Gakkai Kiji Yoon, S. H., and Rhee., J. S. 1982. Composition
49:559-568. of waxes from crude rice bran oil. J. Am. Oil
Taira, H., Taira, H., and Ishihara, M. 1981. Lipid Chem. Soc. 59:561-563.
content and fatty acid composition of lowland Yoshizawa, K., Ishikawa, T., and Noshiro, K.
non-glutinous and glutinous brown rice by 1973. Changes of lipids in saké brewing. I.
Changes of fatty acids composition of lipids Yoshizawa, K., Hasuo, T., and Ishikawa, T. 1980.
by the change in the polishing rate of rice Location of lipids in rice grain and its move-
grains. Nippon Nogeikagaku Kaishi 47:713- ment during rice processing. Nippon Jozo
717. Kyokai Zasshi 75:432-435.
CHAPTER 8
PHYSICAL AND MECHANICAL PROPERTIES

OF RICE
Otto R. Kunze (retired)

Texas Agricultural Experiment Station
and Texas A&M University
College Station, Texas
Yubin Lan
Agricultural Engineering Technology
Fort Valley State University
Fort Valley, Georgia
Finis T. Wratten (deceased)
Agricultural Engineering Department
Louisiana State University
Baton Rouge, Louisiana
Knowledge of the physical and mechanical properties of the rice grain is used in
the planting, harvesting, drying, storing, milling, and processing of rice. Planting,
harvesting, and handling procedures and equipment use the physical characteristics
of the rice grain to make these operations effective and efficient. The rice industry
at all levels is seeking to develop new and better equipment to do specific tasks
faster and with greater accuracy.
Much research has been done with masses or bulk volumes of rice in the areas of
harvesting, drying, storing, and milling. In all of these, milling quality loss is of
great concern and of primary importance. The loss in quality results from individual
grains that break.
Mossman (1986) reviewed basic concepts in rice drying research and concluded
that papers related to the experimental approach with small samples contained some
of the most revealing (and perhaps least appreciated) information. Yet this small-
sample approach seemed “to hold the greatest immediate potential for a quantum
advance in rice drying knowledge.”
Fundamentally, more information needs to be known about the individual rice
grain. What are its physical and mechanical properties? Are the properties the same
for all the grains in a sample? How can the distinction be made between the grains that
fail (break) in the harvesting, drying, storing, and milling chain and those that do not?
Current harvesting, drying, storing, and milling procedures have evolved from
the replication of different approaches to the drying process. For example, heated-
191
air, multipass drying of high-moisture rice has been determined to be a better drying
procedure than heated-air, single-pass drying. This conclusion was established by
trying both methods. Similarly, the rice industry learned quickly that sampling for
grain milling quality should be done several days or weeks after drying rather than
immediately thereafter. Several decades ago, the wisdom of this practice became
apparent (Ban, 1971). The rice grain may fissure (crack) after it has been dried too
fast.
Physical property data on individual grains are in the process of being developed.
However, the information available is often not of the type that can be readily used
by the rice industry. Physical and mechanical property data begin to reach their po-
tential values only when they are put to use (Agnes, 1975).
Individual rice grains are of different kinds or forms. They are normally rough
rice (paddy), as harvested from the field; brown rice, after the hull is removed
(shelled rice); and milled rice, after the grains are milled (bran layers are removed).
Aside from the kinds or forms, there are predominantly three types of rice grains:
long, medium, and short. Some authors, and perhaps even commercial operators,
delineate the types by adding long slender, extra long, and medium slender classifi-
cations.
In India, rice grains are classified as superfine, fine, medium or common, and
coarse (Bhattacharya et al, 1982; Sowbhagya et al, 1984). Procedures to arrive at
these classifications have been based on arbitrary criteria. Superfine and fine grain
varieties command a premium price at the market. Efforts have been made to de-
velop classifications using physical property criteria. Numerous methods have been
proposed to separate rice grains of different sizes and shapes into appropriate classi-
fications, but apparently none have yet been adopted in the commercial market.
The bulk characteristics of rice are important and useful qualities and quantities,
but these must not detract from the important and useful characteristics of the
individual grains that constitute the mass.
The mechanical and physical properties of rice that can be utilized in the design
of harvesters, dryers, storage structures, mills, and other processing machinery need
to be identified and then quantified. However, those properties of rice that can be
most easily quantified are not necessarily the ones most needed by the rice breeder
or the design engineer. A closer interrelationship between physical-property re-
search scientists, rice breeders, and industry design engineers is highly desirable to
achieve the kind of interplay of design and research that will be of optimal benefit
to the rice industry.
The trend has been to catalog physical property data for use by researchers and
the rice industry. This type of catalog for rice could be rather extensive, and more of
the accumulated data probably would be obsolete than current because of the rapid
introduction of new varieties. It may be more realistic to generate standardized
methods for the rapid determination of physical and mechanical properties than to
generate data for a particular variety.
However, the assimilation of physical property data for a product can begin only
when this information is available in the literature. This chapter is an effort to bring
together available information about the physical and mechanical properties of rice,
with the hope that it may inspire current and future researchers to better define the
data in terms of clarity, accuracy, and usefulness. Most of the data concern rice va-
rieties that still exist but are no longer grown commercially. New varieties are now
grown for which few physical property data are available. However, data such as
Physical and Mechanical Properties / 193
size and shape of an obsolete variety of a particular grain type generally can be ap-
plied to a new variety of the same grain type. Other inherent characteristics, such as
fissure resistance of the grain, may be quite different. The goal of the rice re-
searcher should be to develop critically needed data (such as fissure resistance of a
rice variety during its development) so that the breeder can consider this informa-
tion to determine whether a new variety is worthy of being released. If a variety is
characterized after it is released, both the data and the variety may be obsolete be-
fore the information is put to use. The primary author of this chapter (O. R. Kunze)
developed moisture adsorption and fissure response data for six varieties of rice in
his Ph.D. research program. Within three years thereafter, only three varieties were
still grown commercially, and two of these became obsolete within the next three
years.
GRAIN DIMENSIONS
Dimensions of individual grains of long- and medium-grain rough rice with a

moisture range between 13.6 and 21.9% dry basis (db) and of long-grain brown rice
with a moisture range between 5.5 and 23.5% db are shown in Table 1. Dry basis is
defined as the weight of water in a sample divided by its dry matter weight before
being multiplied by 100. The values show that the grain expands in three dimen-
sions as moisture content increases. Increased grain moisture contents are associ-
TABLE 1
Physical Dimensions and Mass Properties of Rough and Brown Rice
Individual Grain Properties
Moisture Bulk Properties
Content Length Width Thickness Volume Density Area Density Porosity
Rice (% db) (mm) (mm) (mm) (mm3) (g/cm3) (mm2) (g/cm3) (%)
Rough–medium grain
(Saturn)a,b
13.6 7.90 3.12 1.96 16.06 1.327 40.19 0.599 58.5
16.3 7.92 3.12 1.96 16.71 1.340 …c 0.619 56.5
19.0 7.95 3.12 1.98 17.53 1.357 … 0.635 55.0
21.9 7.97 3.17 2.01 19.17 1.375 42.45 0.650 53.1
Rough–long grain
(Bluebonnet-50)a,b
13.6 9.68 2.59 1.90 18.36 1.365 … 0.587 59.6
16.3 9.75 2.62 1.93 18.52 1.374 … 0.589 59.3
19.0 9.86 2.64 1.93 19.17 1.380 … 0.607 57.9
21.9 10.03 2.69 1.98 19.66 1.381 … 0.616 56.9
Brown–long grain
(Starbonnet)d,e
5.5 6.85 1.92 1.58 … … … … …
10.5 7.06 1.98 1.62 11.83 1.442 31.86 0.674 53.08
14.4 7.24 2.02 1.67 12.48 1.429 33.14 0.682 52.30
18.5 7.27 2.03 1.68 13.22 1.415 34.52 0.680 51.92
23.5 7.42 2.04 1.69 14.17 1.379 36.16 0.663 51.90
a Each value is the average of five replications.
b Data from Wratten et al (1969).
c Not available.
d Each value is the average of 50 replications.
e Data from Mannapperuma (1975).
ated with increased densities for the Saturn and Bluebonnet-50 varieties but with
decreased density for Starbonnet rice. Volume and surface area increases were con-
sistently observed with moisture content increases.
Kramer (1951) measured grains from 35 varieties that included short, medium,
long, long slender, and medium slender grains as well as red rice. Measurements
were made on both rough- and brown-rice grains. He also established length-
thickness and width-thickness ratios. The coefficient of variation was used to indi-
cate the variability of a dimension or the uniformity of grains within a variety.
Van Ruiten (1979) proposed dimensions for grain classifications of extra long,
long, medium, and short for both rough and milled rice. If universally adopted, such
classifications would greatly improve communications in the international rice
trade. For brown rice, he classified the grains by length-width ratios into slender,
bold, and round types.
Bamrungwong et al (1987) give dimensions and mass data for several Asian rice
varieties, while Sowbhagya et al (1984) give dimensional properties of 16 rice va-
rieties grown in India. The latter authors also give dimensional parameters for six
varieties similar in kernel weight but different in kernel shape.
Kobayashi et al (1972) studied what they believed to be the mechanism that
caused rice kernels to fissure during drying. Moisture distributions and the drying
dimensional changes (strain) in one rice kernel were of particular interest. Their as-
sumption was that complex strains caused by uneven moisture gradients produced
the fissured grains. X-ray photographs were taken and measurements were made of
the shrinkage rate of the rice kernels, but the authors did not report the development
of fissures during drying. Does the rice grain fissure during drying? Grosh and
Milner (1959) observed the water penetration into wheat kernels and reported that
fissures developed in the hard vitreous endosperm of wheat before the water moved
through the grain. Similar fissures from moisture adsorption by the rice grain have
been reported by Kondo and Okamura (1930), Stahel (1935), Kik (1951),
Desikachar and Subrahmanyan (1961), Kunze (1964), and others. Research indi-
cates that some fissures develop during drying, but most rice grains fissure when a
low-moisture grain readsorbs moisture, or they may fissure after being dried too fast
(Crauford, 1963; Ban, 1971; Kunze, 1979). For further discussions of why the rice
grain fissures, see Chapter 9.
Later researchers in Japan (Miwa et al, 1978a) studied the movement of a
marked point and the shrinkage strain on the surface of a kernel of brown rice. X-
ray photographs were taken at regular intervals during drying. They compared
strains in the length-width, length-thickness, and width-thickness planes during
drying but reported no grains that fissured during drying. Several months later,
Miwa et al (1978b) reported that the rice kernels did not fissure during drying but
that fissures developed during the cooling process after drying. Photographs were
used to show these fissures.
Work on the dimensional changes (length, width, and thickness) of three rice va-
rieties (Bengal, Cypress, and Drew) at different harvest moisture contents was done
by Bautista et al (2000). They characterized the moisture content and size distribu-
tions of individual brown rice kernels after they were removed from a panicle and
shelled. The standard deviation of kernel sizes from a panicle decreased with de-
creasing moisture contents. There were dimensional decreases in all kernel dimen-
sions for all three varieties as the kernel moisture contents decreased. However, the
percentage of change for a given dimension was different for each variety. Bengal
(medium-grain variety) had the highest percentage of size reduction. It was fol-
lowed by the Drew variety (long-grain), and Cypress (long-grain) showed the low-
est percentage in size reduction. The changes in dimensions for a given moisture
change could be indicators of the fissuring potential of a grain during moisture ad-
sorption. Kunze (1964) found that bold grains (medium type) fissured more than
long grains for a given moisture-adsorbing environment.
Sun et al (2001) conducted a thermomechanical study of the dynamic shrinkage
of rice kernels during drying. The kernels showed a greater percentage of size re-
duction in the thickness dimension than in the length. They believed that the impact
of a combination of internal stresses during the drying process was a tensile action
along the direction of the longitudinal axis. With this reasoning, the potential fis-
sures that might develop in rice kernels during drying would be perpendicular to the
longitudinal axis. They did not specify whether these grain failures would be on the
surface or the interior of the grain. Rhind (1962) reported that the highest tensile
stresses can be expected in the driest parts of a grain (surface layers) during drying
but in the inner part of the grain during moisture adsorption. Tensile surface cracks
resulting from drying a grain of milled rice are shown in Figure 6 of Chapter 9.
MASS PROPERTIES
Density and Porosity
Values of bulk density and porosity are shown in Table 1. Bulk density increased
with moisture content for the Saturn and Bluebonnet-50 varieties (rough rice) but
showed little variation for the Starbonnet variety (brown rice). Porosity varied in-
versely with moisture content.
Coefficient of Thermal Expansion
The coefficient of cubic thermal expansion (Ck) for milled rice at storage mois-
ture was studied by Arora et al (1973) over a temperature range of 30–70°C. They
found different coefficients above and below a transition temperature of 53°C.
Ck = 2.403 × 10–4 per °C below 53°C
Ck = 3.364 × 10–4 per °C above 53°C
They concluded that a temperature difference greater than 43°C (77°F) may result
in serious cracking of the grains. S. J. Prasad, J. D. Mannaperuma, and F. T.
Wratten (unpublished data, 1975) studied the thermal expansion of brown rice at
eight levels of moisture over a range of 2.2–29.2% db and over a temperature range
of 30–70°C. They found only one coefficient for the complete temperature range
and a given grain moisture content. However, the coefficient varied nonlinearly
with moisture content and was expressed by
Ck = (0.0936 + 0.01097 M 2 + 0.000329 M 3 ) × 10–4
in which M = moisture content percentage (db).
Perdon et al (2000) used thermomechanical analyses to determine thermal volu-
metric coefficients for Bengal (medium-grain) and Cypress (long-grain) rice varie-
ties. The mean thermal volumetric coefficient for both varieties was found to be
4.62 × 10–4/°C in the rubbery (amorphous) state and 0.87 × 10–4/°C in the glassy
(crystalline) state. The temperature at which this transition occurs at a given mois-
ture content is the glass transition temperature, defined as that temperature at which
an amorphous material, at a given moisture content, changes from the rubbery to the
glassy state with decreasing temperatures. This temperature was shown to be higher
for grains at lower moisture contents and lower for grains at higher moisture con-
tents.
Coefficient of Hygroscopic Expansion
Prasad and coworkers (S. J. Prasad, J. D. Mannaperuma, and F. T. Wratten,

unpublished data, 1975) obtained a coefficient of linear hygroscopic expansion for
brown rice at room temperature by averaging the coefficients obtained by mea-
suring length, width, and thickness. The average coefficient of linear expansion was
found to be 0.00405 per percent moisture content (db). They also studied the
variation of volume at different temperatures and found that the coefficient of cubic
expansion could be expressed as
b = 0.0106 + 0.000059T
in which T = temperature (°C) and b = the coefficient of cubic hygroscopic expan-
sion per percent moisture content (db).
Static and Dynamic Coefficients of Friction
Friction coefficients of grains on various surfaces at different moisture contents

have been studied by numerous scientists. Generally, much variation is observed for
a given grain on a surface at a particular moisture content (ASAE, 1972). Specially
cleaned surfaces have a lower coefficient of friction than do uncleaned surfaces. If
surfaces are first conditioned with sliding grain, the variation is less. Most research-
ers have not defined surfaces as conditioned or unconditioned. In practice, grains
may have to slide on a surface only several times a year; such a surface would never
become conditioned. However, surfaces of loading and unloading troughs, chutes,
and spouts may readily become conditioned and have a nearly uniform coefficient
of friction. Among others, variables that influence coefficients of friction are sur-
face moisture content, grain moisture content, grain velocity, and surface tempera-
ture. These have caused such a wide variation in data that the American Society of
Agricultural Engineers has deleted grain friction coefficients from its standards in
engineering practices. This deletion does not detract from the importance of friction
coefficients, but it does imply that the use of available data should be tempered with
caution and judgment.
Static and dynamic coefficients of friction for rough rice on mild steel plates
were studied by M. Kuppuswamy and F. T. Wratten (unpublished data, 1970)
(Table 2). Both static and dynamic coefficients varied with moisture content and
were high for soaked rice. The static coefficient of friction generally decreased with
increased normal load but increased for higher-moisture grains. The dynamic
coefficient of friction generally increased with velocity (cm/sec) at both 12.6 and
21.4% moisture content (mc) (db); also the coefficients were greater for a given
velocity at the higher moisture contents.
TABLE 2
Static and Dynamic Coefficientsa of Friction of Rough Rice of the Dawn Variety on Sheet Steelb
Dynamic Friction Coefficients
Moisture Normal Static
at Speed (cm/sec)
Content Load Friction
(% db) (n) Coefficients 0.64 1.91 3.18 4.45
12.6 89 0.200 0.162 0.176 0.176 0.186
178 0.191 0.159 0.167 0.169 0.181
267 0.182 0.163 0.158 0.170 0.177
356 0.183 0.160 0.166 0.163 0.177
21.4 89 0.255 0.191 0.207 0.198 0.218
178 0.235 0.184 0.188 0.196 0.200
267 0.239 0.181 0.182 0.187 0.191
356 0.203 0.173 0.177 0.183 0.185
36.2c 89 0.302 0.300 0.306 0.306 0.253
178 0.295 0.295 0.265 0.272 0.257
267 0.277 0.310 0.276 0.267 0.275
356 0.270 0.269 0.264 0.277 0.268
a Each value is the average of five replications.
b Courtesy M. Kuppuswamy and F. T. Wratten.
c Soaked.
Resistance to Airflow
In aeration and drying systems, resistance to airflow through a rice mass is an

important consideration in determining the depth of rice to be stored, the fan to
select, and the motor size needed. Investigations by Shedd (1951) provided a
graph of the resistance to airflow of loosely packed rice at 15.5% mc (db).
However, resistance to airflow is known to vary with packing conditions of the
grain. Calderwood (1973) studied the resistance to airflow of rough, brown, and
milled rice under conditions of loose and packed fill (Table 3). Packed fills
showed 14–17% greater density than did loose fills. In all cases, the pressure drop
per meter of rice depth in the packed fills was more than double that in the loose
fills at any of the airflow rates. The greatest difference between the densities of
loose and packed fill (17%) was observed with milled rice, which also showed a
packed-fill resistance to airflow greater than 300% of that of the loose fill. These
values illustrate the sensitivity of airflow resistance to the density of the rice
pack. If an installation is designed for a loose fill but inadvertently receives a
packed fill, the resistance to a given airflow can be off by a factor of two or
more.
Angle of Repose
The angle of repose is defined by Hall (1979) as the maximum angle from the
horizontal that a bulk material assumes in a pile. The angle of repose for rice is
reported to be 36° (Hall, 1957, 1979; Richey et al, 1961; Wimberly, 1983).
Whenever the angle of repose is exceeded, the grains slide over each other to the
base of the pile until the angle of repose is again established. A pile of rice can
have an angle of repose smaller than the reported angle, but under normal
conditions, the angle should never be larger. A larger angle of repose may be
TABLE 3
Resistance to Airflow (pressure drop, in centimeters of water per meter of depth)
of Rough, Brown, and Milled Ricea
Rough Rice, Rough Rice, Brown Milled
Grain/
Test 1 Test 2 Rice Rice
Airflow (m3/sec × 103)
per m2 Loose Packed Loose Packed Loose Packed Loose Packed
of Bin Floor Area Fill Fill Fill Fill Fill Fill Fill Fill
Long grain (Belle Patna)
5.49 0.125 0.292 0.146 0.354 0.167 0.417 0.145 0.442
10.97 0.271 0.625 0.292 0.646 0.312 0.875 0.333 0.896
21.95 0.562 1.375 0.583 1.458 0.604 1.792 0.625 1.896
43.89 1.208 2.917 1.292 3.000 1.292 3.625 1.312 3.896
54.86 1.625 3.771 1.708 3.958 1.729 4.875 1.729 5.104
82.29 2.833 6.250 2.875 6.625 2.875 8.668 2.812 8.458
109.73 4.000 9.167 4.125 9.667 4.167 11.521 4.250 12.250
164.59 7.000 15.833 7.167 16.979 7.167 20.375 7.083 21.083
Fill density, kg/m3 639.1 728.8 626.3 714.4 754.5 879.4 786.5 921.1
Medium grain (Nato)
5.49 0.125 0.375 0.146 0.375 0.187 0.542 0.158 0.521
10.97 0.271 0.750 0.312 0.792 0.396 1.125 0.333 1.125
21.95 0.583 1.625 0.625 1.750 0.771 2.333 0.667 2.312
43.89 1.250 3.500 1.333 3.833 1.708 5.167 1.458 4.437
54.86 1.667 4.500 1.750 4.896 2.250 6.667 1.917 6.417
82.29 2.792 7.542 2.979 8.333 3.729 10.917 3.167 10.333
109.73 4.125 11.125 4.250 11.667 5.562 15.750 4.625 15.083
164.59 7.333 19.500 7.479 20.250 9.542 26.667 8.083 26.000
Fill density, kg/m3 640.7 733.6 623.1 722.4 808.9 935.5 821.7 962.7
a Source: Calderwood (1973); used by permission.
achieved when rough rice (paddy) has a moisture content greater than storage
moisture. Bhattacharya et al (1972) reported an angle of repose of more than 38°
for paddy at 11% mc and nearly 42° for paddy with greater than 25% mc.
Brooker et al (1974) suggest a ±5° variation from the suggested 36° to accom-
modate rough rice from 12–16% mc. Rice that is not clean can have a larger
angle of repose. The hopper at the bottom of a bin usually consists of a cone. The
cone walls must have an angle that is greater than the angle of repose for rice if
the bin is to be self-unloading. Wimberly (1983) reported that spouting for bins is
installed at a minimum of 45° to ensure proper operation. Components of rice
have angles of repose different from that of the rice itself. For example, hulls
ground to a fineness of 16–80 mesh have an angle of repose between 43 and 45°
(Houston, 1972). Factors that may cause variations in the angle of repose include
moisture content, dockage, internal friction, density, bulk density, variety, and
perhaps others. When these factors deviate from the design conditions, the system
may not function properly. A good practice would be to include a factor of safety
of several degrees in the design.
Parboiled rice has a larger angle of repose than does raw rough rice.
Bhattacharya (1985) shows that this angle is about 36° for parboiled rice with a 1%
degree of milling. The angle then increases to more than 46° for parboiled rice with
a 4% degree of milling before decreasing again to about 41° for parboiled rice with
a 7% degree of milling.
THERMAL PROPERTIES
A summary of the thermal properties of specific heat, conductivity, and diffusiv-

ity is shown in Table 4.
Specific Heat
Specific heat of swamp paddy (Gold Coast) rough rice was determined by
Disney (1954) over a moisture range of 10.1–17.5% (db). He used a modified ver-
sion of the Bunsen ice calorimeter method. Haswell (1954) used the same technique
to determine specific heat for rough, brown, and milled rice for a moisture range of
10–17.5%. Bal et al (1970) used the method of mixtures to determine specific heat
of rough rice. Hall (1980) reported the specific heat of rice within the range of 10–
38°C (50–100°F) to be 1.84 kJ/kg × K (0.44 Btu/lb × °F).
Thermal Conductivity and Diffusivity
Wratten et al (1969) determined the bulk thermal conductivity and diffusivity of

rough rice by using a transient heat transfer method. A uniform electric current was
passed through a conductor placed along the axis of a cylinder filled with rice, and
the observed temperatures of the wire at different time intervals were used in the
theoretical solution of the system to obtain the thermal conductivity. They calcu-
lated thermal diffusivity by using experimental values of thermal conductivity, spe-
cific heat, and bulk density.
Thermal properties of brown rice were studied by Mannapperuma (1975), who
also used a transient heat transfer method but made measurements on individual
grains of brown rice. Thermocouple junctions were installed in grains of rice. Tran-
sient temperatures of the grain, which was suddenly changed from one mercury
TABLE 4
Thermal Properties of Rice
Specific Heat (C) Conductivity (k) Diffusivity (a)
Rice Type (J/g °C) (W/m °C) (m2/sec)
Rough
Medium-grain
(Saturn)a C = 1.0509 + 0.03835 Mb k = 0.0894 + 0.000958 Mc a = (12.843 – 0.172 M)c × 108
Swamp paddy
(Gold Coast)d C = 1.1514 + 0.03383 M
Italian origine C = 1.1932 + 0.03257 M
Brown
Long-grain
(Starbonnet)f k = 0.10102 + 0.00308 M a = (5.397 + 0.0456 M) × 108
Italian origine C = 1.2686 + 0.02834 M
Milled
Italian origin C = 1.2477 + 0.02797 M
a Data from Wratten et al (1969).
b M = moisture, % db.
c Bulk properties.
d Data from Disney (1954).
e Data from Haswell (1954).
f Data from Mannapperuma (1975).
bath to another at a different temperature, were used to calculate thermal diffusivity.

Thermal conductivity was calculated by using experimental values of diffusivity,
specific heat, and density.
Thermal conductivity of rice during cooking and drying of cooked rice was
studied by Ramesh (2000). Thermal conductivities of granular rice starches were
studied by Fang et al (2000). Four types of granular rice starches (regular, low-
protein, waxy, and low-protein waxy) were measured at bulk densities of 600 and
800 kg/m3; temperatures of 20, 50, and 80°C; and moisture contents of 4, 20, and
40% (wet basis [wb]), by using a line-source thermal probe system equipped with a
computerized data acquisition system. Moisture content wet basis is defined as the
weight of water in a sample divided by the total sample weight (weight of water
plus the weight of dry matter) before being multiplied by 100. Low-protein and
waxy rice starches had higher thermal conductivities than did regular and low-
protein-waxy rice starches. Thermal conductivity increased significantly with in-
creases in bulk density, temperature, and moisture content.
More recently, the thermal conductivity of rice was studied in Arkansas by
Thielen et al (2001). These researchers were particularly interested in the values of
thermal conductivity of the grain, at a given moisture content, above and below the
glass transition temperature. A line heat source was used to measure the thermal
conductivity at temperatures between 3 and 69°C with grain moisture contents (wb)
between 9 and 17%. The thermal conductivity of rough rice increased with increas-
ing moisture content. For example, at 61°C and 9.2% mc (rubbery state), the aver-
age thermal conductivity value for Bengal rice was 0.112 W/(m·K); while for 17%-
mc rice at the same temperature (rubbery state), the average value was 0.126
W/(m·K). Also for a given moisture content, the thermal conductivity increased
with increasing temperature. For example, at 12.1% mc and 24°C (glassy state), the
average value of thermal conductivity was 0.102 W/(m·K); while at 61°C (rubbery
state), the average value was 0.111 W/(m·K).
Film Heat Transfer Coefficient
The film heat transfer coefficient for rough rice was calculated by Bal et al
(1970) by using an equation recommended for spheres. They obtained values rang-
ing from 57 to 85 W/cm2 over a wide range of temperatures. Mannapperuma (1975)
calculated film heat transfer coefficients for brown rice by using ellipsoidal shapes
and an empirical equation relating Nusselt’s number and a geometry index for the
range of Prandtl numbers usually encountered, as suggested by Smith et al (1971).
The coefficients as calculated increase with increased air velocity and decrease with
increasing temperature (Table 5).
Glass Transition Temperature
Researchers in Arkansas (Perdon et al, 2000) studied glass transition tempera-

tures of Bengal (medium-grain) and Cypress (long-grain) rice varieties. The physi-
cal properties of the rice grain change greatly at this temperature. Properties such as
toughness, modulus of elasticity, tensile and compressive strengths, and hardness
have a higher value in the glassy region, while the magnitude of specific heat, spe-
cific volume, and expansion coefficient are much lower. This change in physical
TABLE 5
Film Heat Transfer Coefficient for Starbonnet Brown Rice (W/m2 °C)a
Air Velocity (m/sec)
Temperature (°C) 0.5 1.0 1.5 2.0 2.5 3.0
0 30.42 44.94 56.40 66.28 75.07 83.19
50 29.90 44.15 55.38 65.09 73.77 81.71
100 29.39 43.41 54.47 64.01 72.58 80.41
150 29.00 42.73 53.68 63.04 71.44 79.16
200 28.54 42.10 52.89 62.13 70.42 78.02
250 28.09 41.48 52.09 61.23 69.40 76.83
300 27.75 40.91 51.35 60.38 68.43 75.81
a Source: Mannapperuma (1975); used by permission.
characteristics at the glass transition temperature could have a significant influence

on how the rice grain responds to drying and processing procedures.
Cnossen and Siebenmorgen (2000) explored the change of state through which
the rice grain may pass between the end of drying and before storage. The starch
granules at high temperatures, at a given moisture content, change from a rubbery
to a glassy state at the glass transition temperature as the grain cools. The
researchers found that fissuring after rapid drying can be greatly reduced or
essentially eliminated if the grain is tempered at the drying temperature before it
is cooled. Initial moisture contents of their grains ranged from 19.6–23.7%.
When, at 60°C and 50% relative humidity (rh), they removed 6.5 percentage
points of moisture from rice (Bengal long-grain variety) harvested at 21.3% mc,
there was only a 5.7% reduction in head-rice yield after the rice was tempered at
the drying temperature for 240 min. To further test the validity of the glass
transition temperature concept, the rice sample could be dried only in the glassy
crystalline region. Examples would be rice at 12% mc dried at 45°C for perhaps
10 hr or rice at 9% mc dried at 50°C for a similar period. Such drying experi-
ments would not cause the grain to pass through the glass transition temperature
during cooling and should not cause any after-drying fissures in the grain, accord-
ing to the rationale used by the Arkansas researchers.
Craufurd (1963) reported that, in fast drying of paddy, cracks do not develop
until drying has ceased. With fast drying, rough-rice grains do not have to pass
through the 15%-mc point in order to fissure. This is known because grains also
fissure after drying when rapidly dried from 13 to 12% mc.
Wasserman et al (1969) reported that high-moisture rice can be rapidly dried to
about 18% mc and then dried to storage moisture with ambient air without head-rice
yield reductions. Ban (1971) reported that high-moisture rice can be rapidly dried to
17 or 18% mc without reductions in head-rice yields. In both of the above cases,
rapid drying was stopped while the rice grain was still in the rubbery state above the
glass transition temperature.
HYGROSCOPIC PROPERTIES
Drying characteristics of grains are largely influenced by their hygroscopic prop-

erties and, to a lesser extent, by their physical, thermal, and aerodynamic character-
istics. Hygroscopic properties include grain characteristics such as equilibrium
moisture content, hygroscopic conductivity, and hygroscopic diffusivity.
Equilibrium Moisture Content
Rice is hygroscopic and, whenever it is moved from one environment to another,

the grains adsorb or desorb moisture. Under the equilibrium condition, moisture is
diffused throughout the hull, bran, and endosperm. Karon and Adams (1949)
showed that the grain components equilibrate to different moisture contents. For
example, at 25°C and 10% rh, the moisture content of rough-rice grains was 4.4%;
hulls, 3.7%; bran, 5%; polish, 5.3%; and milled rice, 5.2%. At 90% rh and the same
temperature, the rough-rice grains had a moisture content of 17.6%; hulls, 15.3%;
bran, 18.0%; polish, 18%; and milled rice, 18.8%. In a given environment, the hull
has the lowest equilibrium moisture content and the milled grain has the highest.
Different rice varieties equilibrate to slightly different moisture contents in a given
environment. Also, the same rice variety equilibrates to a slightly different moisture
content depending on whether the grain is adsorbing or desorbing moisture while
approaching the equilibrium state.
Several researchers have developed equilibrium moisture content values for rice
in its various forms. One of the most comprehensive of these for rough rice was
developed by Wratten and Kendrick (1970) (Table 6). Kunze and Hall (1965),
working with brown rice, superimposed grain equilibrium moisture contents on a
standard psychrometric chart for temperatures between 4 and 33°C. Other research-
ers (ASAE, 2000) have developed equations with which to calculate the equilibrium
moisture content of rough rice for a wide range of temperatures and relative humid-
ities. Equilibrium moisture content values from the above sources are not variety
specific and, therefore, should be used only as an approximation for moisture con-
tent in a given environment.
Siebenmorgen et al (1990) found that individual kernel moisture contents varied
within samples of equilibrated rice. They equilibrated samples at moisture contents
between 9 and 37%. The standard deviation of moisture contents among grains
varied from 0.5% for 10%-average-mc rice to 4.5% for 26%-mc rice. These obser-
vations indicate that rice grains equilibrated at 15% mc may not all be at that mois-
ture content. Some grains may be at a higher moisture content, while others may be
at a lower one. If a rice variety has a critical moisture content above which the
grains will not fissure and below which they will fissure, the observations by
Siebenmorgen et al (1990) may begin to explain why some grains fissure when ex-
posed to a given environment and others may not.
Moisture Movement into the Grain
Zhang and Kunze (1992) showed that moisture was adsorbed into the grain or
evaporated from its surface over the entire grain area and not just from sections of
it. They sealed off portions of the grain and caused the grain to fissure in the ex-
posed area that could adsorb moisture. This indicated that moisture could enter the
grain anywhere. No attempt was made to measure the rates of moisture entry at
different sections of the grain. Nagato et al (1964) determined moisture distributions
in the grain during wetting and drying by taking hardness measurements at different
points in a cross section.
The foregoing observations by different researchers suggest that moisture enters
the grain most easily through the hilar (germ) end. Thereafter, the entry rates appear
to be inversely proportional to the bran layers in which the grain is enclosed. The
greater moisture entry rate appeared to be on the ventral side (with the least bran
layers), and the smaller entry rate appeared to be on the dorsal side (with the most
bran layers). The apical (flowering) end of the grain appeared to adsorb moisture
more slowly than did the hilar (attachment) end.
Hygroscopic Conductivity and Diffusivity
During desorption, moisture is assumed to move by diffusion within the grain to

the grain surface, where it is further assumed to evaporate without any significant
surface resistance. The diffusion equation is a rapidly converging infinite series,
with a number of constants fixed by the geometry and properties of the grain. After
a certain lapse of time, all but the first term of the series become negligibly small
and the equation takes the form of
0 - 0H
= $H-NW
0L - 0H
in which M is moisture content (db) at any time, t; Mi is initial moisture content; Me

is equilibrium moisture content; k is commonly known as the drying constant and
has dimensions of time–1; and A is a dimensionless constant. Henderson and Pabis
(1961) stated that k varies with temperature according to the relationship known as
the Arrhenius equation,
k = kop–C/T
in which ko and C are constants, p = e as above, and T is absolute temperature (K) of
the drying grain.
Henderson and Pabis (1961) used Allen’s (1960) data on rough rice to verify the
Arrhenius equation. They found that the data yielded different sets of constants for
each of three relative humidity conditions. Mannapperuma (1975) also verified the
Arrhenius equation for brown rice. The equation held for a range of temperatures up
to approximately 80°C, at which point a structural change of the grain is presumed
to occur because of gelatinization of the starch. No relative humidity effect on the
equation was noted.
Mannapperuma (1975) differentiated the drying equation with respect to time
and obtained a linear equation: DM/Dt = kMe – kM. A graphic plot of data for DM/Dt
versus M produced a straight line, and the slope and intercept of this line were em-
ployed to determine the drying constant, k, and the dynamic equilibrium moisture
content, Me. (Equilibrium moisture content, as obtained here, is different from that
obtained under static conditions of equilibration.)
Plots of rate of moisture loss versus moisture content for brown rice (Fig. 1)
showed two distinct linear regions with a smooth transition from one region to
another. Mannapperuma (1975) defined these regions as the first and second
drying regions. A determination of k and Me for each region was made for drying
air temperatures of 37.8–107.2°C (100–225°F) in increments of 13.9°C (25°F).
Hygroscopic diffusivity was then calculated from the relationship of the drying
constant with diffusivity, k = GpDm/L2, in which G is a geometric factor, L is a
characteristic thickness, and Dm is the hygroscopic diffusivity. Hygroscopic
conductivity was then determined from its relationship with diffusivity, Dm =
km/pc, in which km is hygroscopic conductivity, p is density, and c is specific

moisture content. A summary of the hygroscopic properties calculated for both
regions is given in Table 7.
Fig. 1. Variation of rate of loss of moisture versus moisture content for brown rice of the Starbonnet
variety. (Reprinted, with permission, from Mannapperuma, 1975)
TABLE 7
Hygroscopic Properties of Starbonnet Brown Ricea
Conditions Dynamic Drying Hygroscopic Hygroscopic
Drying Temperature Relative Humidity EMCb Constant Diffusivity Conductivity
Region (°C) (%) (% db) (hr–1) (m2/hr) (kg/m hr)c
First 37.7 41.0 15.02 1.356 0.185 × 10–6 2.77 × 10–6
51.7 19.0 12.10 2.142 0.290 × 10–6 4.35 × 10–6
65.6 9.0 9.87 3.120 0.417 × 10–6 6.25 × 10–6
79.4 5.8 8.25 3.726 0.497 × 10–6 7.44 × 10–6
93.3 3.6 6.38 5.256 0.694 × 10–6 10.40 × 10–6
107.2 2.2 5.65 6.018 0.792 × 10–6 11.86 × 10–6
Second 37.7 41.0 12.57 0.666 0.091 × 10–6 1.36 × 10–6
51.7 19.0 9.80 1.122 0.145 × 10–6 2.16 × 10–6
65.6 9.0 7.99 1.602 0.201 × 10–6 3.01 × 10–6
79.4 5.8 6.70 2.310 0.287 × 10–6 4.28 × 10–6
93.3 3.6 4.55 2.718 0.331 × 10–6 4.95 × 10–6
107.2 2.2 3.63 3.132 0.379 × 10–6 5.67 × 10–6
a Source: Mannapperuma (1975); used by permission.
b Equilibrium moisture content.
c Expressed for a moisture gradient of 1%/m (db).
MECHANICAL PROPERTIES
Mechanical properties of rice include, among others, compressive, tensile, bending,

and hardness characteristics. Kunze and Choudhury (1972) studied the tensile strength
of Bluebelle rice (long-grain) while the grains were in equilibrium with the environ-
ment and also after they were subjected to a change of relative humidity. Experiments
were conducted with brown and milled rice. Since rice grains fissure during exposure
to a moisture-adsorbing environment, stresses develop in the grain and cause the
fissures. A fissured grain has little if any tensile strength, even though it is difficult to
produce a tensile failure in an equilibrated whole, sound grain. When exposed to a
moisture-adsorbing environment, grain tensile strength degenerates quite rapidly.
Kunze and Choudhury (1972) were able to document this degeneration, which
occurred when milled-rice grains were moved from a 44- to a 100%-rh environment
(Fig. 2). They developed an equation that closely approximated the mean fissure
response time, after exposure, when the average tensile strength of all grains was zero.
Lee and Kunze (1972) conducted similar experiments. An analysis of variance of their
data indicated that tensile strength during moisture adsorption was affected by variety,
Fig. 2. Influence of exposure time at 100% rh at 23°C on tensile strength of long-grain milled rice
(Bluebelle) previously in equilibrium with a 44%-rh environment. Each point represents 50 grains.
Broken lines indicate extrapolated data. (Reprinted, with permission, from Kunze and Choudhury,
1972)
temperature, exposure time, interaction between temperature and exposure time, and
interaction between variety and exposure time.
Temperature and moisture effects were also studied by Lee and Kunze (1972).
They investigated ultimate tensile and compressive strengths, modulus of elasticity,
and modulus of toughness of individual brown-rice grains under moisture-
equilibrated conditions involving three temperatures and four relative humidities.
Tensile Strength
For tension tests, individual grains were bonded inside the ends of two solderless
electrical connectors. Tensile force was applied by pulling on the two free
connector ends. When done carefully, the method gave an alignment sufficiently
accurate to apply a uniaxial tensile force to the specimen (Lee and Kunze, 1972).
Characteristic force deformation curves for tension tests are shown in Figure 3. The
data for tension tests (Table 8) indicate that the average breaking force, deformation,
and energy and the proportional limit all decreased as the moisture content of the rice
grains increased. Tensile strengths (St) ranged from 1,744 to 18,960 kPa (253–2,750
psi). Second-degree surface models were used to develop prediction equations for
tensile strengths. Regression analysis of average ultimate tensile stress values in pas-
cals for long-grain Bluebelle and medium-grain Nato brown rice at temperatures (T) in
degrees C and moisture contents (M) in percent (db) yielded the following equations:
Bluebelle St = 46,234.17 – 357.15 M – 2,700.54 T + 51.38 T 2
Nato St = 18,417.96 – 269.59 M – 642.87 T + 13.40 T 2
in which St is in kilopascals.
Fig. 3. Force deformation curves for Nato (medium-grain) brown rice under uniaxial tension at 20°C
and four moisture levels: (from left to right) 6.8, 12.8, 17.4, and 21.7% (db). (Reprinted, with
permission, from Lee and Kunze, 1972)
TABLE 8
Mechanical Properties of Bluebell (Long-Grain) Brown Ricea,b
Conditions
Cross-
Temper- Moisture sectional Breaking Breaking Breaking Proportional
ature Content Width Thickness Area Force Deformation Energy Limit
Properties (°C) (% db) (mm) (mm) (mm2) (N) (mm) (J) (N)
Tension
26.7 6.0 2.136 1.687 2.497 19.154 0.103 12.5 × 10–4 16.085
11.7 2.136 1.704 2.755 16.659 0.117 12.3 × 10–4 13.607
16.2 2.159 1.745 2.419 11.819 0.083 6.0 × 10–4 11.819
19.3 2.184 1.755 2.329 7.736 0.074 3.5 × 10–4 7.735
33.3 5.5 2.106 1.699 2.381 25.133 0.117 17.1 × 10–4 22.535
11.1 2.131 1.709 2.335 20.929 0.090 10.6 × 10–4 20.315
16.2 2.131 1.727 2.432 14.555 0.110 9.4 × 10–4 14.555
19.5 2.129 1.727 2.323 13.807 0.076 6.1 × 10–4 13.807
Compression
26.7 6.0 2.108 1.707 2.103 237.97 0.326 630.7 × 10–4 167.60
11.7 2.159 1.717 2.484 187.43 0.420 601.1 × 10–4 109.99
16.2 2.187 1.740 2.394 121.68 0.423 408.6 × 10–4 70.63
19.3 2.207 1.750 2.652 92.84 0.474 347.8 × 10–4 52.41
33.3 5.5 2.083 1.661 2.187 288.94 0.382 831.2 × 10–4 181.68
11.1 2.151 1.707 2.258 172.01 0.340 490.5 × 10–4 117.01
16.2 2.177 1.730 2.323 114.38 0.375 337.4 × 10–4 72.41
19.5 2.174 1.720 2.619 87.59 0.462 313.2 × 10–4 48.85
a Source: Lee and Kunze (1972); used by permission.
b Each value is the mean of 30 grains.
Compressive Strength
Core specimens were used for all compression tests. Specimens were prepared
by cutting off grain ends with a razor blade. Cut surfaces on the core were then
placed under magnification while they were lightly sanded with a high-speed elec-
trical grinder to produce smooth and parallel end planes (Lee, 1972). The average
value of the two end areas of a grain was used as the cross-sectional area of the
specimen. Length, width, and thickness dimensions of the specimens were mea-
sured. Typical force deformation diagrams for compression tests are shown in
Figure 4. When the lower portion of the curves was straight, stress and strain were
proportional. Data obtained from compression experiments with Bluebelle brown
rice are shown in Table 8. Grain compression properties were affected by variety
and the environmental condition under which the grain equilibrated. Physical
dimensions of the grain correlated highly with moisture content. Elastic limit was
directly proportional to breaking force. At a given temperature, the higher the
breaking force, the larger the proportional limit. Observed compressive strengths
(Sc) varied from 15.4 × 103 to 188.6 × 103 kPa (2,237–27,348 psi).
Temperature had no significant effect on ultimate compressive strength, as
shown in Table 9.
Prediction equations developed from second-degree surface models for compres-
sion tests were as follows:
Bluebelle Sc = 217,228.97 – 511.32 T – 13,883.28 M + 270.96 M 2
Nato Sc = 240,535.32 – 827.78 T – 15,907.58 M + 339.91 M 2
in which Sc is in kilopascals.
Fig. 4. Force deformation curves for core specimens of Nato (medium-grain) brown rice under uniaxial
compression loading at 33°C and four moisture levels: (from left to right) 5.5, 11.2, 16.3, and 19.8%
(db). (Reprinted, with permission, from Lee and Kunze, 1972)
TABLE 9
Effect of Temperature on Ultimate Compressive Strength (kPa) of Ricea
Rice Variety (Type)
Temperature (°C) Bluebelle (Long-Grain) Nato (Medium-Grain)
20.0 71,354 75,725
26.7 73,360 71,616
33.3 72,843 73,670
a Data from Lee and Kunze (1972).
Bamrungwong et al (1987, 1988) ran compressive tests on rice grains by

letting the grain assume a normal position on the base plate of the testing unit
before lowering the compression indenter on the grain. They superimposed the
load deformation curves of Lee (1972) and Yamaguchi et al (1981) on their
results and found their method to have a larger initial contact deformation before
assuming essentially the same elastic slope in the proportional range of loading.
They found that higher-moisture grains required lower breaking loads and lower
breaking energies and had a lower modulus of elasticity. Essentially the same
characteristics were found in bending tests. Tensile strengths were found to be
from 1/14 to 1/7 the compressive strengths for long-grain rice and from 1/17 to 1/10
for short-grain rice.
Modulus of Elasticity
Modulus of elasticity is the ratio of stress to strain within the proportional limit
of a specimen under tension or compression. It can be calculated from the straight-
line portion of a force deformation curve. Maa (1974) reported the compressive
modulus of elasticity to be about six times the magnitude of the tension modulus.
Lee (1972) reported compression modulus values that were five to 12 times greater
than those of the tension modulus. These variations may be attributable to grain
structure and modes of failure. During compression tests, grains often split along
their longitudinal axes from tensile failure.
Moisture content of the grain influenced the modulus of elasticity values in both
tension and compression tests more than did temperature. The highest values were
observed in low-moisture grains. The modulus of elasticity in compression was
about three times greater for grains with 6% mc (db) than for grains with 20% mc
(db). Compressive moduli of elasticity determined by several researchers are shown
in Table 10. Modulus of elasticity values specified without moisture content have
little comparative value.
Modulus of Toughness
Modulus of toughness is the amount of work per unit volume of a material re-
quired to cause the material to fail under static loading. Lee (1972) reported com-
bined values for Bluebelle long-grain and Nato medium-grain rice (Table 11). The
influence of grain moisture content on the modulus of toughness is similar to that
observed on the modulus of elasticity. As moisture content increased, the modulus
of toughness for tension and compression tests generally decreased. Temperature
effects were small and variable.
Hardness
In the rice industry, grain hardness—or perhaps more accurately, its crushing
strength—is often used to make a rough estimate of moisture content. Nagato et al
(1964) reported that the hardness at any specified point within the endosperm in-
creases or decreases linearly according to the decrease or increase of its moisture
content. They used Vickers hardness measurements (kg/mm2) and divided the
grain’s cross section (perpendicular to the long axis) into central, middle, and outer
TABLE 10
Compressive Moduli of Elasticity of Long-Grain Rice
Equilibrated Conditions EMCa Loading Rate Modulus of
Rice Variety % rh °C (db) (mm/min) Elasticity (MPa) Source
Unknown 45 27.8 13.6 0.51 2,464.24 Arora et al (1973)
Bluebelle 44 26.7 11.7 0.56 1,649.27 Lee (1972)
Bluebelle 44 26.7 12.4 0.51 1,227.80 Maa (1974)
a Equilibrium moisture content.
TABLE 11
Modulus of Toughness: Combined Values for Bluebelle Long-Grain and Nato Medium-Grain Ricea
Modulus of Toughness (J/cm3)
Moisture Content (% db) Tension Compression
6.1 23.51 × 10–2 716.50 × 10–2
11.8 18.96 × 10–2 432.51 × 10–2
16.7 12.07 × 10–2 304.68 × 10–2
20.2 7.38 × 10–2 333.15 × 10–2
a Source: Lee (1972); used by permission.
regions. One hardness reading was taken at the center, and four were taken in each
of the middle and outer regions. A graph plotting Vickers hardness values against
moisture content was developed for each area of the cross section. The graphs had a
negative slope and showed that, for a given moisture content, the middle or inter-
mediate section of the grain was the hardest, the outer region was slightly less hard,
and the central portion was the least hard. The graphs were used to determine the
moisture gradients in grains during moisture adsorption and desorption experiments.
MOISTURE ADSORPTION PROPERTIES—CONSEQUENCES
Rapid moisture adsorption causes low-moisture rice grains to crack or fissure.

This is not unknown in the rice industry, but neither is it completely understood.
Some broken grains can usually be observed in the combine hopper after threshing
and more are apparent after milling. Did these grains break because the combine
and the mill were improperly adjusted, or did they break because the grains were
previously fissured from moisture adsorption? If machines are breaking whole
grains, the machines must be studied and modified. However, if fissured grains are
breaking, the rice industry needs to know what made the rice grain fissure.
Burmistrova et al (1956) reported that cracked grains are found not only after ma-
chine threshing but can already be found on plants in windrows and even in the
standing crop. Harvesting and milling operations were believed to break the grains,
but recent research indicates that these operations simply make the damage appar-
ent. Indudhara Swamy and Bhattacharya (1980) studied the breakage of rice during
milling and concluded that grain breakage resulted primarily from grains with de-
fects. The number of broken grains after milling rarely exceeded the number of de-
fective grains before milling. The ultimate cause of rice breakage was in the grain.
The proportion of defective grains that broke was a function of grain size and shape,
the milling system, and the degree of milling.
Kondo and Okamura (1930) observed that moisture adsorption causes rice grains
to fissure. Rough rice (whole grains) at 12.6% mc was placed into a field environ-
ment at 8:00 A.M. and left exposed for 24 hr. Temperature, relative humidity, grain
moisture content, and percent of fissured grains were recorded every 2 hr. At the
end of the 24-hr period, 72% of the grains had fissured. Brown rice exposed to the
same environment showed 100% fissured grains by midnight. This work has impli-
cations for rice growers and processors today.
Similar work was conducted by Stahel (1935). Dry paddy (rough rice) was re-
moistened. This caused cracks to develop in the grains, which then broke during
milling. The critical point at which fissures began to develop in the grain from
moisture adsorption was about 14% mc. Other researchers have found that wheat
has similar characteristics (Swanson, 1943; Milner and Shellenberger, 1953; Grosh
and Milner, 1959).
Extended stalk drying in windrows has been found to abruptly increase the per-
centage of cracked grain (Burmistrova et al, 1956). Over a four-day period of wind-
row drying, cracking in the lower part increased from 11 to 78% and in the upper
part from 14.3 to 83.6%. Such information has discouraged the development of
similar harvesting procedures in the United States. Desikachar and Subrahmanyan
(1961) studied the formation of cracks in rice during wetting and its effect on the
cooking characteristics of the cereal. Cracks developed more slowly in parboiled
rice than in raw milled rice. In numerous cases, individual raw-rice grains cracked
within 2 min or less after being submersed in water. This rapid response indicates
the high sensitivity of raw milled rice to a moisture-adsorbing environment.
An extensive review of the world’s literature concerning the causes of rice
breakage during milling was made by Rhind (1962). He studied the causes for fissures
and cracks inherent within the grain rather than the mechanics of milling equipment
that might cause grain breakage. Moisture adsorption was of particular interest.
Kunze and Hall (1965) worked with six varieties and two ages of brown rice at
environmental temperatures of 3.3, 20.0, and 33.3°C (38, 68, and 92°F) and several
relative humidities. Grains of low-moisture rice were exposed to increasing relative
humidities, which caused fissures to develop. Figure 5 shows the response times for
three varieties of brown rice at the indicated temperatures and relative humidity
changes. Century Patna 231 was subjected to a vapor pressure difference of 3.2 kPa,
while Fortuna was subjected only to a 2.03-kPa change at the same temperature.
Fortuna gave essentially the same fissure response with only 62% as much vapor
pressure change. The vapor pressure change for Rexoro at 20°C was 1.49 kPa. All
three varieties gave a 98% fissured-grain response within a 6-hr period under the
respective exposure conditions.
McDonald (1967) studied the differences in moisture content of rice grains on
individual panicles. In one case, 20% of the grains on a panicle had dried below
18.5% mc and were cracked. He found fissures in grains that had moisture contents
up to 21%.
Fig. 5. Probability graph of accumulated fissured brown rice grains of long-grain varieties subjected to
various treatments. 1 = Century Patna 231 at 33.3°C, moved from 11 to 75% rh; 2 = Rexoro at 20°C,
moved from 11 to 75% rh; and 3 = Fortuna at 33.3°C, moved from 11 to 51% rh. (Reprinted, with
permission, from Kunze and Hall, 1965)
Milled rice and brown rice exhibit different rates of moisture adsorption (Kunze
and Choudhury, 1972). For a given relative humidity increase, milled rice initially
adsorbed moisture about twice as fast as brown rice (Fig. 6). Milled rice also fis-
sures within a shorter period of time than does brown rice for a given relative hu-
midity increase (Table 12). Kunze and Choudhury (1972) discuss hypothetical
stresses that may cause fissures in rice grains during moisture adsorption.
Rice researchers (Kondo and Okamura, 1930; Stahel, 1935) tried to establish the
critical moisture content above which the rice grain will not fissure when it is sud-
denly exposed to a moisture-adsorbing environment. This critical moisture content
is different for rough, brown, and milled rice of the same lot or variety. The primary
consideration is the rate at which moisture enters the grain. For rough rice, entry is
slow because the moisture must diffuse through the hull and pericarp before enter-
ing the endosperm. For brown rice, the entry is more rapid because the hull has
been removed. For milled rice, the moisture penetration is most rapid. Therefore,
when relatively high-moisture rough, brown, and milled-rice grains are submersed
in the same water, the milled-rice grains may fissure, whereas the brown and rough
rices may not. The milled-rice grain will fissure at a higher moisture content than
will brown- or rough-rice grains. The rate at which moisture enters the grain deter-
Fig. 6. Rates of moisture adsorption for brown and milled Bluebelle rice at 23.3°C (74°F) (initial rh
44%). (Reprinted, with permission, from Kunze and Choudhury, 1972)
TABLE 12
Time (min) to First Fissured Graina
85% rh, 23.3°C 100% rh, 23.3°C
Bluebelle Nato Bluebelle Nato
Long-Grain Medium-Grain Long-Grain Medium-Grain
Replicationb Brown Milled Brown Milled Brown Milled Brown Milled
1 80 26c 62 54 54 25 60 36
2 75 27 62c 44 48 24 60 32c
3 56 33 75 36c 50 21c 52c 35
4 52c 32 66 36 45c 26 56 36
5 70 30 71 58 46 21 57 34
Mean 66.6 29.6 67.2 45.6 48.6 23.4 57.0 34.6
a Source: Kunze and Choudhury (1972); used by permission.
b Fifty grains equilibrated to 23.3°C, 44% rh.
c Minimum time, minutes.
mines the critical moisture content at which the grain will still fissure. This can be
above 20% for milled rice, while for rough rice, it may be as low as 14 or 15%.
Some scientists believe that the rice grain predisposes itself to fissure by devel-
oping small crevices between cell walls during the drying process. These cannot be
seen with the naked eye but are readily visible in electron micrographs. The primary
author and his graduate students studied dozens of micrographs of fractured grain
surfaces where the grains had fissured, were cut with a razor blade, or were manu-
ally broken. The objective was to determine whether a given mode of failure had an
identifying characteristic. Crevices between cell walls of the storage-moisture
grains were often observed. But no characteristic was identified that was unique to a
mode of failure. There was no evidence in the surfaces of the cut or broken grains
that indicated the failure mode to be different from that of the surfaces of the fis-
sured grain.
Scientists have speculated that cell arrangements may be such that certain grains
are more prone to fissure than are others. The primary author’s experience is that
essentially all storage-moisture grains can be made to fissure from rapid moisture
readsorption. This can be from liquid water or from a vapor pressure difference be-
tween the grain and the surrounding air. However, any form of rice can be moved
from the storage moisture level to a higher moisture level through small-step
changes with adequate time periods so that the resulting moisture gradient is never
sufficiently large to fissure the grain.
Before Harvest
After emergence, the rice plant grows and subsequently produces tillers, which
then produce panicles of grain. The first of these tillers emerges when the seedling
is about three weeks old (Stansel, 1975), and other tillers follow for a period of
three to six weeks. Essentially the same sequence occurs during the fruiting and
maturation periods. Therefore, grains maturing in a field can have a wide variation
in moisture content. McDonald (1967) showed that the moisture content difference
between the driest and wettest grains on a panicle at harvest may exceed 10 per-
centage points. The moisture content difference among grains in a field can be
much more. Nguyen and Kunze (1982) collected the 10 driest and the 10 wettest
panicles (by appearance) from a plot of maturing rice. A sample of 10 grains was
taken from the top and from the bottom of each panicle on the day before the rice
was harvested. The driest sample showed 14.9% mc; the wettest sample showed
52%; and the field moisture content of the rice harvested the next day was 22%.
Thus, the moisture content of individual grains may vary considerably from the
measured field moisture content. When rice in the field is at harvest moisture (20–
25%), it already contains some low-moisture grains that will fissure when exposed
to a moisture-adsorbing environment. Kunze and Prasad (1978) harvested the most
mature panicles from a plot and then hand-shelled and inspected what appeared to
be the driest grains. Out of 60 selected grains, 14 were fissured. The field moisture
content of the rice was 29.4%. The damaged grains apparently had dried during a
previous day to a moisture content sufficiently low so that fissures developed when
the grains readsorbed moisture at night.
During Harvest
Immediately after threshing, the high-, low-, and intermediate-moisture rice

grains exist in a mixture, and it is from this that the field moisture content is deter-
mined. According to Wratten and Kendrick (1970), rough rice at 20% mc and
26.7°C (80°F) produces an interstice relative humidity of 94.6%, and 25%-moisture
rice produces an interstice relative humidity of 98.8%. Rice at 14% mc at the same
temperature is in equilibrium with 75.6% rh. Thus, the difference between the equi-
librium relative humidities of the low-moisture grains and the field-moisture grains
is at least 20%. The field-moisture rice produces the high relative humidity that
provides a moisture-adsorbing environment for the low-moisture rice. The percent-
age or proportion of low-moisture grains in the harvested mass varies with the field
moisture content. When the field moisture content is high, few low-moisture grains
are in the mixture. As the field moisture content decreases, more low-moisture
grains exist. During the ripening period, the ratio of low-moisture grains to high-
moisture grains in a field changes from day to day and even within a day. Kunze
and Hall (1965) were able to move brown rice through 10% increments of increas-
ing relative humidity in 24-hr periods without producing fissures in the grains.
When increments greater than 10% were used, grains began to fissure.
Moisture-adsorbing environments that cause low-moisture grains to fissure can
be produced in the combine hopper, field cart, transport truck, or holding bin before
the rice is dried. Thus, one potential source of fissuring in freshly harvested rice
could be avoided if small volumes of low-moisture rice could be separated from
large volumes of high-moisture rice.
During Drying
Column or deep-bed heated-air dryers can provide moisture-adsorbing environ-

ments that produce fissures in low-moisture grains found in freshly harvested rice.
As heated air enters a bed of field-moisture rice, it picks up moisture from the wet
grains and becomes humid and warm. Most of this exchange of heat and mass oc-
curs as the air passes through a few centimeters of the grain bed. This area of heat
and mass exchange in the grain bed is commonly known as the drying front. Air
that has passed through the front is humid and warm, whereas air that is approach-
ing the front is relatively dry and hot.
Attention should be focused on what may happen to low-moisture grains ahead

of the drying front. These grains are in a moisture-adsorbing environment and will
adsorb moisture until the drying front reaches them. If the moisture adsorption pe-
riod is sufficiently long or if the drying bed is sufficiently deep, the grains will fis-
sure before they start to dry. Thus, the damage is due to moisture adsorption but
occurs during the drying process. The fissures develop while low-moisture grains
are in the dryer not because the grains are drying, but because they are adsorbing
moisture. This is a subtle distinction, and researchers need to separate damage that
results because of drying from damage that results during drying because of mois-
ture adsorption. The data of Mannapperuma (1975) and Prasad’s group (S. J.
Prasad, J. D. Mannaperuma, and F. T. Wratten, unpublished data, 1975) show that
the grain expands when moisture is adsorbed. Alternatively, the grain contracts
when it dries. Using this rationale, it is difficult to apply the principles of mechanics
to explain the development of a fissure (large internal crack) while a grain is being
dried to storage moisture.
Kunze and Prasad (1978) mixed low-moisture test grains of milled, brown, and
rough rice with much larger volumes of high-moisture rough rice. These experi-
ments showed that milled rice fissures more readily than does brown rice and that
brown rice fissures more readily than rough rice in a given moisture-adsorbing envi-
ronment.
This is a useful observation for the rice industry. To get some measure of the fis-
suring potential in freshly combined rice, the producer can shell a few rice grains at
storage moisture and bury them in the harvested mass for 24 or 48 hr. If the grains
are not fissured when they are retrieved, the fissuring potential in the harvested
mass is small. If most of the brown-rice grains are fissured, then the mass probably
contains enough moisture potential to fissure some low-moisture rough-rice grains.
If the brown-rice test grains did not fissure, some milled-rice grains at storage
moisture can be used with the same procedure to measure an even smaller moisture
potential. Similar applications can be made by dryer or elevator operators who may
want to mix small volumes of low-moisture grain with larger volumes of high-
moisture grain.
Other experiments were run by Kunze and Prasad (1978), in which they placed
low-moisture grains (brown and rough rice) on top of a bed of high-moisture rice
and then dried the grains with heated air moving in an upward direction. Data from
these experiments are shown in Table 13. Even if the high-moisture grains had only
4 or 5% more moisture than the low-moisture test grains, moisture adsorption was
already sufficient to cause some low-moisture test grains to fissure. When the dif-
ference was between 6 and 7%, nearly 100% of the brown-rice test grains fissured.
Data in the table also show that brown rice is more sensitive to a moisture-adsorbing
environment than is rough rice.
The foregoing experiments simulated freshly harvested rice, with both high-
and low-moisture grains being dried in a continuous-flow column dryer. Low-
moisture grains on the air-exhaust side of the grain column could adsorb moisture
and fissure before the drying front reached them. Turning vanes in the column
dryer are used to reduce such grain damage. Mixing-type continuous-flow dryers
alternately expose low-moisture grains to dry hot air and humid warm air, thus
preventing exposure of low-moisture grains to a continuous moisture-adsorbing
environment.
TABLE 13
Fissures When Low-Moisturea Rice Was Dried on Top of High-Moisture Riceb
Conditions Grains Fissured
Moisture Content of
at Column Top
High-Moisture Rice Temperature Drying Time Column Depth
Variety (% wb) (°C) (hr) (cm) Brown Rough
Long-grain
Lebonnet 14.5 59.4 2.25 30.5c 8 4
15.0 59.4 2.50 40.6d 46 9
28.0 57.8 5.50 40.6 50 50
Medium-grain
Nato 16.8 59.4 3.25 40.6 49 12
22.0 59.4 5.25 40.6 50 39
25.3 58.9 5.75 40.6 50 45
28.0 58.9 6.25 40.6 50 50
Brazos 20.5 56.7 4.75 40.6 50 46
32.0 58.9 5.75 30.5 50 50
a Low-moisture rice = 10–12% moisture content, 50 grains per sample.
b Source: Kunze and Prasad (1978); used by permission.
c 30.5 cm = 12 in.
d 40.6 cm = 16 in.
After Drying
Rice that has been dried to storage moisture (12.5%) either by heated or un-
heated air is subject to fissure when exposed to a moisture-adsorbing environment.
Immediately after drying with heated air, the hull is at a lower moisture content than
the rest of the grain. In most cases, it can readily readsorb moisture from the envi-
ronment. Tempering with aeration after heated-air drying is used to further dry the
grain and to provide a gradual, controlled transition to the storage environment.
Moisture migrating to the grain surface is removed. This prevents moisture buildup
at the surface, which could expand the grain and cause it to fissure.
Cnossen et al (1999) worked with the glass transition temperature concept and
found that rice will not fissure after rapid drying if it is tempered, without
aeration, at the drying temperature. Five to six percentage points of moisture
were removed in one pass with 60°C drying air. When such samples were
tempered at the drying air temperature for 160 min, no head-rice yield was lost.
The milling quality of the rapidly dried sample remained the same as that of the
control sample. Required tempering durations were shorter for long-grain rice
than for medium-grain rice.
The finding of Cnossen et al (1999) has promise, since it is an extension of what
has been observed before. K. Kato and R. Yamashita (unpublished data, 1979)
reduced the fissure development in grains after drying by temporarily storing grains
at 60°C. Nguyen and Kunze (1984) reported that an 11% rh and 45°C storage
environment engendered fewer fissured grains after they were rapidly dried than did
other storage environments that had less tendency to remove surface moisture from
the grains.
Moisture-adsorption environments for rice at storage moisture may exist in the
tops of storage bins, in aeration air, in ambient air during transport, in loading and
unloading, and even in milling operations. The storage-moisture grain is hygro-
scopic and adsorbs moisture whenever there is an opportunity.
WEIGHT–VOLUME RELATIONSHIPS
Rice is traded and sold on a weight basis, but quantities often are expressed on a
volume basis. Assuming that the nominal test weight is 57.664 kg/hL (45 lb/bushel
[bu]), the following relationships are useful in converting from one unit of measure
to another.
1 m3 = 35.31 ft3 = 28.25 bu = 7.84 barrel [bbl] = 1,271.28 lb
= 576.64 kg = 0.577 metric ton [t] = 12.71 bags
1 t = 1,000 kg = 2,200 lb = 22 bags = 48.89 bu = 13.58 bbl
1 bu = 20.43 kg = 45 lb = 0.278 bbl
1 bbl = 73.46 kg = 162 lb = 3.60 bu
SUMMARY AND CONCLUSIONS
Physical properties of the rice grain are becoming increasingly important in the
industry. Appropriate basic information could make field research more productive.
Researchers need to make their data more relevant to the needs of the industry. For
instance, what are the characteristics that the design engineer should or could use to
separate high-moisture from low-moisture grains during harvesting? Available in-
formation is inadequate to provide a satisfactory answer. In another case, current
information indicates that low-moisture grains may fissure in the field, yet some
grains are more resistant to fissures than others. For example, only half of the grains
exposed to a moisture-adsorbing environment may fissure. Which are the grains that
don’t fissure? How can they be characterized? Are they grains from early- or late-
maturing panicles, or are they early- or late-maturing grains from the same panicle?
How does harvest maturity influence the potential for a grain to fissure? These
questions need well-defined answers. Physical and perhaps chemical property data
are needed to characterize fissure-resistant grains.
The study of the physical properties of the rice grain is in its infancy. It must be
expanded and its existing data confirmed and further redefined. Design engineers
must be able to use the data with confidence. Additional research emphasis on the
physical properties of rice can supply the needed inputs to make the industry more
efficient and productive through the development of new concepts for harvesting,
drying, and processing the grain.
REFERENCES
Agnes, J. B. 1975. Deere’s needs for mechanical thermal and mechanical properties. Trans.
properties data. Pages 80-82 in: Proc. Work- ASAE (Am. Soc. Agric. Eng.) 16:320-323,
shop on Design Applications of Mechanical 327.
Properties of Solid Food Materials. Dep. ASAE. 1972. Yearbook. American Society of
Agric. Eng., Penn State University, University Agricultural Engineers, St. Joseph, MI.
Park. ASAE. 2000. Standards. American Society of
Allen, J. R. 1960. Application of grain drying Agricultural Engineers, St. Joseph, MI.
theory to the drying of maize and rice. J. Bal, S., Wratten, F. T., Chesness, J. L., and
Agric. Eng. Res. 5:363-385. Faulkner, M. D. 1970. An analytical and ex-
Arora, V. K., Henderson, S. M., and Burkhardt, perimental study of radiant heating of rice
A. M. 1973. Rice drying cracking versus grain. Trans. ASAE 13:644-647, 652.
Bamrungwong, S., Satake, T., Vargas, D., and wetting and its effect on the cooking charac-
Yoshizaki, S. 1987. Fundamental studies on teristics of the cereal. Cereal Chem. 38:356-
mechanical properties of long grain rice varie- 364.
ties: I. Compressive properties of long grain Disney, R. W. 1954. The specific heat of some
rice. Jpn. J. Trop. Agric. 31:232-240. cereal grains. Cereal Chem. 31:229-239.
Bamrungwong, S., Satake, T., Vargas, D., and Fang, Q., Lan, Y., Kocher, M., and Hanna, M.
Yoshizaki, S. 1988. Fundamental studies on 2000. Thermal conductivity of granular rice
mechanical properties of long grain rice varie- starches. Int. J. Food Prop. 3:283-293.
ties: II. Bending properties of long grain rice. Grosh, G. M., and Milner, M. 1959. Water pene-
Jpn. J. Trop. Agric. 32:6-15. tration and internal cracking in tempered
Ban, T. 1971. Rice cracking in high rate drying. wheat grains. Cereal Chem. 36:260-273.
Jpn. Agric. Res. Q. 6:113-116. Hall, C. W. 1957. Drying Farm Crops. Agricul-
Bautista, R. C., Siebenmorgen, J., and Cnossen, tural Consulting Associates. (Available from
A. G. 2000. Characteristics of rice individual Edwards Bros., Inc., Ann Arbor, MI)
kernel moisture content and size distribution at Hall, C. W. 1979. Dictionary of Drying. Marcel
harvest and during drying. Paper 325 in: Proc. Dekker, Inc., New York.
Int. Drying Symp. (IDS2000), Elsevier Sci- Hall, C. W. 1980. Drying and Storage of Agri-
ence, Amsterdam. cultural Crops. AVI Publishing Co., Inc.,
Bhattacharya, K. R. 1985. Parboiling of rice. Westport, CT.
Pages 289-348 in: Rice: Chemistry and Tech- Haswell, G. A. 1954. A note on the specific heat
nology. 2nd ed. B. O. Juliano, Ed. Am. Assoc. of rice, oats, and their products. Cereal Chem.
Cereal Chem., St. Paul, MN. 31:341-342.
Bhattacharya, K. R., Sowbhagya, C. M., and Henderson, S. M., and Pabis, S. 1961. Grain
Indudhara Swamy, Y. M. 1972. Some physical drying theory. I. Temperature effect on drying
properties of paddy and rice and their interre- coefficient. J. Agric. Eng. Res. 6:169-174.
lations. J. Sci. Food Agric. 23:171-186. Hogan, J. T., and Karon, M. L. 1955. Hygro-
Bhattacharya, K. R., Ramesh, B. S., and scopic equilibria of rough rice at elevated
Sowbhagya, C. M. 1982. Dimensional classifi- temperatures. J. Agric. Food Chem. 3:855-
cations of rice for marketing. J. Agric. Eng. 860.
(India) 19(4):69-72. Houston, D. F. 1972. Rice hulls. Pages 301-352
Brooker, D. B., Bakker-Arkema, F. W., and Hall, in: Rice: Chemistry and Technology, 1st ed.
C. W. 1974. Drying Cereal Grains. AVI Pub- D. F. Houston, Ed. Am. Assoc. Cereal Chem.,
lishing Company, Inc., Westport, CT. St. Paul, MN.
Burmistrova, M. F., Komolkova, I. K., Klemm, Indudhara Swamy, Y. M., and Bhattacharya, K.
N. V., Panina, M. T., Polunochev, I. M., R. 1980. Breakage of rice during milling—
Pyankov, A. I., Solokov, A. F., Tetyanko, N. Effect of kernel defects and grain dimensions.
G., Chaus, V. M., and Eglit, E. G. 1956. J. Food Process Eng. 3:29-42.
Physicomechanical properties of agricultural Karon, M. L., and Adams, M. E. 1949. Hygro-
crops. Translated from Russian. Available scopic equilibrium of rice and rice fractions.
from Office of Technical Services, U.S. De- Cereal Chem. 26:1-12.
partment of Commerce, Washington, DC. Kik, M. C. 1951. Nutritive studies on rice. Agric.
Calderwood, D. L. 1973. Resistance to airflow of Exp. Stn. Bull. 508. University of Arkansas,
rough, brown and milled rice. Trans. ASAE College of Agriculture, Fayetteville.
16:525-527, 532. Kobayashi, H., Miwa, Y., and Hayakawa, K.
Cnossen, A. G., and Siebenmorgen, T. J. 2000. 1972. On the mechanism of the cracking in
The glass transition temperature concept in rice kernels during drying of paddies, particu-
rice drying and tempering: Effect on milling larly moisture distribution and drying strain in
quality. Trans. ASAE 43:1661-1667. one rice kernel. Res. Bull. Fac. Agric. Gifu
Cnossen, A. G., Siebenmorgen, T. J., Reid, J. D., Univ. 33:279-293.
and Yang, W. 1999. Incorporating the glass Kondo, M., and Okamura, T. 1930. Der Durch
transition temperature in rice drying and tem- die feuchtigkeit’s zunahme verursachte
pering to optimize moisture removal rate and querris (Doware) des reiskornes. Ber. Ohara
milling quality. ASAE paper 99-6022. Ameri- Inst. Landwirtsch. Forsch. Kurashiki 4:429-
can Society of Agricultural Engineers, St. Jo- 446.
seph, MI. Kramer, H. A. 1951. Physical dimensions of rice.
Crauford, R. Q. 1963. Sorption and desorption in Agric. Eng. 32:544-545.
raw and parboiled paddy. J. Sci. Food Agric. Kunze, O. R. 1964. Environmental conditions
14:744-750. and physical properties which produce fissures
Desikachar, H. S. R., and Subrahmanyan, V. in rice. Ph.D. thesis. Michigan State Univer-
1961. The formation of cracks in rice during sity, East Lansing.
Kunze, O. R. 1979. Fissuring of the rice grain Nguyen, C. N., and Kunze, O. R. 1984. Fissures
after heated air drying. Trans. ASAE 22:1197- related to post-drying treatments in rough rice.
1207. Cereal Chem. 61:63-68.
Kunze, O. R., and Choudhury, M. S. U. 1972. Perdon, A., Siebenmorgen, T. J., and
Moisture adsorption related to the tensile Mauromoustakos, A. 2000. Glassy state tran-
strength of rice. Cereal Chem. 49:684-696. sition and rice drying: Development of a
Kunze, O. R., and Hall, C. W. 1965. Relative brown rice state diagram. Cereal Chem.
humidity changes that cause brown rice to 77:708-713.
crack. Trans. ASAE 8:396-399. Ramesh, M. N. 2000. Effect of cooking and dry-
Kunze. O. R., and Prasad, S. 1978. Grain fissuring on the thermal conductivity of rice. Int. J.
ing potentials in harvesting and drying of rice. Food Prop. 3(1):77-92.
Trans. ASAE 21:361-366. Rhind, D. 1962. The breakage of rice in milling:
Lee, K. W. 1972. Effects of temperature and A review. Trop. Agric. 39:19-28.
moisture content on mechanical properties of Richey, C. B., Jacobson, P., and Hall, C. W.
brown rice. Ph.D. dissertation. Texas A&M 1961. Agricultural Engineers Handbook.
University, College Station. McGraw-Hill Book Company, Inc., New
Lee, K. W., and Kunze, O. R. 1972. Temperature York.
and moisture effects on mechanical properties Shedd, C. K. 1951. Some new data on resistance
of rice. ASAE paper 72-338. American Soci- of grains to air flow. Agric. Eng. 32:493-495,
ety of Agricultural Engineers, St. Joseph, MI. 520.
Maa, T. J. 1974. The effect of temperature and Siebenmorgen, T. J., Banaszek, M. M., and
time on the mechanical properties of lightly Kocher, M. F. 1990. Kernel moisture content
milled rice. M.S. thesis. Texas A&M Univer- variation in equilibrated rice samples. Trans.
sity, College Station. ASAE 33:1979-1983.
Mannapperuma, J. D. 1975. Analysis of thermal Smith, R. E., Bennet, A. H., and Vacinek, A. A.
and moisture stresses caused during drying of 1971. Convection film coefficients related to
brown rice. M.S. thesis. Louisiana State Uni- geometry for anomalous shapes. Trans. ASAE
versity, Baton Rouge. 14:44-47.
McDonald, D. J. 1967. Suncracking in rice, some Sowbhagya, C. M., Ramesh, B. S., and
factors influencing its development and the ef- Bhattacharya, K. R. 1984. Improved indices
fects of cracking on milling quality of the for dimensional classification of rice. J. Food
grain. M.S. thesis. University of Sidney, Sci. Technol. 21:15-19.
N.S.W., Australia. Stahel, G. 1935. Breaking of rice in milling in
Milner, M., and Shellenberger, J. A. 1953. Physi- relation to the condition of the paddy. Trop.
cal properties of weathered wheat in relation to Agric. 12:255-261.
internal fissuring detected radiographically. Stansel, J. W. 1975. The rice plant—Its develop-
Cereal Chem. 30:202-212. ment and yield. Pages 9-21 in: Six Decades of
Miwa, Y., Kobayashi, H., and Kanamoto, M. Rice Research in Texas. Tex. Agric. Exp. Stn.
1978a. Studies on the strain and the cracking Res. Monogr. 4.
of rice kernels under drying (III)—Particularly Strohman, R. D., and Yoeger, R. R. 1967. A new
on the movement of marked point and the equilibrium moisture content equation. Trans.
shrinkage strain on the surface of kernels of ASAE 10:675.
rough rice. J. Jpn. Soc. Agric. Mach. 41(1):83- Sun, Z., Yang, W., Siebenmorgen, T. J.,
95. Stelwagen, A. M., Jia, C., and Cnossen, A. G.
Miwa, Y., Kobayashi, H., Torii, T., and 2001. A thermomechanical study of dynamic
Kanamoto, M. 1978b. Studies on the strain shrinkage of rice kernels. ASAE paper 01-
and the cracking of rice kernels under drying 6095. American Society of Agricultural Engi-
(IV)—On the cracking of kernels of rough rice neers, St. Joseph, MI.
during cooling-process after drying. J. Jpn. Swanson, C. O. 1943. Effects of moisture on the
Soc. Agric. Mach. 42(1):99-107. physical and other properties of wheat. II.
Mossman, A. P. 1986. A review of basic concepts Wetting during harvesting. Cereal Chem.
in rice-drying research. CRC Crit. Rev. Food 20:43-61.
Sci. Nutr. 25:49-71. Thielen, T. P. H., Yang, W., Siebenmorgen, T. J.,
Nagato, K., Ebata, M., and Ishikawa, M. 1964. and Cnossen, A. G. 2001. Effect of glass tran-
On the formation of cracks in rice kernels sition on thermal conductivity of rough rice.
during wetting and drying of paddies. Nippon ASAE paper 01-6140. American Society of
Sakumotsu Gakkai Kiji 33:82-89. Agricultural Engineers, St. Joseph, MI.
Nguyen, C. N., and Kunze, O. R. 1982. Moisture Van Ruiten, H. T. L. 1979. Physical properties of
content variation among harvested rice grains. paddy and milled rice. Southeast Asia Coop-
Trans. ASAE 25:1037-1040. erative, Post-Harvest Research and Develop-
ment Programme. SEARCA College, Laguna, Working Group, 13th. Texas Agricultural Ex-
Philippines. periment Station, College Station, TX.
Wasserman, T., Houston, D. F., Ferrel, R. E., and Wratten, F. T., Poole, J. L., Bal, S., and Ramarao,
Hamilton, W. E. 1969. Combination system V. V. 1969. Physical and thermal properties of
for rice drying in California. (Abstr.) Pages rough rice. Trans. ASAE 12:801-803.
59-60 in: Proc. Rice Tech. Working Group, Yamaguchi, S., Yamazawa, S., and Wakabayashi,
12th. U.S. Dep. Agric., Washington, DC. K. 1981. On relaxation modulus of rice endo-
Wimberly, J. E. 1983. Paddy Rice Postharvest sperm. J. Jpn. Soc. Agric. Mach. 43:239-245.
Industry in Developing Countries. Interna- (In Japanese with English summary).
tional Rice Research Institute, Los Baños, Zhang, Y., and Kunze, O. R. 1992. Moisture
Philippines. adsorption characteristics of different sections
Wratten, F. T., and Kendrick, J. H. 1970. A fur- of a brown rice grain. (Abstr.) Page 103 in:
ther treatment of hygroscopic equilibrium of Proc. Rice Tech. Working Group, 24th. Texas
rough rice and soybeans at elevated tempera- Agricultural Experiment Station, College Sta-
ture. (Abstr.) Page 38 in: Proc. Rice Tech. tion.
CHAPTER 9
ROUGH-RICE DRYING—
MOISTURE ADSORPTION AND DESORPTION
Otto R. Kunze (retired)

Texas Agricultural Experiment Station
and Texas A&M University
College Station, Texas
David L. Calderwood (retired)

Agricultural Research Service
Beaumont, Texas
Rice is usually harvested at a moisture content higher than that at which it can be
safely stored and, therefore, must be subjected to some type of conditioning. Dry-
ing, the principal method of conditioning rough rice for storage, must be discussed
in terms of its relationship to the nature of the grain, high-moisture rice, different
types of dryers, and various drying procedures. This chapter emphasizes the hygro-
scopic character of the rice grain and the handling of high-moisture or field rice,
and it shows how ambient air conditions before and during harvest can have a
marked influence on subsequent rice quality.
High-moisture rice of a given moisture content and variety is not necessarily uni-
form from one season to the next, nor from one load to the next within the same
season. Head-rice (whole-grain) yields resulting from different lots of rice at the
same moisture content and subjected to the same drying treatments may be quite
different. Freshly harvested field rice must be characterized for quality so that fur-
ther quality losses during drying, storing, and milling operations can be credited to
the appropriate process or procedure.
Rice grains may acquire different kinds of defects while they are still in the field.
One defect that has an influence on quality is a cracked, or fissured, grain. Many
terms are used by different authors to indicate defective grains, but often the
terminology is not sufficiently precise to convey an exact meaning. Terms such
as surface cracks, stress cracks, sun cracks, sun cracking, cracks, transverse
lines of cracks, checks, sun checks, faults, internal faults, air spaces, layering,
splits, fractures, partial fractures, vacuoles, crack rings, fissures, and others are
readily found in the literature. Which of these terms are intended to convey the
same meaning and which should convey a defect that is quite different? More
importantly, what is the source of the defect that is being described? If a char-
223
acteristic of the defect clearly defined its source, then research emphasis could be
applied where it was needed.
Since rough-rice drying is a universal necessity, it is practiced by the rice indus-
try throughout the world. The desire is to dry rough rice, but much more happens in
the process. The rice grain begins to dry during maturation before harvest. All the
rice grains on a panicle are not at the same moisture content. There may be a 10-
percentage-point difference among the moisture contents of grains on one panicle
(McDonald, 1967), more of a difference among the grains on one plant, and much
more of a difference among the grains in a field. The plants fruit over a period of
time, and the grains mature over a period of time and then dry over a period of time.
On any given day during drying, the individual grain moisture contents are very
dissimilar. This dissimilarity is extremely important because the most-mature grains
in the field become sufficiently dry to readsorb moisture at night, while the least-
mature grains may still be in the hard-dough stage. Thus, rice drying in the field
inherently involves moisture desorption and readsorption.
“Rice drying in the field” describes a natural phenomenon, but the phrase is not a
comprehensive description of what happens. As the rice dries, the most-mature (dri-
est) grains begin to readsorb moisture during humid nights.
Since the early 1920s, the focus has been on rough-rice drying, and limited
progress has been made because the research and the literature have been restricted
to that focus. In the 1930s, interesting and important research on moisture
readsorption by low-moisture rice grains was in the process of being developed, but
that research was soon phased out and redirected to specifically focus on rice
drying. That was the pressing need at the time. Most scientists and essentially all
research administrators focused on the drying aspect. But this narrow focus does not
represent what happens in the field while rice is drying on the stalk, when it has
been cut and is drying on the stubble, or when it has been bundled and shocked.
Neither does it represent what happens in the combine hopper, the holding cart, the
transport truck, or finally, the rice dryer. When rough-rice drying is the goal,
moisture adsorption and desorption potentials must be considered with respect to
the condition of each grain during every stage of the process. To do otherwise is to
ignore one of the most important components that influence rice milling quality,
namely the potential for moisture readsorption by the low-moisture grains.
This chapter attempts to emphasize the moisture readsorption aspect that has
received little attention or perhaps been overlooked in previous efforts to provide
guidelines for rough-rice drying. In 2000, the American Society of Agricultural
Engineers recognized the concept of “the rice fissuring mechanism” as one of the
outstanding agricultural engineering contributions to the rice industry during the
twentieth century (Cuello and Huggins, 2000).
In this chapter, the term fissure designates the large fracture that usually forms
perpendicular to the long axis of the grain and is caused by an increase in moisture
content within the surface cells or possibly a moisture loss from the cells in the grain
interior. When cited authors used other terminology, their terms are often enclosed in
parentheses.
SELECTED HISTORICAL ASPECTS
Rice production practices that have evolved over the centuries require that the
grain be harvested before it is dry enough to be stored. Historically, producers have
Rough-Rice Drying / 225
found that rice loses milling quality if it is left in the field to dry to storage moisture.
This loss in quality has been consistently observed, but its cause has not been ade-
quately determined and explained. When rice in the field reaches field equilibrium
moisture content (EMC), its moisture content moves in cycles but lags behind that
of the field environment. (EMC is discussed in Chapter 8.) Daytime environments
may reach 37°C (100°F) with less than 60% rh (under such conditions, the EMC for
rough rice is 10–12%), while the night environments may be 27°C (81°F) with 97%
rh or more (22–24% EMC for rough rice) (Wratten and Kendrick, 1970). Thus, a
dried grain in the field fluctuates within the span of 12–22% moisture content (mc)
every day, but it may never reach either extremity. The cycle for a maturing grain
has a smaller amplitude. The grain does not have to be on a standing stalk; it can be
on cut stalks drying on stubble or in bundles in a shock. The reader must recognize
and appreciate the daily adsorption and desorption potentials of the low-moisture
grain in the field while reviewing the historical aspects of rice drying.
The hygroscopic nature of the rice grain has always been a challenge to rice pro-
ducers. As early rice production increased through mechanization, the loss of mill-
ing quality was soon recognized, and its importance was magnified. Milling quality
quickly emerged as a chief consideration in rice production. If research could
improve head-rice yield by only 1%, it would increase the world head-rice yield by
5.89 × 106 t (U.S. Census Bureau, 2002).
One of the first books on rice was written by E. B. Copeland (1924), the dean of
agriculture at the University of the Philippines, Los Baños. His book was a compre-
hensive treatise covering all phases of rice production. Copeland wrote that the rice
grain cracked in the field from exposure to the sun and from rapid drying during
harvest and even in shocks. His term sun-cracks was a first effort to describe what
happened to rice grains if they were left too long in the field to dry. He defined the
sun crack as a fine, crosswise cracking of the grain caused by rapid drying in the
sun. The sun crack was observed to be more prominent in the early crop of rice than
in the late crop. The term seemed to be a reasonable explanation for what was hap-
pening. Good drying weather in the early-season crop produced more sun-cracked
grains than did less favorable drying weather in the late-season crop. The term
described the kernel failure so well that its use was readily adopted in the rice
industry throughout the world. When artificial drying was introduced, heated air
took the place of the sun, and the term conveniently carried over into the heated
forced-air drying of rice.
In 1961, Kunze (1964) initiated research to determine properties of the rice grain
that could be useful in harvesting, drying, and processing. Among others, obvious
properties to be determined were tensile, compressive, and shear strengths. To
obtain tensile strength data, long-grain rough rice was shelled to get brown rice.
Broken, cracked, or otherwise imperfect grains were discarded. Insulation was
removed from no. 12 electrical wire and then cut into short sections of 1/8 to 3/16 in.
A water-based adhesive was used to mount the brown rice grain into the hollow
cylinders of the insulation to prepare a specimen for a tensile test. Rigorous obser-
vations determined that the grains, when inserted with the adhesive into the cylinder
of insulation material, would fissure (crack). Such grain failures could be observed
above the adhesive line shortly after the insertion.
Good grains were difficult to break by hand, but a fissured grain would break
easily. The literature indicated that rice grains fissure (crack) from drying; there-
fore, it appeared that the water-based adhesive was drying the grain. This did not
seem to be a rational conclusion. Further pursuit of the observation indicated that

the grain adsorbed water from the adhesive. Finally, the rice grain was ground off at
one end and set on a moist paper towel, and within minutes it fissured from mois-
ture adsorption. Such an observation was unprecedented, and interest was diverted
from physical and mechanical properties of rice to environmental conditions and
physical properties that produce fissures in rice. The literature search was changed
to focus on moisture adsorption properties of the rice grain. Little information was
available. A careful search did show some interesting but unexplained data
concerning rice harvested in the field and left to dry in individual bundles or in
shocks (Dobelmann, 1955, 1961). Rice samples were collected after every night of
exposure. When the samples were dry, they were milled. Samples from single
bundles in the field consistently showed more broken grains after milling than did
samples from bundles in shocks. The data were reported for nights of exposure
rather than for days in the sun. Rice from bundles dried in the shade consistently
showed the fewest broken grains after milling. This work was notable because it
moved the emphasis for milling quality from ambient drying in the sun to the
ambient exposure that the grain received at night. The day provided drying
potential, while the night provided the potential for the rice grain to readsorb
moisture. Ambient relative humidity influenced the rice grain’s potential to desorb
or adsorb moisture.
Specific information on moisture adsorption by the rice grain was scarce and
seemed to be nearly nonexistent. When a related source was found, the listed
references were often not available and could not be readily obtained. No body
of information had been developed for moisture adsorption by rice. The work
of Kunze (1964) was completed in three years, but the literature search
continued for more than three decades. The information that follows has evolved
from the continuing search for moisture adsorption effects on the dried rice
grain.
Six years after Copeland (1924) published his book, two Japanese scientists
(Kondo and Okamura, 19301) published work concerned with moisture adsorption
by rice at storage moisture. The title of their article, translated into English, was
“Fissuring of the Rice Grain from Moisture Adsorption.” These scientists produced
fissured (sun-cracked) grains through moisture adsorption. Their work is a 19-page
treatise that is the most enlightening, revealing, detailed, and comprehensive article
ever written in the area of moisture adsorption. They summarized their work with
eight conclusions. The last one, translated into English, was
It is without any doubt that moisture adsorption by the dried grain is the definite cause
of fissured grains. It follows then that it is very important that dried grains be quickly
placed into an air-tight storage to prevent moisture readsorption.
The work was done in Japan but was published in German because that was the lan-
guage in which to publish scientific research. The work was not widely published
nor was it widely read, because the mainstream of rice research was in drying and
few postharvest rice researchers could read German.
The research of Kondo and Okamura (1930) was original. They reported only
their findings and cited no references. Seemingly, they were unaware of the book
that Copeland (1924) had written.
1 Article obtained from Japan in 1975.

The fissure (crack) phenomenon is the common thread that links the work of
Kondo and Okamura (1930) with that of Copeland (1924). Since the two were so
diametrically opposite, no one perceived that they could be related. The two causes
of fissured rice grains seemed to be completely independent of each other; they
were at opposite ends of the spectrum. The need throughout the world was to dry
rice effectively and efficiently, while moisture readsorption by dried rice was of
little interest, of no practical value, and hence not worthy of financial support for
research. The perception that fissured grains resulted from drying led to a focus on
conditioning rice after harvest for storage.
The problem of attaining high head-rice yields persisted, and researchers were
challenged to improve the milling quality of rice. Stahel (1935), in Surinam, was
unaware of the work by Kondo and Okamura (1930) while he was doing his own.
However, his conclusions followed those of Kondo and Okamura. When low-
moisture rice grains were left exposed to a moisture-adsorbing environment during
the night, fine crosswise cracks developed. These were responsible for most of the
breakage during milling. Stahel was unable to produce sun cracks by drying thin
layers of rice in the sun to less than 10% mc. He then proceeded to dry rice in the
sun to various low moisture contents before remoistening the grains. This effort was
to determine at what level of low moisture the readsorption of moisture would cause
fissures to develop in the grain. He found that fissured (cracked) grains would
develop when grains were first dried to 14 or 15% mc before readsorbing moisture.
Stahel (1935) cited Copeland (1924) as a reference in his work. He took excep-
tion to the use of the sun-crack term and called it a misnomer, because the cracks in
the rice grain did not develop from drying in the sun but rather from moisture read-
sorption after having been dried.
Thus, within slightly over a decade, the misconceptions about sun cracking of the
rice grain had been identified and put into proper perspective. The three works came
from the Philippines (Copeland, 1924), Japan (Kondo and Okamura, 1930), and
Surinam (Stahel, 1935). Copeland’s book apparently received the greatest accep-
tance and distribution. The works of Kondo and Okamura and of Stahel were deeply
buried in the literature by the volumes of research papers published on rice drying.
The sun-crack theory flourished, essentially without challenge, into the 1970s and
even into the 1980s. During the 1960s and for years thereafter, the concept was that
the rice grain fissured during drying but could also be made to fissure from moisture
readsorption. Desikachar and Subrahmanyan (1961) wrote:
The development of checkings or cracks in rice and other grains during dehydration as
a result of quick or uneven removal of water is well known, but the occurrence of
similar cracks during hydration or wetting is very little understood.
The two causes of fissured rice grains were still perceived to be distinctly
independent and different.
In the ensuing years, research on rice drying continued, but wheat researchers
were also busy drying wheat. It is unfortunate that the rice researchers did not
communicate with the wheat researchers nor wheat researchers with the rice
researchers. Yet, the drying-crack phenomenon did infiltrate the thinking of the
wheat researchers. Sharp (1927) found that when the air-dried wheat kernel was
remoistened, it lost some of its density. He suggested that removal of the regained
moisture (drying) caused air spaces to develop in the grain. Later, Swanson (1943)
studied the effects of rewetting wheat during harvest. He found that nonweathered
wheat samples at the millers were, on average, 1.81 kg (4 lb) heavier than weath-
ered samples. He concluded that after the water was again removed (drying), vacu-
oles were left in the grain. These vacuoles decreased the grain’s specific gravity and
hence its test weights. Milner and Shellenberger (1953) studied the physical proper-
ties of weathered wheat and found that natural weathering caused the formation of
internal fissures that could be detected radiographically. The fissures appeared when
the grain was dried in a manner that produced stresses within the kernel. They also
confirmed the loss of density and the air spaces or vacuoles postulated by Sharp and
Swanson. (Milner and Shellenberger [1953] still attributed the development of
fissures to drying.) Grosh and Milner (1959) tracked the penetration of water into
the wheat kernel. They observed that the fissures (cracks) occurred in the hard
vitreous endosperm of wheat before the water moved through the kernel. Such
evidence that fissures (cracks) develop in wheat because of wetting the endosperm
alone had not been reported in previous literature. Thus, by 1959, the wheat
researchers had also independently clearly defined that moisture adsorption, not
drying, caused the wheat grain to fissure.
The moisture adsorption-fissure phenomenon was clearly defined in the rice
industry by 1935 and in the wheat industry in 1959. But “moisture adsorption” was
never the catch phrase that could garnish funds for a research proposal. However,
the moisture adsorption-grain fissure concept was a fundamental truth. It stood on
its own merits and occasionally exposed itself to unsuspecting researchers.
Subsequent moisture adsorption work was incidental. Kik (1951), working in
Arkansas on the nutritive qualities of rice, observed fissures (cracks) in rice grains
before they were milled. He reported that these fissured grains were largely respon-
sible for the breakage of rice during milling. According to Kik, the cracks resulted
from remoistening the grains and not from drying. He was aware of the works of
Kondo and Okamura (1930) and of Stahel (1935) and agreed with them. But the
observations by Kik were not central to his work, and neither were they central or of
interest to the scientists who read his work. The information was not published in a
manner or a publication that caught the attention of scientists whose interest was
centered on the drying area.
During the first half of the twentieth century, rice was being grown on five of the
seven continents. Every rice-growing continent had researchers working on prob-
lems peculiar to their area along with rice-drying problems that were common to
every production area. Researchers on a continent communicated among them-
selves, but few were aware of what was happening elsewhere. The research on one
continent was not necessarily well-known on another.
In Australia, Langfield (1957) reported that it was well-known that much of the
grain breakage during harvesting and milling was the result of fissures (fine cross-
wise cracks) that developed in the grains during the ripening process. The fissuring
(cracking or checking) commonly referred to as “sun-checking” resulted from rapid
fluctuations in atmospheric humidity during the ripening process. Langfield recog-
nized when milling quality was lost but was unable to specifically tell why it was
lost.
Four years later, Desikachar and Subrahmanyan (1961) reported peripheral
observations made while rewetting and cooking raw milled and parboiled rice.
Transverse lines of cracks developed in the grains. It took longer for these fissures
(cracks) to develop in parboiled rice than in raw milled rice. They found that rice
grains (raw milled or parboiled) were very susceptible to fissuring during moisture
adsorption. Development of the fissures was dependent on the length of the soaking
period and the temperature of the water. The fissures were a direct result of the
soaking and started to develop in 3 or 4 min after the grains were submersed in
water. Their conclusion was that hydration precedes fissuring. After soaking, both
raw milled and parboiled rice required a shorter cooking time.
Other comments by Desikachar and Subrahmanyan (1961) imply that fissures
(cracks) resulting from drying were an accepted phenomenon, but fissures resulting
from wetting were very little understood. A person could conjecture that these sci-
entists had to acknowledge the then-accepted fissure-from-drying dogma before
they could propose that such a fissure also resulted from wetting.
Researchers at the time continued to sense that their explanation was incomplete,
because the rice grains were fissuring in the field during maturation, but they were
unable to specifically identify the cause of the fissures. Rhind (1962) studied what
was happening and developed the rationale that the highest tensile stresses could be
expected in the driest parts of the grain (surface layers) during drying but would be
in the inner part of the grain during moisture adsorption.
The rice-drying problem was not solved, and the ambient environment continued
to take its devastating toll on all producers who thought they could leave their rice
in the field to dry. Huysmans (1965) reported that rice that was harvested too late
had a great number of fissured (sun-cracked) grains with a subsequent high percent-
age of broken rice in the milled product. Fissured-grain percentages for a given
variety of medium-grain rice during 1962-1963 were below 10% up to 45 days after
flowering (DAF), with head-rice yields of more than 50%. Fifty days after
flowering, the percentage of fissured grains had increased to 25.5% and the head-
rice yield had dropped to 36.8%. A 15.5% increase in fissured grains (10 to 25.5%)
caused a head-rice yield reduction of essentially the same magnitude (from more
than 50 to 36.8%).
The foregoing paragraphs are a brief historical review of moisture adsorption
technology in rice and wheat from 1930 to 1964. It was compiled over many years
and is sufficiently complete and detailed to convey a convincing story that moisture
readsorption by the low-moisture rice grain is among the chief causes of fissured
grains and milling quality losses. Moisture adsorption cannot occur until the EMC
of the grain drops below the environmental conditions that would produce a higher
EMC in the grain. The individual grains on a rice plant pass through this stage in the
field as the rice matures and loses moisture (dries). Therefore, moisture readsorp-
tion and drying in the field cannot be separated from each other. The moisture
adsorption fissure and the perceived drying fissure are the same. Milling quality
data collected from rice drying in the field on the stalk, stalks drying on the stubble,
stalks in single bundles, or stalks in shocks (composed of bundles) cannot be
explained unless the moisture adsorption phenomenon is considered. The moisture
adsorption potential and its subsequent results must be considered at every drying
and processing step through which the grain must pass before it is packaged for
commercial consumption.
RICE IN THE FIELD
Drying of rice starts in the field as the grain begins to mature. During the matu-
ration period, the grain receives water and nutrients from the plant under normal
growing conditions of sunlight, fertility, and moisture. At maturity and the begin-
ning of drying, the grain gradually decreases its dependence on the plant for water
and nutrients and begins to increase its interactions with the field environment.
During this period, the grain transpires more moisture to the air than it receives
from the plant, with the consequence that the grain moisture content gradually
decreases. Whenever the grain moisture content drops below the equilibrium condi-
tions of temperature and relative humidity in the ambient environment, the rough-
rice grain, which is hygroscopic, readsorbs moisture. Thus, the moisture content of
a grain becomes subject to the ambient environment in which it exists. Over a
period of perhaps a week or more, the rice grain changes from plant dependent to
plant independent. The temperature and moisture of the ambient air as well as the
maturity of the grain largely determine the rate at which a grain dries. After the
initial drying, the grain continues to be hygroscopic and responds to every environ-
ment to which it is exposed.
Nagato et al (1964) studied the development of the rice grain in the field from
heading until 55 days thereafter. They divided the time of grain development into three
periods: 1) that in which moisture content decreases physiologically with the increase
of dry matter weight, 2) that in which moisture content stays at approximately 28% for
about 10 days, and 3) that in which moisture content decreases physically under the
influence of weather. Fissures (cracks) do not occur during the first and second periods
because grain moisture content in these stages is not affected by weather conditions.
However, the grains may fissure in the field during the third period. Yoshida (1981)
gives another description of how the rice grain develops.
Although rice grains in a field all follow the same sequence of flowering, filling,
maturing, and drying, each grain follows this sequence on its own time schedule.
Rice florets on a panicle require three to five days to flower (Stansel, 1975),
whereas all the panicles on a plant may require one to several weeks. The greater
the number of tillers per plant, the longer its flowering period. When all plants are
considered, the length of the flowering period in a field may be quite long. Grain
maturation and drying on the panicles follow essentially the same schedule. The
measured moisture contents of rice samples collected from a field may vary only a
few percentage points, whereas the moisture contents of grains on a single panicle
that has matured and is drying may vary 10 percentage points or more (McDonald,
1967), and the moisture contents of grains in a field may vary from the moisture
content of storage to that of grains in the milk or soft-dough stage. If the moisture
contents of these grains were plotted on a statistical histogram for a particular
day, the storage-moisture grains and those in the milk or soft-dough stage would
be on the extremities of an approximately normal distribution. Early in the
season, the distribution would be skewed toward the higher moisture values, and
later it would be skewed toward the field EMC. The mean would be somewhere
between the two extremes and could be in the harvest-moisture-content range
between 18 and 25%.
The work of Chau and Kunze (1982) shows the moisture contents of grains from
a rice plot during a harvesting or field-drying period (Fig. 1). More information was
extracted from these data to portray the moisture content variation within the entire
population of grains in the field on a given day. The high- and low-moisture lines
from Figure 1 were used as a basis from which to project extensions of grain mois-
ture contents both before and after the dates on which moisture content determina-
tions had been made from samples in the field. These conditions are depicted as an
envelope, the boundaries of which represent the high- and low-moisture grains on a
series of days (Fig. 2). A broken line within the envelope estimates the field mois-
ture content of rice during the same period. Moisture content variations within a
grain during a single day can be several percentage points, but this variation is not
shown. A harvest season without rain was also assumed. Most grains in a field sam-
ple should have moisture contents close to the measured field moisture content, but
the envelope boundaries indicate the span of grain moisture content on that day.
Figure 2 shows that all grains are initially at a relatively high moisture content.
During maturation or ripening, both high- and low-moisture grains exist in the field
on a given day. If they were left in the field after ripening, eventually there would
be little spread in moisture content. However, a considerable period of time elapses
between the beginning of field drying and its end. The day of harvest would nor-
mally occur in the region to the right of the center of the envelope. The height of the
envelope would not vary much from one variety to another, but the width of the
Fig. 1. Average moisture content of rough rice grains from the top (t), middle (m), and bottom (b)
positions of the wettest (w) and driest (D) panicles during the indicated date of the harvesting season.
(Reprinted, with permission, from Chau and Kunze, 1982)
envelope would be greatly influenced by variety (early- or late-maturing), weather

conditions, and the plant population density in the field. The width of the envelope
is important because rice grains that dry to 13 or 14% mc during the day may fis-
sure when they readsorb moisture at night. As an illustration of envelope width,
Figure 2 shows 22 days between the times when the first and last rice grains reach
30% mc. The envelope also shows that the field moisture content dropped from 25
to 18% in about six days. Stahel (1935) reported that the moisture content of rice
standing in the field was 4–5 percentage points lower in the afternoon of a dry,
sunny day than it was at 10:00 or 11:00 a.m. of that day.
Finally, rice harvested with a field moisture content of 22% may contain grains
having moisture contents between 15 and 45%. This does not consider the daily
drying cycle, which may cause the low-moisture grains to reach 12 or 13% mc.
Researchers in Arkansas have developed more information for rice grains within
the envelope. Kocher et al (1990) measured moisture contents of individual grains
to obtain the distribution of kernel moisture contents for different planting dates and
harvest periods. The variance of kernel moisture content was large early in the har-
vest season and decreased as the average moisture content of the grains decreased
(Siebenmorgen et al, 1990). Siebenmorgen et al (1992) correlated head-rice yield
with distributions of individual-kernel moisture contents during harvest. There was
little change in head-rice yields for three varieties of rice when the average moisture
content ranged from 15 to 22%. Rapid rewetting by rain caused reductions in head-
rice yield when the average grain moisture contents decreased to 15% or lower
before a rain. Further work by Siebenmorgen et al (1998) related rice milling
quality changes from moisture adsorption to individual-kernel moisture content
Fig. 2. Envelope boundaries estimating the highest and lowest moisture of grains in a rice plot during a
harvesting season.
distributions. Head-rice yields began to decline because of moisture adsorption

when the bulk critical moisture content ranged between 12.5 and 14.9%. This
moisture content range is in general agreement with that reported by Stahel (1935)
in Surinam. He removed grain samples while rice was drying in the sun and then
placed them into water to determine at what moisture content the rice grains began
to break when processed through a Smith sheller. Grains with lower moisture
contents before rewetting showed much larger reductions in whole-grain rice.
The work of Kunze et al (1988) relates how highly fertilized and less-fertilized
rice (Lemont variety, long-grain type) dried in the field when plots were planted on
the same day and the same cultural practices were used. Sample harvests were
started on August 4, about two weeks before the normal harvest date, and continued
well beyond the harvest season (September 4, 1987). Harvests were made as shown
in Figure 3, which also shows how the rice dried in the field. Sample grains were
threshed, screened, aspirated, and weighed, and their moisture contents were
determined before grains were placed on screen trays (two grain layers deep or less)
to dry in a conditioned air laboratory. After drying, the moisture content was
measured again.
Three hundred grains from each sample were shelled by hand, and the resulting
brown rice was inspected for fissures (Fig. 4). The remainder of the sample was
Fig. 3. Grain moisture content at each harvest date for plots that received the indicated nitrogen appli-
cations. Code numbers are 1) the plot number, 2) preplant nitrogen (kg/ha) applied, and 3) nitrogen (kg/ha)
top-dressed at the time of grain heading. (Reprinted, with permission, from Kunze, 1988)
mechanically shelled and milled. When the field moisture content was between 22
and 24%, the percentage of fissured grains in a sample was low (2–3%).
The research shows the relationships among 1) grains that fissured in the field
while the rice was maturing, 2) grain field moisture content, 3) rainfall, and 4) fer-
tilization rates. Cultural practices and fertilization rates influenced the percentage
of fissured grains largely to the extent that they influenced the moisture contents of
the rice grains. Low-moisture grains (storage moisture) under the given cultural
practice and from both fertilization rates experienced a high percentage of fissured
grains when they were exposed to daily relative humidity cycles of 40% or more.
Fissured grains usually break during milling.
The highly fertilized rice normally produced more tillers, fruited over a longer
period of time, and produced greater yields. Field samples from these plots had
higher moisture contents at a given sampling date than did samples from plots
that received less fertilizer. Milling quality comparisons of rice harvested at a
given moisture content (on different dates) from highly or less-fertilized plots
should be made with great care. The results may not be comparable if the
desorption and adsorption potentials of the preharvest days were not essentially
the same.
Fig. 4. Fissured grains observed in rice samples harvested on the indicated dates from plots that
received the nitrogen applications as shown. Code numbers are 1) the plot number, 2) preplant
nitrogen (kg/ha) applied, and 3) nitrogen (kg/ha) top-dressed at the time of grain heading. (Reprinted,
with permission, from Kunze, 1988)
The less-fertilized rice matured and dried earlier than did the highly fertilized
rice (Fig. 3); it also fissured more and sooner in the field. At early harvests, the less-
fertilized rice had a higher head-rice yield (relatively more mature grains) than did
the highly fertilized rice (relatively more immature grains). When percentage of
fissured grains (Fig. 4) was plotted against dates of harvest, the plots for the less-
fertilized rice were well above those for the highly fertilized rice (the reverse of the
moisture content curves in Fig. 3). Grains from the less-fertilized plots reached 50%
fissured grains five to six days before that percentage was reached in the highly
fertilized plots. Yet when grains from both levels of fertilization reached a moisture
content essentially in equilibrium with the field environment (two to three weeks after
the normal harvest period), the percentage of fissured grains at both levels of
fertilization was nearly the same (between 84 and 88%). The two groups had slightly
different percentages of fissured grains at a given moisture content during drying.
Since the two groups did not approach the same low-moisture level in the field at the
same time, the difference could be due to the difference in weather conditions.
Fertilization rates had little influence on the final percentages of fissured grains.
Producers often feel that highly fertilized rice fissures less than does less-
fertilized rice when both are planted and harvested on the same day. The reason for
fewer fissures in the highly fertilized rice may well be that it has a higher moisture
content (because of a longer fruiting season). For a more-comprehensive review of
relationships among grain moisture content, percentage of fissured grains,
fertilization rates, and milling quality of rice in the field, the reader should refer to
the work by Kunze et al (1988).
The phenomenon of grains fissuring in the field on the stalk during maturation
before harvest is alluded to by numerous researchers: Copeland (1924), Burmistrova
et al (1956), Langfield (1957), Huysmans (1965), and others. The allusions are cor-
rect, but there has been a misconception about what causes the grains to fissure. The
general and almost universal belief has been that the grains sun crack during dry-
ing, whereas the real cause is moisture readsorption by the grains after they have
dried. The difference is that the grains do not fissure during the day but rather in the
evening and at night while the environment is humid.
In California, Jenkins (1989) used a modified harvesting technique. Rice panicles
were cut and swathed on the stubble in the field before being covered with chopped
straw. With this treatment, the rice grain did not dry as much during the day and
neither did it readsorb as much moisture at night from high humidity and dew. This
method of drying showed essentially a 20-percentage-point improvement in head-
rice yield over open windrow treatments.
Peasants in India are aware that rice grains fissure on the stalk in the field. The
primary author (O. R. Kunze) was a consultant at the G. B. Pant University,
Pantnagar, India, in 1985. The university arranged to let him and his adviser
periodically harvest panicles from a peasant’s plot in a bar ditch within the
university campus. The rice had headed out and was maturing. From day to day,
they picked 10 of the most-mature panicles and from them selected the most-mature
grains. These were shelled and inspected for fissures. For a week or more, there
were no fissured grains. Fissured grains then began to appear. A weekend
intervened, and when they got back to the plot, the owner had cut and gathered the
rice and carried it home to dry in the shade. Why did he harvest it? Because the
most-mature grains were beginning to fissure from moisture readsorption at night,
and the researchers had not yet paid him for his paddy of rice.
The rice that the peasant gathered was in the early-harvest stage. These are the
rice seeds that the combine harvests and deposits into a hopper. They have a mix-
ture of moisture contents. However, when they are placed into a moisture meter, the
specific grain moisture contents are averaged to imply a uniformity of moisture
content for a mass that may be very nonuniform. If a problem exists because of
high- and low-moisture grains, the field moisture content of the grains does not
reflect it. Although the field moisture content of the grains is a meaningful measure,
it might be so only because more meaningful measures are not available. Would the
producer and the dryer operator react differently if, in addition to the field moisture
content, the readout from the moisture meter included moisture contents of the
highest- and the lowest-moisture grains? Would the drying procedure be changed if
the sample from the moisture meter could be diverted into a fissured-grain detector
that would count the fissured and unfissured grains? Would such information attract
the interest of the rice breeder, the agronomist, and the equipment manufacturer?
Would this cause the producer to change the production strategy and the dryer
operator to explore different drying procedures? Modern electronic technology is
bringing these measures into our grasp.
Today, the field moisture content of rice is generally known as the grain is har-
vested or as it enters the dryer. The goal is to maintain quality and to dry efficiently.
The harvesting and drying decisions that must be made should be worth more and
better information.
OPTIMUM HARVEST MOISTURE
The U.S. Department of Agriculture (USDA, 1973) states that the proper stage
of maturity for harvesting rice is when grains are fully mature in the upper portions
of the panicle and in the hard-dough stage at the base. This is a subjective definition
that implies that fully mature grains are not sufficiently low in moisture content to
fissure from rapid moisture adsorption. The optimum harvest moisture content is
defined as a stage of maturity for grains on a single panicle of rice. However, the
growth of panicles is dependent upon the development of tillers, which emerge in
sequence over time. The plant produces tillers from a bud on the culm at a joint
called a node (Stansel, 1975). Additional tillers may arise from each successive
node. They develop alternately on each side of the culm. A new tiller develops for
each two new leaves on the culm (Nagai, 1959). Hence, only a few panicles per
plant may have grains at the optimum harvest moisture content at any time during
the harvest season. At harvesttime, some panicles may already have passed the
optimum harvest moisture content, while others may still be approaching it.
Additionally, all rice plants in a field are not at the same stage of maturity, as
shown by the fact that field samples of rice taken at different locations often vary in
moisture content by several percentage points. The producer has to use judgment to
ascertain when the maximum number of panicles are at the optimum harvest
moisture content. At this stage of maturity, there are immature panicles that are still
developing, as well as overmature panicles that have already passed the optimum
harvest moisture stage. The combine harvests the immature, optimally mature, and
overmature grains into the hopper, and the producer uses the moisture content
measurement of such a field sample to help determine when to start the harvest.
Optimum harvest moisture content can also be defined in terms of yields and the
amount of energy needed to dry the rice. Head-rice yields are usually reported to be
highest when rice in the field is at a high moisture content (Kester et al, 1963), and
therefore, energy expenditures are high because much moisture must be removed
before the grain can be stored. Some researchers report that field yield is still
increasing when head yield has begun to decline and total milling yield is still
approaching its peak (Bose and Chattopadhyay, 1976). During this period, there is
little trade-off among head, field, and total yield because their respective peaks may
be separated by only a few days. Later in the season, head, field, and total yields
decrease, along with the amount of energy required for drying. The trade-off is
essentially between high yields of head, field, and total rice with high drying energy
requirements early in the season or reduced yields with lower drying requirements
later in the season. Still another approach is to maximize total yield with less
emphasis on head rice and broken grains. Prices of the grain components and sav-
ings in drying costs would be considerations in this approach. The decision of when
to harvest is made from judgment and experience. The variety of rice, area to be
harvested, size or capacity of equipment, weather at harvesttime, and drying energy
requirements are a few of the considerations that may cause the choice of optimum
harvest moisture content to vary from one region to another. Lu et al (1995) discuss
the effects of rice moisture content at harvesttime on the economic returns to the
producer.
The situation that the producer must work with is shown in Figures 3 and 4. In
Figure 3, rice is initially at a high moisture content in the field before the grains
begin to dry quite rapidly. The same research showed that head-rice yield for the
same harvest dates initially increased because more grains were maturing into fully
grown kernels. At maturity, the grains began to dry rather rapidly (Fig. 3), and the
potential for the grains to fissure begins to increase. The period from August 11 to
August 17 appeared to be the optimum harvesttime for a maximum head-rice yield. In
this case, the producer had a window of about six days in which to harvest. After
August 17, the percentage of fissured grains increased from 12 to 28% in two days in
the lesser fertilized rice; the increase for the fertilized rice was less dramatic, but the
upward trend had started. The trend for increases in percentage of fissured grain was
only two to three days behind the trend for decreases in moisture content. The data for
Figures 3 and 4 were collected on the Texas Gulf Coast. Most rice-growing areas have
similar climates, but the dramatic changes in temperature and relative humidity may
be very different for inland river basins than for coastal plains. Each area has its
unique characteristics, and the rice grower must determine from experience what these
characteristics are and how detrimental they may be to the ripening rice.
Researchers have reported different optimum harvest moisture contents for dif-
ferent varieties in a region. In Arkansas, rice with a moisture content between 17
and 24% provided the highest head yield for Zenith, a medium-grain variety
(McNeal, 1950), whereas a moisture content between 16 and 22% provided the
highest head yield for Rexark, a long-grain variety. Kester et al (1963) reported
head-rice yields for medium- and short-grain varieties to be maximum at moisture
contents between 25 and 32%. Morse et al (1967) found that total grain yield of
Caloro, a short-grain variety, increased as the moisture content decreased to about
20% and that the percentage of head rice peaked at a harvest moisture content
between 26 and 30%. The effects of harvest moisture content on milling quality
were studied by Geng et al (1984) in California. They concluded that the required
moisture content at harvesttime for maximum head-rice yield can be defined only in
the wide range of 25 ± 5%.
In energy-related research (Calderwood et al, 1980), rice initially at 25% mc was

harvested at 11 intervals during a time span of about 30 days. During each of sev-
eral tests with four varieties, dry weather caused the rice moisture content to drop
steadily, but rain and dew caused variable amounts of rewetting. Rice eventually
dried to moisture contents that varied from 13 to 16% in different tests. Grain yields
were not greatly affected by delayed harvesting but tended to decrease before the
tests were concluded. The percentage of total milled rice (whole and broken grains
combined) increased with the later harvest dates, but the percentage of head rice
(whole grains of milled rice) reached a maximum at an intermediate harvest date
and then declined rapidly with further delays in harvesting. Both the long-grain
variety, Labelle, and the medium-grain variety, Nato, maintained a high percentage
of head rice when harvested at 16% mc. Other varieties showed significantly
reduced head-rice yields when harvested at 18% mc. The energy expenditure for
drying rice to 12.5% mc was estimated to be reduced about 50% by harvesting rice
at 16% rather than at 22% mc.
In countries where the producer has less access to moisture-measurement
devices, the optimum time for harvest may be specified by days after flowering
(DAF). Bose and Chattopadhyay (1976) determined the optimum time for the har-
vest of Ratna rice grown in India to be between 26 and 29 DAF. At this time, the
field moisture content was between 22.5 and 25%. They also worked with Jaya and
Pankaj varieties and found that rough rice harvested at a moisture level of 23% had
better total milling yield and a better percentage of head yield than did grain har-
vested at lower moisture levels. Still other researchers (Steffe et al, 1980) specified
that crops of average yield in California should be ready for harvest between 45 and
55 days after the first heading. According to Stansel (1975), the first heading for a
rice field occurs when about 15% of the panicles have emerged from the leaf sheath
and embryos are just beginning to be pollinated and fertilized. Optimum harvest
moisture contents specified by days after heading or after flowering must be related
to specific rice varieties and weather conditions.
FUNDAMENTALS OF RICE DRYING
Moisture content in a grain sample is measured on either a wet or a dry basis.

Both measures are used in the rice industry. Moisture levels on a wet basis are most
commonly used in the rice trade, whereas dry-basis measurements are more often
used in technical papers to report research.
Percent moisture content, wet basis (Mwb), is the weight of water (Ww) in a sam-
ple divided by the total sample weight (water and dry matter) and multiplied by 100.
In equation form, this is
Ë Ww Û
% M wb = ÌÌ ÜÜ100
Í Ww + Wdm Ý
in which Wdm is the weight of dry matter.
Percent moisture content, dry basis (Mdb), is the weight of water in a sample
divided by the dry matter weight in the sample and multiplied by 100. In equation
form, this is
ËW Û
% M db = ÌÌ w ÜÜ100
Í Wdm Ý
To convert percent moisture content from wet to dry basis, or from dry to wet
basis, the following equations can be used.
100 (% M wb )
% M db =
100 - % M wb
100 (% M db )
% M wb =
100 + % M db
Drying takes place when the vapor pressure in a rice grain is greater than that in
the surrounding air. Since moisture is removed at the surface of a rice grain, a
moisture gradient develops within the grain, with the center having a higher mois-
ture content than the surface. Initially, the drying rate at the surface is rapid, but
after a brief time, the drying rate at the surface is limited by the rate of moisture
movement from the interior to the surface. The rate of internal moisture movement
can be increased by increasing the vapor pressure difference between the rice grain
and the ambient air. Ordinarily this is accomplished by heating the air, which in turn
heats the grain.
Different cultivars of rice have different grain geometries, and the different con-
figurations respond differently to given conditions. The ratios of surface area to
grain volume are different. Therefore, the various rice types (long, medium, and
short) and the various rice forms (rough, brown, and milled) tend to dry at different
rates under a given drying condition. Similar reasoning holds for moisture read-
sorption by the different types and forms of rice. Kunze (1964) worked with six
varieties of brown rice and found that larger, thicker grains of brown rice fissured
more from moisture readsorption than did smaller, thinner grains. He also found
that different varieties of rice with similar grain configurations had different fissur-
ing susceptibilities. In the final analysis, the producer, dryer and storage operator,
miller, and processor must have some basic knowledge about the rice type that is
being grown, dried, stored, milled, and processed. General rules hold across all rice
types and forms, but each variety, type, and form has a unique response to a given
process or condition.
Since the rice grain is hygroscopic, it is dynamic and physically responds to the
moisture and temperature changes to which it is exposed. Researchers in 1975 (S.
Prasad, J. D. Mannaperuma, and F. T. Wratten, unpublished data) found that a 1%
change in moisture content produces stresses in the grain that are 100 times the
magnitude produced by a 1°C temperature change. Other researchers (Kobayashi et
al, 1972) studied stresses in the rice grain during drying. They concluded that
fissured (cracked) rice was produced by complex strains from moisture gradients.
X-ray photographs were taken of rough-rice grains during the drying period. These
showed that the grain contracted as drying proceeded. But the photographs did not
show that fissures developed during drying. If they had taken the X-ray photographs
while the low-moisture grains were adsorbing moisture, successive pictures would
have shown fissures as they developed in the grains. Grosh and Milner (1959) used
X-ray techniques to show that cracks, radial and transverse to the crease, developed
in the hard vitreous endosperm of a wheat grain in advance of water movement into
the kernel. Kondo and Okamura (1930) show pictures that illustrate the type of
fissures caused in storage-moisture rice grains when they readsorb moisture.
Kunze and Choudhury (1972) hypothesized that stresses might develop in the
rice grain during drying and suggested that moisture removal from the grain would
cause tensile stresses at the grain surface. Since the grain is a free body, equal and
opposite compressive stresses would be produced in the grain interior (Fig. 5).
According to this rationale, rapid drying would cause tensile failures at the grain
surface. Such failures are seldom observed in rough and brown rice but are readily
observed in milled rice (Fig. 6).
Henderson (1954) studied fissures (checks) in the rice grain and reported that
partial fractures expanded into fully grown fissures as the rice grain was dried. This
is rational, because moisture adjacent to a fissure in a grain can move more easily to
the grain surface than can moisture in an unfissured section. The fractured surfaces
would dry more readily and contract faster than would other portions of the grain.
Fig. 5. Hypothetical distributions of drying stresses within a rice kernel. (Reprinted, with permission,
from Kunze and Choudhury, 1972)
Fig. 6. Surface cracks that resulted from drying milled rice grains for 20 min at 60°C. Twelve grains
are shown in a rosette; grain surfaces appear to have a network of fine cracks that are believed to be
tensile failures.
Such partial fractures have been observed in rice grains, but no effort has been
made to identify their source. No data seem to be available that document the
development of partial fissures. What causes them? The partially fissured grain is
often quite strong, and those that are observed in white rice have survived the mill-
ing process.
Information available today gives a reasonable explanation for their develop-
ment, even though no hard data are available to support the concept. As mentioned
earlier, the rice grain cycles daily through a moisture span as it ripens in the field.
Moisture is lost during the day, and some is regained at night. There is a critical
low-moisture point at which the drying rice grain fissures, perhaps only partially,
when it readsorbs moisture in the field. But the successive day would cause the
grain to dry a little more, and the moisture readsorption during the following night
could cause the partial fissure to expand completely across the grain. Partial fissures
observed in milled rice may have needed only one more night of exposure in the
field to have become full-fledged fissures.
The primary author can support this rationale from related experiences. While
serving overseas as a rice process consultant, he would demonstrate how easily a
fissured grain (milled or brown) could be broken in the palm of his hand under
thumbnail pressure. He was rather disconcerted, however, to find that, when the
grain was only partially fissured, it would not break. The solution to this
potential problem was to set the demonstration grains out into the environment
during the night. The humid night exposure caused all partial fissures to become
full-fledged ones. The next day, the demonstrations could be performed with
confidence.
When rice grains, at storage moisture, are exposed to a moisture-adsorbing envi-
ronment, the resulting fissure is full and complete. Craufurd (1963) rewetted whole
grains of brown rice and tried to determine the origin of the fissure (crack). All fis-
sures observed were complete fissures, and these developed so fast that he was
unable to determine their points of origin. If a physical action occurs with sufficient
speed, it makes an audible sound. Stermer and Kunze (1994) amplified this sound
so that fissured grains could be automatically detected and counted.
Sharma (1980) reported that rough rice generally does not fissure during
drying. Kunze (1979) found some fissured grains in rice samples immediately
after they were dried. Older rice samples showed more fissured grains after
drying than did younger samples. The fissured grains may have had imperfections
before drying started. In earlier work, Kunze (1964) found that more rice grains
fissured during drying from storage moisture to dry-matter weight than during
drying from field moisture to storage moisture. The observations implied that the
external grain volume (bulk volume) became essentially constant when the grain
surface reached a moisture content slightly below storage moisture. As the
internal portion of the grain (with higher moisture levels) continued to dry, the
interior section contracted, developed tension because of moisture loss, and
fissured. Efforts were made (O. R. Kunze, unpublished data) to confirm this
hypothesis with measurements using an air-comparison pycnometer. However,
when a grain fissures, the pycnometer measurement shows decreased volume
because the opening between the fissured planes is filled with air, even though
the apparent volume of the grain may not have changed. This phenomenon also
could take place whenever rice is overdried. Data are needed to quantify this
hypothesis so it can be confirmed or rejected.
Tests by Stipe et al (1972) showed that reduction in head-rice yield was influ-
enced by the amount of moisture removed within a time interval, rather than by the
temperature of the drying air. Only a small reduction in head-rice yield resulted
from 1.0% moisture removal with heated-air temperatures of 232°C (450°F) during
a 0.125-min time interval. A greater reduction was observed when 2% moisture was
removed with 49°C (120°F) air during a 5-min time interval.
While the primary author was a rice processing consultant at the Indian Institute
of Technology at Kharagpur, India, in 1975, he was asked to conduct rice research.
He suggested simple experiments such as drying rice in a bed that was too deep,
drying rice with too high a temperature, drying rice while using too long a pass, or
drying rice initially at too high a moisture content. The consistent reply to these
suggestions was “This has already been done and we know that the results are
negative.” His final suggestion was “Well, let’s do everything wrong and see what
the results will be, i.e., use freshly harvested rice at too high a moisture content
(29.8%), use a drying bed that is too deep (63.5 cm [25 in.]), use a drying
temperature that is too high (59°C [138°F]), and use a drying pass that is too long
(12 hr).” After drying, the rice was milled and produced a total milling yield of 76%
and a head-rice yield of 74% (Kunze, 1979). But soon after milling, the grains
began to fissure, and within a few hours, all the grains had one or more fissures.
This type of response from rice that had been dried was unprecedented and most
intriguing.
Kunze (1979) pursued the information further and concluded that 1) the grains
were not fissured at the end of drying, 2) the grains fissured after drying, and 3) a
time interval after drying was required before the fissures developed (Fig. 7).
Although the moisture gradient was not measured, it is believed that the subsiding
moisture gradient after rapid drying caused the fissures. The fissuring continued
until the moisture gradient essentially equalized across the grain.
A continuing literature search revealed that the foregoing observation was not
completely new. Craufurd (1963), working in Sierra Leone, West Africa, was the
first to report that fissures develop in the rice grain after rapid drying. He worked on
adsorption and desorption characteristics of raw and parboiled rice (paddy). His
observation of fissures after fast drying was for parboiled rice. In a given sample, no
fissured grains developed during the first half hour after drying, 8% of the grains
were fissured after the first hour, 29% after the second hour, 49% after the third
hour, and 62% after the fifth hour. Henderson (1954) had observed half-cracks as
well as whole cracks in rice grains. The half-cracks were located at the center of the
grains and progressed toward the outside as further cracking developed. Therefore,
Craufurd (1963) reasoned that the cracks originated at the center of the grain. This
was presumed to occur while the center of the grain was losing moisture to the
dryer outer layers. Such a moisture migration would cause the outside of the grain
to expand and elongate while the inside of the grain would contract and shrink.
These simultaneous actions would cause the grain to fissure across its long axis
from the center to the surface.
Eight years later, Ban (1971), working in Japan, reported the same phenome-
non—the grains did not fissure during rapid drying but rather after drying. Indica-
tions are that Ban was not aware of Craufurd’s work, since Craufurd was not listed
as a reference. His observation caused the rice industry in Japan to propose that the
number of cracked grains, in a sample taken from a dryer, should not be determined
until 48 hr after drying. He dried rice from 20 to 12.5% mc with a moisture-removal
rate of 1.5% per hour, and few fissures developed at any time after drying. Faster
rates were possible, without fissure damage after drying, if drying was stopped at
17–18% grain moisture content. Other rice that was rapidly dried to storage
moisture was sealed under airtight conditions in which neither drying nor moisture
adsorption occurred, and the grains fissured with time for 48 hr.
Fig. 7. Fissures that developed in brown rice after rough rice in equilibrium with ambient air was dried
(10 hr at 60°C) and hulled. Brown rice was formed into a rosette. Light from the center of the rosette
(under the disk) diffuses through the entire length of unfissured grains. A fissure in a grain breaks the
light transmission, resulting in less illumination after the fissure than before it. Picture at top left was
taken 1 hr after drying; top right, 4 hr; middle left and right, 6 hr (note grain in eight o’clock position);
bottom left, 6.5 hr; and bottom right, 24 hr. (Reprinted, with permission, from Kunze, 1979)
There was, however, a much earlier observation of the fissuring-after-drying

phenomenon reported by Earle and Ceaglske (1949) with the drying of macaroni.
When drying was stopped with a large moisture gradient in the product, the
macaroni would crack after drying. This phenomenon puzzled the commercial
operators, because the macaroni looked and felt strong as long as the drying was
in progress. Soon after the drying fans were stopped, the macaroni began to
crack. So the fissuring-after-drying observation by Earle and Ceaglske (1949)
with macaroni predates the observations of Craufurd (1963), Ban (1971), and
Kunze (1979) with rice.
Sharma and Kunze (1982) found that most grains fissure within 48 hr after
drying but that additional fissures develop at a slower rate for another 72 hr there-
after.
The foregoing observations have a strong message for dryer operators and their
drying procedures. The rice industry has the universal practice of not taking milling
quality samples immediately after a lot of rice is dried. Why? The milling quality of
rice may change after drying. The potential for this change is dependent on how the
rice was dried.
Nishiyama et al (1979) developed a “crack generation equation” that, with
appropriate constants, was used to fit the data points developed by Sharma and Kunze
(1982) for rice grains that fissured after drying. K. Kato and R. Yamashita
(unpublished data, 1979) used temperature-controlled environments to store rice after
rapid drying. They found that fewer grains fissured at higher storage temperatures.
Other storage atmospheres, including both temperature and relative humidity,
were investigated by Nguyen and Kunze (1984). They found that storage at 10% rh
and 45°C helped to further reduce the fissuring-after-rapid-drying phenomenon, but
they did not achieve the goal of eliminating the fissures.
Further work was done by Cnossen and Siebenmorgen (2000). It is discussed
under “Glass Transition Relationships” later in this chapter.
Other cereal grains with a hard vitreous endosperm show the same fissure-after-
drying characteristics. White et al (1982) studied stress cracks in popcorn that
developed as a result of drying. They concluded that stress cracks in popcorn are
caused by the redistribution of moisture after the popcorn has been removed from
the dryer. The delay before the stress cracks developed was related to the moisture
gradient in the kernel when the drying was stopped.
Few, if any, good rice grains fissure during drying under normally accepted good
drying practices. Rice grains that may appear to fissure during drying are usually
imperfect grains with scars, scratches, insect damage, impact damage, or other
defects that allow moisture to exit the defective area faster than from other portions
of the grain. They may be cracked because they were impacted or subjected to
mechanical stress during harvesting, handling, and transport operations. Other
grains may already have fissured in the field before they were harvested.
However, drying creates the potential for low-moisture grain to readsorb
moisture even while the drying is proceeding. A low-moisture grain located ahead
of the drying front in a dryer may fissure because it adsorbs moisture from the
warm, humid drying air. The grain fissures in the dryer not from drying but from
moisture adsorption; moisture is initially added to the low-moisture grain and not
extracted from it. To test this phenomenon, a dryer operator can simply place good,
whole, dried rice grains of any form into the warm, humid exhaust air of a dryer for
an hour or more and then inspect the grains for fissures.
In general, the major (most prevalent) contributor to fissured rice grains is

moisture readsorption by low-moisture grain 1) in the field, 2) in the combine hop-
per, 3) in the field cart, 4) while in transport, and 5) in the dryer, as well as 6) by
rapid readsorption of moisture any time after drying. It is crucial for rice producers,
dryer operators, storage managers, and processors to be keenly aware of moisture
readsorption effects on low-moisture rice grains. The potential for and the effect of
moisture readsorption by rice must be understood at every step of the processing
chain from before harvest in the field until it is packaged in an airtight container for
commercial sale.
The foregoing information is focused primarily on the desorption and adsorption
of moisture by the rice grain. Little emphasis is given to the properties of drying air
(psychrometrics) and the principles of air movement. Excellent discussions of such
topics can be found in books on drying written by Loewer et al (1994), Brooker et al
(1992), and Hall (1980).
MODELING GRAIN MOISTURE
When rough rice is dried slowly or in passes, the magnitude of the resulting
moisture gradient within the kernel is small. If the grain is tempered with aeration
after drying, the moisture moving to the grain surface is removed as the moisture
gradient declines. Low or small moisture gradients do not cause a rice grain to
develop surface cracks or internal fissures. Drying stresses cause a grain to be under
tension at the surface and under compression at the center. These tensile stresses
could cause cracks to develop in the surface of the grain during drying. Such cracks
can be easily produced by rapidly drying milled rice. However, under recommended
drying procedures, the rough-rice grain, with its husk and bran layers, normally
does not dry fast enough to produce these tensile cracks in the grain surface.
The grain failure that is more significant and of greater interest is the fissure that
develops across the longitudinal axis of the grain after, not during, rapid drying.
Such a drying procedure causes a high moisture gradient in the grain at the end of
drying. Accurate descriptions of grain drying rates are dependent on accurate
descriptions of moisture movements in the grain. Scientists have assumed that the
moisture diffusion mechanism controls the drying rate.
Gradients
Sarker (1993) conducted research to determine the moisture gradients in rice

while the grain was drying. The finite element method (FEM) was used to predict
these transient gradients and their distribution in the rice grain. Gradients were also
modeled for a period after drying had ceased. This technique was specifically
developed to help understand the moisture transport behavior in rice, but it could be
used with other cereal grains as well. It has the potential from which to develop a
system that would minimize energy consumption for drying and reduce the potential
for fissures to develop in the grain after drying has ceased.
The analysis by FEM (Sarker, 1993) showed that a higher moisture gradient existed
along the short axis of the grain than along the long axis. The moisture removal rate
was greatest perpendicular to the long axis of the grain. Peak gradients occurred within
15–60 min after drying started. Higher moisture gradients developed for higher initial
grain moisture contents and also for higher drying air temperatures.
A low moisture gradient in a low-moisture grain may cause that grain to fissure
after drying, while a gradient of the same magnitude in a higher-moisture grain does
not cause a fissure to develop. The high-moisture grain may still be sufficiently
elastic as the moisture gradient in it declines. A high moisture gradient in a high-
moisture grain (above 15%) is usually not detrimental. When 45°C drying air was
used to dry rice with an initial moisture content of 22.7% for 2 hr to a moisture
content of 12.95%, 36% of the grains fissured after drying; but when the same dry-
ing air temperature was used to dry rice with essentially the same initial moisture
content for 18 hr to a moisture content of 7.07%, then 74% of the grains fissured
after drying. Samples dried for 24 hr under the same conditions experienced 94%
fissured grains (Sarker, 1993). If rapid drying is stopped after a short drying period,
the fissures develop soon after drying but may not develop at all. If rapid drying is
stopped after longer periods of drying, more time is required before the fissures start
to develop, and the fissures then develop over a longer period of time—but they do
develop.
If the drying air temperature is high enough and the initial grain moisture content
is also high, the beginning of a parboiling effect may be produced, and fewer grains
fissure. There are some indications that high-moisture grains subjected to a very-
high-temperature drying environment seem to develop a case-hardened, stable outer
surface. Further drying causes the grain to pull itself apart (fissure) on the inside.
This phenomenon is not observed very often. But when it does occur, it presents a
problem to the processor.
According to Sarker (1993), 2 hr of drying at 55°C produced a large moisture gradient
but removed less moisture than did 24 hr of drying at 35°C. Yet the grains dried at 55°C
for 2 hr fissured much more (98%) than those dried for 24 hr at 35°C (4%).
After drying, residual moisture gradients influenced the retardation time and the
rate at which the fissures subsequently developed in the grains. Gradients of greater
magnitude caused grains to fissure soon after drying (short retardation time), or the
grains did not fissure at all. In this case, the fissure response was fast, but the num-
ber of grains that fissured was small. Longer drying periods caused gradients of a
smaller magnitude and more time was required (longer retardation time) before the
fissures started to develop, and their rate of development was slower (fissured over
a longer period of time), but more grains fissured. When rapidly dried grains are
exposed to the ambient air, the dry surface cells readsorb moisture from the envi-
ronment and expand. This additional moisture on the grain surface complements
moisture migrating from the grain center to the surface to produce the surface com-
pressive stresses that cause the grain to fissure. However, if rapidly dried grains are
sealed in an airtight container immediately after drying, they fissure without the
assistance of the environmental moisture (Ban, 1971; Kunze, 1979).
The residual moisture gradient in the grain after drying and the grain’s final
moisture content appear to be the primary factors that determine whether a grain
will fissure or not. Research to date is inadequate to define the fissuring-after-
drying phenomenon for the infinite number of drying conditions that exist. Possible
factors that determine the magnitude of a high moisture gradient after drying
include drying-air temperature and relative humidity, drying airflow rate, grain ini-
tial and final moisture contents, length of the drying period, grain dimensions, vari-
ety and/or type of rice, and perhaps others. These variables are not independent of
each other. Dryer operators need to be aware of the fissuring-after-rapid-drying
phenomenon in order to adjust their drying and tempering procedures for a
maximum volume output and a minimum of after-rapid-drying fissures in the grain.

For more and detailed information on the fissuring-after-rapid-drying phenomenon,
the reader should refer to work by Sarker (1993), Sharma and Kunze (1982), Kunze
(1979), and Ban (1971).
Stresses
Milling quality losses must be related to their particular sources. Numerous

sources have the potential to fissure grains or to produce milling quality losses. The
loss must be identified with the source if the source is to be mediated or eliminated.
Lan (1994) used the FEM to study fissures related to moisture adsorption by a low-
moisture grain. This is the type of grain failure that develops in the field before the rice
is harvested and also the type that develops after drying when the grain is dried too fast
and a steep moisture gradient remains in the grain after drying has ceased.
Lan (1994) developed numerical solutions to the moisture diffusion equation and
the stress–strain relationship equation. One finite element model was used to deter-
mine moisture diffusion within the grain, while another was used to simulate the
internal expansion and contraction in terms of strains and stresses. He found axial
stresses to be the primary reason for the development of fissures across the longitu-
dinal axis of the grain. Tangential stresses were found to be of sufficient magnitude
to produce secondary fissures parallel to the longitudinal axis of the grain. The larg-
est tensile stresses from moisture adsorption were found to be at the center of the
grain, where fissures usually start.
According to Lan (1994), milled rice fissured the fastest (retardation time was
the shortest) after exposure to a moisture adsorption environment; brown rice
required twice as much time to fissure as milled rice; and rough rice required three
times as much time to fissure as brown rice. Moisture adsorption rates were com-
mensurate with retardation times; fast moisture adsorption resulted in a short retar-
dation time, and slow moisture adsorption resulted in a long retardation time before
fissures began to develop. Moisture gradients in a grain increased rapidly after the
grain was exposed and reached a peak within 5-min after the exposure was started.
This indicated a high moisture level at the grain surface while there was yet no
moisture change in the grain interior. High-moisture starch granules at the grain
surface were under compression, but the total area under compression was not large
enough to fissure the grain. After more time, as moisture moved deeper into the
grain, the moisture gradient gradually declined, while the area within the grain
under compression increased. The interior area under tension decreased. When the
total compressive stresses near the grain surface exceeded the tensile strength of the
grain interior, the grain fissured. Equations were developed to calculate the mois-
ture adsorption rates for the Lemont variety of rice in rough, brown, and milled
form for a given exposure time, exposure relative humidity, initial moisture content,
and final moisture content at a temperature of 21°C (70°F). Similar equations were
developed to calculate the cumulative percentage of fissured kernels at any time
after exposure in a sample under the same initial conditions and a similar exposure
to moisture. For more detailed information, the reader may refer to work by Lan
(1994), Lan and Kunze (1996), and Lan et al (1999).
Other researchers have modeled the preharvest stress cracking of the rice grain in
the field. Such works have been reported by Lague (1990) and Lague and Jenkins
(1991a,b).
Glass Transition Relationships
Cnossen and Siebenmorgen (2000) studied the change of state of brown rice
kernels over a range of grain moisture contents. For medium- and long-grain rice
with high moisture content, they found that 5 to 6 percentage points of moisture
content could be removed per drying pass if the rice was tempered sufficiently
immediately after drying and at temperatures above the glass transition temperature.
Kernels could then be cooled to 21°C (70°F) without loss in head-rice yield. Needed
tempering durations were shorter for long-grain rice than for medium-grain rice.
According to Cnossen and Siebenmorgen (2000), a rice grain may remain in the
rubbery region during high-temperature drying and then cross the glass transition
line into the crystalline (glassy) state during cooling. Or with low-temperature
drying, the rice grain may change from the rubbery to the crystalline state during
the drying process. The glass transition temperature line separates the states of the
grain into rubbery and crystalline regions.
In the case of slow, low-temperature drying, the rice grain may never leave the
crystalline state as it dries to storage moisture content. In the case of fast, high-
temperature drying, the grain initially moves from the crystalline to the rubbery
state as it is heated and then moves back to the crystalline state before it reaches
storage moisture content at ambient temperature. Currently, there is considerable
interest in determining how milling quality during drying may be affected as the
lower-moisture grain crosses the glass transition line in its approach to storage
moisture content and temperature.
No research has been done to specifically relate the fissuring of rice grains after
moisture readsorption to the transition of the grain from the crystalline to the
rubbery state (adsorption) as fissures develop. Fissures develop in the rice grain
while it is in the field whenever the moisture content drops to 15% or below. In this
case, the rice grains are in the crystalline state well below the glass transition
temperature line.
Considerable progress has been made in recent years to more clearly define what
happens when the rice grain is dried. More needs to be done. Until the fundamentals
of rice drying are brought into clear focus, there is little prospect that significant
advances will be made. Once participants in the rice industry relate these
fundamentals to problems arising from grain fissuring, opportunities for new and
exciting advances will be greatly increased.
COMMERCIAL RICE DRYING
Equipment and procedures for commercial rice drying vary from one country to
another because of different levels of mechanization as well as social and economic
practices. Generally, large, centralized rice-drying installations are found in the
United States, and smaller, more decentralized installations are in countries where
rice harvesting is less mechanized. Commercial rice-drying facilities in the United
States include farmer-owned cooperatives, rice mills that purchase and dry freshly
harvested rice, and privately owned dryers.
Each facility is generally equipped with one or more continuous-flow, heated-air
dryers; numerous storage bins; and conveying systems for moving rice, as required,
for receiving, drying, and shipping. At some drying facilities, the identity of each lot
of rice is preserved, but at others, lots of similar variety, quality, and moisture con-
tent are commingled. Preserving identity requires many relatively small bins,
whereas commingling results in better utilization of bin space and less unproductive
time otherwise needed for emptying and refilling dryers.
Since the 1960s, the Japanese rice-drying and -storage industry has adopted the
country elevator concept, as it was formerly known in the Midwestern United
States; these are installations similar to those of the farmer cooperatives. More than
170 such units, with about 2,000 t of drying and storage capacity, were built
between 1963 and 1980 (T. Ban, unpublished data).
Handling Undried Rice
Commercial dryers are often a considerable distance from fields where rice is
grown, and green (undried) rice may be hauled a distance of 25 km or more by
truck to a drying facility. Usually, trucks cannot be loaded directly from the com-
bine, so self-propelled, self-unloading field carts are used to transport the rice from
the combine to a truck. Procedures for receiving rice at commercial dryers in Lou-
isiana and Texas were studied by Slay and Hutchison (1961).
Storage of Undried Rice
Green rice is often received at the commercial dryer at a much faster rate than
drying can be accomplished. To accommodate the maximum amount of rice, the
drying facility must store undried rice for a considerable period of time. Green or
high-moisture rice can heat spontaneously because of respiration and the activity
of microorganisms (molds) when stored in bulk. McNeal (1960) conducted
experiments with small, nonaerated containers. His work showed that a decline in
grade from heating damage occurred within 24 hr in rice with a 24% mc when the
storage temperature was as low as 10°C (50°F). Rice at 15–18% mc showed
damage after only five days when stored at temperatures between 10 and 38°C
(50–100°F).
Dillahunty et al (2000) studied the effect of moisture content and temperature on
the respiration rate of rough rice. Their work showed that the respiration rate
increased exponentially as the rice moisture content increased. The respiration rate
increase was particularly dramatic for rice at or above 15% mc.
Aeration at a suitable rate dissipates heat caused by respiration and microorgan-
ism activity in undried rice and keeps its temperature close to ambient. Arkansas
tests (Aldred, 1953) indicated that an airflow rate of 1.4 cubic meters per minute per
tonne (1.4 m3/min·t = 1 ft3/min·bushel [bu] = 2.2 ft3/min·cwt) was adequate to pre-
vent spoilage in 18- to 24%-mc rice for 7–10 days. Studies in California by
Henderson (1955) showed that adequate aeration allowed rice with 18% mc to be
carried through a wet winter period without quality reduction or a decrease in
viability.
Storage tests of high-moisture rice in Texas by Calderwood (1966) provided data
for the prediction of safe storage times based on the allowable percentage of dam-
aged grains for a grade of No. 1 by the U.S. Standards for Rough Rice. Relation-
ships between initial moisture content, average ambient temperature, airflow rate,
and the predicted days of safe storage are shown in Figure 8. Aflatoxin assays from
rice samples in these storage tests indicated that samples that graded No. 1 seldom
contained more than a trace of aflatoxins (Calderwood and Schroeder, 1968).
Fig. 8. Safe storage time for moist rice, in aerated bins, based on maintaining Grade No. 1 as specified
in the U.S. Standards for Rough Rice. (Reprinted, with permission, from Calderwood, 1966)
The use of chemical preservatives to maintain the quality of undried rice was
studied by Calderwood and Schroeder (1975). Both aerated and nonaerated storage
bins were used to determine effective dosage rates of preservatives in relation to
rice moisture content and storage time.
Continuous-Flow Dryers
A commercial continuous-flow dryer is commonly a large upright structure into

which rice is loaded at the top. The rice then flows down by gravity, being exposed
to the drying air during its descent. Wasserman and Calderwood (1972) described
two types of continuous-flow dryers. When rice in the drying column flows down-
ward in a straight path, the dryer is classified as a nonmixing type; when rice in the
drying section is continuously diverted by baffles, the dryer is classified as a mixing
type (Fig. 9). The rate at which rice flows through the dryer is regulated by feed
rolls or oscillating rocker arms at the dryer outlet.
The nonmixing type of dryer has a simple design that permits rice to flow
between two parallel screens 25–30 cm (10–12 in.) apart. The central section of the
dryer serves as a hot-air plenum from which the heated air flows horizontally
through the column of rice being dried. The rice is completely encased, and no
grains or chaff can be conveyed out of the system with the drying air. This permits a
high airflow, and rates of 112–262 m3/min·t (180–420 ft3/min·cwt) are commonly
used. Drying air temperatures often range up to 54°C (130°F). Since the grains tend
to flow straight downward, rice adjacent to the hot-air plenum is dried the most and
rice adjacent to the exhaust screen is dried the least. Dryer operators have com-
mented that the milling quality of rice depends on where the sample is taken from in
the dryer. In the nonmixing-type dryer, rice grains taken from adjacent to the
Fig. 9. Two types of continuous-flow, heated-air dryers commonly used for the commercial drying of
rough rice. (Reprinted, with permission, from Calderwood, 1975; copyright Tropical Products Institute
[now Tropical Development and Research Institute])
plenum screen may have dried too fast too long and hence could fissure after dry-
ing. Alternatively, low-moisture grains taken from adjacent to the humid-air exhaust
screen may have adsorbed moisture and fissured from moisture adsorption before
the drying pass was completed. Such results appear possible, but no research data
are available to confirm these theories. Grains from all locations across the column
are mixed when rice is discharged from the dryer. Some dryers are equipped with a
turnflow device that reverses the position of the rice grains about midway during
their descent from the top to the bottom of the column. Samples of grain from
which to determine moisture content and milling quality usually are taken from the
discharged mixture.
A mixing-type dryer commonly used for rice drying consists of a vertical com-
partment across which rows of baffles (sometimes called racks or air channels) are
installed. One end of each is open, and the other end is closed. Alternate rows are
open to heated air, and intervening rows are open to the exhaust plenum. Rows of
inlet and exhaust baffles are offset so that the grain stream is split as it descends and
the grains are alternately exposed to hot inlet air and then to warm, humid exhaust
air, and no grain is continuously exposed to the hottest air. Airflow rates of 44–97
m3/min·t (70–155 ft3/min·cwt) and air temperatures of 66°C (150°F) are typical for
mixing-type dryers. The air velocities used are usually high enough to convey chaff
and some immature grains out of the drying column. These materials must be cap-
tured and collected to prevent air pollution.
Concurrent-flow dryers with either two or three drying stages per pass have been
constructed to dry rice (Bakker-Arkema et al, 1982; Fontana et al, 1982). Heated air
flows in the same direction as the rice, thus causing the hottest air to contact the
wettest grain. After passing through the first drying stage, the rice is tempered and
then passed through a second drying and tempering stage. Cooling is accomplished
with air flowing counter to the direction of grain flow. The concept for these units is
interesting, but further development is needed to make them competitive with other
dryers on the market.
Multipass Drying
In multipass drying, rice flows through the dryer several times with a
tempering period between each of the passes. During a pass through the dryer, a
moisture gradient is established in each grain because moisture from the grain
surface has been removed but moisture content in the central portion of the grain
has changed little. The tempering period allows moisture to diffuse from the
central portion to the surface of the grain. During the next pass, the surface
moisture is again removed. As a consequence, the moisture gradient is never
large. Steep or large moisture gradients (caused by rapid drying or by rapid
moisture adsorption) are generally detrimental to grain quality.
In a typical multipass operation, rice is exposed to heated air for 15–20 min
for each pass, and moisture content is reduced by 2–3% of the dry weight.
Between passes, the rice is moved to a bin for tempering for 4–48 hr. The elapsed
time for multipass drying generally exceeds 24 hr, but it also may extend over
several days. However, residence time of the grain in the dryer seldom exceeds
1–2 hr.
Chen (1999) developed exponential equations to predict the drying behavior of
rough rice for an intermittent drying pattern with short drying periods. A two-
equation model was used to fit the predicted points to the experimental points on the
drying curve.
Tempering
The tempering period following a dryer pass provides time for the moisture
gradient within a grain to equalize, with a movement of moisture from the
interior to the external surface of the grain. Observations by Dachtler (1959)
indicated that a continuous gain in drying efficiency occurred when rice was
tempered up to 24 hr. The gain was very rapid during the initial 6 hr and then
became progressively slower, with only a minor gain in drying efficiency during
the final 12 hr of a 24-hr period. Wasserman et al (1964) showed that rice
temperature affected the time required for satisfactory moisture equilibration.
Rice that was cooled to 24°C (75°F) immediately after drying required l6 hr for
adequate tempering, but only 4 hr was required to get equivalent tempering in
rice stored at 40°C (105°F).
Based on small-scale tests of tempering rice in a laboratory, Steffe et al (1979)
determined that a tempering time of less than 3 hr was adequate in multipass drying
of medium-grain rough rice. They expressed the opinion that the tempering times
used in California rice drying were longer than necessary.
Cnossen and Siebenmorgen (2000) used the glass transition temperature relation-
ship to temper rice after rapid drying without incurring subsequent reductions in
head-rice yield. Grains were tempered at the same temperature as that used for
drying. The rice variety Bengal (medium-grain) required 160 min of tempering,
while Cypress rice (long-grain) required only 80 min.
Some dryers are equipped with an ambient-air cooling section directly below the
heated-air section. Cooling before a tempering period is not a recommended prac-
tice when the cooling period is of a short duration. Whenever the next dryer pass
must be delayed for a considerable time, the rice should be cooled in a bin with
aeration to inhibit molding, souring, and yellowing of the grain.
Supplemental Drying with Aeration
Tempering periods are extended deliberately by some dryer operators to take

advantage of the additional drying that accompanies slow cooling by aeration. This
process was investigated in Texas tests (Calderwood and Hutchison, 1961). Rice is
cooled by aeration while it is stored in a bin after each pass through the dryer. As the rice
cools, it gives up moisture even when the ambient air is at a high relative humidity. The
amount of drying is a function of both the amount of cooling (difference between rice
temperature and ambient air temperature) and the moisture content of rice before it is
cooled. A statistical analysis (Calderwood, 1970) showed that the average amount of
drying per cooling period exceeded 1% of the dry weight of rice in observations that
spanned a 10-year period. Each cooling period should be long enough for rice
temperatures at all locations in a bin to approach the average temperature of the ambient
air. An estimate of cooling time versus airflow rate is provided in Figure 10.
Dryer Adjustments
Laboratory-scale experiments in California (Thompson et al, 1955) to determine

the effects of drying air temperature and humidity, airflow rate, and number of passes
Fig. 10. Estimated cooling time for aerated rough rice after a pass through a heated-air dryer.
(Reprinted, with permission, from Calderwood and Hutchison, 1961)
on milling yields and drying time resulted in a graphic summary of these factors (Fig.
11). Methods of applying the diagram to commercial facilities are described in several
publications (Thompson et al, 1955; Wasserman et al, 1956a,b, 1957), and several
reports on the use of this diagram at commercial facilities have been published
(Wasserman et al, 1958a,b; Pominski et al, 1961; Faulkner and Hill, 1962). In general,
it was concluded that 1) when air temperature is increased without changing the
number of passes, drying is faster but head yields are lower; 2) when the number of
passes is increased without changing air temperature, drying is faster and head-rice
yields are higher; and 3) when a proper combination of high temperature and number
of passes is chosen, both high head yield and fast drying are achieved.
An experimental facility, designed to simulate a commercial installation but
smaller in size, was operated in Texas to determine the effects of dryer adjustments
on drying rate and milling quality. In tests with retention times of 15 and 25 min
and with rice temperature and airflow constant, the 15-min retention time resulted
in less dryer usage than did a 25-min retention time for rice lots of equal size
(Calderwood and Webb, 1971).
In tests with a constant temperature control setting for each pass of a particular lot of
rice, with feed rolls set for a 15-min retention time, and with an airflow rate of 154
m3/min·t (247 ft3/min·cwt), drying time was reduced 40% as airflow temperatures were
increased from 43 to 78°C (110 to 172°F). However, the average head-rice yields were
reduced 1.6% by drying at the higher air temperature (Calderwood and Webb, 1971).
When retention time and air temperature were held constant at 15 min per pass
and 66°C (150°F), respectively, and the airflow varied from 112 to 182 m3/min·t
(180 to 291 ft3/min·cwt), the drying rate increased with the rate of airflow, but
milling quality remained unchanged (Calderwood and Webb, 1971).
Fig. 11. Effect of number of passes on head yield and drying time. Airflow rate, 27 m3/min·m2; air
humidity, 5.95–7.93 g of water per kilogram of dry air; bed depth, 5.1 cm; initial moisture content,
23.6%; final moisture content, 13.0%; total milled rice, 74% (dockage-free basis). Equal amounts of
water were removed in each pass. (Reprinted, with permission, from Thompson et al, 1955)
Adjusting the air temperature control to dry at a particular rice temperature

between 35 and 49°C (95 and 120°F), with retention time constant at 15 min per
pass and an airflow rate of 154 m3/min·t (247 ft3/min·cwt), resulted in an average
moisture removal per pass of 1.0 and 2.7% (db) for rice temperatures of 35 and
49°C, respectively (Calderwood, 1972c). Drying treatments during which rice
attained maximum temperatures of 50°C (122°F) appeared to have no adverse
effects on cooking characteristics according to physicochemical tests (Calderwood
and Webb, 1971).
Studies at commercial rice dryers in Louisiana and Texas (Calderwood and
Louvier, 1964) showed that increased drying capacity could be achieved by effec-
tively using aeration for moisture removal following dryer passes and by
reducing rice retention time in dryers while increasing the heated-air temperature.
Drying capacity at some facilities was increased by improving handling methods
(Louvier and Calderwood, 1968). At dryers where the flow of rice to drying units
had been interrupted to receive or ship other rice, drying capacity was increased
by installing additional conveyors so that all operations could function
simultaneously. Garner bins, installed above the drying units, were recommended
to save time required for filling a dryer.
Preheat Drying
A prototype infrared preheater for use with the Louisiana State University-type
dryer was developed by Faulkner and Wratten (1966, 1970) following laboratory-
and pilot-scale tests. With the preheat method, rice was heated rapidly while falling
between gas-fired infrared heaters on its way to the continuous-flow dryer. With
preheating to 60°C (140°F) and retention time in the continuous-flow dryer of 2–3
min with a drying air temperature of 49°C (120°F), rice moisture content was
reduced approximately 2% in each pass with no apparent damage to milling yield.
Resident time for drying from 20 to 12% mc was estimated to be a total of 12 min,
resulting from four 3-min passes. This compared with total residence times of 60–
120 min to do the same amount of drying by conventional methods.
Combination System
Plant capacity can be increased by combining heated-air multipass drying with

deep-bed ambient-air drying (Wasserman et al, 1969). High-moisture rice is first
dried with the multipass method to about 18% mc, and then the partially dried rice
is transferred to a deep-bed drying bin where drying is continued until the moisture
content is reduced enough for safe storage. This system has the potential to produce
high-quality rice because 1) drying is faster and rice breakage lower when heated-
air drying is used only at high moisture contents, and 2) breakage is lower when
ambient air drying is used at low moisture contents. Some deep-bed dryers hold up
to 4,500 t (100,000 cwt), with rice depths of 9–11 m (30–35 ft). Several dryer oper-
ators have encountered difficulties when rice was improperly cleaned or leveled,
when airflow was too low, when rice was stored at too high a moisture content,
when air humidities were too high for extended periods of time, or when chaff was
allowed to segregate during filling of the deep-bin dryer. In general, however, the
results have been good, and the combination system of drying is practiced
extensively in California.
BIN DRYING
Bin drying, in which air is moved through rice in a deep bed, generally 2.5–3.7
m (8–12 ft), is fundamentally different in principle from drying with continuous-
flow dryers, in which the thickness of the rice pack is limited to 25–30 cm (10–12
in.). The bed depth limits the airflow rate because of power requirements. The
ambient air is unheated or only slightly heated to avoid overdrying or damaging the
quality of the rice that is first exposed to the drying air, and dangerous moisture
gradients within individual grains are thus avoided. Bin drying is widely used as an
on-farm method and has limited application in commercial drying.
Equipment
Bins built of steel, wood, or concrete have been adapted for drying. The main
consideration is to provide a tight structure, thereby preventing the leakage of air
and moisture through bin floors and walls. An ordinary grain storage bin can be
converted into a bin dryer by installing an air-distribution system. Three types of
air-distribution systems are shown in Figure 12. A round, steel bin with a
perforated floor is the kind of structure most commonly used for bin drying.
Either axial-flow (propeller) or radial-flow (centrifugal) fans are suitable for
bin drying, but axial-flow fans are most commonly used because of their simple
construction, ease of installation, and low cost. Automatic gas heaters commonly
are installed between the fan outlet and the air plenum chamber of the bin.
Airflow Requirements
The direction of air movement has little effect on dryer performance; however,
fans are commonly installed for upward air movement for the following reasons
(Sorenson and Crane, 1960): 1) after some drying, the wettest layer of rice is at the
top, where sampling is easily accomplished, 2) heat from the motor and fan can be
utilized in drying, and 3) the wettest rice is most frequently the first loaded into the
bin and the first to be dried.
The airflow rate required for successful drying depends on the initial moisture
content of the rice and on weather conditions. Rice must be dried to 15% mc within
a relatively short time to prevent quality damage (discolored milled grains) from
molds. Research by Sorenson and Crane (1960) in Texas indicated that the allow-
able time for drying to 15% mc is about six to eight days when rice temperature
ranges between 26.7 and 30°C (80 and 86°F). Drying from 15% mc to 12.5% mc
for marketing can be done over a long period of time because rice within this range
is less susceptible to molds. An airflow rate of 2.7 m3/min·t (4.4 ft3/min·cwt) was
found to be the lowest satisfactory rate in Texas for rice having an initial moisture
content of 20%. The recommended airflow rate is 3.5 m3/min·t (5.6 ft3/min·cwt),
which provides a factor of safety for drying wetter rice during unfavorable weather
conditions.
In the California rice-growing area, lower airflow rates are recommended for bin
drying because the climate is normally less humid and cooler. Drying tests
(Henderson, 1958) indicated that satisfactory airflow rates were 1.6 and 2.5
m3/min·t (2.5 and 4.0 ft3/min·cwt) for rice having initial moisture contents of 20 and
25%, respectively.
Fig. 12. Types of air-distribution systems used in a deep-bed dryer. A, Perforated false floor; B, main duct
and laterals; and C, perforated ducts. (Reprinted, with permission, from Sorenson and Crane, 1960)
Depth of Rice
With the recommended airflow rates of 3.5 m3/min·t (Sorenson and Crane,
1960), rice to be dried should usually be no deeper than 2.5–3 m (8–10 ft) for eco-
nomical use of power. If the initial moisture content exceeds 20%, the depth should
be reduced to obtain a higher airflow rate. If the initial moisture content of rice is
well below 20%, the design depth of 2.5–3 m may be exceeded. Greater depths also
may be used after rice in the top layer is reduced to 15% mc. For example, 15%-mc
rice from another bin could be added to the top of 15%-mc rice in a given bin,
resulting in a depth greater than 3 m. With this greater depth, the airflow rate is
reduced and a longer period of fan operation is required before the moisture content
is reduced to 12.5% in all parts of the bin.
Fan-Operating Schedule
In California, Henderson (1958) recommended that the fan be operated continu-

ously, except during heavy rains and dense ground fog, until the moisture content
was reduced to 15% or less. He suggested that after the third week of October, fans
be operated only when the sun was shining.
The following fan-operating schedule was recommended for unheated-air drying
in Texas (Sorenson and Crane, 1960). Start the fan as soon as the distribution sys-
tem is completely covered with rice; operate the fan continuously as the bin is filled
and continue its operation until the moisture content in the top 0.3 m (1 ft) is
reduced to 15%; and then operate the fan only when the relative humidity is less
than 75% and continue drying until the moisture content of the top 0.3 m of rice is
reduced to 12.5%.
Air Delivery and Fan Power
An estimate of air delivery in a bin dryer can be made by measuring the static
pressure under the perforated floor and the grain depth in the bin. The airflow rate
for different static pressures and depths of rice is shown in Figure 13, along with the
Fig. 13. Static pressure required to develop different airflow rates through rice depths from 1.8 to 3.7 m
and fan power requirements in a 7.3-m-diameter bin. Power in a bin of another size is proportional to
the ratio of its cross-sectional area to that of a bin with a 7.3-m diameter. (Reprinted, with permission,
from Calderwood, 1975; copyright Tropical Products Institute [now Tropical Development and
Research Institute])
fan power required for a bin with a 7.3-m (24-ft) diameter. The airflow rates are
based on data published by Shedd (1953).
The static pressure in the plenum includes both the pressure drop in the bed
of rice and an additional pressure drop of 60 Pa (0.25 in.) through the air-
distribution system (Sorenson and Crane, 1960). A static fan efficiency of
62.8% was assumed in calculating fan power. In a bin having a diameter other
than 7.3 m (24 ft), the fan power is directly proportional to the ratio of its
cross-sectional area to 42 m 2 (452.4 ft2), the cross-sectional area of a bin with a
7.3-m diameter.
Supplemental Heat
Supplemental heat may be used to improve the effectiveness of bin dryers during
periods of high relative humidity and cold weather. With supplemental heat, the
drying operation can be continuous without nightly interruptions. However, over-
drying and quality loss may result, along with the additional disadvantages of
greater equipment and fuel costs and the need for closer management. When natural
or liquid petroleum gas is used as fuel, the combustion products are mixed with the
drying air. Appropriate safety controls, which are usually dependable and inexpen-
sive, must be used with all gas burners.
Recommendations for California (Henderson, 1958) and Texas (Sorenson and
Crane, 1960) are 1) use supplemental heat only when the relative humidity is above
75% for a prolonged period, 2) provide only enough heat to raise the air temperature
6.7°C (12°F), and 3) limit air temperatures to 29°C (84°F) in California and 35°C
(95°F) in Texas.
In the 1980s, much effort was expended throughout the world to promote the use
of the rice hull as a fuel for supplemental heat in rice drying (Beagle, 1978).
Progress was made, but rice hulls are not a convenient source of energy in
developed nations. Until the supply of petroleum fuels becomes more critical and
expensive in these areas, the rice hull will probably not be used as a primary source
of energy for supplemental heat in rice drying. In countries where petroleum fuels
are scarce, rice hulls have become a common energy source for both primary and
supplemental heat in the drying of rice. Hulls burn at a temperature between 800
and 1,000°C and yield between 11,630 and 13,956 kJ/kg (5,000 and 6,000 Btu/lb).
With a burner efficiency of 30%, the hulls provide more than adequate energy to
dry field rice from 25 to 12% mc.
The use of solar energy to provide supplemental heat for bin dryers was studied
by Calderwood (1981). During midday hours, when substantial amounts of solar
heat were collectible, the relative humidity of ambient air generally was lower than
75%, so supplemental heat was neither needed nor desirable, based on the recom-
mendations of Sorenson and Crane (1960). In some of the tests, heat collected in the
daytime was stored in a rock bed and then extracted at night or at other times when
the relative humidity was above 75%. This practice was determined to be impracti-
cal, because storing the heat directly in the bed of rice was more economical and
caused little or no reduction in head-rice yield when the average daily heat input did
not exceed 2.2 kWh/m3 of rice. This amount of heat reduced drying time by 25–
65% and electrical energy for fan operation by 24–34% compared with the levels
used in drying with unheated air.
Stirring Augers
Vertical stirring augers have been installed in a large number of rice-drying bins.
Operation of these augers causes rice at a low level in a bin to be elevated and then
deposited on the surface of the rice stack. These augers generally are about 5 cm (2
in.) in diameter and require only a small amount of power for operation. They are
designed to move through a grain bin in a set pattern and contact grain at every
location at frequent intervals. This movement tends to loosen grain that may have
been packed excessively during loading, and the mixing action results in a uniform
moisture content throughout the entire depth. A uniform moisture content is benefi-
cial because all the stored rice then experiences a measure of drying within a short
time, and a zone in which no drying takes place for a long time is eliminated. Thus,
dryers equipped with stirring augers may be loaded to a greater depth or may con-
tain rice at a higher moisture content than was recommended by Sorenson and
Crane (1960) for dryers without stirring augers.
Handling of Materials
Mechanical unloading of round-bin dryers is accomplished by discharging rice

through an opening in the center of the perforated floor to a conveyor under the
floor. Rice that will not flow by gravity is brought to the center opening by a
sweep auger. The screw conveyor installed below the perforated floor moves rice
from the bin. A variety of unloading equipment is available for different bin
systems.
A common arrangement is to group several drying bins around a large circle with
a cone-shaped dump pit in the center. The bins are filled by a portable loading auger
that moves grain from the pit and discharges into any of the bins or into a truck. The
bins are unloaded through a center opening, and rice is conveyed to the dump pit for
further handling by the portable loading auger (Fig. 14).
Another arrangement is to install the bins in a straight line together with two
horizontal conveyors, a dump pit, and a bucket elevator. Rice is moved from the
Fig. 14. Rice drying bins installed on the diameter of a large circle with a portable loading auger to
move rice from a centrally located dump pit.
dump pit with the bucket elevator, which discharges into a horizontal conveyor or
a truck. This conveyor is located above the bin roofs and spouts into any of the
bins. The other conveyor, near ground level, moves rice from any of the bins to
the dump pit. A variation of this method is to provide a bucket elevator tall
enough to spout directly into any of the bins and eliminate the top horizontal
conveyor.
Spreaders are used to eliminate the manual labor required to level rice in bins.
Some spreaders sling grain at high velocities and may increase packing because heavy
material tends to fall to the outside and light material tends to settle in the center.
Nonuniform airflow and drying may result if slingers are not operated carefully.
CORN AND SMALL-GRAIN DRYERS
Several different models of heated-air dryers that are widely used for drying corn
and small grains were used for drying experiments with rough rice at the request of
the manufacturer or his representative. Each of these dryers was found to be adapt-
able for rice drying. Head-rice yields of dried samples were comparable to those of
air-dried control samples when moderate drying temperatures were used. Moderate
air temperatures dried rice more slowly than did the hot temperatures that generally
were suitable for drying corn and small grains.
OTHER DRYING METHODS
Fluid-Bed Drying
Fluid-bed drying uses an air velocity high enough so that rice is suspended in,
totally surrounded by, and sometimes transported by drying air. Faulkner and
Wratten (1970) used a pilot-scale, fluid-bed dryer to dry rice preheated with infra-
red energy. An industrial-type, fluid-bed dryer was used by Calderwood (1972b).
Drying from 17.4 to 12.6% mc was accomplished in a single pass by using an air
temperature of 66°C (150°F) and a rice flow of 0.84 kg (1.85 lb) per minute. Rice
discharged from the dryer at a temperature of 49°C (120°F). No significant drop in
milling yield was noted, compared with that of unheated-air-dried control samples;
however, the fan power requirement was excessive.
Infrared Drying
Heating rice by infrared radiation resulted in rapid drying without excessive

breakage in laboratory rice-drying experiments (Schroeder and Rosberg, 1959,
1960; Schroeder, 1960, 1961). Grains spread in a single layer were dried at rates in
which the loss of moisture content ranged from 0.49 to 3.60% per minute. Germi-
nation decreased as temperature and rate of moisture removal increased. Faulkner
and Wratten (1966, 1970) used infrared radiation to rapidly preheat rice before it
was passed through a heated-air dryer.
Dielectric Drying
Radio frequency energy within the range of 5–40 MHz was used in Louisiana as
the source of heat for drying rice in laboratory experiments (Dachtler, 1959). Rice
was dried from 21 to 13.5% mc in three passes, each of 5-min duration. The advan-
tages of this method are rapid drying and uniform application of heat in the rice
mass. This method has not been used for large-scale drying because of excessive
equipment costs.
Microwave energy at a frequency of 2,450 MHz was used as the heat source for
laboratory-scale drying experiments by Calderwood (1972a). The temperature of
rice increased rapidly as a result of exposure to microwave energy, but little or no
drying occurred during this time. Drying took place afterward as rice was cooled to
ambient air temperature by aeration. Drying per cycle of heating and cooling varied
from 1% of the dry weight of rice for a 14°C (25°F) temperature change to 5% for a
50°C (90°F) temperature change. Using high temperatures to increase the amount of
drying reduced milling quality. In this respect, drying by heat generated internally
by microwave energy was no better than conventional heated-air drying methods in
which heat is applied externally.
Sack Drying
A sack dryer consists of a fan, heater, and plenum chamber with openings covered
by bags of rice. This kind of dryer may be used when handling is exclusively in bags.
It is an effective method of avoiding varietal mixing of seed rice, but the higher labor
requirements involved in sacking and other handling have caused many seed producers
to abandon this method in favor of a bulk handling system. Research in Texas (Aldred,
1953) indicated that a high milling yield and good germination resulted when a sack
dryer was operated with a maximum air temperature of 43°C (110°F) and an airflow
rate of 4 m3/min (140 ft3/min) per 50-kg (110-lb) sack of dried rice. Turning the bags
of rice once during the drying operation was recommended.
DRYING IN COUNTRIES OTHER THAN THE UNITED STATES
Since rice is a universal food, grown throughout the world wherever the climate
is suitable, the crop is harvested and dried in many ways. These methods range from
the intensive use of human labor in developing countries to highly mechanized sys-
tems in developed countries. Much could be written about the labor-intensive and
partially mechanized systems of harvesting and drying, but this has already been
done.
Araullo et al (1976) are the editors of a publication that extensively treats the
harvesting and threshing, drying, storing, handling and conveying, and milling
operations for rice. The publication is thorough and comprehensive in its coverage,
with many contributors from both developing and developed countries. The princi-
ples of rice drying are largely the same whether the grain is sun-dried in a thin layer
on a village street or dried in a sophisticated heated-air dryer at a large grain eleva-
tor. The goal in both cases is to maintain milling quality while removing sufficient
moisture to store the grain.
Another appropriate reference is the technical handbook for the postharvest rice
industry in developing countries by Wimberly (1983). It treats such subjects as the
cleaning, drying, conveying, storing, and milling of rice. The publication is well
illustrated and provides meaningful information for those who wish to broaden their
knowledge of equipment used for the harvesting and drying of rice throughout the
world.
CONCLUSIONS
In 2001, world rough-rice production was 588.9 million tonnes (U.S. Census
Bureau, 2002.). India and the People’s Republic of China produced over one half
(53%) of the total crop. Rice is the staple food for one half of the world’s
population. The United States produces about 1.6% of the world crop but is one of
the primary exporters of rice.
World rice production figures are massive and difficult to comprehend. The
annual production would build a highway of rice 1 m deep and over 26.4 m wide
around the world. Yet the basic component that composes the mass is the small rice
grain. If individual grains of rice are lost in fields throughout the world, such losses
can amount to many tonnes of rice being lost; if individual grains become fissured
in the field, then hundreds of tonnes of rice can become fissured; if individual
grains become defective because of improper drying at a dryer, then tonnes of rice
are likely to experience this degradation; and if individual grains fissure after
drying, many tonnes of rice may experience these fissures. Whatever the loss or
degradation, the individual grain is the basic component that reacts to a condition.
There have been no great changes in rice-drying practices since the second
edition of this book. The emphasis has been on more effective and efficient
moisture removal. Now another very important factor has entered the rice-drying
arena. That is moisture readsorption. It was previously a factor of no interest and,
therefore, was slighted, disregarded, and ignored. It was on the opposite end of the
spectrum of the goal to be achieved. Moisture readsorption has risen to the forefront
and has become a major consideration in the rice-drying process. While the rice grain
is in the field maturing and ripening, drying and moisture readsorption are
inseparable. Drying during the day is what causes and makes possible moisture
readsorption during the evening and night. Moisture readsorption provides the answer
to the question of milling quality loss in the drying process, an answer that scientists
have searched for. Moisture readsorption by the dried rice grain has been the missing
component that explains the inconsistent and incoherent results obtained in rice-
drying research since its inception. Moisture readsorption potentials need to receive
predominant consideration in future field and forced-air rice-drying research.
A second rice-drying phenomenon that has received inadequate attention is the
fissuring-after-drying aspect of the drying process. The practice of not taking sam-
ples immediately after a lot of rice has been dried is adhered to throughout the
world. However, the justification for this practice has only recently become
obvious and understood. If a rice grain is dried rapidly to near storage moisture or
below and the drying is stopped while a steep moisture gradient exists in the grain,
the grain will fissure after it has been dried.
The sources for perhaps the greatest milling quality losses when drying rice to
storage moisture have been clearly identified, but the body of knowledge that may
be used to eliminate or even ameliorate these sources for milling quality losses is
meager. To the author’s knowledge, no major sources of research funding have yet
set their focus on these targets. These should be initiatives of high priority to those
who have the privilege and the responsibility of appropriating research funding.
Field-scale studies with rice are necessary to provide the data that the industry
needs for the adoption of a process or procedure. Most current postharvest rice
technology is the result of field studies. Yet much about the individual rice grain
can be learned from studies in the laboratory.
How does the individual rice grain react and why does it react to a specific
condition? These questions are difficult to answer in the field with masses of
rice because some grains react and others do not. The time has come for
laboratory and field studies to be more closely integrated to produce a
synergistic effect on our knowledge. Observations in the laboratory must be
confirmed in the field, and field observations must be confirmed in the
laboratory. On a broader basis, academia needs to communicate more with
industry and industry more closely with academia if significant advances are to
be made in rice-drying research.
Earlier, this chapter noted that terminology relating to postharvest rice technol-
ogy needs better definition. Wimberly (1983) has a glossary of such terms in his
handbook, but the list is less comprehensive than needed. Even in the mid-1930s,
Stahel (1935) pointed out that the term sun-crack is a misnomer since fissures
(cracks) in the rice grain are not due to rapid drying in the sun but rather to a rise in
moisture content; remoistened rough rice developed fissures (sun-cracks) at a criti-
cal moisture content of 14% or lower. Even though the term is a misnomer, it has
received wide acceptance throughout the world. Later, McDonald (1967) made a
study of “suncracking” in rice, in which he found that grains did not become vul-
nerable to fissuring (suncracking) damage until they reached a moisture content of
21%, after which, the proportion of affected grains increased with decreasing
moisture content. Little can be done about terminology that is already in the litera-
ture, but a continuous effort should be made to make the future use of terms more
descriptive and specific.
Another phrase of concern in the literature can be paraphrased as “the rice
cracked during drying.” Kunze (1979) reported an experiment in which rice was
harvested at 29.8% mc before being dried to 9.1% mc in a single 12-hr pass with
59°C air. The dried rice produced a milled head yield above 70% shortly after
drying. Yet soon after milling, the rice grains began to fissure and continued to
do so until nearly all the grains had multiple fissures. Some rice grains do fissure
during drying. These have been identified as 1) low-moisture grains, adsorbing
moisture ahead of the drying front; 2) defective grains or others that may have a
partial fracture before drying starts; and 3) grains that are overdried. Other
factors may cause rice grains to fissure during drying, but these still need to be
identified and defined. Data are needed to show the fissured or defective grains in
a sample 1) before harvest, 2) after harvest but before drying, 3) immediately
after drying, 4) several days after drying while in storage, and 5) after storage but
before milling. Thereafter, the total and head-rice yields should be determined for
a milled sample. Such information would direct the researcher to areas in which
corrective action is needed.
Innovative research approaches are often restrained by previously reported
research concerning temperatures, moisture contents, residence times, airflow
rates, and other factors that have produced reductions in head rice and/or total
yields in the drying process. Duplication of research is not recommended, but
we must continuously prove that these restrictions are still valid with new
varieties, new and better test procedures, and more sophisticated
instrumentation. The researcher involved with the drying of rice must look at
the horizon and beyond and must imagine what could be, while carefully
sorting and discarding restraints that are no longer valid because of newly
developed concepts and information.
REFERENCES
Aldred, F. L., Ed. 1953. Recent research on dry- Calderwood, D. L. 1981. Rice drying techniques
ing and storage of rough rice. Southern Coop. with solar heat. Pages 36-40 in: Agricultural
Ser. Bull. 29. Texas Agricultural Experiment Energy, Vol. 1. ASAE Publ. 3-81. American
Station, College Station. Society of Agricultural Engineers, St. Joseph,
Angladette, A. 1964. Agricultural engineering, MI.
rice drying principles and techniques. Informal Calderwood, D. L., and Hutchison, R. S. 1961.
Bull. 23. Food and Agriculture Organization Drying rice in heated-air dryers with aeration
of the United Nations, Rome. as a supplementary treatment. Agric. Market-
Araullo, E. V., DePadua, D. B., and Graham, M., ing Res. Rep. 508. U.S. Department of Agri-
Eds. 1976. Rice Postharvest Technology. culture, Washington, DC.
IDRC-053e. International Development Re- Calderwood, D. L., and Louvier, F. J., Jr. 1964.
search Centre, Ottawa, Canada. Procedures for increasing drying capacity of
Bakker-Arkema, F. W., Fontana, C., Brook, R. C., commercial rice dryers. Rice J. 67(8):11-15,
and Westelaken, C. M. 1982. Concurrent-flow 18.
rice drying. ASAE Paper 82-3068. American Calderwood, D. L., and Schroeder, H. W. 1968.
Society of Agricultural Engineers, St. Joseph, MI. Aflatoxin development and grade of undried
Ban, T. 1971. Rice cracking in high rate drying. rice following prolonged storage in aerated
Jpn. Agric. Res. Q. 6:113-116. bins. ARS 52-26. U.S. Department of Agri-
Beagle, E. C. 1978. Rice husk conversion to culture, Washington, DC.
energy. FAO Agric. Serv. Bull. 31. Food and Calderwood, D. L., and Schroeder, H. W. 1975.
Agriculture Organization of the United Chemical preservatives for maintaining moist
Nations, Rome. rice in storage. Tex. Agric. Exp. Stn. Prog.
Bose, S. P., and Chattopadhyay, P. 1976. Har- Rep. 3322C.
vesting and drying of high moisture paddy. Calderwood, D. L., and Webb, B. D. 1971. Effect
Rice Process Eng. Cent. Rep. 2:33-37. of the method of dryer operation on perform-
Brooker, D. B., Bakker-Arkema, F. W., and ance and on the milling and cooking charac-
Hall, C. W. 1992. Drying and Storage of teristics of rice. Trans. ASAE (Am. Soc. Agric.
Grains and Oil Seeds. Van Nostrand Reinhold, Eng.) 14:142-146.
New York. Calderwood, D. L., Bollich, C. N., and Scott, J.
Burmistrova, M. F., Komolkova, I. K., Klemm, E. 1980. Field drying of rough rice: Effect on
N. V., Pania, M. T., Polunochev, I. M., grain yield, milling quality and energy saved.
Pyankov, A. I., Solokov, A. F., Tetyanko, N. Agron. J. 72:649-653.
G., Chaus, V. M., and Eglit, E. G. 1956. Chau, N. N., and Kunze, O. R. 1982. Moisture
Physicomechanical Properties of Agricultural content variation among harvested rice grains.
Crops. Translated from Russian. Available Trans. ASAE (Am. Soc. Agric. Eng.) 25:1037-
from Office of Technical Services. U.S. 1040.
Department of Commerce, Washington, DC. Chen, Y.-L. 1999. Intermittent drying of rough
Calderwood, D. L. 1966. Use of aeration to aid rice with short drying periods. ASAE Paper
rice drying. Trans. ASAE (Am. Soc. Agric. 99-6021. American Society of Agricultural
Eng.) 9:893-895. Engineers, St. Joseph, MI.
Calderwood, D. L. 1970. Analysis of pilot-plant Cnossen, A. G., and Siebenmorgen, T. J. 2000.
and commercial rice-drying operations. (Abstr. The glass transition concept in rice drying
70) Cereal Sci. Today 15:306. and tempering: Effect on milling quality.
Calderwood, D. L. 1972a. Microwave energy for Trans. ASAE (Am. Soc. Agric. Eng.) 43:
drying rough rice. Pages 60-61 in: Proc. Rice 1661-1667.
Tech. Working Group, 14th. Texas A&M Copeland, E. B. 1924. Rice. MacMillan and Co.,
University, College Station, TX. Ltd., London.
Calderwood, D. L. 1972b. Rice drying and stor- Craufurd, R. Q. 1963. Sorption and desorption in
age studies. Rice J. 75(7):63. raw and parboiled paddy. J. Sci. Food Agric.
Calderwood, D. L. 1972c. Effect of rice tem- 14:744-750.
perature on continuous-flow dryer operation. Cuello, J. L., and Huggins, L. F. 2000. Out-
Pages 32-34 in: Rice Research in Texas, 1971. standing agricultural engineering achieve-
Consolidated PR 3092-3105. Texas A&M ments of the 20th century. Resource (St.
University, Texas Agricultural Experiment Joseph) 7(4):18.
Station, College Station, TX. Dachtler, W. C., Ed. 1959. Research on Condi-
Calderwood, D. L. 1975. Rough rice (paddy) tioning and Storage of Rough and Milled Rice.
drying methods in the United States. Trop. ARS 20-7. U.S. Department of Agriculture,
Stored Prod. Inf. 30:13-22. Washington, DC.
Desikachar, H. S. R., and Subrahmanyan, V. 1961. Kobayashi, H., Miwa, Y., and Hayakawa, K.
The formation of cracks in rice during wetting 1972. On the mechanism of the cracking in
and its effect on the cooking characteristics of rice kernels during drying of paddies, particu-
the cereal. Cereal Chem. 38:356-364. larly moisture distribution and drying strain in
Dillahunty, A. L., Siebenmorgen, T. J., Buescher, one kernel. Gifu Daigaku Nogakubu Kenkyu
R. W., Smith, D. E., and Mauromoustakos, A. Hokoku 33:279-293.
2000. Effect of moisture content and tem- Kocher, M. F., Siebenmorgen, T. J., Norman, R.
perature on respiration rate of rice. Cereal J., and Wells, B. R. 1990. Rice kernel moisture
Chem. 77:541-543. content variation at harvest. Trans. ASAE
Dobelmann, J. P. 1955, 1961. Cited by (Am. Soc. Agric. Eng.) 33:541-548.
Angladette, 1964. Kondo, M., and Okamura, T. 1930. Der durch die
Earle, P. L., and Ceaglske, N. H. 1949. Factors feuchtigkeits zunahme verursachte querris
causing the checking of macaroni. Cereal (Doware) des reiskorns. Ber. Ohara Inst.
Chem. 26:267-286. Landwirtsch. Forsch. 4:429-446.
Faulkner, M. D., and Hill, L. C. 1962. A demon- Kunze, O. R. 1964. Environmental conditions
stration of improved rice drying techniques. and physical properties which produce fissures
Rice J. 65(10):10, 11, 14. in rice. Ph.D. thesis. Michigan State Univer-
Faulkner, M. D., and Wratten, F. T. 1966. A new sity, East Lansing, MI.
system for drying rice. Rice J. 69(13):8-10. Kunze, O. R. 1979. Fissuring of the rice grain
Faulkner, M. D., and Wratten, F. T. 1970. The after heated air drying. Trans. ASAE (Am.
Louisiana State University infrared-preheat Soc. Agric. Eng.) 22:1197-1207.
rice drier. Page 41 in: Proc. Rice Tech. Work- Kunze, O. R., and Choudhury, M. S. U. 1972.
ing Group, 13th. Texas A&M University, Moisture adsorption related to the tensile
College Station, TX. strength of rice. Cereal Chem. 49:684-696.
Fontana, C., Bakker-Arkema, F. W., and Kunze, O. R., and Peralta, E. K. 1993. Practical
Westelaken, C. M. 1982. Concurrent-flow vs. relative humidity systems that fissure rice and
cross-flow drying of long grain rice. ASAE other grains. Appl. Eng. Agric. 9:85-89.
Paper 82-3569. American Society of Agricul- Kunze, O. R., and Prasad, S. 1978. Grain
tural Engineers, St. Joseph, MI. fissuring potentials in harvesting and drying of
Geng, S., Williams, J. F., and Hill, J. E. 1984. rice. Trans. ASAE (Am. Soc. Agric. Eng.)
Harvest moisture effects on rice milling qual- 21:361-366.
ity. Calif. Agric. 38(11/12):11-12. Kunze, O. R., Peralta, E. K., and Turner, F. T.
Grosh, G. M., and Milner, M. 1959. Water pene- 1988. Fissured rice related to grain moisture,
tration and internal cracking in tempered weather and fertilization rates. ASAE Paper
wheat grains. Cereal Chem. 36:260-273. 88-6510. American Society of Agricultural
Hall, C. W. 1980. Drying and Storing of Agri- Engineers, St. Joseph, MI.
cultural Crops. AVI Publishing Company, Lague, C. 1990. Modeling preharvest stress-
Inc., Westport, CT. cracking of rice kernels. Ph.D. dissertation.
Henderson, S. M. 1954. The causes and charac- University of California, Davis.
teristics of rice checking. Rice J. 57(5):16, 18. Lague, C., and Jenkins, B. M. 1991a. Modeling
Henderson, S. M. 1955. Deep-bed rice drier per- pre-harvest stress-cracking of rice kernels. Part I.
formance. Agric. Eng. 36:817-820. Development of a finite element model. Trans.
Henderson, S. M. 1958. Deep-bed grain drying ASAE (Am. Soc. Agric. Eng.) 34:1797-1811.
on the ranch with unheated air. Leaflet 103. Lague, C., and Jenkins, B. M. 1991b.
California Agricultural Experiment Station Modeling pre-harvest stress-cracking of rice
Extension Service, Davis, CA. kernels. Part II: Implementation of the
Huysmans, A. A. C. 1965. Milling quality of model. Trans. ASAE (Am. Soc. Agric. Eng.)
paddy as influenced by timing of the harvest. 34:1812-1823.
Int. Rice Comm. Newsl. 14(3):4-12. Lan, Y. 1994. Fissure characteristics related to
Jenkins, B. M. 1989. A modified swath harvesting moisture adsorption stresses in rice. Ph.D. dis-
technique for rice: Part I. Influence on moisture, sertation. Texas A&M University, College
quality and harvester performance. Trans. ASAE Station.
(Am. Soc. Agric. Eng.) 32:1399-1408. Lan, Y., and Kunze, O. R. 1996. Fissure charac-
Kester, E. B., Lukens, H. C., Ferrel, R. E., teristics related to moisture adsorption in rice.
Mohammad, A., and Finrock, D. C. 1963. Trans. ASAE (Am. Soc. Agric. Eng.) 39:2169-
Influence of maturity on properties of Western 2174.
rices. Cereal Chem. 40:323-336. Lan, Y., Kunze, O. R., Lague, C., and Kocher, M.
Kik, M. C. 1951. Nutritive studies on rice. Agric. F. 1999. Mathematical model of the distribu-
Exp. Stn. Bull. 508. University of Arkansas, tion of stress in a rice kernel from moisture
College of Agriculture, Fayetteville, AR. adsorption. J. Agric. Eng. 72:247-257.
Langfield, E. C. B. 1957. Time of harvest in Schroeder, H. W. 1961. Infrared drying of rough

relation to grain breakage on milling in rice. J. rice. III. Medium-grain type: Nato and Mag-
Aust. Inst. Agric. Sci. 23:340-341. nolia. Rice J. 64(1):11, 12, 24-27.
Loewer, O. J., Bridges, T. C., and Buckling, R. Schroeder, H. W., and Rosberg, D. W. 1959.
A. 1994. On-Farm Drying and Storage Sys- Drying rough rice with infrared radiation. Tex.
tems. American Society of Agricultural Engi- Agric. Exp. Stn. Misc. Pub. 354.
neers, St. Joseph, MI. Schroeder, H. W., and Rosberg, D. W. 1960.
Louvier, F. J., Jr., and Calderwood, D. L. 1968. Infrared drying of rough rice. I. Long-grain
Drying and handling rice at commercial rice type: Rexoro and Bluebonnet-50. Rice J.
dryers. Rice J. 71(5):12, 14, 17, 18, 20. 63(12):3-5, 23-27.
Lu, R., Siebenmorgen, T. J., Costello, T. A., and Sharma, A. D. 1980. A two-compartment drying
Fryar, E. O. 1995. Effect of rice moisture model related to the fissuring in rough rice.
content at harvest on economic return. Appl. Ph.D. dissertation. Texas A&M University,
Eng. Agric. 11:685-690. College Station.
McDonald, D. J. 1967. “Suncracking” in rice, Sharma, A. D., and Kunze, O. R. 1982. Post-
some factors influencing its development and drying fissure developments in rough rice.
the effects of cracking on milling quality of Trans. ASAE (Am. Soc. Agric. Eng.) 25:465-
the grain. M.S. thesis. University of Sydney, 468, 474.
Sydney, NSW, Australia. Sharp, P. F. 1927. Wheat and flour studies. IX.
McNeal, X. 1950. When to harvest rice for best Density of wheat as influenced by freezing,
milling quality and germination. Ark. Agric. stage of development, and moisture content.
Exp. Stn. Bull. 504. Cereal Chem. 4:14-46.
McNeal, X. 1960. Rice storage—Effect of moisture, Shedd, C. K. 1953. Resistance of grains and
temperature and time on grade, germination and seeds to air flow. Agric. Eng. 34:616-619.
head rice yield. Ark. Agric. Exp. Stn. Bull. 621. Siebenmorgen, T. J., Banaszek, M. M., and
Milner, M., and Shellenberger, J. A. 1953. Physi- Kocher, M. F. 1990. Kernel moisture content
cal properties of weathered wheat in relation to variation in equilibrated rice samples.
internal fissuring detected radiographically. Trans. ASAE (Am. Soc. Agric. Eng.) 33:
Cereal Chem. 30:202-212. 1979-1983.
Morse, M. D., Lindt., J. H., Oelke, E. A., Siebenmorgen, T. J., Counce, P. A., Lu, R., and
Brandon, M. D., and Curley, R. E. 1967. The Kocher, M. F. 1992. Correlation of head rice
effect of grain moisture at time of harvest on yield with individual kernel moisture content
yield and milling quality of rice. Rice J. distribution at harvest. Trans. ASAE (Am.
70(11):16-20. Soc. Agric. Eng.) 35:1879-1884.
Nagai, I. 1959. Japonica Rice—Its Breeding and Siebenmorgen, T. J., Perdon, A. A., Chen, X.,
Culture. Yokendo Ltd., Tokyo, Japan. and Mauromoustakos, A. 1998. Relating rice
Nagato, K., Ebata, M., and Ishikowa, M. 1964. milling quality changes during adsorption to
On the formation of cracks in rice kernels individual kernel moisture content distribu-
during wetting and drying of paddies. Nippon tion. Cereal Chem. 75:129-136.
Sakumotsu Gakkai Kiji 33:82-89. Slay, W. O., and Hutchison, R. S. 1961. Receiv-
Nguyen, C. N., and Kunze, O. R. 1984. Fissures ing rice from farm trucks at commercial dry-
related to post-drying treatments in rough rice. ers. Agric. Marketing Res. Rep. 499. U.S.
Cereal Chem. 61:63-68. Department of Agriculture, Washington, DC.
Nishiyama, Y., Satoh, M., and Shimizu, H. 1979. Sorenson, J. W., Jr., and Crane, L. E. 1960. Dry-
Crack generation of rough rice after drying. J. ing rough rice in storage. Tex. Agric. Exp. Stn.
Fac. Agric. Iwate Univ. 14:277-288. Bull. B-952.
Pominski, J., Wasserman, T., Spadaro, J. J., Stahel, G. 1935. Breaking of rice in milling in
Decossas, K. M., and Dore, A. B., Jr. 1961. relation to the condition of the paddy. Trop.
Improvements in commercial drying of Agric. Lond. 12:255-260.
southern-grown rice. I. Zenith—A medium- Stansel, J. W. 1975. The rice plant—Its develop-
grain variety. Rice J. 64(9):10, 12, 13, 16, 17. ment and yield. Pages 9-21 in: Six Decades of
Rhind, R. 1962. The breakage of rice in milling: Rice Research in Texas. Tex. Agric. Exp. Stn.
A review. Trop. Agric. 39(1):19-28. Res. Monogr. 4.
Sarker, N. N. 1993. Fissuring of rough rice grains Steffe, J. F., Singh, R. P., and Bakshi, A. S. 1979.
due to transient moisture gradients from Influence of tempering time and cooling on
heated air drying. Ph.D. dissertation. Texas rice milling yields and moisture removal.
A&M University, College Station. Trans. ASAE (Am. Soc. Agric. Eng.) 22:1214-
Schroeder, H. W. 1960. Infrared drying of rough 1218, 1224.
rice. II. Short-grain type: Calrose and Caloro. Steffe, J. F., Singh, R. P., and Miller, G. E., Jr.
Rice. J. 63(13):6-8, 25-27. 1980. Harvesting, drying and storage of rough
rice. Pages 311-359 in: Rice Production and Wasserman, T., Ferrel, R. E., Brown, A. H., and
Utilization. B. S. Luh, Ed. AVI Publishing Smith, G. S. 1957. Commercial drying of
Company, Inc., Westport, CT. Western rice. Cereal Sci. Today 2:251-254.
Stermer, R. A., and Kunze, O. R. 1994. Acousti- Wasserman, T., Ferrel, R. E., Kaufman, V. F.,
cal detection of fissures in rice. Appl. Eng. Smith, G. S., Kester, E. B., and Leathers, J. G.
Agric. 10:267-271. 1958a. Improvements on commercial drying
Stipe, D. R., Wratten, F. T., and Miller, M. F. of Western rice. I. Mixing type dryer. Louisi-
1972. Effects of various methods of handling ana State University design. Rice J. 61(4):30-
brown rice on milling and other quality 32, 34-36, 38; 61(5):40, 42-46.
parameters. Pages 113-153 in: Annual Pro- Wasserman, T., Ferrel, R. E., Kaufman, V. F.,
gress Rep., 64th. Univ. of Louisiana Rice Exp. Smith, G. S., Kester, E. B., and Pierce, S.
Stn., Crowley. 1958b. Improvements in commercial drying of
Swanson, C. O. 1943. Effects of moisture on the Western rice. II. Non-mixing columnar-type
physical and other properties of wheat. II. dryer. Rice J. 61(7):9-12, 14.
Wetting during harvest. Cereal Chem. 20:43- Wasserman, T., Ferrel, R. E., Houston, D. F.,
61. Breitwiser, E., and Smith, G. S. 1964. Tem-
Thompson, J. H., Brown, A. H., Ramage, W. D., pering Western rice. Rice J. 67(2):16, 17, 20-
Roberts, R. L., and Kester, E. B. 1955. Drying 22.
characteristics of Western rice—Colusa 1600 Wasserman, T., Houston, D. F., Ferrel, R. E.,
variety. Rice J. 58(11):14-19, 52. and Hamilton, W. E. 1969. Combination
U.S. Census Bureau. 2002. Statistical Abstracts system for rice drying in California. Pages
of the United States: 2001. 122nd ed. The 59-60 in: Proc. Rice Tech. Working Group,
Bureau, Dep. of Commerce, Economics and 12th. U.S. Department of Agriculture, Wash-
Statistics Admin., Washington, DC. ington, DC.
USDA. 1973. Rice in the United States: Varieties White, G. M., Ross, I. J., and Poneleit, C. G.
and production. ARS Agric. Handb. 289. U.S. 1982. Stress crack development in popcorn as
Department of Agriculture, Washington, DC. influenced by drying and rehydration proc-
Wasserman, T., and Calderwood, D. L. 1972. esses. Trans. ASAE (Am. Soc. Agric. Eng.)
Rough rice drying. Pages 140-165 in: Rice: 25:768-772.
Chemistry and Technology, 1st ed. D. F. Wimberly, J. E. 1983. Technical Handbook for
Houston, Ed. Am. Assoc. Cereal Chem., St. the Paddy Rice Postharvest Industry in Devel-
Paul, MN. oping Countries. The International Rice
Wasserman, T., Brown, A. H., Houston, D. F., Research Institute, Los Baños, Laguna, Philip-
and Ferrel, R. E. 1956a. Drying characteristics pines.
of Western rice—Caloro variety. I. Equal Wratten, F. T., and Kendrick, J. 1970. Hygro-
moisture removal and constant drying air tem- scopic equilibrium of rough rice and soybeans
perature in all passes. Rice J. 59(3):12-16. at elevated temperature. Proc. Rice Tech.
Wasserman, T., Brown, A. H., Houston, D. F., Working Group, 13th. Texas A&M University,
and Ferrel, R. E. 1956b. Drying characteristics College Station, TX.
of Western rice—Caloro variety. II. Varying Yoshida, S. 1981. Fundamentals of Rice Crop
drying air temperature and amount of moisture Science. The International Rice Research
removed in each pass. Rice J. 59(4):41-45. Institute, Manila, Philippines.
CHAPTER 10
ROUGH-RICE STORAGE
Terry A. Howell, Jr.

McKee Foods Corporation
Collegedale, Tennessee
Robert R. Cogburn (retired)

Hamshire, Texas
In temperate areas of the world, rough rice is harvested once a year, usually over a
two-month period, and stored for one year or longer. End-use processors and con-
sumers expect a high-quality product with minimal variations throughout the year, so
rough-rice storage practices must be optimized to produce a stable, consistent product.
In tropical regions, where rice is harvested more often than once a year, rice-storage
periods are shorter because the environmental conditions are more extreme.
Proper storage of rough rice is critical for maintaining its quality and value. Rice
is one of the few cereal crops that is primarily consumed whole rather than as flour;
thus, its value is highly correlated with intact kernels. Every processing step is criti-
cal to maintaining kernel integrity, and sound storage techniques are a vital part of
postharvest rice processing.
Rice quality may be defined in many ways, often depending on the characteris-
tics a particular end user desires. Characteristics of interest include head-rice yield
(HRY), color, microbial load, cooking quality (texture, flavor, and others), and
pasting properties of rice flour slurries. In some cases, grading systems are used to
quantify these characteristics objectively; however, in other cases, subjective mea-
sures are used. Although rice may be stored with the goal of improving rice qualities,
the primary objective of storage is to provide an environment that does not cause
quality deterioration.
In this chapter, storage structures and practices are discussed in terms of factors
that influence final rice quality: aging of rice, as well as the effects of storage
insects, microflora, rodents, and birds.
STORAGE TECHNOLOGY
Structures
The structures used to store rice throughout the world are extremely varied. The
primary purpose of the structures is to protect the rice from negative environmental
conditions such as high heat and humidity, insects, and other pests. Therefore, stor-
269
age structures are designed to adapt to several variables, including available materi-
als, facility size, storage duration, finances, and convenience.
FARM-SCALE STORAGE
In developed countries, most rice is stored in bins and flat bunkers made of steel or
concrete. These structures vary in size according to the size of the farming operation.
Individual farmers may store from 20,000 bu to more than 100,000 bu (400–2,000
tonnes [t]). In the United States, it is becoming more cost effective for producers to store
rice themselves; thus, more on-farm grain storage facilities are being built every year.
Before selecting a design, producers must consider financial, convenience, and
capacity issues. Most new facilities are designed in one of two formats: 1) bins in a
semicircle around a center “dump pit,” or 2) bins in straight lines with or without a
grain-moving device. Figures 1 and 2 are representative of these types of facilities,
Fig. 1. Farm-scale bins in a semicircular arrangement. (Courtesy Dennis Gardisser, University of

Arkansas Cooperative Extension Service)
Fig. 2. Farm-scale bins in straight lines. (Courtesy Dennis Gardisser, University of Arkansas
Cooperative Extension Service)
Rough-Rice Storage / 271
respectively. Bins configured in either style are usually equipped with perforated
floors that allow the rice to be dried in the bin and/or aerated by fans during storage.
Many new facilities are equipped with grain pumps, unique grain-handling systems
that allow rice to be moved from one bin to another with relative ease. One reason
that individual producers are moving toward owning their own storage facilities is
the convenience afforded by the bins at harvest. Bringing harvested rice to a storage
facility on a producer’s own property is much easier and less time-consuming than
is taking it to a cooperative, where producers may encounter long lines.
In less-developed countries, storage structure are less sophisticated, primarily
because of financial limitations and size differences, although they serve the same
purposes. In many areas, individual farmers produce smaller quantities of rice; as a
result, their individual storage needs are smaller. Storage structures range from
dugout pits to “storage rooms” in private homes. Figures 3 and 4 are examples of
Fig. 3. A pit for storage of rice in India, circa 1969. (Courtesy D. L. Calderwood)
Fig. 4. A variety of storage structures in India, circa 1969. (Courtesy D. L. Calderwood)

Fig. 5. Industrial-scale round silos for rice storage. (Courtesy Dennis Gardisser, University of Arkansas
Fig. 6. Industrial-scale flat storage system. (Courtesy Dennis Gardisser, University of Arkansas
storage structures found in less-developed countries. Regardless of the sophistica-

tion of the storage facility, rice should be kept off the floor and dry.
LARGE-SCALE STORAGE
Throughout the world, the need exists for large-scale storage of rice (>100,000
bu [2,000 t]). Rice cooperatives and end-use processors use these facilities to pro-
vide a steady, constant supply of rice, typically associated with large drying and
milling operations. The massive structures required to handle these large volumes
are usually constructed of galvanized steel or concrete. Round bins are most com-
mon, but large, flat storage buildings are used as well. These structures must be
designed to include systems for loading and unloading the grain, movement of
air, and protection from pests. Figures 5 and 6 show two large-scale storage facil-
ities, one using round silos and the other flat, respectively.
Fig. 7. Bulk bag warehouse storage system for rice. (Courtesy Terry J. Siebenmorgen, University of
Arkansas)
WAREHOUSES
Warehouses are used to store milled rice (a topic addressed only briefly in this
chapter.) Milled rice is a delicate commodity and should be treated carefully. Many
millers use bags to store rough, brown, and milled rice. These bags may vary in size
from 22.7 kg (50 lb) to 1 t. Figure 7 shows a 1-t bag containing rice in a warehouse
storage facility. Warehouses used to store these bags must provide the same
protection for the rice as the bins discussed previously: temperature and relative
humidity must be controlled and pest populations minimized. Space should be
allowed between stacks for fumigation, if necessary. Warehouses are much less
efficient than bins and have higher costs associated with storage, but they are very
useful for higher-valued products such as milled and specialty rice.
Management Practices
CONTROL OF RESPIRATION
Rice management during storage can often be related to the effects of respiration
on rice quality. Rice, insect predators, and potentially harmful bacteria are all typi-
cally governed by their ability to respire. Therefore, many decisions and strategies
used during storage can be evaluated for their effectiveness in providing an envi-
ronment that controls respiration. Equation 1 states that carbohydrates and oxygen
are consumed to produce carbon dioxide, water, and heat:
+ 2 &2 ++ 2+HQHUJ\
& + 2 (1)
For the rice kernel itself, respiration involves the conversion of carbohydrates
(starch) and oxygen (from the storage environment) into water and carbon dioxide
(with heat). As with most reactions, the speed of the process depends on several
factors, including temperature, moisture content, and the presence of reactants and
other products. Dillahunty et al (2000) showed that, at moisture contents below
14%, respiration rates were negligible, but the rates increased exponentially as
moisture content increased. Warm temperatures also enhance respiration, and the
combination of high temperatures and high humidity found in many rice-growing
areas can be very harmful to rice quality if the rice is not managed properly. When
respiration is allowed to proceed at a high rate, the reaction is spurred by its own
products of moisture and heat, leading to the exponential increases seen by
Dillahunty et al (2000).
The effects of respiration on insects and bacteria are discussed in later sections.
DRYING
Many aspects of drying are affected by storage practices. Rice is typically har-
vested at a moisture content of 18–24%. Under these conditions, rice is susceptible
to spoilage through a variety of mechanisms. At temperatures above 70°F (21°C),
rice may be kept at these moisture levels only up to about two weeks. Drying is
necessary to preserve the grain for longer periods.
In the United States and other developed countries, drying may be done at the
farm level or at larger drying facilities. In farm-scale operations, rice is usually
dried gently in a bin, at temperatures of 90–105°F (32.2–40.6°C). To remove mois-
ture from the grain, fans supply hot air at flow rates of 0.1–1.5 cubic feet per minute
[cfm]/bu (0.1–1.6 m3¼min–1¼t–1). External heat is sometimes applied, but this is not
absolutely necessary in hot climates. When the rice reaches acceptable moisture
contents—around 13%—drying is stopped, and storage managers can begin their
storage practices. In large-scale drying operations, dried rice is transferred to
storage bins and kept until needed for milling. Properly dried and stored rice may be
kept up to five years without quality degradation (Calderwood et al, 1984).
In some tropical countries, rice is sun-dried on roads and other dry, flat areas.
Drying in this manner is not ideal, but it offers an inexpensive solution to those
without an abundance of resources. Usually, this drying technique is able only to
produce rice with a moisture content near 14%, which is sufficient in tropical cli-
mates for short-term storage of two to three months but not for longer-term storage
(Wimberly, 1983)
A more detailed discussion of drying is available in Chapter 12 of this book.
HANDLING
Grain held in the same position for long periods can become susceptible to
problems caused by temperature and moisture gradients. Rotating rice from one bin
to another (or within a bin), a practice called “turning,” allows it to be “refreshed.”
The rice is turned, or mixed, by being rotated from one bin to another to break up
hot spots and redistribute moisture and temperature gradients. The movement also
provides opportunities for insect control, if necessary. Many bins are equipped with
augers, or stirring devices, so that rice within a bin can be turned without leaving
the bin. These augers rotate slowly through the bin to move rice from the bottom
layers to the top and vice versa.
AERATION
Once rice has been properly dried, it may be stored for many months; however,
the rice should not be ignored until it is needed for milling. Aeration has become a
valuable tool for storage managers during the storage season. It uses low-volume
airflow to condition the rice during storage. The airflow helps to reduce grain tem-
peratures throughout the storage season, equilibrate temperature and moisture gra-
dients, and remove odors.
Many of the grains produced in the United States are aerated with very low air-
flow volumes of 0.1–0.3 cfm/bu (0.1–0.3 m3¼min–1¼t–1). In large rice-storage
facilities, these rates are common, as fans are sized for this purpose alone. In
contrast, the fans used in most farm-scale rice bins are sized for drying and generate
flow rates around 1.0 cfm/bu (1.1 m3¼min–1¼t–1). Previously, little attention was paid
to these differences, but in recent years, high-volume aeration has been shown to be
beneficial in reducing grain temperatures during the fall and winter months to levels
that are useful in combating insects and bacterial growth (Ranalli and Howell, 2002;
Ranalli et al, 2002).
A storage manager may determine fan flow rates using fan pressure curves sup-
plied by the fan manufacturer. For a particular size fan, these data (also available in
tabular form) show the flow rate (usually in cubic feet per minute) as a function of
the measured static pressure in the plenum beneath a grain bin floor. Once the
overall flow rate is determined, the flow rate per bushel may be calculated by
dividing by the amount of rice in the bin. The flow rate is valuable because it may
be used with equation 2 to roughly calculate how much time is required to pass a
temperature front through a bin using aeration. The equation states:

W = (2)
)5
where time (t) is in hours and flow rate (FR) is in cfm per bushel. A more conser-
vative estimate can be found by replacing 12 with 15 in the numerator. According to
equation 2, a flow rate of 1 cfm/bu would require 12 hr to pass a temperature front
through the grain, whereas a flow rate of 0.1 cfm/bu would require 120 hr.
Periodic aeration through the fall and winter months gradually reduces grain
temperature and moisture gradients. Storage managers must use caution not to
rehydrate rice by aerating it with high-humidity air. However, with low flow rates,
the addition of moisture to the rice would require sustained aeration at high
humidity to significantly increase the average moisture content. Likewise, aeration
under very dry conditions for short periods of a few hours at a time will not dry the
rice further; however, sustained aeration at either extreme will change the grain
moisture content.
MAINTENANCE OF RICE QUALITY
As stated previously, rice quality is defined in many ways, including field yields,
HRY, size and color characteristics, cooking quality, and sensory and eating quality.
For the purposes of this chapter, rice quality relates to the milling and cooking char-
acteristics of rice. During storage, these quality characteristics are directly affected
by storage temperature and relative humidity and the presence of damaging insects,
microorganisms, and other pests. The milling and cooking quality of rice may be
affected by all of these conditions, but rice quality also changes during storage as a
function of the storage time, a phenomenon referred to as “aging.”
Aging
The physicochemical properties of rice (such as HRY and the pasting properties
of rice flour slurries) and cooking properties (such as volume expansion and water
absorption) change during storage, even under constant, stable conditions. The exact
mechanisms of these changes are largely unknown; however, most evidence shows
that the values of these properties increase during storage for the first two to four
months after harvest and then taper off or return to their original values during pro-
longed storage. Storage managers and processors must be aware of this and should
use aging to their benefit. Many processors will not use new-crop rice until it has
been aged for 60 days.
HRY is the primary measure of rice quality for most postharvest operations;
thus, the value of a rice crop is closely tied to this property. For years, anecdotal
evidence has claimed that HRY improves during the first months of storage.
Since the mid-1990s, a concerted effort has been made to investigate these claims
and determine the source of the changes, and both laboratory work and field
research have substantiated these claims. In a laboratory-controlled setting,
Daniels et al (1998) found that the HRY of cv. Cypress increased 10 percentage
points during the first two months of storage when multiple drying methods were
used. This large increase is not typically seen in field research or practice. Ranalli
et al (2003) found that HRY increased during the early months of storage in
farm-scale tests; however, the magnitude of the change was smaller than that in
the laboratory work of Daniels et al (1998). Nevertheless, even with a change of
only 2 percentage points, aging becomes a very valuable phenomenon for the rice
industry.
Pasting properties of rice flour slurries are useful indicators of how rice flour
will perform in baking applications and also for characterizing starch and protein
activity in rice. For several years, scientists have been studying the effects of aging
on pasting properties, and, typically, they have found that these properties follow
the same trend as for HRY. Using a Rapid ViscoAnalyser or a Brabender
Amylograph, one can measure the viscosity of a slurry with controlled heating so
that information can be gathered about the slurry’s gelatinization and other fluid
properties. A peak viscosity (usually correlated with the slurry’s response to
gelatinization) and a final viscosity are usually reported. Perdon et al (1997)
examined the pasting properties and cooking properties of rice stored at different
temperatures and durations. The peak viscosity increased during the first two
months and remained fairly constant throughout six months. Higher storage
temperatures showed a more significant increase. These results agree with those of
other scientists (Villareal et al, 1976; Indudhara Swamy et al, 1978; Hamaker et al,
1993). Farm-scale results showed a split response, with some rice showing a
gradual increase in peak viscosity during storage up to 12 weeks and some showing
a steady response during storage (Ranalli et al, 2003). Pasting properties did not
exhibit a trend during storage, according to the work of Daniels et al (1998). So,
although some research shows pasting properties to be constant during storage, the
consensus opinion is that viscosity values have a typical aging trend—increasing
slightly over the first 8–12 weeks of storage and then returning to near-original
values.
Volume expansion and water absorption (hereafter jointly referred to as cooking
properties) have been evaluated in many of the studies cited previously, among oth-
ers. Perdon et al (1997) did not find a strong correlation between storage conditions
and cooking properties. Only at high storage temperatures was there any activity;
values increased during the first two to three months, then decreased to original
values. The work of Daniels et al (1998) was similar. At storage times of one year,
Indudhara Swamy et al (1978) found that water absorption increased slightly and
then remained flat. In a field study, Ranalli et al (2003) found that water absorption
increased over the first two months of storage before returning to near-original
values. The effects of storage conditions on cooking properties remains somewhat

vague, with most research showing little or no correlation between the two.
Protection of Rice from Degradation
The second major factor involved with rice quality, other than maintaining its
inherent characteristics, is protecting rice from degradation caused by microbial
contamination, insect damage, and rodent and bird invasion. All of these potential
problems can be overcome with good preparation and prevention. These practices
begin with good storage structures, sanitary environments, and good drying prac-
tices. The following sections discuss how each predator can cause rice degradation
and how it can be resisted.
MICROFLORA
Numerous microorganisms can cause deterioration of grain if conditions are
suitable for their growth. Perhaps the most common of these are various fungi.
These are often classified as field fungi and storage fungi, the former being those
organisms that infect the commodity while it is growing in the field and the latter
those that normally are manifested after harvest and during storage. This terminol-
ogy is not absolute, however. Under some conditions, fungi usually associated with
storage can infect the grains in the field, and with sufficient moisture and tempera-
ture, field fungi can grow on stored grain. Moisture contents below 14% (water
activity below 0.75) are generally regarded as sufficient to prevent the growth of
mold and fungus.
Improper drying of rice can lead to many problems with microbial contamination
in stored rice. For instance, if the average moisture content of a rice sample is 16%,
one can be reasonably sure that pockets of higher moisture content exist within the
grain, some at dangerous levels that allow microbial contamination. If fungal
growth were to take place within this pocket, a chain reaction might occur that
would be difficult to stop. First, the high moisture content in the presence of
microbes would permit microbial growth. Then, these microbes would increase the
grain temperature through respiration (see equation 1), causing increases in tem-
perature and moisture (both products of respiration) and, potentially, clumping of
the grain. If undetected, severe damage might occur. However, good storage
management periodically monitors grain temperatures and moisture contents and
aerates consistently to prevent these processes from occurring. Ranalli and
Howell (2002) showed that yeast and mold counts in farm-stored rice were
primarily dependent on the moisture content of the rice rather than on tem-
perature.
The most negative result of having pockets of high-temperature, high-moisture
rice is the potential for “stack burn,” or yellowing, of rice. Dillahunty et al (2001)
showed that storage of rice at high temperatures causes severe problems with rice
color. Because kernel color is a primary tool for evaluating rice quality, rice value
can drop dramatically when discolored kernels are present.
INSECTS
The infestation of rice by stored-grain insects can be an economically crippling
problem. Not only do the insects consume kernels, they also cause rice grades to
drop, may increase the presence of microorganisms, and can aid in the development
of hot spots. Insect trouble can be avoided through several means, but insect control is
best understood by examining the types of insects that cause problems in stored rice.
Species. The origin of insects within a bin has been debated for years, and several
theories are available to describe how they arrive in the grain. Cogburn and Vick
(1981) and Cogburn et al (1984) showed that Angoumois grain moths, Sitotroga
cerealella (Olivier), and lesser grain borers, Rhyzopertha dominica (F.) (Fig. 8), can
be found in fields before harvest. However, residual populations within bins are
generally accepted as the primary cause of infestations. When a storage facility is
emptied, the residual dust and kernels provide ample breeding grounds for insects to
establish their presence. Thus, proper cleaning and treatment of empty facilities should
be the first, and is perhaps the best, treatment to prevent insect damage in stored rice.
Most species of storage insects are distributed throughout the world. The same
species that attack rice in the United States attack rice and other grains around the
globe and inflict the same type of damage everywhere. Storage insects often are
discussed in terms of the damage they inflict, their feeding habits, and their method
of reproduction. One group destroys whole-grain rice. The most destructive of
these, usually found in rough or brown rice, are the lesser grain borer; the Angou-
mois grain moth; and the Sitophilus complex: the rice weevil, S. oryzae (L.) (Fig.
9), the maize weevil, S. zeamais Motschulsky, and the granary weevil, S. granarius
(L.). The immature stages develop inside the grain, and, except for the moth, the
adults feed on whole, intact grains. Adult weevils are equipped with a snout at the
front of the head, with the mouthparts located at the tip of the snout. Female weevils
bore a hole into a grain and deposit an egg at the bottom of the hole. The remaining
space is filled with a mucouslike substance that dries into a protective plug. The
larva feeds on the endosperm as it grows, then pupates, and emerges to the adult
stage before leaving the interior of the grain. Lesser grain borers and Angoumois
grain moths do not embed eggs inside grains. They deposit eggs in the grain mass,
and the resultant larvae move about, select a grain, and chew their way inside, after
which they develop as weevils do. Moths infest only the surface layer of rice,
whereas the weevils and borers can be found at any depth in a grain mass.
The second, and larger, group of stored insects lays eggs in or near the commod-
ity, and larvae and adults move about freely in the grain but do not penetrate the
grains. Coleoptera (beetles) in this group include the cadelle, Tenebroides mauri-
Fig. 8. Lesser grain borer in rice. (Courtesy John Murdoch, University of Arkansas)
tanicus (L.); the sawtoothed grain beetle, Oryzaephilus surinamensis (L.); the flat-
grain-beetle complex, Cryptolestes spp.; the red flour beetle, Tribolium castaneum
(Herbst) (Fig. 10); and the cigarette beetle, Lasioderma serricorne (F.). External
feeders usually cause less physical damage than internal feeders, but they can exist
under a broader set of conditions and attack rice in all phases of storage and distri-
bution. They infest facilities such as mills, warehouses, silos, rail cars, and cargo
ships. They do not usually damage the endosperm of whole grain, feeding instead
on the bran coat or the germ or on grain dust and debris anywhere in the grain mass
or storage facility. Infestation is often a problem because of contamination rather
than because of destruction of the commodity. However, heavy infestations of some
species can cause heating, which makes moisture content rise, promoting mold
growth and resulting in spoilage and the physical loss of the grain.
Stored-product Lepidoptera (moths) infest the surface of stored bulk rice and
attack rice in bags. They are extremely common pests in rough-rice storages, mills,
Fig. 9. Rice weevil in rice. (Courtesy John Murdoch, University of Arkansas)
Fig. 10. Red flour beetle. (Courtesy USDA-ARS Grain Marketing and Production Research Center
website).
warehouses, and transportation facilities. Although they do not usually damage the
endosperm, the larvae spin profuse webbing, often matting its entire surface
(Cotton, 1956). This can result in large financial losses. Fumigation kills the insects,
but the commodity then must be recleaned or remilled to eliminate the webbing and
insect carcasses. Many sound grains that are webbed together are removed in the
cleaning process and are therefore lost. Thus, these moths might be classified as
both destructive and contaminating pests. These include the Indian-meal moth, Plo-
dia interpunctella (Hübner), and the almond moth, Cadra cautella (Walker).
Insect Control. Control of storage insects begins before rice is harvested and
continues until the rice is emptied from its container. In the discussion of insect
feeding habits, it was noted that insect populations are better able to colonize when
broken kernels, dust, and other grain fragments are present. Therefore, proper
maintenance of storage facilities should include the thorough cleaning of the storage
areas. In addition, the rice should be harvested, handled, and dried as gently as pos-
sible to maintain kernel integrity. After rice has been loaded into its container,
spilled rice and other trash should be disposed of so that the areas around the bins
do not attract insects, rodents, or birds. Reducing the opportunities for insect har-
borage around storage facilities can minimize potential problems.
Rice typically is protected from stored-product insects through standard proce-
dures. In addition, insect control can be attained through alternative methods such as
modified atmospheres, grain cooling, or irradiation.
A typical set of standard procedures for preventing insect outbreaks and damage
in stored rice begins with a thorough cleaning of the storage facility before the rice
is binned. Next, the walls and other surfaces of the storage facility may be sprayed
with a residual chemical such as malathion, or granular protectants may be applied
directly to the rice to prevent initial insect populations from taking root. Granular
protectants are applied throughout the grain mass or to the top surface only,
depending on the severity of any expected problems. The efficacy of protectants
may be reduced as insects gain resistance to the chemicals. Fumigants and aerosols
are also used as a defense against insects. Because fumigants are deadly gases that
destroy insects, they must be applied by trained personnel in a sealed area, typically
in conjunction with an existing aeration system. After the gas is held in the storage
container for a set period (typically three days for methyl bromide and one week for
phosphine), the gas is vented. Methyl bromide and phosphine are the two fumigants
labeled for stored-grain use in the United States, but methyl bromide is being phased
out as a result of the 1995 Montreal Protocol. This agreement among several nations
calls for the phase-out of methyl bromide by 2005 because it depletes ozone.
Therefore, several alternative fumigants are being studied and are intended to be
approved by 2005. Additionally, aerosols (atomized mixtures of insecticides and
solvents) are used in open spaces in warehouses and bins to treat flying insects such as
moths.
Alternative or new technology for battling insects during storage may be chosen
for several reasons. These may include a desire to find techniques and technologies
that are more efficient or less time-consuming, or, in the case of organic producers,
a desire to store rice without the use of chemicals. Regardless of motive, opportuni-
ties are plentiful to experiment with alternatives to standard practices. Modified
atmospheres, grain cooling, and irradiation are alternatives discussed here.
Modified atmospheres, sometimes called controlled atmospheres, have been used
in a manner similar to that for fumigation. With fumigation, a toxic chemical is
used to kill insects, whereas, with modified atmospheres, a reduction in oxygen is

used to kill them. As the name implies, modified atmospheres require that the air
around the stored rice be modified in some way that causes respiration to become
difficult for the insects. Common modifications include increasing carbon dioxide
levels in the storage headspace or replacing the headspace gases completely with
nitrogen to minimize the respiration reaction. These types of treatments are popular
because both gases are naturally present in the environment and thus less toxic to
human beings than are chemicals. However, users should exercise care with these
techniques because prolonged exposure to these environments can be harmful to
humans. The major drawback to modified atmospheres is that the settings must be
maintained for extended periods, much longer than those used with traditional
fumigation. This can be difficult in storage bins that are hard to seal.
Another effective alternative treatment for insect control is grain cooling or
heating. Because the use of excessive heat for extended periods can damage kernels,
this method has not been widely implemented; however, cooling rice, with or with-
out refrigeration, has been shown to be effective against insects. The premise of this
treatment is that insects cannot survive and thrive outside a temperature range of
55–100°F (13–35°C), so when temperatures are reduced below this level during
storage, insect viability is severely hampered (Fields, 1992). Most U.S. production
areas have a cool climate in the fall, allowing for aeration with ambient air to
reduce grain temperatures. In other areas, external refrigeration units may be used
temporarily to reduce the grain temperatures. Ranalli and Howell (2002) had suc-
cess in eliminating three insects, using ambient aeration to cool rice. Aeration
controllers that automatically reduce rice temperatures during the storage season are
available commercially, and some cooperative extension service engineers are able
to consult on their construction.
Irradiation, the exposure of a product to radiation, is growing in use and popu-
larity in the United States, although its use in destroying insects in rice has not been
documented. It is used against insect populations either by exposing the contami-
nated rice directly to irradiation or by releasing males sterilized by irradiation into
the population (Harein and Davis, 1992). Organic producers do not consider this
method acceptable, and the U.S. Food and Drug Administration has labeled irradia-
tion as a food ingredient. These two factors have slowed the acceptance of irradia-
tion; however, its use is slowly growing in other food groups (meats, poultry, fruits,
and vegetables, among others).
BIRDS AND RODENTS

Although rodents and birds present a different set of problems, the damage these
predators cause may be minimized with proper planning and protective devices.
Rats and mice gain entry to storage facilities through any opening or weak spot in
the storage structure and can cause significant damage to the rice, pallets, sacks, and
wooden structures. Rodent damage is prevented by minimizing their access to the
rice (rodent-proofing), poisoning the rodents, or trapping them. Rodent-proofing is
relatively easy in metal or concrete structures. Shrubs and other harboring areas
should be trimmed and set back from any processing facility. Poisons are usually
effective; however, they can become a problem if the poisoned rodents are found in
parts of the storage or processing facility. Rodent traps are a common and economi-
cal method of protecting stored rice. Traps must be checked routinely and system-
atically for best results. Birds mostly cause problems in rice that is not protected
with a secure roof. Much like rodents, birds can penetrate any weakness in the stor-
age structure. However, with good seals and with screens covering vents, their
impact can be minimized.
SUMMARY
Maintaining a safe, secure, high-quality environment for rice during storage

before it is milled is accomplished by understanding a few key principles. The rice
must be stored in a secure structure to prevent damage from environmental factors,
microbial contamination, insects, and rodents or birds. During storage, the rice must
be monitored to ensure that it is being held at proper temperatures and relative
humidities. Rice quality is maintained or improved throughout the storage process
by observing these conditions and quickly correcting any deviations.
REFERENCES
Calderwood, D. L., Cogburn, R. R., Webb, B. D., six months after harvest. Ark. Farm Res.
and Marchetti, M. A. 1984. Aeration of rough 42(1):8-9.
rice. I. Long-term storage. Trans. ASAE Harein, P. K., and Davis, R. 1992. Control of
27:1579-1585. stored-grain insects. Pages 491-534 in: Storage
Cogburn, R. R, and Vick, K. W. 1981. Distribu- of Cereal Grains and Their Products, 4th ed.
tion of Angoumois grain moth, almond moth, D. B. Sauer, Ed. Am. Assoc. Cereal Chem., St.
and Indian meal moth in rice fields and stor- Paul, MN.
ages in Texas as indicated by pheromone- Indudhara Swamy, Y. M., Sowbhagya, C. M.,
baited sticky traps. Environ. Entomol. and Bhattycharya, K. R. 1978. Changes in
10:1003-1007. physicochemical properties of rice with aging.
Cogburn, R. R., Burkholder, W. E., and J. Sci. Food Agric. 29:627-639.
Williams, H. J. 1984. Field tests with the Perdon, A. A., Marks, B. P., Siebenmorgen, T. J.,
aggregation pheromone of the lesser grain and Reid, N. B. 1997. Effects of rough rice
borer (Coleoptera: Bostrichidae). Environ. storage conditions on the amylograph and
Entomol. 13:162-166. cooking properties of medium-grain rice cv.
Cotton, R. T. 1956. Pests of Stored Grain and Bengal. Cereal Chem. 74:864-867.
Grain Products, rev. ed. Burgess Publishing Ranalli, R. P., and Howell, T. A., Jr. 2002. Total
Co., Minneapolis, MN. aerobic and yeast/mold bacterial counts of
Daniels, M. J., Marks, B. P., Siebenmorgen, T. J., rough rice during on-farm storage. J. Food Sci.
McNew, R. W., and Meullenet, J. F. 1998. 67:807-810.
Effects of long-grain rough rice storage history Ranalli, R. P., Howell, T. A., Jr., Arthur, F. H.,
on end-use quality. J. Food Sci. 63:832-835. and Gardisser, D. R. 2002. Controlled ambient
Dillahunty, A. L., Siebenmorgen, T. J., Buescher, aeration during rice storage for temperature
R. W., Smith, D. E., and Mauromoustakos, A. and insect control. Appl. Eng. Agric. 18:485-
2000. Effect of moisture content and tem- 490.
perature on respiration rate of rice. Cereal Ranalli, R. P., Howell, T. A., Jr., and
Chem. 77:541-543. Siebenmorgen, T. J. 2003. Effects of con-
Dillahunty, A. L., Siebenmorgen, T. J., and trolled ambient aeration on rice quality during
Mauromoustakos, A. 2001. Effect of tem- on-farm storage. Cereal Chem. 80:9-12.
perature, exposure duration, and moisture Villareal, R. M., Resurreccion, A. P., Suzuki, L.
content on color and viscosity of rice. Cereal B., and Juliano, B. O. 1976. Changes in phys-
Chem. 78:559-563. icochemical properties of rice during storage.
Fields, P. G. 1992. The control of stored product Staerke 28:88-94.
beetles and mites with extreme temperatures. Wimberly, J. E. 1983. Technical handbook for
J. Stored Prod. Res. 28:89-118. the paddy rice postharvest industry in devel-
Hamaker, B. R., Siebenmorgen, T. J., and oping countries. International Rice Research
Dilday, R. H. 1993. Aging of rice in the first Institute, Los Baños, Laguna, Phillipines.
CHAPTER 11
RICE MILLING
Norris Bond
Satake USA, Inc.
Houston, Texas
The process of rice milling is probably nearly as old as agriculture itself, but
mechanized rice milling traces its roots back to the invention of an abrasive type of
bran removal equipment, invented by Douglas and Grant around 1860. Since then,
rice milling has evolved into a complicated industrial process that employs sophisti-
cated technologies for controlling the operation.
DEFINITION
To better understand this complicated process, it is necessary to define it in terms

of the specific actions that are performed and the purpose behind those actions.
Rice milling is a process in which people use machines to remove foreign material,
husks, bran, and broken kernels from rice to prepare the grain for a variety of commer-
cial purposes.
When this definition is broken down into its important parts, several observations
become evident.
People are listed in the definition before machinery because the actions and the
knowledge of the operators involved in rice milling are more important to the effi-
ciency of the operation than the design of the machinery involved.
Although the milling process is complicated, it consists of only four significant
actions: removal of foreign material; removal of the husk from the rice kernel;
removal of the bran layers from the rice kernels; and removal of broken grains.
The objective of modern rice milling is not technical but economic. This prin-
ciple determines the degree of sophistication and automation used in the milling
process.
RELATIONSHIP TO OTHER PROCESSES
For the purposes of this overview, milling is treated as an independent process.

Keep in mind, however, that it is only one stage in the conversion of rice in the field
to rice on the consumer’s table. Factors that affect the milling process, such as uni-
formity of length and thickness, are determined by the practices of the individual
farmer. Similarly, the manner in which rice is dried and stored before milling has a
dramatic effect upon the efficiency of the process. Finally, the milling process must
283
produce a product that meets the changing expectations of industrial and consumer
customers. Thus, optimization of the milling process can take place only when the
effects of each of these stages upon the others are considered. For this reason, mill-
ers should participate in and encourage research that takes a coordinated view of the
overall process of rice production.
STAGES OF THE MILLING PROCESS
Like any process, rice milling consists of several different stages, each of which
must be successfully completed in sequence for the overall process to be successful.
These stage consist of 1) cleaning—the removal of foreign material from rice, 2)
brown rice production—the removal of the husk from the rice kernel, 3) bran
removal—the selective removal of germ and bran layers from the rice kernels, 4)
classification—the removal of broken grains from milled rice, and 5) blending and
packaging—the controlled addition of broken grains and nutrients and placement of
measured weights of finished products into consumer packages. This chapter dis-
cusses the first four stages.
Cleaning
PADDY
Depending upon harvesting methods and upon the specifications of the milled
rice product, several different types of impurities may need to be removed from the
rice entering the milling facility. Stones, mud balls, straw, metal, and glass may be
present as contaminants. In addition, some lots of rice may contain grass and weed
seeds, red rice, and grains other than rice. Finally, in the production of parboiled
rice and other specialty-milled rice products, immature rice grains and chalky
kernels may be considered objectionable. The process of removing these impurities
from rice during milling is referred to as “cleaning.”
Cleaning rice is accomplished by separation based upon differences in physical
properties. These include size, density, magnetic conductance, frictional force, and
Fig. 1. Drum sieve. (Courtesy Hayes and Stolz Industrial Manufacturing Co., Inc., Fort Worth, TX)
Milling / 285
optical characteristics. A succession of increasingly efficient separating machines is

used for optimal performance.
The first step in cleaning involves a machine generically called a “drum sieve”
or a “scalperator.” The drum sieve, illustrated in Fig. 1, removes impurities that are
considerably longer than rice. Paddy rice is placed inside the drum, which is made
of wire mesh. The rice falls through the openings in the drum, while objects that are
much larger than rice remain in the drum and are discharged at its end.
The scalperator, illustrated in Fig. 2, combines the ability to remove objects
longer than rice with removal of those that are lighter or less dense than rice. A
moving column of air lifts lighter materials such as dust and empty husks from the
stream of rice. The air then enters an expansion chamber, in which its velocity
decreases dramatically. The dust and other light impurities then settle from the air
stream and are discharged from the machine.
The next step in the cleaning process involves a machine referred to as a “paddy
cleaner.” In the paddy cleaner (Fig. 3), paddy rice falls through an aspiration col-
umn onto a series of decks that are covered with perforated steel sheets. On the top
decks, objects that are longer or wider than rice, such as longer straws, soybeans,
and large seeds, are removed. On the lower decks, objects that are shorter than rice,
such as sand and small weed seeds, are removed and discharged from the machine.
Clean paddy rice then encounters a second aspiration column. As was the case with
the scalperator, the lighter-weight impurities are removed from the air in a settling
chamber and exit the machine.
Following the paddy cleaner, the next step is to remove short straws that are
longer than rice but not so much longer as to enable efficient separation in the paddy
cleaner. For this purpose, a type of disc grader, referred to as a “stick machine”
Fig. 2. The scalperator. (Courtesy Carter-Day International, Minneapolis, MN)

(Fig. 4) is often used. In a disc grader, a series of circular discs rotates on a central
shaft within a housing. When paddy rice is introduced into the housing, the rice
kernels fit into the indentations in the faces of the rotating discs. The kernels are thus
lifted until they fall out of the indentations and are discharged from the front of the
machine. Straws, being too long to remain in the indentations, remain in the housing
and are then conveyed to the end of the machine, where they are discharged.
After foreign material that is longer, shorter, wider, and lighter than rice has
been removed, it is necessary to remove impurities such as stones, mud balls, and
dense glass that might be present in the lot of paddy rice. To accomplish this, a
machine called a “destoner” is used. The destoner (Fig. 5) is a type of density sepa-
rator. Rice is fed onto an inclined, textured deck made of woven wire or perforated
sheet metal. Air is pulled through the rice from beneath the deck. The air pressure
Fig. 3. Paddy cleaner. (Courtesy Carter-Day International, Minneapolis, MN)
Fig. 4. Stick machine. (Courtesy Carter-Day International, Minneapolis, MN)

Milling / 287
causes the bed of rice to be fluidized (to lift away slightly from the textured surface
of the deck). Denser objects such as stones stay on the textured surface of the deck
and are conveyed by the vibratory motion of the machine to the high side of the
Fig. 5. Destoner. (Courtesy Satake Corp., Houston, TX)
Fig. 6. Drum magnet. (Courtesy Bunting Magnetics Co., Newton, KS)

deck, where they are discharged from the machine. The rice, which “floats” above
the vibratory action of the deck, falls under the force of gravity to the lower end of
the separating deck, where it is discharged.
Finally, metal that might have been present in the rice at the outset or that could
have been introduced into the rice stream during milling must be removed. For this
task, magnetic separators are employed. In cases involving high flow rates, drum
magnets (Fig. 6) and magnet hump separators (Fig. 7) are often used. These types of
separators commonly use ceramic magnets, which are capable of producing a strong
magnetic field that penetrates deep into a bed of rice.
Fig. 7. Double-hump magnetic separator. (Courtesy Eriez Magnetics, Erie, PA)
Fig. 8. Grate magnets. (Courtesy Magnetic Products, Highland, MI)

Milling / 289
Another type of commonly used magnetic separator is the grate magnet separator
(Fig. 8). When equipped with rare-earth magnets, the grate magnet produces an
extremely strong magnetic field at the surface of the magnet.
So far in the milling process, foreign materials that are metallic or that differ
from paddy rice by being much less dense, much more dense, longer, wider, or
shorter have been removed. In some milling applications, this degree of removal
of foreign materials may be sufficient. But if paddy rice is parboiled, it is ex-
posed to high temperature and moisture levels for extended periods of time.
Immature kernels, when exposed to these conditions, may become discolored and
damaged. In this case, it is necessary to remove immature kernels before the
parboiling process.
The machine used most often for this purpose is commonly called a “precision
grader,” although it would be more accurately described as a thickness grader (Fig.
9). In these machines, rice is tumbled inside a cylinder that contains rectangular
openings or slots. As the rice continually tumbles inside the cylinder, individual
grains eventually line up with the opening of the slot. Grains that are thinner than
mature kernels fall through the slot and are discharged from the machine, while the
remaining kernels are conveyed to the end of the cylinder, where they are collected
and discharged. The separating action of a thickness grader is shown in Fig. 10.
Fig. 9. Thickness grader. (Courtesy Carter Day, Minneaplis, MN)
Fig. 10. Thickness and width grading. Left, thickness grading using oblong perforations; right, width
grading using round perforations. (Courtesy Satake Corp., Hiroshima, Japan)
BROWN RICE AND MILLED RICE

At this point in the milling process, all of the foreign materials that can be
removed in a practical manner have been separated from the paddy rice. However,
other impurities can be removed later, after the husk and the bran layers have been
removed from the rice kernels. The rice kernel that remains after the husk is
removed is referred to as “brown rice.” Once both the husk and the bran layers have
been removed, the remaining rice is called “milled rice.”
When a thickness grader is used for cleaning brown rice or milled rice, the
functional operation of the machine is similar to that used for removing immature
grains from paddy before parboiling. In this case, however, the healthy brown rice
Fig. 11. A paddy kernel, showing the major parts. (Courtesy Satake Corp., Hiroshima, Japan)
Fig. 12. A modern paddy husker. (Courtesy Satake Corp., Hiroshima, Japan)
Milling / 291
kernels fall through the slots in the cylinder, while other seeds such as corn, wheat,
paddy, and “red rice” remain in the cylinder and are discharged separately. This
removal of foreign objects also protects the vital parts of the equipment that is to
follow.
Brown Rice Production
The production of brown rice from cleaned paddy rice involves the removal of
the husk from the rice kernel. This is performed in two separate but interrelated
stages. In the first stage, called “paddy husking,” the husk is broken loose from the
paddy kernel (Fig. 11) and removed from the rice. This process may result in dam-
age to the grain, and it is always abrasive to those parts of the machines responsible
for the husking action. Removing the husk from every paddy kernel is not always
desirable, because doing so might be economically infeasible. To account for this
“planned inefficiency,” a second stage is used, in which grains that have not had
their husks removed are separated from the brown rice grains and returned to the
husker.
In the first stage, a paddy husker (Fig. 12) is used to remove the husks from most
of the paddy kernels. Paddy rice is fed, at a controlled rate, through a gap between
two rubber-covered rollers turning at different speeds in opposite directions.
Fig. 13. Operation of the paddy husker. (Courtesy Satake Corp., Hiroshima, Japan)
Fig. 14. Paddy separator. (Courtesy Satake Corp., Hiroshima, Japan)
Because each side of the paddy attempts to travel at the same linear speed as its
corresponding roller, the husk is sheared from the brown rice kernel beneath the
husk. The efficiency of this operation is a function of the force with which each
roller presses into the paddy kernel, the quantity and shape of the stream of paddy
grains between the rollers, the linear speed of the surface of the two rollers, and the
hardness of the rubber covering the rollers.
After the husk has been broken loose from the brown rice, it is removed by
aspiration. The combined stream of brown rice and husks falls, in a thin layer,
through a column of air. The husks, being less dense than brown rice, are lifted
by the air and carried into a settling chamber, where they are collected and
discharged separately from the brown rice. The operation of the paddy husker is
illustrated in Fig. 13.
The normal husking efficiency of a modern paddy husker in commercial opera-
tion is approximately 90%. This means that 10% of the throughput consists of
paddy grains. The removal of these grains from the brown rice is the task of the
paddy separator.
The paddy separator (Fig. 14) is a high-capacity density separator. When the
mixture of paddy grains and brown rice kernels on a separating deck of the paddy
separator is subjected to an oscillating motion, the kernels stratify, with the less-
dense paddy grains “floating” on the top of the denser brown rice kernels. Each of
the separating trays has a texture that, combined with the oscillatory motion of the
tray, causes the brown rice kernels to move toward the high end of the tray, where
they are discharged. The lighter paddy kernels remain on top of the brown rice, and
do not come in contact with the textured surface of the separating tray. Thus the
paddy grains fall, under the influence of gravity, to the lower end of the separating
tray, where they are discharged separately from the brown rice. The separating
action of a paddy separator is shown in Fig. 15.
Milling / 293
Fig. 15. Principles of a paddy separator. (Courtesy Satake USA, Inc.., Houston, TX)
This type of paddy separator usually has three discharges. One, on the high
side, discharges almost pure brown rice. A second, on the low side, discharges
rice that has a high concentration of paddy. A third discharge, in the middle of
the separating tray, discharges a mixture of paddy and brown rice that is usually
returned to the inlet of the separating trays. As a result, the elevators and con-
veyors that serve these machines often carry twice the rated throughput of the
machines themselves.
Bran Removal
Once all husks have been removed from the brown rice, the next step is to
remove the germ and the bran layers from the underlying starchy endosperm. In the
United States, a large portion of the rice crop is milled and then exported. Different
markets require different degrees of bran removal during milling, so bran removal
must be precise and controllable. For most milled-rice products, the value of the
finished product is determined largely by the degree to which the product contains
broken kernels.
Thus, the objective of the milling process is to selectively remove the germ and
the bran layers from rice, leaving the milled kernel largely intact and with a specific
utility and appearance.
ABRASIVE AND FRICTION METHODS

Two primary types of milling actions are commonly used to remove the germ
and bran layers from brown rice. These two milling actions are illustrated in Fig.
16. The first is an abrasive cutting action. The other uses pressure and movement
between the grains to enable the force of friction to tear away bran layers. Each of
these actions has its inherent advantages and disadvantages.
The abrasive action uses low pressure. As a result, it produces fewer broken
grains, or “brokens,” and is a more efficient use of power. On the other hand, abra-
sive milling produces a relatively rough finish and often reduces the size of the
germ rather than removing it from the kernel. The abrasive cutting surfaces used in
this method are likely to cut into the starch endosperm while removing the last bran
layers. As a result, some starch may be lost with the bran that is removed. For this
reason, it is difficult to achieve the optimum total milled-rice yields using only this
method.
The friction method, because of its higher pressure, tends to produce more bro-
kens, and it takes more power per unit of capacity than does abrasive milling. Using
this method, however, one can precisely control the amount of bran removal. It pro-
duces a smoother, more lustrous appearance than does the abrasive method. Finally,
when excessive milling pressure can be avoided, this method produces the greatest
chance for optimized total milling yields.
Since each of the two methods has its limitations, modern rice mills usually
obtain the best results by combining them. The first layers of the bran are typically
removed using the low pressure and high power efficiency of abrasive milling. The
bran removal is completed using the rubbing action of the friction method, resulting
in superior appearance and optimum total milled-rice yields.
An abrasive bran removal machine is commonly referred to as an “abrasive
whitener.” The whitener most commonly used in the United States is pictured in Fig.
17A. In this machine, brown rice is introduced into the annulus formed between an
Fig. 16. Models of milling action. Left, tearing action; right, grinding action. (Courtesy Satake Corp.,
Hiroshima, Japan)
Fig. 17. Vertical whiteners: abrasive (left) and friction (right). (Courtesy Satake Corp., Hiroshima, Japan)
Milling / 295
outer perforated metal cylinder and an inner rotor composed of rotating abrasive
wheels. As the brown rice comes into contact with the abrasive wheels, some of the
bran is abraded from the kernels and is removed by a negative air system through
the perforations in the outer cylinder. It is critical to this type of machine that
individual grains change position with respect to the abrasive wheels and the
perforated cylinder during the milling process. Different mechanical devices,
including resistance bars and rotating blades, are used for this function.
Following the removal of the outer bran layers using abrasives, the remaining
layers are removed using one or more machines that primarily use the rubbing or
friction type of milling action. Today, the most commonly used of this type of
machine is a “friction whitener,” shown in Fig. 17B. In this machine, the bran is
removed from the grain primarily by rubbing each grain against other grains, with
the bran again being removed by suction through a perforated cylinder. The rubbing
action is produced by an eccentric rotor, which turns through the grain within the
multisided perforated cylinder.
The vertical friction whitener also has the ability to add water to the rice just as it
enters the milling chamber. The water removes loose bran and decreases the rise in
temperature of the rice during the process.
For markets that value highly polished milled rice, a machine called a “mist pol-
isher” is used to complete the bran-removal process. The mist polisher (Fig. 18) is
similar in function to the friction whitener. In the mist polisher, a combination of
compressed air and water produces an atomized mist that is applied to the rice at the
entrance to the milling chamber. As with the vertical friction whitener, the water
serves to clean and cool the surface of the rice as it is being milled. Because of its
long milling chamber and relatively large clearance between rotor and perforated
cylinder, the mist polisher is capable of producing extraordinary finishes for the
product (milled rice) without generating significant breakage.
VERTICAL VERSUS HORIZONTAL MACHINES

In recent years, modern rice mills have been moving away from horizontal bran-
removal equipment in favor of vertical milling machines. It is widely believed that
vertical milling machines inherently produce fewer brokens than horizontal machines.
Fig. 18. Mist polisher. (Courtesy Satake Corp., Hiroshima, Japan)

Fig. 19. Horizontal and vertical bran removal machinery, abrasive and friction types. (Courtesy Satake
USA, Inc., Houston, TX)
Many of the vertical units being offered by today’s manufacturers do indeed break
less rice, but this is more likely due to improvement in the application of milling
principles than to simple verticality. Today’s machines make more efficient use of
abrasive and textured surfaces, employ longer milling chambers, and promote
greater grain-to-grain movement than their predecessors. It is logical, then, that they
would produce fewer brokens than their earlier horizontal counterparts.
Vertical machines do have some inherent advantages over horizontal machines,
however. In horizontal machines, the grain density in the lower part of the milling
chamber and in the side of the chamber on the leading edge of the grain rotation
tends to be significantly higher than that in the upper and trailing sections. In
vertical machines, the grain density is inherently more uniform, since all of the
milling chamber is normally filled. This means that vertical machines use all
available abrasive and perforated surfaces, while horizontal machines do not. Thus,
vertical machines tend to mill more uniformly, at higher capacities, and with greater
efficiency of power utilization. This principle is illustrated in Fig. 19.
Classification—Removal of Broken Grains
MECHANICAL MEANS
Even though modern rice milling systems are very efficient, some rice kernels
are inevitably broken sometime during the milling process. After the rice has been
milled, it is necessary to separate those broken kernels from the whole grains.
Depending upon the level of broken grains allowed by the customer’s specification,
these broken grains may then be blended in a precise ratio with unbroken kernels to
maximize the profitability of the mill.
The first machine used for the removal of broken kernels is a high-capacity
grading sifter, like the one shown in Fig. 20. The grading sifter uses a rotary or
vibratory motion to expose grains to a surface that is made of perforated metal or
woven wire cloth. Grains shorter than a certain size fall through the openings in the
cloth, while longer kernels remain on the sifting surface. This principle can be used
to remove foreign objects like bran balls from rice, while at the same time removing
small broken grains and some whole grains. The grading sifter’s principle of opera-
tion is illustrated in Fig. 21.
The grading sifter isolates small broken grains and some of the whole kernels.
To segregate the remaining broken grains and whole kernels, a more precise type of
length grader, called the “indented cylinder grader,” is used. The indented cylinder
Milling / 297
Fig. 20. Grading sifter. (Courtesy Sweco, Florence, KY)
Fig. 21. Sifting operation. (Courtesy Sweco, Florence, KY)
is pictured in Fig. 22. In this machine, rice is placed inside a cylinder that has
indentations punched into its inside surface. As the cylinder rotates, rice kernels are
lifted by the indentations. Depending upon the length of the individual grains, they
fall from the indentations at different points. The broken grains, which are shorter
than the unbroken kernels, remain in the cylinder longer, and are lifted to greater
heights. An adjustable collecting trough is positioned so as to select the size and
amount of brokens to be removed. This principle is illustrated in Fig. 23.
ELECTRICAL SORTING
Through magnets and mechanical means, modern rice millers are able to remove
almost all foreign objects from rice during the milling process. But there may be
some nonmagnetic objects that are similar in size and density to whole or broken
rice kernels, and these are very difficult to remove using mechanical means.
To remove such foreign objects, it is common to use devices that can differ-
entiate between rice kernels and the foreign objects on the basis of their appearance.
These devices are commonly referred to as “color sorters,” although in most cases
these machines cannot detect differences between colors.
In an “optical sorter” (Fig. 24), rice kernels are made to fall at a specific speed
down a path, causing them to pass through the focal point of a camera. The camera
gathers light, which is reflected from an adjustable plate known as the “back-
ground.” A photoelectric sensor then converts the gathered light into an electrical
signal, the magnitude of which is proportional to the intensity of the reflected light.
Any grain that appears darker than the background results in a negative voltage. A
trigger mechanism, called the “sensitivity control,” allows the operator to define
how dark an object must be in order to be considered “objectionable.” A
microprocessor then analyzes the signal, compares it to the magnitude that is
considered objectionable, and if appropriate, sends a signal to an electromagnetic
valve. The valve is energized, opens, and allows a blast of compressed air to escape.
The compressed-air blast deflects the undesirable kernel from the rice stream,
removing it from the finished product.
As a final protection for the finished milled rice, it is common to employ metal
detectors, which can remove even nonferrous metals from rice. Metal detectors can
Fig. 22. Indented cylinder grader. (Courtesy Carter Day International, Minneapolis, MN)
Fig. 23. Operation of an indented cylinder. The unbroken kernels (which are larger than the inden-
tions) fall out, while the broken kernels stay in the indentions and fall into the collecting trough.
(Courtesy Satake Corp., Hiroshima, Japan)
Milling / 299
Fig. 24. Optical sorter. (Courtesy Satake Corp., Hiroshima, Japan)
Fig. 25. Metal detector. (Courtesy Fortress Technology, Inc., Scarborough, ON, Canada)
be used in spouts, with rice flowing in bulk through the detector, or after the rice is
in packages. The “drop-through” metal detector is illustrated in Fig. 25. In such a
device, a magnetic field is generated by a radio frequency broadcast. Anything
magnetic or electrically conductive that enters this field causes a disturbance in the
strength of the magnetic field. The disturbance is amplified, and a signal is sent to a
diverter valve, which then removes the object that produced the disturbance.
CONTROLLING THE RICE MILLING PROCESS
This chapter has given a brief overview of the functions of the primary machines
of the rice milling process. Although the machinery is important, controlling the
operation of the machinery is even more critical. For this reason, it is appropriate to
mention some of the principles used in establishing a control philosophy for the
modern rice mill. These principles should be consistent with optimizing the eco-
nomics of the milling process, rather than focusing only on technical measurements.
For any given milled-rice appearance, optimizing the milling process is largely a
matter of avoiding the creation of broken kernels and the loss of starch to the by-
products. Of these two, avoiding the loss of total milling yield is more important
than avoiding breakage.
Optimization of the milling process requires timely knowledge of the distribution
of milled rice products and the by-products being produced. It is critical that the
miller be aware of both the nature and the quantity of every product and by-product
stream that enters and leaves the facility. Analysis of this “material balance” is what
alerts the miller to the existence of problems in the milling process and provides
guidance toward the appropriate solution.
CHAPTER 12
IMPACT OF DRYING, STORAGE, AND MILLING

ON RICE QUALITY AND FUNCTIONALITY
Terry J. Siebenmorgen
Jean-Francois Meullenet
Rice Processing Program
University of Arkansas
Fayetteville, Arkansas
Rice “quality,” in its many definitions, can be affected at almost any point in the
postharvest processing chain. The rice processing industry faces a continual chal-
lenge to prevent various forms of chemical and physical degradation in order to
maintain quality at its highest level. Given the status of a rice lot delivered at harvest,
overall quality can generally not be improved, short of specialized processes that im-
prove particular quality attributes (e.g., parboiling can improve milling quality tre-
mendously). It is more generally the case that processors are expected to maintain
quality at the delivered levels. Given the volume of product typically handled by
processing facilities, as well as the wide range of sources and production practices
often experienced with rough rice, maintaining quality from the first point of de-
livery through final packaging is indeed challenging.
Chapters 9–11 in this text address the postharvest operations of rice drying, stor-
age, and milling. They concentrate on the theory and current understanding of the
processes that make up these unit operations, as well as the commercial procedures
and equipment currently in use. This chapter emphasizes the effects these operations
can have on milling performance and other aspects of rice quality and functionality.
Both physical and chemical indices of rice quality are included in this chapter.
The physical quality indices characterizing milling quality are heavily emphasized,
due in large part to the economic importance of milling quality to the industry.
However, quality indices dictated more by chemical structure and behavior, such as
pasting properties, are also given due attention, to reflect the increasing importance
of these factors to end-use processors. Finally, the emerging fields of rheology and
sensory science have allowed the effects of postharvest operations on consumer
quality indices to be preliminarily assessed.
PRE-DRYING ISSUES
Discussion of the effects of postharvest operations on quality must begin with

consideration of the handling practices and the subprocesses that rice undergoes
before drying and storage. The harvesting “window” for rice is relatively short,
301
particularly considering the volume of rice that must be dried within a short time.
Modern combine harvesters have greatly increased the speed at which rice can be
harvested; along with larger and faster grain carts, trucks, and trailers for
transporting rice from combines to driers, they have enabled a much greater deliv-
ery rate to driers. In some cases, the delivery rate can exceed the drying and han-
dling capacity. In such instances, questions have arisen as to the effects on quality
of temporary “wet holding.”
Respiration
Rice, like any living organism, respires in the presence of oxygen. The equation
describing this process is (Mohsenin, 1980):
C6H12O6 + 6O2 6CO2 + 6H2O + 677.2 kcal
As the equation indicates, carbon dioxide, water, and energy are produced by the
oxidation of carbohydrates. Along with the rice kernel itself, microbes associated
with the rice also respire and greatly contribute to the overall respiratory activity in
a rice bulk, particularly under the conditions of moisture content (MC), relative
humidity, and temperature that promote microbial growth.
Several deleterious effects can be incurred from high respiration rates, especially
if respiration is allowed to proceed over extended periods. Kernel discoloration,
sometimes referred to as “yellowing” or “stackburn,” is the most commonly rec-
ognized negative effect of advanced respiration in rice and typically results from
storage at high MC. Yellowing can be caused by delayed or improper drying of
rice (Sahay and Gangopadhyay, 1985) or by improper handling of fresh rice after
harvest or during storage such that respiration is increased (Aibara et al, 1984).
“Dry matter losses” from grain are also incurred, as indicated by the equation. Dry
matter losses are sometimes reported as a function of storage MC and temperature
for grains; an example is given for barley and wheat by Burrell (1982). Alterna-
tively, allowable storage durations under various grain MCs and storage tempera-
tures can be estimated for maintaining dry matter losses less than given levels, by
using knowledge of the respiration rate. Brooker at al (1974) present an example
for corn.
Experimentally, the rate of respiration is often measured by the rate of carbon
dioxide production. However, in commercial practice, high respiration rates are in-
dicated by elevated temperature zones, or “hot spots,” in a grain mass. Hot spots
typically result from localized high-MC areas, since the rate of respiration is ex-
ponentially related to MC (see below). Foreign materials, such as leaf and stalk
sections or weed seed or other weed plant material, also respire at high rates be-
cause of the typically high MC of this material. Siebenmorgen et al (1994) mea-
sured the MC of rough rice, as well as that of accompanying stalk and leaf material,
throughout a harvest season. They reported an MC of 66.1% for a rice stalk-leaf
mixture when the rice grain MC was 19.8%. Certainly, there is considerable merit
in properly adjusting combines to minimize levels of material other than grain and
in scalping or cleaning rough rice before drying and storage to avoid the unwanted
effects of respiration from this material.
Impact of Drying, Storage, and Milling / 303
Factors Affecting Respiration
Both the MC and the temperature of rice dictate the rate at which respiration
occurs. Bailey (1940) reported the rate of respiration of rough rice over a limited
MC range of 12–17%. Dillahunty et al (2000) measured the respiration rate of
rough rice for both medium-grain (Bengal variety) and long-grain (Cypress variety)
rough rice over a range of temperatures and MCs. Figure 1 presents the results of
Dillahunty et al as well as those of Bailey (1940); the respiration rates of both stud-
ies were in close agreement. Figure 1 also shows the exponential response of respi-
ration rate to MC. The rate begins to significantly increase above a MC of 14%,
which indicates why rough rice must be dried to approximately this MC level
quickly after harvest. The figure further shows that, in order to minimize respiration
rates and resultant dry matter losses over long storage durations, the MC is typically
lowered to 12–13%.
Figure 2 indicates the effect that temperature has on respiration rates. As the
storage temperature of rough rice increases, respiration rate increases to a certain
point, after which it decreases. The decrease is believed to be due to the thermal
retardation of respiration of the kernel and associated microbes present on the ker-
nels. Figure 2 shows that the temperature at which respiration peaks decreases as
the MC increases. These trends were similar across both long- and medium-grain
varieties.
Quality Effects Resulting from Respiration
Postharvest respiration processes can produce yellowing of rice that affects rice
quality, appearance, flavor, and yield (Singaravadivel and Raj, 1983; Phillips et al,
1988; Misra and Vir, 1991). Although yellowing occurs in paddy rice, the endo-
sperm itself becomes discolored; thus, yellowing is not apparent until the rice is
milled.

3UHGLFWHG&\SUHVV
3UHGLFWHG%HQJDO
5HVSLUDWLRQ5DWH
%DLOH\
VGDWD
PJ&2 NJKU

0RLVWXUH&RQWHQW
Fig. 1. Predicted respiration rate curves as reported by Bailey (1940) and for rice cultivars Cypress and
Bengal generated from experiments in which conditioned moisture content was varied and temperature
was maintained at 30°C. (Reprinted, with permission, from Dillahunty et al, 2000)
Yellowing does not necessarily produce yellow-colored kernels. Colors can

range from yellow to orange to reddish. Some factors cited as being responsible for
yellowing include various species of fungi or mold (Schroder, 1963); grain water
activity; and surrounding air temperature, oxygen, and carbon dioxide content (Bason
et al, 1990). Some of these factors may also work together.
Dillahunty et al (2001) investigated the effects of temperature and heating dura-
tion on medium-grain (Bengal variety) and long-grain (Cypress variety) rough rice
at high (~21%) and low (~18%) harvest MC levels. Results, as measured by the
color values chroma and hue angle, showed that, as exposure duration and tem-
perature increased, the occurrence of yellow rice also increased (Figs. 3 and 4). Vi-
sual observations of discolored samples indicated that certain kernels were discol-
ored more than others. The viscosity of samples measured after exposure to various
temperatures (Fig. 5) showed that only rice exposed to the most severe combina-
tions of temperature and exposure duration had reduced peak viscosity. Less severe
treatments produced yellowed rice but not a reduction in peak viscosity.
Fig. 2. Respiration rates for Bengal (A) and Cypress (B) rice at different temperatures and moisture
contents (MCs). Each data point is the mean of the measurement of three separate replicates. (Re-
printed, with permission, from Dillahunty et al, 2000)
Wet-Holding Effects on Rice Functionality
Daniels et al (1996, 1998) and Meullenet et al (1999) studied the effects on rice
quality of holding rough rice at high MCs. These experiments were designed to simu-
late extended delays in drying, as might occur during the peak of the harvest season.
Daniels et al (1998) showed that head-rice yield (HRY) did not seem to be affected
by wet-holding conditions (20.5% MC for 86 hr at 20°C) for long-grain rice. How-
ever, the delayed-drying condition, particularly for samples dried at low tempera-
ture, resulted in significantly lower water absorption and volume expansion com-
pared to the samples that were dried immediately after harvest (Daniels et al, 1996).
Fig. 3. Chroma values of milled rice resulting from exposing rough rice from each cultivar-location
combination to the temperatures and durations indicated. Each point represents combined data for
harvested samples at high and low moisture content levels (mean value of 16 color measurements).
(Reprinted, with permission, from Dillahunty et al, 2001)

K

K

H
O
J
Q
D

H
X K
+

R
7HPSHUDWXUH &
Fig. 4. Hue angle values of milled rice resulting from exposing rough rice to the temperatures and
durations indicated. Lower values indicate darker-colored rice. Each point represents combined data of
Bengal and Cypress samples harvested at high and low moisture contents (mean value of 32 color
measurements). (Reprinted, with permission, from Dillahunty et al, 2001)
&
&
&
&

&RQWURO &

&

7LPHPLQ
Fig. 5. Viscosity profiles for Cypress rice exposed to the indicated temperatures for 72 hr. (Reprinted,
with permission, from Dillahunty et al, 2001)
Meullenet et al (1999) reported slight but significant effects of wet holding (us-
ing the same holding conditions described by Daniels et al, 1998) on the sensory
quality of long-grain Cypress rice. Wet holding significantly affected both the
clumpiness and hardness of cooked rice. Samples for which drying was delayed
yielded a significantly clumpier and less firm cooked rice. Recently, the impact of
holding paddy rice (Akitakomachi and M202) with moisture contents in the 17–27%
range on the aroma and flavor of white rice following drying and milling was
examined (E. T. Champagne, J. Thompson, K. L. Bett-Graber, R. Mutters, J. A.
Miller, and E. Tan, unpublished data). The undesirable flavor note “sour/silage”
significantly increased in milled rice (i.e., for both cultivars) with an increase in
paddy moisture content after 48 hr of storage at a temperature of 40°C. Sour/silage
and alfalfa/green bean notes significantly increased in intensity in milled rice with
paddy storage time for paddy stored for up to 48 hr in the range of 24–27% MC.
Astringent mouthfeel significantly increased in milled rice with time of storage of
paddy at 27% MC. Other undesirable flavor attributes such as sewer-animal,
haylike musty, and silverlike metallic in milled rice did not significantly change
with high-moisture storage of paddy rice. These results point out the necessity of
evaluating postharvest handling as a system, as each postharvest operation has the
potential to influence the functionality and quality of the final product.
DRYING
The previous section indicated that if high-MC rice is not dried within a given
period, respiration processes could cause reductions in quality in terms of discolora-
tion, functionality, and sensory properties. Drying in a timely fashion is thus critical to
maintaining quality at its highest possible level at the start of postharvest operations.
Lowering MC to acceptable levels involves heat and mass transfer subprocesses,
which are fairly well established. However, changing the MC of rice, as of most
hygroscopic materials, causes corresponding changes in many physical properties,
the summation of which can cause hygroscopic stress within kernels. If localized
stresses reach levels greater than the kernel material’s strength, fissures can form
due to material failure. The challenge associated with rice drying is thus not only to
reduce MC, but to reduce it without compromising the physical integrity of the ker-
nel. Additionally, drying should be accomplished without imposing chemical dam-
age that would degrade rice functionality.
Chapter 9 in this text addresses the fundamentals of rice drying. Additionally,
chapters in other texts address the subject; these include Kunze and Calderwood
(1980) and Wang and Luh (1991). A related chapter by Siebenmorgen (1994) pre-
sents the importance of MC, from preharvest through final laboratory assessment, in
affecting milling quality of rice. Finally, a recent chapter by Siebenmorgen et al
(2003) focuses on milling quality effects that can occur as a result of the drying
process.
Effects of Drying on Milling Quality
The most notable negative quality effect of drying is HRY reduction. The maxi-
mum possible HRY associated with a harvested rice lot, dried under the most gentle
conditions, is determined by several factors, including the rice variety, growth con-
ditions, and harvest MC. Siebenmorgen et al (1992) and Pan et al (2002) have
shown that harvesting at too high a MC, in which a high percentage of immature
kernels are present, or at a very low MC, in which dried kernels may have fissured
due to rapid moisture adsorption, can cause reductions in HRY relative to harvest-
ing at the optimal MC.
Given a maximum possible HRY for a lot, the goal of rice driers, particularly
commercial driers, is to maximize the drying rate while maintaining a HRY as close
as possible to the maximum. As such, research has been conducted to quantify the
HRY reduction associated with certain drying conditions. In particular, Fan et al
(2000) conducted a large drying study to elucidate the effects of several harvest and
drying conditions. In this study, one medium-grain and two long-grain varieties
were harvested over a range of MCs and locations and dried in thin layers under
three air conditions (A, 43.5°C, 38% rh; B, 51.7°C, 25% rh; and C, 60.0°C, 17% rh)
with corresponding equilibrium moisture contents (EMCs) of 9.5, 7.3, and 5.8%, re-
spectively. After drying for a certain duration, samples were immediately cooled in
a chamber held at 21°C, 50% rh, which corresponds to an EMC of approximately
12.5%. As shown by Cnossen and Siebenmorgen (2000), this cooling treatment can
cause HRY reduction if a sufficient MC gradient is present in kernels at the end of
drying. Fan’s study was thus indicative of a drying situation in which cooling is
experienced by the rice immediately after drying.
Figure 6 represents the HRY reductions that occurred when Bengal medium-
grain rice was dried under the three indicated air conditions over a range of dura-
tions before being cooled to 21°C. The data show that under the 9.5% EMC air con-
dition, continuous drying for periods up to 3 hr did not affect HRY. However, for
more-severe, lower-EMC air conditions, HRY decreased dramatically after shorter
drying durations. Figure 7 gives a comparison of the varietal responses when drying
under the two more severe drying conditions; the medium-grain variety was more

$&5+(0&

+HDGULFH\LHOG

%&5+(0&

&&5+(0&

'U\LQJGXUDWLRQPLQ
Fig. 6. Head-rice yield (each point is the average of two replications) of Bengal variety rice (harvested
at 22.5% moisture content, from Stuttgart, AR) versus drying duration under the three indicated drying
air conditions before being cooled to 21°C. RH = relative humidity, EMC = equilibrium moisture
content. (Reprinted, with permission, from Fan et al, 2000)
susceptible to HRY reductions than the long-grain varieties. Fan et al (2000) further
reported that, as the harvest MC of a rice variety is increased, more moisture can be
removed before HRY reduction occurs under a given drying air condition. This
point is summarized by Figure 8, which concurs with the findings of Pan et al
(2002), who dried medium-grain samples of M202.
Fig. 7. Head-rice yield of three rice varieties, each at the indicated harvest moisture content (HMC),
versus drying duration when dried under air conditions of 51.7°C and 25% rh (A), and 60.0°C and
17% rh (B) before being cooled to 21°C. (Reprinted, with permission, from Fan et al, 2000)

&\SUHVV
3HUFHQWDJHSRLQWVRI0&
'U\LQJFRQGLWLRQ&
%HQJDO
'U\LQJFRQGLWLRQ%

%HQJDO
'U\LQJFRQGLWLRQ&

+DUYHVWPRLVWXUHFRQWHQW
Fig. 8. Percentage points of moisture content (MC) removed before head-rice yield reduction occurred in
relation to harvest moisture content. Drying conditions: B = 51.7°C, 25% rh, 7.3 equilibrium moisture
content (EMC); C = 60.0°C, 17% rh, 5.8 EMC. (Reprinted, with permission, from Fan et al, 2000)
While beyond the scope of this chapter, further work in quantifying HRY reduc-
tions was conducted by Cnossen and Siebenmorgen (2000), who showed the effects
of tempering for various durations after drying. In related work, Cnossen et al
(2003) determined the effect of drying and tempering treatments on kernel fissuring.
Their work indicated that tempering durations based on minimizing fissures due to
the drying process may be longer than tempering durations based on minimizing
HRY reduction.
Effects of Drying on Sensory Quality
According to Daniels et al (1996), samples dried under high-temperature condi-

tions (54°C, 22% rh) had a significantly higher overall sensory impact and grain
flavor note. Samples dried under low-temperature conditions (33°C, 68% rh) were
found to be rougher and harder and exhibited a lower cohesiveness of mass. No
significant differences were found between samples for the stale grain flavor note,
sulfur flavor note, clumpiness, gluiness, moisture absorption, or geometry of slurry.
Champagne et al (1997) reported no significant changes in the flavor of rice (Ben-
gal, M-401, and Koshihikari) dried at temperatures ranging from 18 to 60°C. In the
same study, Lyon et al (1999) reported no effect of these drying conditions on the
sensory texture characteristics of the same three cultivars grown in Arkansas and
California. After instrumental testing to assess rice texture, few effects of drying
temperature were reported (Champagne et al, 1998), with the exception of cohesive-
ness, which was found to be lower in rice dried at lower temperature than in that
dried at higher temperature.
Meullenet et al (1999) reported significant effects of drying on the sensory char-
acteristics of long-grain Cypress rice. Although no effects of drying temperature
(33–55°C) were reported immediately after drying, the aging process seemed to be
affected by this initial processing step. After four weeks of storage at temperatures
ranging from 4 to 38°C, the effect of drying temperature became apparent. The
more gentle drying treatment resulted in significantly greater roughness and hard-
ness values and lower starchy note, cohesiveness of mass, and overall flavor impact.
After 20 weeks of storage, similar trends were observed and significant differences
reported for sulfury aroma, clumpiness, and moisture absorption. Samples dried at
the lower temperature exhibited significantly greater clumpiness and lower starchy
notes and moisture absorption.
Effects of Drying on Other Functional Characteristics
Drying conditions have been reported to affect cooking properties and peak past-
ing viscosity of long-grain rice (Daniels et al, 1998). Rice dried at low temperature
(33°C) had greater water absorption and volume expansion than did rice dried at
high temperature (54°C). For both water absorption and volume expansion, drying
condition was a significant contributor as a main effect and also as an interaction
with storage duration and temperature. The peak viscosity of rice flour slurries, as
evaluated by the Rapid Visco-Analyser (RVA), was greater for low-temperature-
dried rice than for high-temperature-dried rice.
Fan et al (1999) conducted a study to investigate the effects of several post-
harvest operations, including high- and low-temperature drying treatments, on the
gelatinization and retrogradation properties of long-grain rice. Drying treatment was
significant only in interaction with the storage and pre-drying treatments in predict-
ing the conclusion temperature of differential scanning calorimetry thermograms of
gelatinization and retrogradation. Drying treatment was not significant in predicting
the enthalpy or onset and peak temperatures for either gelatinization or retrograda-
tion data.
The exact influence that drying has on the physicochemical properties respon-
sible for sensory and functional characteristics is largely unknown. However, Teo
et al (2000) reported that modification of the protein component, rather than the
starch, was primarily responsible for the rheological changes associated with rice
flour aging. These authors also reported that the protein oryzenin was denatured at
temperatures between 47 and 66°C. Based on this, it could perhaps be speculated
that oryzenin can be denatured in some drying situations and could be, at least in
part, responsible for some of the effects on functionality described above. This ques-
tion is currently being addressed.
STORAGE
Effects of Storage Conditions on Functional Properties of Rice
Research has shown that rough-rice storage history can affect properties such as
HRY and cooking quality of rice (Villareal et al, 1976; Chrastil, 1990; Hamaker et
al, 1993; Tamaki et al, 1993). Changes during storage include increases in grain
hardness (Tsugita et al, 1983; Sajwan et al, 1989; Dhaliwal et al, 1991), water ab-
sorption and volume expansion (Villareal et al, 1976; Tsugita et al, 1983), and peak
viscosity (Villareal et al, 1976; Tsugita et al, 1983; Hamaker et al, 1993). Daniels et
al (1998) and Pearce et al (2001) reported on the effects of multiple postharvest
factors (including drying temperature, storage moisture content, and storage tem-
perature and duration) on end-use quality parameters such as HRY, water absorp-
tion, and volume expansion and on amylography. These authors found that HRY
increased with storage duration at all storage conditions. The greatest increases
were seen during the first three months of storage. This confirmed results from
Perez and Juliano (1981), Barber (1972), and Villareal et al (1976), showing that the
most significant changes during rice storage occur during the first 90 days. Rice
stored at 10% MC exhibited significantly higher HRY than did rice stored at 12 and
14% MC. On average, the water absorption and volume expansion for all samples
increased slightly with storage duration, although the changes were not drastic. This
confirmed results reported by Chrastil (1992), Villareal et al (1976), and Tsugita et
al (1983). Pearce et al (2001) reported that storage temperature had a significant
effect on the cooking properties of rice. Both water uptake and volume expansion
increased with increasing storage temperature, particularly when rice was stored at
38°C. According to Daniels et al (1998), peak viscosity did not follow a pattern
over time. This did not support the results of Hamaker et al (1993), Perdon et al
(1997), and Pearce et al (2001), which showed significant increases in peak viscos-
ity during the first three months of storage. Pearce et al (2001) also reported that the
rate of increase in paste viscosity over time was greater for rough rice stored at
higher temperatures.
Overall, results reported for studies performed by the Rice Processing Program
at the University of Arkansas (Perdon et al, 1997; Daniels et al, 1998; Pearce et al,
2001) showed that rough-rice postharvest parameters affect the end-use functional-
ities of rice through complex relationships. The data showed that the functional
properties of rice are not related to postharvest factors by simple linear relationships
and that significant interactions between the factors are present. This supports the
hypothesis that physicochemical changes in rice are complex functions of posthar-
vest conditions.
Effects of Storage Conditions on Sensory Properties of Rice
Several recent studies have reported the effects of storage conditions on the sen-
sory properties of rice. These studies have focused on the evaluation of the effects
of rice MC (Champagne et al, 1997, 1998; Lyon et al, 1999, 2000; Meullenet et al,
1999, 2000), storage temperature (Meullenet et al, 1999, 2000), and storage dura-
tion (Meullenet et al, 1999, 2000). All these studies have clearly demonstrated the
importance of storage conditions in determining rice sensory quality.
Lyon et al (1999) found that rice dried to 12% was stickier to lips and chewier
than rice dried to 15% MC. The effect of MC seemed to be dependent on the culti-
var studied. Champagne et al (1998) reported that the texture properties of rice,
measured instrumentally, were also affected by rough-rice MC. In general, instru-
mental values for hardness were higher (and those for cohesiveness, adhesiveness,
and springiness were lower) in regular-milled rice dried to 15% moisture than in
rice dried to 12%. Conversely, hardness was lower and cohesiveness, adhesiveness,
and springiness were higher in deep-milled rice at the higher MC.
Results reported by Meullenet et al (2000) showed that rice stickiness was not af-
fected by storage MC, while hardness decreased with increasing storage MCs. The
trends in the effects of MC on hardness are in agreement with the work of Tamaki
et al (1993), who reported that rice stored at 12% MC was initially firmer than rice
stored at 15 or 18% MC. Other sensory characteristics affected by rough-rice stor-

age MC were toothpull and toothpack. These characteristics both decreased with
increasing storage MC. As storage progressed, stickiness was found to decrease
while hardness increased during the first 20 weeks of storage (Meullenet et al,
2000), although the kinetics of these changes were affected by both storage MC and
temperature.
Champagne et al (1997) also reported effects of MC on rice flavor. Sulfurlike
aroma was higher in samples dried to 15% MC than in samples dried to 12% MC.
However, this result was not supported by the results reported by Meullenet et al
(2000). Starchy aroma intensities were also affected by storage MC (Meullenet et
al, 2000), with 12%-MC rice being higher in starchy aroma. This is in agreement
with Champagne et al (1997), who reported higher starchy notes in samples with
lower MC.
Although storage MC has been shown to affect the sensory quality of cooked
rice, the effects of storage temperature and duration have been shown to be more
pronounced (Meullenet et al, 2000). Meullenet et al reported that storage tempera-
ture and duration affected virtually all flavor and texture attributes of rice. Of par-
ticular interest was the increase in oxidation notes in rough rice stored at elevated
temperatures and the effects of storage duration on texture properties. Cooked-kernel
hardness reached a maximum between 15 and 22 weeks of storage depending on
rough-rice storage MC. Storage temperature and duration significantly affected rice
stickiness (Meullenet et al, 2000). Increasing storage temperatures decreased rice
stickiness. Tamaki et al (1993) reported similar results on Japanese rice cultivars
using instrumental measurements to assess stickiness. In the study of Meullenet et al
(2000), rice stickiness reached a maximum after 20 weeks of storage and decreased
significantly after 36 weeks of storage. A shorter storage duration was necessary for
the samples stored at higher temperature (38°C) to reach the maximum perceived
stickiness. This is in partial disagreement with Tamaki et al (1993), who reported
that rice stickiness decreased consistently during the first 90 days of storage, re-
gardless of storage MC or temperature.
MILLING
Most rice is milled for direct consumption or for subsequent utilization as an

ingredient in end-use products. The primary purpose of milling is to remove
the germ and bran layers from the kernel endosperm. The extent to which bran
has been removed from the kernel endosperm is referred to as the “degree of
milling” (DOM). To a large extent, the milled-rice customer and the intended
use of the rice dictate the target bran removal level (e.g., most rice milled for
breakfast cereal processing is not milled to the same extent as that used for
“table” rice).
The four morphological layers surrounding the rice kernel endosperm (the peri-
carp, seedcoat, nucellus, and aleurone layers [Luh, 1980]) and the germ are collec-
tively referred to here as “bran.” The bran contains approximately 18–20% lipids
and 14–15% protein, while milled rice, comprising primarily the kernel endosperm,
is generally much lower in lipids (approximately 0.3–0.5%) and protein (approxi-
mately 7%). These values can vary greatly due to varietal, environmental, or pro-
cessing variability (see Chapter 4). Because of these stark differences in composi-
tion between the bran and endosperm, the DOM can affect the functionality of
milled rice. In addition to having functional effects, the bran remaining on kernels
after milling can affect sensory characteristics. Both of these quality effects are dis-
cussed below.
As milling removes the outer protective layers of the rice caryopsis, the endosperm
of milled rice becomes relatively more prone to moisture sorption from its surrounding
environment. This is particularly true for rice that has just been milled, since rice
exiting a mill is typically at higher temperatures than the surrounding air. A situation
often experienced in the milling industry that can be very costly in terms of broken
rice and mill downtime is the fissuring and subsequent breaking of kernels during
postmilling operations, often referred to as “residual breakage.” An extensive research
effort conducted to address this phenomenon is described at the end of this section.
Kernel Size Uniformity
A rice bulk inherently comprises kernels with varying dimensions (Bautista and
Siebenmorgen, 2000). As a rice kernel develops, it reaches its maximum length and
width before reaching its maximum thickness (Wadsworth et al, 1979). The nar-
rower the range of thickness of kernels in a batch of rice, the more uniform in ma-
turity the rice is considered to be. Since the highest-MC kernels are usually the thin-
ner kernels, rice with a narrower range of thickness may indicate less MC variation
within the rice bulk as well (Wadsworth and Matthews, 1986). A narrow range of
MCs is desirable because this can minimize moisture transfer in the rice kernels of
different MCs as the rice is transported or held before drying. Rice with a wider
range of MCs can result in increased breakage of the larger, lower-MC kernels as
they absorb moisture from the higher-MC kernels (Wadsworth et al, 1982; Calder-
wood, 1984; Siebenmorgen and Jindal, 1986; Jindal and Siebenmorgen, 1994).
Of the primary kernel dimensions, kernel thickness is the one for which variation
in a batch of rice can particularly lead to problems in rice processing (Wadsworth et
al, 1979). While there are other reasons for breakage in rice kernels besides kernel
thickness (Matthews and Spadaro, 1976), it has been shown that the thickest and
thinnest kernels of rough rice contribute to the greatest percentage of broken kernels
in the final product (Sun and Siebenmorgen, 1993). Thinner kernels may break
more easily because they are not as strong as the thicker kernels, while thicker ker-
nels may break more easily because of fissures previously induced within the kernel
by moisture absorption (Matthews et al, 1981a). The percentage of bran, germ, and
rice chips increased as kernel thickness decreased (Matthews and Spadaro, 1976).
Chalky and damaged kernels also increased as the kernel thickness decreased, con-
tributing to low milling yields (Kocher et al, 1990; Wadsworth and Hayes, 1991).
In general, starch content increased with increasing kernel thickness while pro-
tein, lipid, fiber, and ash contents all decreased with increasing kernel thickness
(Matthews et al, 1981b; Wadsworth and Matthews, 1986). Amylose content also
increased with increasing kernel thickness (Wadsworth and Matthews, 1986). Al-
though the starch content was higher in the thicker kernel fractions, Matthews et al
(1981b) found the thinner kernels to have higher setback viscosities. Wadsworth et
al (1979) found thinner kernels to have lower peak viscosities.
Since such a major difference exists between the thicker and thinner kernel frac-
tions (in both physical and chemical properties), it has been suggested that the thin-
ner kernel fractions be separated out at some point in the processing chain for other
uses. This could possibly aid in drying and milling efficiency, as well as increasing
Fig. 9. Surface fat concentration (SFC) of milled-rice kernel thickness fractions at three degree of
milling (DOM) levels for cultivars Alan (A), Newbonnet (B), and Katy (C) milled as an unfractionated
bulk. Each point is the average of two SFC determinations. (Reprinted, with permission, from Chen
and Siebenmorgen, 1997)
the overall quality of the rice (Wadsworth and Hayes, 1991). The thinner kernels
may possibly be advantageous for other purposes since they do have some unique
properties such as higher protein content (Wadsworth and Hayes, 1991).
Milling System Effects on Kernel Quality
Given the inherent variability in kernel sizes that make up a bulk, Chen and
Siebenmorgen (1997) conducted a study using a single-break friction mill (Satake,
model BA-7) to elucidate the effects of milling system operation on the various
Fig. 10. Surface lipid content (SLC) of kernel thickness fractions from Kaybonnet long-grain rice
milled as an unfractionated bulk to three degree of milling (DOM) levels in two commercial systems:
single-break milling system (A) and multibreak milling system (B). (Reprinted, with permission, from
Chen et al, 1998)
kernel thickness fractions. Three unfractionated long-grain varieties were milled to

three overall DOM levels. The milled bulk head rice was subsequently separated
into thickness fractions, and the surface lipid (fat) concentration, whiteness, and
transparency values were determined for each fraction (Figure 9). For each overall
DOM, surface fat concentration decreased with increasing kernel thickness up to a
thickness of 1.67 mm, after which it remained constant. As the overall DOM in-
creased, the difference in surface fat concentration between thin and thick kernels
lessened. This study indicated that if rice is milled only to low DOMs, thin kernels
are not milled to the same DOM as thicker kernels.
Chen et al (1998) conducted a follow-up study of a similar type but with full-
scale commercial systems. In this study, Kaybonnet long-grain rice was milled to
three overall DOMs in both a single-break friction mill (Satake, model BA-15) and
Fig. 11. Protein content of kernel thickness fractions at the indicated degree of milling (DOM) levels
for Kaybonnet long-grain rice milled in two commercial systems: single-break milling system (A) and
multibreak milling system (B). (Reprinted, with permission, from Chen et al, 1998)
a triple-break abrasion and friction system (Satake, models VTA, KB-40, and KB-
40 with water mist). The milled rice was fractionated into five thickness fractions,
and the surface lipid content and protein content of the head rice in each fraction
were measured. Figure 10 indicates findings with both milling systems similar to
those reported by Chen and Siebenmorgen (1997) with the smaller friction system.
Additionally, Figure 11 shows the trends in protein content across thickness frac-
tions. Thinner kernels (thickness <1.7 mm) had greater protein content than thicker
kernels, which concurs with the conclusion of Matthews et al (1981b). Figure 11
also shows that overall DOM did not greatly influence protein content, whereas
protein content significantly varied across thickness fractions. This was explained
by noting that the bran has a higher protein content than the endosperm, but the
mass of bran relative to endosperm is low. By increasing DOM, bran is removed,
but little to no endosperm is removed. Thus, increasing the DOM slightly decreases
the overall protein content of milled rice because of the greater proportion of endo-
sperm to bran in milled versus undermilled rice. However, the protein content of the
thickness fractions decreased significantly with increasing kernel thickness due to
the inherent differences in protein content associated with kernel thickness.
Chen et al (1998) also showed that less rice was broken in the multibreak system
than in the single-break system at any DOM (Fig. 12). Rice breakage progressively
occurred as the rice was milled to greater DOM levels in the single-break system,
whereas most of the breakage that occurred in the multibreak system appeared to be
caused in the early milling stages.
Another quality aspect reported by Chen et al (1998) concerns the distribution of
kernel thicknesses produced by the two milling systems. Figure 13 shows that the
single-break system produced rice containing a lower percentage of thin kernels and
a higher percentage of thick kernels than did the rice produced in the multibreak
system. These results suggest that the gentler milling action of the multibreak sys-
Fig. 12. Relationship between rice breakage and surface lipid content of unfractionated rice milled in
two commercial systems. DOM = degree of milling. (Reprinted, with permission, from Chen et al,
1998)
Fig. 13. Mass distribution of the head rice produced from milling Kaybonnet long-grain rice in two
commercial systems. Levels shown are the average across degree of milling levels. (Reprinted, with
permission, from Chen et al, 1998)
tem (Fig. 12) prevents breakage of thinner kernels that would be broken in a single-
break system.
The above studies were conducted with long-grain varieties. However, Chen et al
(1999) report similar findings when milling a medium-grain variety, Bengal, in a
multibreak milling system.
In summary, these studies indicate that differences exist in DOM and protein
levels of kernels within a bulk of milled rice and that these differences are due in
large part to differences in kernel thickness. This observation is particularly true for
rice milled only to low DOM levels. Differences in DOM among kernel thickness
fractions diminish as the DOM level increases, whereas protein level differences
remain. The relative proportion of thin to thick kernels in a milled rice bulk was
shown to be influenced by the type of commercial milling system used.
Degree of Milling Effects on Functionality
End-use processors have been aware of the influence of DOM on processing per-
formance for some time. The extent of this influence can be attested to by the fact
that most customers of milled rice have DOM index specifications that must be
satisfied; some customers have multiple DOM indices. These indices can include
color, total and/or surface lipid content, bran streak percentage, and free-fatty-acid
content. Some of the indices result from the customer’s need for the rice to have
a specific appearance, while others are required as indicators of some functional
and/or sensory performance.
While processors qualitatively know the influences of DOM on functionality,
only limited data quantify these effects in the scientific literature. One of the rea-
sons for this has been the lack of a uniformly accepted method for reporting DOM
for comparison studies, a topic currently being addressed by the Rice Quality and
Milling Technical Committee of the American Association of Cereal Chemists.
Perdon et al (2001) conducted a study directly quantifying the effects of DOM

on rice pasting behavior. In this study, samples of two medium-grain and two long-
grain cultivars were milled for 15, 30, 45, and 60 sec in a McGill no. 2 laboratory
mill to produce samples with varying DOMs. The resulting head-rice DOM was
measured using both a petroleum ether extraction method to obtain surface and total
lipid content, and a commercial milling meter (Satake, model MM-1B). The peak
(Pv) and final (Fv) viscosities of the head rice were measured using both a Bra-
bender amylograph and an RVA. In all cases, Pv linearly increased as DOM in-
creased (i.e., the amount of remaining bran on kernels decreased), as indicated in
Figure 14. The rate at which Pv increased with DOM was greater in the medium-
grain cultivars than in the long-grain cultivars. The DOM did not have a consistent
or highly significant effect on Fv. Perdon et al concluded that, in general, the
amount of bran left on kernels by the milling process significantly affected the vis-
cosity of milled rice paste. While it is debatable that Pv and Fv are entirely accurate
indicators of rice functionality, this study did show that DOM certainly affects Pv
and Fv and would be expected to have a strong influence on the functional perform-
ance of rice in most processes.
From this work, it is apparent that, when measuring milled rice pasting properties,
some measure of the rice DOM should be made and reported. This is particularly
important given the fact that the DOM of rice milled for set durations in laboratory
mills such as the McGill no. 2 can vary tremendously due to rice MC (Andrews et
al, 1992), rice temperature (Archer and Siebenmorgen, 1995), or rice cultivar or
mill settings (Andrews et al, 1992; Bennett et al, 1993). Since in most instances,
milling conditions are not altered to achieve a given DOM, it is apparent that the
DOM of a sample should be reported when conducting paste viscosity tests.
Milling Effects on Sensory Quality
Since rice DOM is an indication of the amount of lipids remaining on the surface
of the rice kernel and lipids represent the most unstable of the three main compo-
nents of a kernel, DOM plays a role in determining rice sensory quality. Champagne
et al (1997) reported that the effects of DOM on flavor attributes of nonstored rice

2ULRQ
< ;

8
9
2ULRQ
5

%HQJDO < ;
\
LW
V < ;

R
F
%HQJDO
V
L < ;
9

N
D

H
3

'HJUHHRI0LOOLQJ6DWDNH00%'20
Fig. 14. Effect of degree of milling on RVA peak viscosities of medium-grain rice cultivars harvested
in 1995 and 1996. (Reprinted, with permission, from Perdon et al, 2001)
were dependent on cultivar (Koshhikari, Bengal, and M401), MC, and growing lo-
cation. The DOM significantly affected the intensities of the corn, grain-starchy,
and waterlike metallic attributes in the combined rice samples. However, significant
DOM × MC interactions were reported for attributes including corn, grain-starchy,
dairy, floral, and popcorn. Deep-milling (giving high DOM) contributed to small
increases in the intensities of the dairy, floral, and waterlike metallic notes and a
small decrease in the popcorn note in rice dried to 12% MC. Deep-milling had the
opposite effect on the intensities of dairy and floral notes in rice dried to 15% MC.
The effects of milling × MC interaction were exhibited for most flavor attributes,
rendering the analysis of trends for specific cultivars difficult. One limiting aspect
of the study by Champagne et al (1997) is the fact that oxidation notes such as those
reported by Meullenet et al (1999, 2000) were not included in the lexicon of terms
used to describe the flavor of cooked rice, most probably because the samples were
not stored and oxidation notes were not detected by the panel. The evaluation of
attributes such as cardboard (i.e., slightly oxidized) would certainly yield some dif-
ferences between samples of varying DOM for rice that is stored for any extended
period of time.
Postmilling Physical Quality Reduction
Once rice has been milled to meet DOM specifications, additional quality indices
must be met. Two of the most critical of these are the percentage of broken kernels
remaining in head rice and the percentage of kernels with stress cracks or fissures.
These factors are somewhat related as internal stress cracks can, under sufficient
mechanical force and impact, cause a kernel to break apart. Both of these indices
affect millers, who must produce a milled-rice product that meets specifications, as
well as the end user, who can incur waste if stress cracks are present.
Broken kernels can be produced by the mechanical rigors of hulling and bran
removal. However, stress cracks, and the possible subsequent breaking of kernels,
can occur after milling. The conditions leading to this milling quality phenomenon
are the subject of a recent research effort described below.
The term “residual breakage” is used in the rice industry to describe the phenom-
ena in which milled rice fissures and subsequently breaks in postmilling operations.
Because practically all milled-rice customers have upper limits on the percentage of
brokens in head rice, milled-rice lots that were screened to remove brokens to cer-
tain levels sometimes must be rescreened to remove brokens that appeared after
even short storage durations. This has been a very sporadic, yet costly, problem both
in terms of converting head rice to brokens and in occupying valuable screening and
cleaning equipment in the mills, thus limiting mill output.
In a collaborative industry-university study, a large research project was initiated
in the University of Arkansas Rice Processing Program to identify the cause of this
sporadically occurring problem. Several facets of an experimental design and setup
were developed to address this problem. First, a pilot-scale hulling and milling sys-
tem was installed to approximate a commercial milling system (Siebenmorgen et al,
1998). Control equipment for air temperature and relative humidity was also util-
ized in a system for exposing freshly milled rice to closely controlled and replicable
air conditions. This system is shown in Figure 15 and discussed by Siebenmorgen et
al (1998). Finally, an instrument was designed and assembled that would uniformly
apply a consistent force to kernels that had been exposed to certain experimental
conditions. This instrument is shown in Figure 16. Siebenmorgen et al (1997) report

the design and testing of this “roller apparatus.” The design of the instrument was
based on the premise that, if a kernel contained one or more fissures, an imparted
mechanical force would cause it to break. The testing described by Siebenmorgen et
al (1997) showed that the design objective was satisfied in that 91.8% of fissured
Fig. 15. Rice exposure apparatus. (Reprinted, with permission, from Siebenmorgen et al, 1998)
Fig. 16. Roller mechanism used to measure the extent of fissured kernels in milled rice. (Reprinted,
with permission, from Siebenmorgen et al, 1997)
kernels broke when passed through the rollers while 5.2% of unfissured kernels
broke.
Five long-grain rice cultivars were procured from farm bins; the MCs of the lots
varied from 13.2 to 15.7%. A test was conducted by hulling and milling approxi-
mately 45 kg from a given cultivar in the pilot-scale system. Four milled rice sub-
samples of approximately 10 g each were then placed in the four chambers of the
exposure system (Fig. 15) to simulate air conditions surrounding the rice kernels in
a commercial milling operation during postmilling screening and cleaning opera-
tions. After the kernels were exposed to an air stream with a set temperature and
relative humidity for a certain duration, they were removed and subsequently passed
through the roller apparatus (Fig. 16) to determine the mass percentage of kernels
that broke due to fissures having been formed during the exposure period.
Figure 17 shows that milled rice breakage increased dramatically at low (<40%
in most trials) and high (>75%) relative humidity levels. Milled rice with MCs
<14% exposed to air at mid-range relative humidity conditions (40–75%) and
temperatures of 20 and 30°C incurred little more damage than milled rice re-
ceiving no exposure, which showed a breakage level of approximately 10% when
passed through the roller mechanism. Congruent trends were reported by Lloyd and
Siebenmorgen (1999) in a similar study using a medium-grain variety. However,
when the temperature was increased to 40°C, the relative humidity range with
minimal breakage was narrowed to 40–65%. In general, the overall level of break-
age, as well as the rate of breakage, increased with increasing air temperature. This
is due to the fact that the mass diffusivity of rice increases directly with temperature
(Lu et al, 1994).

H
J
D
N
D
H
U
%

H
F
L

5

G
H
O
O
L

0

5HODWLYH+XPLGLW\
&\ SUHV V /HPR QW 1HZER QQHW
.DW\ $ODQ .DW\ GU\
Fig. 17. Milled rice breakage incurred by 20 min of exposure to air at 30°C and a range of relative
humidities for five varieties of rice at the indicated milled rice moisture contents (%, wb). (Reprinted,
with permission, from Siebenmorgen et al, 1998)
These results indicate that, for a given rice MC and air temperature, there is a
safe relative humidity range to which milled rice can be exposed without producing
an increase in fissured or cracked kernels that will likely break in subsequent han-
dling operations. This could be important in milling situations where the postmilling
treatment typically includes sufficient exposure to air; if the air is outside the safe
relative humidity and temperature range, it could produce stress cracks. Other
situations, occurring during transport or in end-use processing plants, also could
produce stress cracks if kernels are exposed to these high or low relative humidity
levels.
As Figure 17 indicates, rice MC played a major role in determining breakage for
a given air condition. At low relative humidity levels (<40%), milled rice at higher
MC levels incurred more damage than that at low MC levels for a given cultivar.
This trend was reversed at high relative humidity levels in that milled-rice breakage
increased with decreases in rice MC for all varieties.
Because of the observed strong influence of MC on milled rice breakage, addi-
tional tests were conducted to isolate the effects of MC on breakage. Figure 18
shows the relationships between milled-rice breakage and milled-rice MC for three
varieties exposed for 40 min to low-relative-humidity (25%) air at 30°C. The data
show a clear trend of increasing breakage with MC at this low relative humidity
level. This figure indicates the sensitivity of milled rice breakage to MC level, e.g.,
an increase of one percentage point in MC corresponded to an increase of approxi-
mately 40–50 percentage points in breakage for MC levels in the range of 12–14%.
Figure 18 also indicates varietal differences in breakage under certain exposure
conditions. Cypress and Lemont varieties experienced more damage that Newbon-
net. Cypress and Lemont both tend to have larger kernels than Newbonnet, which
suggests that breakage may be a function of kernel size, with thicker kernels being
more susceptible to moisture sorption than thinner kernels. Jindal and Siebenmorgen
(1994) also showed that, for a given lot of rough rice, the thicker kernels experi-

&\SUHVV /HPRQW

H
J
D
N
D
H
U
%

H
F
L
5

G
H
O
O
L
0

1HZERQQHW

0RLV WXUH&RQWHQW ZE
Fig. 18. Breakage vs. milled rice moisture content for rice of the indicated varieties exposed to air at
30°C and 25% rh for 40 min. (Reprinted, with permission, from Siebenmorgen et al, 1998)
enced a greater reduction in HRY than thinner kernels when subjected to moisture
adsorption. Varietal differences were also shown by Lan and Kunze (1996), who
quantified the amount of fissured kernels in rough-rice samples exposed to air at
24°C and 100% rh. Their results showed that after this exposure, samples of Le-
mont contained 70–80% fissured kernels while only 20–25% of the Newbonnet ker-
nels were fissured.
This study revealed that stress cracking of kernels, with probable subsequent
breakage, can occur if the difference between the MC of the kernel and the EMC
corresponding to the surrounding air becomes great enough to cause rapid moisture
sorption that results in fissure formation. This difference in moisture potentials can
occur in situations of low relative humidity and high rice MC or in situations of
high relative humidity and low rice MC.
SUMMARY
In many instances, the work reported herein is a presentation of observed trends

in quality indices rather than a full scientific explanation. It is obvious that more fun-
damental research is needed to understand the physicochemical properties respon-
sible for these trends and, just as important, how preharvest and postharvest pro-
cesses affect these properties. Advances are being made in this light. Current work
incorporating glass transition temperature and its influence on drastically changing
kernel properties (Perdon et al, 2000) as a basis for explaining drying behavior and
fissure formation (Cnossen et al, 2000) is one such example. The recent finding of
Teo et al (2000) regarding aging of flour is a further indication of a possible fun-
damental explanation for changes occurring during postharvest operations. This has
prompted research on protein-starch interactions and the effect they have on rheo-
logical and sensory behavior. All of this and related work is necessary so that sci-
entists can not only describe, but also explain, how postharvest operations affect the
many aspects of rice quality.
REFERENCES
Aibara, S., Yamashita, H., Ismall, I. A., Bason, M. L., Gras, P. W., Banks, H. J., and
Sekiyama, F., and Morita, Y. 1984. Post- Esteves, L. A. 1990. A quantitative study of
harvest changes in the respiratory activity of the influence of temperature, water activity
and ethanol accumulation in fresh rice and storage atmosphere on the yellowing of
grains. Agric. Biol. Chem. 48:1293-1301. paddy endosperm. J. Cereal Sci. 12:193-201.
Andrews, S. B., Siebenmorgen, T. J., and Bautista, R. C., and Siebenmorgen, T. J. 2000.
Mauromoustakos, A. 1992. Evaluation of the Rice kernel properties affecting milling
McGill no. 2 rice miller. Cereal Chem. 69: quality at harvest. B. R. Wells Research
35-43. Studies. R. Norman and J. F. Meullenet, Eds.
Archer, T. R., and Siebenmorgen, T. J. 1995. Univ. Ark. Agric. Exp. Stn., Fayetteville, AR.
Milling quality as affected by brown rice Bennett, K. E., Siebenmorgen, T. J., and
temperature. Cereal Chem. 72:304-307. Mauromoustakos, A. 1993. Effects of
Bailey, C. G. 1940. The handling and storage McGill no. 2 miller settings on surface fat
of cereal grains and flaxseed. Plant Physiol. concentration of head rice. Cereal Chem. 70:
15:257-274. 734-739.
Barber, S. 1972. Milled rice and changes dur- Brooker, D. B., Bakker-Arkema, F. W., and
ing aging. Pages 215-263 in: Rice: Chemis- Hall, C. W. 1974. Drying Cereal Grains. AVI
try and Technology, 1st ed. D. F. Houston, Publishing Co., Inc., Westport, CT.
Ed. Am. Assoc. Cereal Chem., St. Paul, Burrell, N. J. 1982. Refrigeration. Pages 407-
MN. 435 in: Storage of Cereal Grains and Their
Products, 3rd ed. C. M Christensen, Ed. Am. Dillahunty, A. L., Siebenmorgen, T. J., Buescher,
Assoc. Cereal Chem., St. Paul, MN. R. W., Smith, D. E., and Mauromoustakos,
Calderwood, D. L. 1984. Milling yield of rough A. 2000. Effect of moisture content and tem-
rice blended at different moisture contents. perature on the respiration rate of rice. Ce-
Trans. ASAE 27:248-249, 254. real Chem. 77:541-543.
Champagne, E. T., Bett, K. L., Vinyard, B. T., Dillahunty, A. L., Siebenmorgen, T. J., and
Webb, B. D., McClung, A. N., Barton, F. Mauromoustakos, A. 2001. Effect of tem-
E., II, Lyon, B. G., Moldenhauer, K. A., perature, exposure duration, and moisture
Linscombe, S., and Kohlwey, D. E. 1997. content on the yellowing of rice. Cereal
Effects of drying conditions, final moisture Chem. 78:559-563.
content, and degree of milling on rice flavor. Fan, J., Marks, B. P., Daniels, M. J., and
Cereal Chem. 74:566-570. Siebenmorgen, T. J. 1999. Effects of post-
Champagne, E. T., Lyon, B. G., Min, B. K., harvest operations on the gelatinization and
Vinyard, B. T., Bett, K. L., Barton, F. E., II, retrogradation properties of long-grain rice.
Webb, B. D., McClung, A. N., Moldenhauer, Trans. ASAE 42:727-732.
K. A., Linscombe, S., McKenzie, K. S., and Fan, J., Siebenmorgen, T. J., and Yang, W.
Kohlwey, D. E. 1998. Effects of postharvest 2000. A study of head rice yield reduction of
processing on texture profile analysis of long- and medium-grain rice varieties in re-
cooked rice. Cereal Chem. 75:181-186. lation to various harvest and drying condi-
Chen, H., and Siebenmorgen, T. J. 1997. Effect tions. Trans. ASAE 43:1709-1714.
of rice thickness on degree of milling and as- Hamaker, B. R., Siebenmorgen, T. J., and
sociated optical measurements. Cereal Chem. Dilday, R. H. 1993. Aging of rice in the
74:821-825. first six months after harvest. Ark. Farm Res.
Chen, H., Siebenmorgen, T. J., and Griffin, K. 42:8-9.
1998. Quality characteristics of long-grain Jindal, V. K,. and Siebenmorgen, T. J. 1994.
rice milled in two commercial systems. Ce- Effects of rice kernel thickness on head rice
real Chem. 75:560-565. yield reduction due to moisture adsorption.
Chen, H., Siebenmorgen, T. J., and Du, L. 1999. Trans. ASAE 37:487-490.
Quality characteristics of medium-grain rice Kocher, M. F., Siebenmorgen, T. J., Norman,
milled in a three-break commercial milling R. J., and Wells, B. R. 1990. Rice kernel
system. Cereal Chem. 76:473-475. moisture content variation at harvest. Trans.
Chrastil, J. 1990. Chemical and physicochemi- ASAE 33:541-548.
cal changes in rice during storage at different Kunze, O. R., and. Calderwood, D. L. 1980.
temperatures. J. Cereal Sci. 11:71-85. Systems for drying of rice. Pages 209-233
Chrastil, J. 1992. Correlations between the in: Drying and Storage of Agricultural
physicochemical and functional properties of Crops. C. W. Hall, Ed. AVI Publishing Co.,
rice. J. Agric. Food Chem. 40:1683-1686. Inc., Westport, CT.
Cnossen, A. G., and Siebenmorgen, T. J. 2000. Lan, Y., and Kunze, O. R. 1996. Fissure resis-
The glass transition temperature concept in tance of rice varieties. Appl. Eng. Agric. 12:
rice drying and tempering: Effect on milling 365-368.
quality. Trans. ASAE 43:1661-1667. Lloyd, B. J., and Siebenmorgen, T. J. 1999.
Cnossen, A. G., Jimenez, M. J., and Environmental conditions causing milled
Siebenmorgen, T. J. 2003. Rice fissuring rice kernel breakage in medium-grain varie-
response to high drying and tempering tem- ties. Cereal Chem. 76:426-427.
peratures. J. Food Eng. 59(1):61-69. Lu, R., Siebenmorgen, T. J., and Archer, T. R.
Daniels, M. J., Marks, B. P., Meullenet, J.-F., 1994. Absorption of water in long-grain
and Siebenmorgen, T. J. 1996. Effects of rice rough rice during soaking. J. Food Process
storage history on the end-use quality of Eng. 17(1994):141-154.
long-grain rice. Pages 152-157 in: B.R. Luh, B. S. 1980. Rice: Production and Utiliza-
Wells Arkansas Rice Research Studies. R. tion. AVI Publishing Co., Inc., Westport, CT.
Norman, Ed. Univ. Ark. Agric. Exp. Stn., Lyon, B. G., Champagne, E. T., Vinyard, B. T.,
Fayetteville, AR. Windham, W. R., Barton, F. E., I I , Webb,
Daniels, M. J., Marks, B. P., Siebenmorgen, T. B. D., McClung, A. N., Moldenhauer, K. A.,
J., McKnew, R. W., and Meullenet, J.-F. Linscombe, S., McKenzie, K. S., and
1998. Effects of long-grain rough rice stor- Kohlwey, D. E. 1999. Effects of degree of
age history on end-use quality. J. Food Sci. milling, drying condition, and final moisture
63:832-835. content on sensory texture of cooked rice.
Dhaliwal, Y. S., Sekhon, K. S., and Nagi, H. P. Cereal Chem. 76:56-62.
1991. Enzymatic activities and rheological pro- Lyon, B. G., Champagne, E. T., Vinyard, B. T.,
perties of stored rice. Cereal Chem. 68:18-21. and Windham, W. R. 2000. Sensory and in-
strumental relationships of texture of cooked Phillips, S., Widjaja, S., Wallbridge, A., and
rice from selected cultivars and postharvest Cooke, R. 1988. Rice yellowing during post-
handling practices. Cereal Chem. 77:64-69. harvest drying by aeration and during stor-
Matthews, J., and Spadaro, J. J. 1976. Breakage age. J. Stored Prod. Res. 24:173-181.
of long-grain rice in relation to kernel thick- Sahay, M. N., and Gangopadhyay, S. 1985.
ness. Cereal Chem. 53:13-19. Effect of wet harvesting on biodeterioration
Matthews, J., Wadsworth, J. I., and Spadaro, J. J. of rice. Cereal Chem. 62:80-83.
1981a. Rough-rice breakage in relation to Sajwan, K. S., Mittra, B. N., and Pande, H. K.
kernel thickness for hand- and combine- 1989. Effect of storage environment on mill-
harvested rice. Trans. ASAE 24:255-258. ing out-turn of modern high yielding varie-
Matthews, J., Wadsworth, J. I., and Spadaro, J. J. ties. Int. J. Trop. Agric. 7:202-207.
1981b. Chemical composition of Starbonnet Schroder, H. W. 1963. The relationship be-
variety rice fractionated by rough-rice kernel tween storage molds and damage in high-
thickness. Cereal Chem. 58:331-334. moisture rice in aerated storage. Phytopathol-
Meullenet, J.-F., Marks, B. P., Griffin, V. K., ogy 53:804-808.
and Daniels, M. J. 1999. Effects of rough Siebenmorgen, T. J. 1994. Role of moisture
rice drying and storage conditions on sen- content in affecting head rice yield. Pages
sory profiles of cooked rice. Cereal Chem. 341-380 in: Rice Science and Technology.
76:483-486. W. E. Marshall and J. I. Wadsworth, Eds.
Meullenet, J.-F., Marks, B. P., Hankins, J. A., Marcel Dekker, Inc., New York.
Griffin, V. K., and Daniels, M. J. 2000. Sen- Siebenmorgen, T. J., and Jindal, V. K. 1986.
sory quality of cooked long-grain rice as af- Effects of moisture adsorption on the head
fected by rough rice moisture content, stor- rice yields of long-grain rice. Trans. ASAE
age temperature, and storage duration. Ce- 29:1767-1771.
real Chem. 77:259-263. Siebenmorgen, T. J., Counce, P. A., Lu, R., and
Misra, A. K., and Vir, D. 1991. Assessment of Kocher, M. F. 1992. Correlation of head rice
losses due to discoloration of paddy grains. yield with individual kernel moisture content
Indian J. Mycol. Plant Pathol. 21:277-278. distribution at harvest. Trans. ASAE 35:
Mohsenin, N. 1980. Thermal Properties of 1879-1884.
Foods and Agricultural Materials. Gordon Siebenmorgen, T. J., Andrews, S. B., and
and Breach, Science Publishers, Inc., New Counce, P. A. 1994. Relationship of the
York. height rice is cut to harvesting test parame-
Pan, Z., Wei, L., and Thompson, J. F. 2002. ters. Trans. ASAE 37:67-69.
Effect of harvest and drying conditions on Siebenmorgen, T. J., Nehus, Z. T., and Archer,
rice milling quality and moisture distribution T. R. 1997. A mechanical method to deter-
with simulated column drying. ASAE Paper mine fissures in milled rice kernels. Appl.
026140. American Society of Agricultural Eng. Agric. 13:637-639.
Engineers, St. Joseph, MI. Siebenmorgen, T. J., Nehus, Z. T., and Archer, T.
Pearce, M. D., Marks, B. P., and Meullenet, J.-F. R. 1998. Milled rice breakage due to environ-
2001. Effects of postharvest parameters on mental conditions. Cereal Chem. 75:149-152.
functional changes during rough rice storage. Siebenmorgen, T. J., Yang, W., Bautista, R. C.,
Cereal Chem. 78:354-357. and Cnossen, A. 2003. Rough rice drying and
Perdon, A. A., Marks, B. P., Siebenmorgen, milling quality. Pages 567-594 in: Rice; Ori-
T. J., and Reid, N. B. 1997. Effects of rough gin, History, Technology, and Production.
rice storage conditions on the amylograph C. Smith and R. Dilday, Eds. John Wiley &
and cooking properties of medium grain (cul- Sons, Inc., Hoboken, NJ.
tivar Bengal) rice. Cereal Chem. 74:864-867. Singaravadivel, K., and Raj, S. A. 1983.
Perdon, A. A., Siebenmorgen, T. J., and Changes in parboiled rough rice caused by
Mauromoustakos, A. 2000. Glassy state improper drying and microbial infection. J.
transition and rice drying: Development of a Food Biochem. 7:15-21.
brown rice state diagram. Cereal Chem. 77: Sun, H., and Siebenmorgen, T. J. 1993. Milling
708-713. characteristics of various rough rice kernel
Perdon, A. A., Siebenmorgen, T. J., thickness fractions. Cereal Chem. 70:727-733.
Mauromoustakos, A., Griffin, V. K., and Tamaki, M. T., Tahiro, M., Ishidawa, M., and
Johnson, E. R. 2001. Degree of milling ef- Ebata, M. 1993. Physico-ecological studies
fects on rice pasting properties. Cereal Chem. on quality formation of rice kernel: IV. Ef-
78:205-209. fect of storage on eating quality of rice. Jpn.
Perez, C. M., and Juliano, B. O. 1981. Texture J. Crop. Sci. 62:540-546.
changes and the storage of rice. J. Texture Teo, C. K., Karim, A. A., Cheah, P. B., Norziak,
Stud. 12:321-333. M. H., and Seow, C. C. 2000. On the roles of
protein and starch in the aging of non-waxy III. Milling performance and quality charac-
rice flour. Food Chem. 77:229-236. teristics. Trop. Sci. 31:27-44.
Tsugita, K T., Ohta, T., and Kato, H. 1983. Wadsworth, J. I., Matthews, J., and Spadaro, J. J.
Cooking flavor and texture of rice stored un- 1979. Physical and physicochemical prop-
der different conditions. Agric. Biol. Chem. erties of Starbonnet variety rice fractionated
47:543-549. by rough rice kernel thickness. Cereal Chem.
Villareal, R. M., Resurreccion, A. P., Suzuki, 56:499-505.
L. B., and Juliano, B. O. 1976. Changes in Wadsworth, J. I., Matthews, J., and Spadaro, J. J.
physicochemical properties of rice during 1982. Moisture content variation in freshly
storage. Starch 28:88-94. harvested rice associated with kernel thick-
Wadsworth, J. I., and Matthews, J. 1986. ness. Trans. ASAE 25:1127-1130.
Variation in rice associated with kernel Wang, C. Y. and Luh, B. S. 1991. Harvest, dry-
thickness. I. Chemical composition. Trop. ing, and storage of rough rice. Pages 311-
Sci. 26:195-212. 346 in: Rice, Vol. 1—Production, 2nd ed. B.
Wadsworth, J. I., and Hayes, R. E. 1991. Varia- S. Luh, Ed. Van Nostrand Reinhold, New
tion in rice associated with kernel thickness York.
CHAPTER 13
PARBOILING OF RICE
Kshirod R. Bhattacharya
Rice Research and Development Centre
Mysore, India
Paddy, or rough rice, is unique among cereals in that it is, by and large, milled as
whole grain, not flour. It is also unique in that milled rice can be of two forms. In
most parts of the world it is used as white, or raw (meaning nonparboiled), rice. In
other parts, it is used as parboiled rice, obtained by treating paddy rice with water
and heat before it is dried and milled.
Parboiled rice is the major staple throughout South Asia (the Indian subconti-
nent), where over 90% of the world’s parboiled rice is produced and consumed. All
or most of the rice in Bangladesh (Roy, 2001), Sri Lanka (Malkanthi et al, 2001),
and many rice-growing states of present-day India is parboiled; 55–60% of India’s
rice and a substantial portion in Nepal is parboiled (Fig. 1). It also is eaten widely
elsewhere, as is discussed below. It is estimated that as much as a fifth of the
world’s rice is parboiled (Kik and Williams, 1945; Tata, 1962; Gariboldi, 1974).
Use of parboiled rice seems to have been increasing in recent times.
HISTORY
Origin of Parboiling
Parboiled rice is widely thought to have originated in ancient India (Ghose et al,
1960; Tata, 1962), but exactly when and how has been shrouded in mystery. The
following considerations may provide some answers. Achaya (1998a) recently
noted, “Parboiled rice, pulungalarisi in Tamil, first occurs in the Sirupan-arrupudai
of slightly later Sangam literature”. The latter period, according to Achaya (1998b),
corresponds to the fourth to fifth century A.D. Srinivasa Aiyangar (1985), whom
Achaya quotes, writes, “The chief cereal used by the Tamils was the
paddy…Ordinarily paddy was boiled before it was husked; this was called
pulungalarisi,” which leaves no doubt that ancient Tamils ate parboiled rice. Thus,
there is a clear record that parboiled rice has been an established product for
somewhat less than two thousand years (at least).
According to one theory, Tamils are among the original inhabitants of India;
therefore, the above reference might suggest that parboiling originated in south
India, perhaps centuries before Christ. However, another possibility is suggested by
a historical parallel. Bulgur is essentially nothing but parboiled wheat. According to
329
the literature (Haley, 1960; Bayram, 2000; Sunnyland Mills, 2001), bulgur
originated in the Mediterranean region, Middle East, and North Africa; Biblical
references cite its use by ancient Babylonians, Hittites, and Hebrew populations
some four thousand years ago; Arab, Israeli, Egyptian, and Roman civilizations
record eating dried, whole cooked wheat as early as 1000 B.C.; and it still is used
widely in Middle Eastern countries, particularly for making pilaf. Interestingly, it
was referred to as Arisah in the Old Testament, a word strikingly similar to Arisi,
which stands for “rice” in Tamil (a very old Indian language). If bulgur happens to
be the older of the two, then one might wonder whether the Aryans, who are
supposed to have migrated from Central Asia to India some 1,500–2,000 years
before Christ, carried the knowledge of bulgur with them and applied it to rice.
However, the theory of Aryan migration is now contested in India.
Fig. 1. Raw and parboiled rice areas in South Asia (impressionistic drawing). South Asia is the home
of parboiled rice, because over 90% of the world’s parboiled rice is made and consumed here.
Parboiling / 331
The other point with regard to the history of parboiled rice is exactly how and
why this process was started. No doubt, the fact that it makes the dehusking of
paddy easy and reduces the breakage of rice during milling would have reinforced
its continued practice once the process was discovered. But one more factor may
have favored its prevalence in South Asia.
Over 90% of the world’s rice is grown and consumed in the southern and eastern
parts of Asia. There is an interesting, gradual shift in the chemical composition and
sensory character of rice in the region (Fig. 2). People of northeast and east Asia
grow and prefer low-amylose rice, which cooks soft and sticky. Those in southeast
Asia prefer intermediate-amylose rice, which cooks neither hard nor soft, remains
integral, yet does not harden after cooking. People of South Asia, at the other
extreme, grow and prefer high-amylose rice that cooks hard and fluffy. Parboiling,
as we shall presently see, renders cooked rice yet more hard and fluffy. Because South
Asia is the home of parboiled rice, one is tempted to speculate that some South Asians
probably were not satisfied with the available high-amylose rice but showed a
preference for rice that was hardened further by parboiling. In that sense, parboiling
may be considered an artificial means to increase the amylose content in rice.
Few outside South Asia heard of parboiled rice until the turn of the nineteenth
century. Two long-drawn sagas of scientific effort—one mediconutritional and one
technological—introduced it and spread it worldwide.
Fig. 2. Home of the world’s rice (impressionistic drawing). More than 90% of the world’s rice is grown
and consumed here. The arrow shows the direction of gradation in the sensory quality of cooked rice
from very soft and sticky to very firm and fluffy.
Discovery of the Nutritional Benefit of Parboiling
The mediconutritional story of parboiled rice is intimately connected with the

stories of machine-milled rice, the new scourge of beriberi, the science of nutrition,
the birth of the concept of vitamins, and the discovery of thiamin. A full account of
this saga is available in Aykroyd et al (1940); a brief account follows.
Braddon (1907) noted that the Chinese immigrants in Malaya, who consumed
milled raw rice, suffered from beriberi, but the Tamil immigrants, who preferred
parboiled rice, did not. He then showed, by an epidemiologic survey of Asia, that
the consumption of raw milled rice, but not of parboiled rice, was associated with
beriberi. McCarrison and Norris (1924) similarly noted that, in India, where rice
was the main staple, beriberi was largely confined to a small area in the eastern
coast where raw rice (apparently machine milled) was the staple (Fig. 1). Direct
evidence of the anti-beriberi property of parboiled rice also was found. Fletcher
(1907) observed that beriberi in a mental asylum in Kuala Lumpur was largely
eradicated when parboiled rice was substituted for highly milled raw rice. Fraser
and Stanton (1909) divided a batch of road laborers into two groups, one group
being given raw milled rice and the other parboiled rice. The former reported
beriberi within six months. When the rice was interchanged, beriberi in the first
group ceased but appeared in the second. Taylor et al (1928) noted that, when the
raw-rice diet of a Burma lighthouse crew was changed to parboiled rice in 1911, the
incidence of beriberi dropped dramatically. Vedder (1913) critically reviewed this
and other evidence and proposed a theory of beriberi. Finally, Aykroyd (1932),
using a bioassay with birds, and Acton et al (1933), using a chemical method,
experimentally demonstrated that milled parboiled rice contains much more thiamin
than milled raw rice.
With this dramatic discovery, parboiled rice became a worldwide favorite of public
bodies, nutritionists, and researchers alike. International bodies strongly recommended
its wider popularization as a preventive against beriberi. Researchers then observed a
second virtue of parboiled rice, namely, vastly reduced grain breakage during milling,
and a third property, a marked change in cooking quality. The decades of the 1930s
and 1940s saw intense research activity directed to a study of these three properties
(see Bhattacharya, 1985, for the numerous early references).
Modernization of the Parboiling Industry
The call for popularization of parboiled rice drew attention to the need to
modernize the primitive parboiling industry. No doubt the improved vitamin,
milling, and cooking properties of parboiled rice attracted the attention of
entrepreneurs in Europe, who saw in it a useful product, and a series of patents for
production of parboiled rice was issued in the 1930s and 1940s. Modern plants for
producing parboiled rice were started in Europe and the United States after World
War II (Kik and Williams, 1945; Gariboldi, 1974, 1984) and have been pro-
gressively improved. However, these processes were too sophisticated, and the new
industry too microscopic, to be of any relevance to the vast traditional parboiling
industry in South Asia.
Although the age-old home-scale parboiling processes gave a wholesome
product, the commercial process adopted at the turn of the nineteenth century did
not. The process involved soaking paddy in ambient or warm water for prolonged
Parboiling / 333
periods, which led to pronounced microbial fermentation and a characteristic

offensive odor (Ghose et al, 1960; Tata, 1962). Several decades were devoted to
sporadic research on the origin of the odor and its amelioration. The process
apparently was adopted very widely in Burma for export to India and Ceylon.
Therefore, the problem was especially noticed there and studied by Charlton (1923),
the title of whose paper (“The Prevention of Nuisances Caused by the Parboiling of
Paddy”) gives a vivid description of the problem. His pioneering findings, all of
which still stand, were that 1) the odor evidently was caused by microbial
fermentation; 2) the fermentation apparently was caused by anaerobic bacteria,
because it could be prevented by blowing air through the mix (which, however,
caused germination) (cf. the use of chromate [Pillaiyar, 1977] and the gradual loss
of dissolved oxygen in the mix [Vasan and Kumaravel, 1982]); 3) the unwanted
odor could be prevented by soaking either in bleaching powder (cf. Desikachar et al,
1955), in acid pH (cf. Paul, 1954) or, better still, in sulfurous acid (acid calcium
sulfite sludge + 0.18% sulfuric acid), which had the added advantage of yielding a
light-colored rice (cf. Mecham et al, 1961; Jayanarayanan, 1964). However, all of
these methods were impractical. The fourth and most practical prevention method
was to employ hot soaking (above 55°C), with recirculation of water within the tank
to prevent temperature stratification, in which case the smell was completely
eliminated. Unfortunately, nothing followed this remarkable work, probably partly
because it was published as a bulletin and was virtually lost, so that researchers had
to rediscover the same findings more than two, or even five, decades later.
The thread was taken up intermittently from the late 1930s by several workers,
mainly in India (Sanjiva Rao, 1948; Sanjiva Rao and Guha, 1952; Paul, 1954; Ghose
et al, 1960) and especially at the Central Food Technological Research Institute
(CFTRI) in Mysore (Subrahmanyan et al, 1955; Desikachar et al, 1955, 1957) and at
the Jadavpur University at Kolkata (Mazumder et al, 1960a,b) starting in the 1950s.
After traversing the same route of fermentation, bleaching powder, and acetic acid,
researchers concluded that the best method was to adopt soaking in hot water at
~70°C, which not only eliminated the unwanted odor but had the additional advantage
of reducing the soaking time from a few days to a few hours. An improved hot-
soaking system finally was released in the late 1950s (CFTRI, 1969).
Thus, the long saga of efforts to modernize one of the most extensive traditional
food industries in the world came near to its successful end.
Let it be noted in passing that this largely Indian effort, spread initially over a
period of some 40 years and continued ever since, made a signal contribution to the
entire parboiling industry in the world, which today handles perhaps 125 million
tonnes (t) or more of paddy per year. The few companies operating in Europe and
the United States (in the 1940s and 1950s) no doubt had their own source of
protected information, but they represented only a miniscule fraction of this vast
industry. The overwhelming bulk of the industry in the world, including the
European and American industries post-1950s, benefited profoundly from this
knowledge bank, including the simple yet effective CFTRI process, which was
available freely without any patent restriction. In this whole episode, the pioneering
contribution of B. Sanjiva Rao and V. Subrahmanyan at Bangalore in the 1930s and
1940s, of B. C. Guha and A. N. Bose in Kolkata in the 1940s and 1950s, of V.
Subrahmanyan and H. S. R. Desikachar at the CFTRI starting in the 1950s, and of
V. Subrahmanyan in the Paddy Processing Research Centre (PPRC), in Thiruvarur
and Thanjavur, starting in the 1970s cannot be overemphasized.
Present Status of the Parboiling Industry in the World1
A rough picture of the production and trade of parboiled rice in the world, at the
turn of the new millennium, is given in Table 1. As already mentioned, the lion’s
share of the world’s parboiled rice is produced in South Asia, almost entirely for
local consumption (Fig. 1). Perhaps half to three-fourths of this is still being
produced by primitive, large-scale (“single-boiling” or “double-boiling”), home or
small-scale systems. However, the improved hot-soaking process of CFTRI is
steadily replacing the traditional systems, very much so in India but in a limited way
in Sri Lanka, Bangladesh, and Nepal as well. These processes (and others men-
tioned below) are described later, in the section on production systems.
The Punjab region of undivided India is the home of the famous basmati rice.
Small amounts of basmati paddy traditionally were parboiled here by a sand-heating
system, called the sella process (Narayanswami, 1956; Tata, 1962). One surmise (S.
Z. Ali, CFTRI, Mysore; personal communication) is that this sella process
originally was invented or practiced only for basmati rice, because rice other than
basmati hardly was grown in this region until the introduction of the high-yielding
1
The discussion below is based largely on information generously furnished by Antonio Bianchi (F. &
P. SpA, Robbio, Italy), Salvador Barber (Espanola De I+D, Valencia, Spain), Steven D. Danforth
(Agri Process Innovations, Stuttgart, AR, USA), Vichai Sriprasert (Riceland International Ltd., Bang-
kok, Thailand), John Shifflet (Riceland Foods Inc., Stuttgart, AR, USA), Vipul and Shilen Thakrar
(Tilda Ltd., Rainham, Essex, UK), D. del Castillo (Instituto de Investigaciones del Arroz de Cuba,
Havana, Cuba), Devinder K. Mehan (previously Siamati Ltd., Bangkok, Thailand, currently at Picric
Ltd., Sonepat, Haryana, India), Toshinori Kimura (Institute of Agricultural and Forest Engineering,
University of Tsukuba, Tsukuba, Japan), and S. Chakraborty (Induss Food Products and Equipments
Ltd., Kolkata, India). Any error in understanding, interpretation, omission, or commission is, of
course, the responsibility of the author.
TABLE 1
Approximate Annual Production and Trade (million tonnes, milled equivalent)
in Parboiled Rice in the Worlda
Area/Country Production Export Import
Asia
India 50 1 –
Bangladesh 23 – 0.5
Sri Lanka 2 – 0.2
Thailand 1.8 1.8 –
West Asia – – 1
North America
United States 0.8 0.5 +b
Central and South America and Caribbeans 2.4 + +
Brazil 2 – –
Africa
West Africa
Southern Africa
+
–
–
–
{1.5
Europe 0.6 + 0.2
Italy 0.3 + –
Spain 0.15 + –
Total 81 3.5 3.5
a Estimated from various sources, impressions, and guesses as of 2000.
b The + sign indicates a definite but minor quantity (<100,000 t).
Parboiling / 335
dwarf varieties (during the “green revolution”) in the early 1970s. Little was known
about this process outside the region for the same reason. Yet the name sella stayed.
Today, the process has disappeared, but parboiled rice is called by the generic name
sella in the northern Indian states.
After the partition of India (in 1947), a very similar process started up in the Sind
province of Pakistan as an extensive industry for supplying large quantities of
parboiled rice to then East Pakistan (now Bangladesh). Not having a tradition of
parboiled rice, especially the use of a boiler, and yet finding a market, producers
made use of the same sand-heating process, called in Sind the chatti process,
because the clay or iron pots in which paddy was soaked were called chatties. This
short-lived, flourishing industry seems to have died after the separation of
Bangladesh in 1971. Today, hardly any parboiling industry exists in Pakistan,
although much effort is being made now to establish a modern parboiling industry
to produce rice for export.
Export of basmati rice from India and Pakistan started as a trickle in the 1970s,
mainly to West Asia, and increased gradually (after the “world oil shock”). The
demand in Saudi Arabia was, to a large measure, for parboiled basmati, and the
demand began to spurt; therefore, a large number of new parboiling plants sprang
up in the northern region of India. All these units originally started with the
traditional double-boiling system but, upon complaints of bad odor, now have
shifted to a new, unique system, which may be called the “modernized double-
boiling” system. Approximately 10 large units (parboiling perhaps 200–500 t of
paddy per day in season) and numerous medium and small units exist at present,
confined mainly to the Haryana state of India (a part of original Punjab). Today,
more than a half-million tonnes of basmati paddy is parboiled by this semimodern
technique in India.
Thailand has been a top exporter of rice for a long time. With the demand for
parboiled rice in the international market, Thai exporters started (approximately in
the 1930s) parboiling small quantities by traditional Indian systems (Iamsuri, 1977),
but they began modernizing their plants by the 1960s and 1970s. Today, Thailand is
the largest exporter, shipping close to 2 million tonnes of milled parboiled rice
annually, some—of the very highest quality—to West Asia, Europe, and the United
States and the rest as bulk commodity to various countries in Africa and Asia. As of
now, there are ~70 production units with a combined monthly capacity of ~400,000 t.
These units operate at various levels of technology from rather primitive to
improved to semimodern, with perhaps one or two units having the latest modern
technology, although the drying, by and large, is done entirely mechanically. The
first modern, continuous parboiling plant of U.S. design was introduced in the early
1980s at Riceland International, Bangkok. It is noteworthy that the entire production
of Thailand is for export.
Brazil is probably the largest producer of parboiled rice outside South Asia,
processing ~3 million tonnes of paddy, 25% of the country’s annual production, in
2000. Yet, barely 4% of the crop used to be parboiled 25 years ago. More
surprising, almost the entire amount is for internal consumption. Parboiled rice in
Rio de Janeiro is known as “rice of the restaurants,” which gives an indication of
the major type of use. It also is used for school feeding programs. There is fairly
extensive home consumption as well, especially in the south. Over 100 plants
produce parboiled rice, mainly in two states: Santa Catarina (>50% of production,
with more than 50 producing units) and Rio Grande de Sul (~40% of production and
20 producing units). The majority are small plants. Technology generally is

semimodern at best. Soaking used to be in ambient water but now is in hot water.
Systems using steam for cooking are either batch or the original, manually operated
Malek system. Another is the so-called “Stove process,” a kind of dry-heat
parboiling, which yields a somewhat heterogeneous product. The stove process is
more common in Santa Catarina, whereas the steam types are common in Rio
Grande de Sul. The sixteen largest units are in a group called Association Brasilera
de Arroz, with capacity ranging from 1.5 to 25 t/hr each. There are also a few
American-type continuous plants; Uncle Ben’s, Inc. has a modern processing plant
in Brazil. Most of the rice processed is long grain.
The next major producer is the United States, processing ~1.2 million tonnes of
paddy. Surprising as it may seem, all of this paddy is processed in just six plants:
three in Stuttgart (Arkansas), one in Houston (Texas), and two in Greenville
(Mississippi). Obviously, these are giant plants, with capacity ranging from 20 to
more than 30 t/hr. Technology is almost entirely the latest “American system,” with
Uncle Ben’s in Greenville apparently doing some brown rice parboiling.
Approximately half the production is exported (~0.5 million tonnes, milled).
Italy is another substantial producer of parboiled rice, handling approximately a
half-million tonnes of paddy, which accounts for ~30% of the country’s production,
with japonica rice accounting for ~55%. Eight companies operate in Italy, all using
modern technologies supplied by Gariboldi, Schule, Buhler, or the United States.
The average processing capacity seems to be 5–10 t/hr. The major portion is
shipped to other countries within the European Union (EU).
Spain parboils ~200,000 t of paddy per year. Only one company is located in the
south (Arrocerias Herba S.A.); it has two plants, processing ~4 and 20 t/hr,
respectively, both using Gariboldi technology. The latter plant is probably the
largest in Europe. The domestic market is small, with most of the production
exported within the EU, especially to central Europe. The rice is generally long
grain but not necessarily indica. Parboiled rice is used mainly for making quick-
cooking, canned or frozen rice products, including ready-to-eat (RTE) dishes. Two
companies are engaged in these processes. One more plant using Gariboldi
technology is now being set up in southwest Spain.
France had one Gariboldi plant. But the company has now changed over to
produce quick-cooking (5 min) parboiled rice, both indica and japonica, with a
capacity of ~80,000 t per year.
Europe consumes a total of ~1.7 million tonnes of rice, of which ~0.6 million
tonnes is in the parboiled form. The major consumers of the latter are in France,
Germany, the United Kingdom, and Italy, in that order.
Small quantities of rice are parboiled in many South and Central American
countries and the Caribbean, apart from Brazil. The practice started as early as the
1950s. Surinam has one old Indian-type plant and one new Gariboldi plant, both of
which produce ~2.5 t/hr. Uruguay has one modern unit using U.S. technology, with
a capacity of 15 t/hr, that is owned by five major rice mills in the country. Guyana
has three Indian-type units, all of 2.5-t/hr capacity. The major part of the production
in these three countries is for export. Production in other countries of the region
(Argentina, Cuba, Honduras, Venezuela, Colombia, and Belize) is mostly for
internal use. Argentina has two U.S. continuous-type plants of ~6-t/hr capacity
each. There are two Indian-type plants of 2.5-t/hr capacity each in Honduras, while
Venezuela has one Buhler-type 5-t/hr plant. In Cuba, parboiled rice is used mainly
Parboiling / 337
in catering centers—schools, hospitals, and universities. Total production is less

than 20,000 t per year. Cuba had no plants until 1985, but five exist today, with
rated capacities of 1–5 t/hr, although only one or two are working. Four are of
Gariboldi technology and one of Schule (initial drying in a fluidized bed dryer).
Large amounts of parboiled rice are consumed in West Africa and southern
Africa, both by native people and people of Indian origin. Small amounts are
parboiled in West Africa by relatively primitive or semimodern technologies
(Craufurd, 1962; Sluyters, 1963). The bulk is imported. Nigeria alone imports no
less than 0.8 million tonnes. Both South Africa and Zimbabwe consume a large
amount of parboiled rice, all obtained through import (no rice is grown in southern
Africa).
West Asia, especially Saudi Arabia, is another region where a large amount of
parboiled rice is consumed, all of it imported. The quantity involved is approx-
imately three-fourths of a million tonnes. Much of it is of the highest quality,
imported from the United States and Thailand, plus basmati from India.
Vietnam seems to be planning a parboiling industry, apparently for export.
Schule (2001) supplied parboiling plants to Greece and Egypt. The latter, at Desouk
near Alexandria, is a continuous system and has been operating successfully for
many years. Interestingly, a large amount of work on rice parboiling has been going
on in Japan for several decades, especially in the Institute of Agricultural and Forest
Engineering, University of Tsukuba (Kimura, 1983, 1989, 2001; Kimura et al, 1976,
1993, 1995; Islam et al, 2001, 2003; Malkanthi et al, 2001; Roy 2001; Shimizu et al,
2001), no doubt with an eye on the future.
Gariboldi (now called Tecnorice F.G., Milan, Italy); F. H. Schule Mühlenbau
GmbH (Reinbek, Germany); SCMS Engineering, Philip Rahm, Agri Process
Innovations (all in the United States); Buhler (Uzwil, Switzerland); Satake
Corporation (Hiroshima, Japan); and Induss Foods Products and Equipments Ltd.
(Kolkata), along with many other smaller units in India, are the major suppliers of
technology.
Approximately 25 million tonnes of milled rice (~6% of world production,
compared with 18% in the case of wheat) is internationally traded. Of this, par-
boiled rice constitutes ~3.5 million tonnes (~15%). The main exporters are Thai-
land, India, and the United States; West Africa, southern Africa, Saudi Arabia, and
Bangladesh are the major importers (Table 1).
The above account shows that the consumption of parboiled rice is steadily
widening in the world. No doubt, the production and consumption in South Asia is
so large and so stable that this widening of consumption in other parts of the world
is not reflected in a noticeable increase in the total production.
People of Asia, the home of rice, have strong traditional preferences for either
raw or parboiled rice. People of other regions, on the other hand, do not have any
traditional preference and are willing to adopt whatever type is convenient or
available. It is among these nontraditional consumers that the use of parboiled rice
definitely is increasing, for a variety of reasons.
One reason is
the increased interest of consumers to diversify meals, to taste new food products. The
availability of money, the extended practice of traveling to exotic countries, the
spreading of exotic restaurants in all large cities, the huge amount of books published
and TV programs on the art of cooking with special reference to exotic dishes, etc.
have contributed to this consumer interest in the so called “ethnofood”. (Salvador

Barber, personal communication)
This increased interest in rice dishes brings us to the most important reason: that
parboiled rice is more convenient to cook. In fact, some modern companies market
their parboiled rice as “easy-cook” rice, which is true in the sense that it is safer to
cook (i.e., easier to prevent overcooking), although not in the sense of being faster
to cook. Thus,
many housewives in the West (USA and EU) still find rice difficult to cook (“it
always sticks!”). Parboiled rice offers “guaranteed performance”. (Shilen Thakrar,
personal communication)
This property is important not only to household consumers but also to large
users:
The Caterer (hospitals, schools, industry meals suppliers) needs rice that does not
stick, no matter if side dish or main dish. I know parboiled rice…that will “hold” for
more than 2 hours in bain-marie after cooking, without sticking, and the cooking is
done up to 400 kg per batch! (Antonio Bianchi, personal communication)
Further, out-of-home eating in the West has increased considerably in the last 10
years…. Parboiled rice lends itself to catering/bulk processing requirements. With the
increase in catering, a [trend of chefs having less skill] is occurring and this again
encourages greater use of parboiled rice (guaranteed performance rice). (Shilen
Thakrar, personal communication)
Third, there is an increasing tendency in the world market to further process rice
into quick-cooking (10-min cook), instant (2-min cook), frozen cooked rice, and
RTE rice. Parboiled rice lends itself better to these products. Also,
more and more ready meals (both chilled and frozen) are eaten (in home). Parboiled
rice is preferred for this as ready meals are bulk processed. (Shilen Thakrar, personal
communication)
THE PROCESS: PRODUCTION OF PARBOILED RICE
Processing Conditions
Parboiled rice is “par”-tially “boiled” (i.e., partially cooked) rice. In other words,
parboiling means precooking of rice within the husk without disturbing its size and
shape. To avoid overimbibition and deformation of the grain, water and heat must
be kept separate. Paddy first is hydrated, then heated to cook the rice, and finally
dried.
SOAKING
Empirical Observations. The first step in parboiling is to hydrate the paddy
sufficiently by soaking it in water to enable it to be gelatinized on subsequent
heating. Soaking was first studied by Mecham et al (1961) and Bandopadhyay and
Ghose (1965), who found that the hydration rate increases with increasing
temperature. While confirming this fact, Bhattacharya and Subba Rao (1966a)
Parboiling / 339
showed two distinct patterns of hydration, one below the gelatinization temperature
(GT) and the other above (Fig. 3). At low temperatures, the paddy not only absorbs
water slowly but also eventually comes to an equilibrium at ~30% moisture (wet
basis [wb]). At high temperatures, in contrast, the rate of hydration, after an initial
lag period, increases exponentially, because the starch begins to gelatinize, which
causes the hydration to continue unabated. Then, when the grain moisture exceeds
~30–32% (wb), the husk splits open, for the “container” (husk) is no longer able to
contain the expanded “content” (endosperm), resulting in a further spurt to
hydration (see the 75, 80, and 90°C curves in Fig. 3). The result is overimbibition,
leaching, and deformation of the grain. This is a crucial point that comes up
repeatedly in the discussion of soaking.
The curve of volume expansion of paddy during soaking (Fig. 4, left) shows a
distinct break at ~30–35% moisture (wb; ~50% db). (The break also is found in the
data of Bakshi and Singh [1980], even though not noticed by them.) This break can
be explained on the basis of a difference in the change in grain volume (or density
and bulk density) between brown rice and paddy with increasing moisture. The
density of milled, or brown, rice decreases with increasing moisture content as
expected; but that of paddy rice paradoxically increases (Bhattacharya et al, 1972)
(Fig. 4, right). This is because of the void space between the endosperm and the
husk, which enables the endosperm (“content”) to expand on being hydrated
without significantly affecting the overall grain volume (which is bounded by the
“container” husk). From Figure 4 (right), one can calculate that, for hydration from
13 to 30% moisture, the increase in grain volume is ~30% for brown rice but only
Fig. 3. Hydration of paddy at different temperatures. RT = room temperature; dotted line = rice soaked
at 75°C for 2 hr, then at 65°C; * = original rice. (Reprinted, with permission, from Bhattacharya and
Subba Rao, 1966a; ¸ American Chemical Society)
~9% for paddy rice. Interestingly, this latter value agrees almost exactly with the
value in Figure 4 (left) up to the break point. When, however, ~32% grain moisture
is exceeded, caused by the temperature being above the GT, the husk splits open
and the rate of volume expansion no longer is circumscribed by the husk volume
and should be the same as for brown rice. This explains the steeper expansion in the
later portion of the curve in Figure 4 (left).
The above data reveal that, in normal soaking for parboiling, the volume
expansion of paddy rice in the tank is on the order of only 10%.
Bandopadhyay and Ghose (1965) studied the change in porosity and pressure
drop across the paddy bed during soaking. They found the porosity to decrease
marginally from 0.50 to 0.48 in one variety during soaking at up to 75°C, but it fell
significantly to 0.42 when soaked for 3 hr at 80°C. Similarly, there was very little
pressure drop across a column of grain during soaking below 75°C, but it became
perceptible at 80°C. The changes at 80°C clearly occurred because of gelatinization,
followed by splitting of the husk, resulting in greater rounding of the grain (reduced
length:breadth [L/B] ratio) as well as more adhesion among them. Bhattacharya et
al (1972) showed that porosity is positively related to L/B ratio in rice.
Theoretical Considerations. The kinetics of the absorption of water by paddy
rice during soaking have been studied by Bandopadhyay and Ghose (1965), Ali and
Ojha (1976), and Bandyopadhyay and Roy (1976, 1978). They adopted the
formulation of Becker (1960), who used Fick’s law to describe the diffusion of
moisture in wheat during soaking. Accordingly, they showed that the amount of
water absorbed by paddy during soaking at any given time can be represented by
Ë Û Ë6Û
[ - [R = ÌÌ Ü([V - [R )Ì Ü 'q
Ü Í9 Ý
Í p Ý
or [ - [R = N P q
when km replaces the term (2/»p)(xs – xo)(S/V)»D, being a constant for a given set of
conditions. In this formulation, xo = the initial, uniform moisture content (g/g, dry
Fig. 4. Left, swelling of paddy during soaking at various temperatures. (Reprinted, with permission,
from Bandyopadhyay and Roy, 1977) Right, relationship between moisture content and density and
bulk density of brown rice and paddy. (Adapted from Bhattacharya et al, 1972; © Society of Chemical
Industry. Permission is granted by John Wiley & Sons on behalf of the SCI.)
Parboiling / 341
basis [db]), x̄ = average moisture content for a given absorption period q (g/g, db),
xs = effective moisture content at the bounding surface at times greater than zero
(g/g, db), S = the exposed surface area of the grain (cm2), V = volume of the grain
(cm3), D = diffusion coefficient (cm2/sec), and q = absorption time (sec).
In other words, the absorption of water should be approximately proportional to
the square root of soaking time. Indeed, this relationship holds reasonably well (Fig.
5, left). However, the lines in the figure do not pass through the origin but intersect
the x̄ – xo ordinate at a point. This intersection suggests that there is an initial
instantaneous hydration (xi) when paddy is first put in water, which can be observed
in the soaking curves as well (Fig. 3). It was observed in wheat as well by Becker
(1960), who attributed it to rapid filling of the capillaries in the pericarp. For paddy,
all researchers have assumed the initial hydration to be due to rapid filling of the
space between the husk and the endosperm. This is incorrect. It is the husk that
provides the main barrier to hydration of paddy, not the endosperm. The bran layer
is a second barrier, as shown by the comparative hydration curves of milled rice,
brown rice, and paddy (Indudhara Swamy et al, 1971) (Fig. 6). If the husk is slightly
split open by a blade, paddy hydrates as fast as brown rice (Fig. 6). Hence, the rapid
initial hydration of paddy probably is nothing but quick filling of empty spaces and
capillaries in the uneven, hairy husk surface.
The extent of the initial hydration (~5% db) has been found to vary somewhat
from variety to variety and temperature to temperature (just as in wheat). This fact,
although much speculated upon in the above literature, may be no more than a
reflection of the grain surface texture, prevailing surface tension and vapor pressure
of water, and experimental variables.
Taking this initial moisture gain into consideration, one can write a semi-
empirical hydration equation as x̄ – xo = km»q + xi.
The relation between water absorption and square root of time is valid only up to
a grain moisture of ~55% (db; ~32% wb) (x̄ – xo = 0.40) (Fig. 5, left). This is
Fig. 5. Left, plot showing that moisture gain during soaking is proportional to the square root of
soaking time (q). Right, break in Arrhenius plot for soaking of paddy. The constant, km, is a measure of
the amount of moisture absorbed per root unit time. (Reprinted, with permission, from Bandyopadhyay
and Roy, 1976)
evident even in Figure 3, which shows that, beyond this moisture, attained only at
soaking temperatures above 70°C, the curve takes an upturn caused, as mentioned
before, by ensuing gelatinization and the attendant splitting of husk.
For this reason, it might be more appropriate to analyze the hydration of paddy
as a case of simultaneous diffusion and gelatinization. Bakshi and Singh (1980,
1982) examined this aspect. They studied the hydration of paddy in water at 50–
120°C. Based on earlier literature, they thought that soaking of paddy for parboiling
was complete when it attained 45% (db) (~30% wb) moisture. This assumption is
fallacious, because 45% (db) moisture is sufficient for parboiling only if it is
distributed evenly throughout the grain; it is insufficient when the paddy is soaked at
high temperatures and a steep moisture gradient prevails, as is discussed below.
Further, these authors reported as high as 50% gelatinization after soaking paddy at
50°C for 6 hr, which seems untenable. Further work in this area is necessary.
The Indian researchers cited above thought that the primary purpose of deriving
equations of soaking lay in calculating the appropriate soaking time and
temperatures. The author of this chapter believes that hydration is, in practice,
affected by so many variables (e.g., surface area and volume of grain, texture of
husk [including its hairs], tightness of lemma-palea closing, and gelatinization
temperature) that it should be far more practical to determine soaking conditions
empirically rather than by calculation. (The simplest method of determining the
optimal soaking conditions is to soak paddy at different temperatures for different
times, drain it, open steam it for 5–10 min, dry it, and mill. The minimum time at
the highest temperature at which the husk does not split and the rice shows no
Fig. 6. Hydration of milled and brown rice, paddy, and split paddy (husk slightly split open with a
blade) at room temperature. (Adapted from Indudhara Swamy et al, 1971; used by permission)
Parboiling / 343
opaque central core (i.e., has no “white belly”) are the optimum time and
temperature [Table 2]). The true value of equations lies in providing a better insight
into the phenomenon. Any equation is based on many assumptions. By testing and
gradually refining these assumptions from the fit of the experimental data to the
predicted data, one is led to a better understanding of the phenomenon. To illustrate,
one would like to know why the very low- and very high-temperature curves in Figure
5 (left) show a good fit with the experimental values but the middle curves do not.
The temperature dependence of hydration is related to the effect of temperature
on diffusion, which can be expressed by an Arrhenius-type equation,
-(
' = 'R H 57
in which Do = diffusion constant (cm2/sec), E = activation energy (cal/mol), T =
absolute temperature (Kelvin), R = the universal gas constant, and therefore,
Ë-(Û
NP = . Ü H[SÌ
Í 57 Ý
where K replaces the constant part (2/»p)(xs – xo)(S/V)»Do.
Using the above relation, Bandyopadhyay and Roy (1975, 1976) found that the
Arrhenius plot shows a sharp break at ~70°C (Fig. 5, right), the temperature of the
break varying somewhat from variety to variety. The value of activation energy (E)
calculated from the slope is in the range of 4–8 kcal/mol at temperatures below the
break point and 25–32 kcal/mol above. Clearly, this phenomenon too is related to
the change brought about by initiation of gelatinization at higher temperatures of
soaking discussed above.
Practical Soaking. Soaking, in practice, must be considered in terms of the above
factors. Two broad alternatives emerge. On the one hand, not much care or control
of the soaking time is necessary if the temperature is maintained at less than 60–
65°C, because the equilibrium moisture content under these conditions does not
exceed 30–32% wb (Fig. 3); thus, there can be no oversoaking or any splitting of
husk. The disadvantage here is that too much time may be required and
fermentation may take place if the temperature is too low. Conversely, soaking is
fast and no fermentation can occur at high temperatures. However, the danger here
is that oversoaking with consequent splitting of husk and attendant leaching and
grain deformation may occur. This is promoted by another factor. A moisture
gradient forms in the grain during soaking (Mecham et al, 1961). As shown in
TABLE 2
Moisture Content of Paddy When Optimally Soaked at Different Temperaturesa,b
Optimal Soaking Timec (hr) Moisture (% wb)
Soaking Temperature
(°C) BS RC BS RC
Room temperature 60 52 28.3 28.9
50 9.5 7.5 28.5 …
60 6.0 5.0 28.7 30.8
70 4.0 2.8 30.7 35.0
80 3.0 2.5 35.0 43.8
a
Source: Bhattacharya and Subba Rao (1966a); © American Chemical Society. Used with permission.
b Two varieties (BS and RC).
c Minimum time of soaking needed for rice (after steaming, drying, and milling) to show no “white belly.”
Figure 3, the moisture gradient would be relatively low and would progressively
decrease and eventually disappear at low temperatures. The situation is radically
different at high temperatures, because the gelatinized shell would continue to
absorb water freely, leading to a sustained high moisture gradient, which would
make the total grain moisture exceed 30–32% before the grain center was
sufficiently hydrated (Table 2), causing husk splitting and its attendant problems.
Hence, practical soaking at reasonably high temperatures can be done in one of
the three following ways: 1) at >75°C, the soaking time should be strictly monitored
and the water drained out when the grain moisture reaches 30–32%, after which the
paddy is tempered to equalize its moisture (Mecham et al, 1961); 2) paddy should
be soaked at ~70°C with good control of time; or 3) soaking should be started at
~75°C and the batch allowed to cool naturally during soaking. The efficacy of the
last procedure is shown by the 75–65°C curve in Figure 3. The actual temperature
must be adjusted depending on the GT of the variety, however. Otherwise, low-GT
varieties would be overhydrated even at 70°C (Biswas and Juliano, 1988).
It is not essential that the paddy be soaked to saturation for parboiling, although
that is the usual practice. Approximately 24% moisture evenly spread throughout
the grain is enough to bring out its gelatinization after normal atmospheric steaming
(Ali and Bhattacharya, 1982; Pillaiyar et al, 1993; Velupillai, 1994). Even lower
moisture content would enable the grain to be parboiled if it is steamed under
elevated pressure (see below).
Hydration of paddy can be hastened by adopting certain measures. Certain
modern parboiling systems operating in the United States or Europe were long
believed to use vacuum and pneumatic pressure during soaking, but no data were
known. Research information on these aspects now has become available. In his
FAO bulletin, Gariboldi (1984) quoted from his own early work (Fig. 7), which
Fig. 7. Water absorption by paddy in 45°C water: A = at ambient pressure; B = vacuum treatment
before soaking; C = vacuum treatment before soaking followed by hydrostatic pressure (8 kg/cm2).
(Reprinted, with permission, from Gariboldi, 1984)
Parboiling / 345
shows that both vacuum and pneumatic pressure hasten hydration of paddy.
Velupillai and Verma (1982) and Jagannadha Rao et al (1997) confirmed that
applying air pressure during soaking reduces the soaking time. Kulkarni and Bal
(1984, 1986) confirmed that deaerating paddy before putting in water hastens its
hydration.
Pillaiyar et al (1998) recently have shown that grains from different tillers or at
different positions within a panicle hydrate at slightly different rates; younger grains
are slow to hydrate. Therefore, grain-to-grain variation is a definite hazard during
soaking of a paddy lot. Appreciable variation in the initial grain moisture is said not
to affect its hydration significantly (Ali and Ojha, 1976).
Incidentally, the hydration curve of the split-husk paddy in Figure 6 provides a
challenge to future technologists. If one can devise a technology by which the husk
can be slightly cracked before soaking, the entire soaking operation will become
very simple, performable at room temperature.
STEAMING OR HEATING
The purpose of steaming the soaked paddy is to gelatinize the starch. If the grain
has been hydrated adequately and evenly, steaming for just 2 min at atmospheric
pressure is enough to gelatinize it (Bhattacharya and Subba Rao, 1966a). However,
the heat input has an effect on the milling quality. The ability of the parboiled grain
to withstand adverse conditions of drying without giving rise to cracks increases
with increasing severity of heat treatment (Table 3). In fact, Mecham et al (1961),
who dried paddy after parboiling it in a cross-flow dryer, needed steaming under
high pressures to get a good yield of head rice. This may be one reason why all
high-technology processes from the United States and Europe traditionally used
steaming under elevated pressure.
The husk splits during steaming, probably as a result of the swelling caused by
absorption of the adhering water or the condensate. No splitting of husk occurs in
processes where the grain moisture remains low (“short-soaking tempering”, dry-
heat, and “pressure” parboiling; described below in the section on production
systems).
TABLE 3
Effect of Processing and Drying Conditions on the Breakage of Parboiled Rice During Millinga
Soaking Temperature Steaming Pressureb Milling Breakage
(°C) (kpscg) Drying Methodc (%)
Room temperature 0.0 Air 1
Sun 36
60°C 94
1.05 Air 1
Sun 2
60°C 6
80d 0.0 Air 1
Sun 8
60°C 15
a Source: Bhattacharya and Subba Rao (1966a); © American Chemical Society. Used with permis-
sion.
b Steamed for 10 min at kg/cm2 gauge (= kpscg) indicated.
c Dried continuously to 11–13% moisture.
d Soaked for 2.5 hr and then tempered at the same temperature for 0.5 hr.
Steaming, although used almost exclusively in practice, is by no means

essential for parboiling. Other systems of heating, such as mild heating at 80°C in
a closed box immersed in a bath (Pillaiyar and Mohandoss, 1981a,b) or heating in
a closed rotating drum by flue gases in the jacket (Pillaiyar et al, 1977), by
thermic fluid (Pillaiyar et al, 1996), by electrical resistance (Vasan and Ganesan,
1981), by ohmic heating (LSU, 1998), by microwave (Velupillai, 1994), and even
by hot sand or air, can be used. The relevant systems are discussed under
Production.
DRYING
After regular parboiling, paddy contains ~35–38% moisture. (This is the
absorbed grain moisture. In addition, there is ~5% adhering water.) It must be dried
quickly to 12–14% moisture for safe storage and milling. Types of dryers and
systems are discussed elsewhere in this volume. Another aspect of drying is the
drying protocol, which exerts a profound influence on the milling quality of the
product. If properly processed parboiled paddy is air dried, it yields nearly 100%
whole grains after milling; but if it is dried in the sun or with heated air, breakage
can be very high. The method of drying is the kingpin of the parboiling process
insofar as the milling quality of rice is concerned.
When wet parboiled paddy is dried in the sun or with hot air, no damage occurs
until the moisture content drops to ~16% (Craufurd, 1962, 1963; Sluyters, 1963;
Bhattacharya and Indudhara Swamy, 1967), after which the amount of breakage
rises steeply (Fig. 8, left). Interestingly, cracks appear in the grain not during but
Fig. 8. Effect of drying conditions on the milling quality of parboiled paddy. Left, paddy dried
continuously at various temperatures in one-stage drying. Right, paddy dried at 60°C to 12.4%
moisture all at once, then held (“conditioned”) in closed bottles at 50°C for different time intervals
(indicated), cooled to room temperature (RT), and finally milled after holding at RT for different times
(abscissa). (Reprinted, with permission, from Bhattacharya and Indudhara Swamy, 1967)
Parboiling / 347
after drying, when the rice has cooled, and over some period of time. In fact, if the
grain is tempered hot immediately following drying, no damage occurs, not even
upon subsequent cooling (Fig. 8, right). The conclusion is that a steep moisture
gradient and cooling are both necessary for cracking to occur following drying;
neither condition alone is sufficient. This phenomenon may be related to the effect
of glass transition temperature recently propounded at the University of Arkansas
(Cnossen and Siebenmorgen, 2000; Perdon et al, 2000).
In practical terms, because the cracks start only after the moisture drops to
16%, the solution is to dry the paddy as fast as feasible to 16–18% moisture, and
then to temper it for 4 hr (if warm) to 8 hr (if cooled) to enable the steep moisture
gradient to relax. Then the drying can be completed in a second pass without any
adverse effect (Bhattacharya and Indudhara Swamy, 1967). The efficacy of this
principle of two-stage drying with an intermediate tempering at 16–18% moisture
has been demonstrated in mill-scale drying both in the sun (Bhattacharya and Ali,
1970) and with heated air in a Louisiana State University (LSU) dryer
(Bhattacharya et al, 1971). With this method, breakage during milling of even
mildly parboiled rice (steamed at atmospheric pressure) can be virtually
eliminated. This finding has been confirmed (Rama Rao and Bonde, 1977). The
importance of the demonstrated effectiveness of this technique for sun drying is
not to be underestimated, because sun drying is by no means as inefficient a
system as is often claimed (Bhattacharya, 1985), and the bulk of parboiled paddy
in the world, especially in South Asia (running into tens of millions of tonnes) is
still being dried in the sun, and perhaps will continue to be so dried for some time
to come.
Drying the just-cooked parboiled paddy directly in vertical continuous-flow
dryers is problematic because the grain is soft and not easy-flowing and also
because a very large amount of moisture must be removed (from ~65 to 15% db).
Nevertheless, this process is widely practiced in India as a batch system. The
overall heat-utilization efficiency is very good (Bhattacharya et al, 1971). Rotary
dryers (Maung Maung, 1977; Gariboldi, 1984; Velupillai, 1994) and fluidized bed
dryers (Schule, 2001) are said to be in use for the first stage in all modern processes,
but no data have been published on their use.
Apart from drying with heated air or sunlight, the same effect can be achieved by
other means. One of the production systems (Gariboldi process) is said to use
drying under vacuum (along with heat) (Gariboldi, 1984). However, no details are
available. Pillaiyar (1989) showed that wet parboiled paddy in a container can be
dried at a fast rate by immersion in a hot bath and application of vacuum, with no
adverse effect on any property. Velupillai et al (1989) and Wadsworth et al (1990)
tested microwave-vacuum drying of wet parboiled paddy in the laboratory with
success. Because the drying is caused by direct heating of the moisture within the
grain, there is no falling rate of drying until the moisture reaches 18% (i.e.,
diffusion of water is not rate limiting). It has been claimed that parboiled paddy can
be dried in a single pass by this system to safe-storage levels.
Bal et al (1976) and Rama Rao and Bonde (1977) studied drying parameters of
parboiled paddy in the laboratory. Igathinathane and Chattopadhyay (1999a,b)
studied the moisture diffusion in fluidized-bed drying of parboiled paddy, brown
rice, and milled rice, from which the diffusion in husk, bran, and endosperm were
calculated. Palani Muthu and Chattopadhyay (1992) studied the thin-layer drying
characteristics of parboiled milled rice.
Production Systems
Parboiling of rice can be done in diverse ways. Apart from ancient home-scale
systems, the various actually existing and potential systems are summarized in
Figure 9.
HOME- AND SMALL-SCALE METHODS

Originally, parboiled rice was made at home and, because homes had no separate
source of steam, was made by hot water. Endless variations existed (Tata, 1962;
Gariboldi, 1974). Perhaps paddy would be put in water in a vessel, heated until the
water boiled, and left to cool overnight; it might be soaked in moderately hot water
for several hours, after which the water was drained out and the rice heated with a
little water at the bottom to generate some steam; it might be soaked in moderately
hot water for a few hours and briefly boiled; and so on.
The results of Bhattacharya and Subba Rao (1966a,b), who found that mere
soaking of paddy at a high temperature imparted many of the properties of parboiled
rice (Table 4), explain the rationale of these home-scale methods. No doubt the
product from these processes would either contain some “white belly”
(ungelatinized opaque core) or undergo a certain amount of grain deformation.
However, it yielded a wholesome product free from the microbial fermentation that
bedeviled the process when large-scale parboiling was started.
Certain small-scale methods do use separate soaking and steaming but do not
require the use of a boiler (Gariboldi, 1984; Bhattacharya, 1985). Essentially, paddy
is taken into an iron tank or drum on a false bottom and is first soaked in hot water
with the heat supplied by a fire from below. The water is drained out below the
level of the false bottom, and the water is heated again, this time to steam the rice.
Alternately, paddy is soaked in hot water separately in pots by the customer, and
then brought in and steamed in the above tank for a fee. This method is still widely
used in pockets throughout South Asia.
Processes using crude homemade boilers also are in use. Roy (2001) describes
the various home and small-scale processes in vogue in Bangladesh even today,
including both single- and double-steaming systems (see below) adapted to such
small-scale operations.
SOAK-DRAIN-STEAM-DRY METHODS
With the beginning of marketing, large-scale manufacture of parboiled rice was
needed. With the advent of boilers, the soaking and heating (steaming) could be
separated. Thus started the soak-drain-steam-dry methods (Fig. 9, process 1), the
backbone of the parboiling industry throughout the world.
Single-Boiling Process. In the initial phase, entrepreneurs took the easiest way
out, namely, soaking paddy in ambient water, because heating such a large bulk of
water is not only costly but also causes stratification of the water—hotter on top and
colder at the bottom—and consequent differential hydration. This method of
parboiling (Fig. 9, process 1a) was called “single boiling” (because paddy is
steamed only once) (Ramiah, 1937; Narayanswami, 1956; Ghose et al, 1960).
Paddy is soaked in ambient water in masonry tanks (each holding 5–30 t) for
approximately three days. Then the water is drained out, and the soaked grain is
lifted manually in baskets into aboveground, upright, cylindrical steaming tanks
(each holding 200–500 kg of paddy), each of which has a conical bottom closed
Parboiling / 349
Fig. 9. Different actual and potential processes of parboiling. kpscg = kg/cm2 gauge; m = moisture; HTST = high-temperature, short-time. All degrees are Celsius.
with a hatch, and into which passes a steam pipe. The rice is steamed for a few
minutes as the condensate drains, after which it is taken out and dried in the yard
(Gariboldi, 1984; Bhattacharya, 1985).
Double-Boiling Process. Meanwhile, some enterprising millers discovered that,
if paddy was steamed before being put into water, the water became heated, and the
soaking time could be reduced to 24–36 hr. The soaked paddy was again steamed as
before. Thus arose the “double-boiling” method (so called because paddy is
steamed twice) (Fig. 9, process 1b). Most units in India operating either of these
two processes have a fairly large capacity: 50–100 t of paddy per day, or more.
CFTRI (Hot-Soaking) Process. As explained earlier, the prolonged soaking in
cold or warm water in the above processes generally causes a lot of microbial
fermentation, especially if the tanks are not regularly cleaned and if unfavorable
weather causes a delay in drying, resulting in a characteristic offensive odor that
remains in the product. Loss of up to 3% dry matter during soaking also has been
reported (Anthoni Raj and Singaravadivel, 1980; Pillaiyar et al, 1980). Several
decades were devoted to research on the origin of the odor and its amelioration,
culminating ultimately in the CFTRI hot-soaking process (Subrahmanyan et al,
1955; Desikachar et al, 1955, 1957).
Having established the efficacy of hot soaking (Fig. 9, process 1c), researchers at
CFTRI soon realized that its adoption required certain minimal changes in the
equipment. The equipment devised (CFTRI, 1969) consists of upright, conical-
bottom, aboveground, mild-steel tanks, each holding 4 t of paddy and fitted with
steaming manifolds (Fig. 10). Water first is taken into the tank and heated by
injecting steam to ~90°C, or hot water from a separate storage tank is pumped into
the soaking tanks. Paddy previously cleaned and lifted to an overhead bin then is
dropped into the water; the floating chaff is removed, and the rice is allowed to soak
for 3–4 hr at ~70°C, with occasional recirculation of water within the tank to
TABLE 4
Effect of Soaking Alone and of Full Parboiling of Paddy on Properties of Resulting Milled Ricea
Soaking Milling Yieldd (%)
Water Thiamin
Temperature Steamingc Total Head Uptakee Content
(°C) Timeb (kpscg) Rice Rice (g/g) (mg/100 g)
Untreated 71.0 7.5 3.12 35
Room temperature b … 71.5 9.5 … 29
50 b … 71.5 12.5 … 80
60 b … 71.5 23.0 … 117
70 b … 72.5 31.0 3.20 204
c … 73.0 65.5 2.72 174
80 a … 72.5 25.5 3.34 226
b … 73.0 70.5 2.68 209
c … 73.0 70.0 2.20 185
70 b 0 74.0 73.5 2.32 264
0.7 74.0 74.0 1.90 233
1.4 74.0 74.0 1.78 186
a Source: Compiled from Bhattacharya and Subba Rao (1966a,b) and Subba Rao and Bhattacharya
(1966); © American Chemical Society. Used with permission.
b a = Undersoaking, b = optimum period of soaking for the temperature, c = oversoaking.
c Steamed for 10 min at kg/cm2 gauge pressure indicated.
d Well milled with McGill miller no. 1.
e Rice cooked at 96°C for 30 min.
Parboiling / 351
equalize the temperature between the top and the bottom. Some millers, of
“doubling-boiling” vintage, let paddy into the tank, steam it, and then pump in cold
water; the resulting temperature after circulation is ~60–65°C, which requires ~6 hr
of soaking (Sarda, 1965).
After soaking, the water is drained out and steam is passed through the steaming
manifold, with the drain cock kept open to drain the condensate. Steaming is
stopped after excess steam comes out from the bottom and the top. If the rice is
dried mechanically, it is discharged onto a belt or screw conveyor and fed to the
dryer through an elevator. Kimura (1989) studied the distribution of steam inside
the tank. He found that propagation of heat is fastest in the upward direction, next
fastest downward, and least fast in the horizontal direction. On this basis, he
recommended a horizontal arrangement of steam pipes at the bottom of the tank.
This is an extremely simple yet effective process, although entirely manually
controlled. It is a batch process but can be made semicontinuous by using a number
of tanks and staggering their operation. The process has been adopted fairly widely
in India, although the older processes and sun drying have by no means disappeared.
Its simplicity, inexpensiveness, flexibility, and absence of any patent protection
Fig. 10. Parboiling tank designed by Central Food Technological Research Institute. Left, schematic
drawing; right, tank at an installation. (Courtesy Sri Valli Engineering Works, Tiruvarur, Tamil Nadu)
have helped its fairly wide adoption even outside India, including South America
(Steven D. Danforth, personal communication), and Bangladesh, Sri Lanka, Nepal,
Nigeria, and Vietnam (S. Chakraborty, personal communication).
Jadavpur University Process. Similar studies were done at the Jadavpur Univer-
sity, Kolkata, and a sophisticated, continuous process was developed (Mazumder et
al, 1960a,b). Paddy is propelled by paddles in a long, trough-shaped, covered
soaker-steamer containing water at ~75°C, with hot water being circulated in a
jacket. The speed is adjusted so the rice reaches the end after 2.5–3 hr. The rice is
automatically scooped out and put into a small steaming chamber, where it is
exposed to live steam for a few minutes. Next, it is passed through a cup-and-cone
cooling tower, then fed into a rotary dryer housing a nest of steam pipes to heat the
paddy, with a concurrent flow of hot air. Unfortunately, this process of excellent
potential never moved beyond the pilot-plant scale.
Improved Processes in Thailand. By the 1960s and 1970s, plants gradually began
to be modernized in Thailand. Although a few very modern plants are in operation,
the majority may be called semimodern. A typical process follows (author’s
personal observation). Hot soaking, after initial cleaning and washing of paddy, is
done in huge ground-level pits holding 150–200 t of rice, and hot water (70°C) is
poured continuously from pipes fitted on the surrounding walls. Recirculation of
water within the tank is difficult or impossible; hence, a certain amount of latent
fermentation may occur in the bottom layers. The soaked rice is taken out with
power shovels and loaded into trucks fitted with steam coils. Each truck moves to
the boiler, where it is connected to the steam line; the rice is steamed in the truck
itself. Then the truck moves on and dumps the paddy at the dryer plant or, rarely
these days, at the drying yard for sun drying.
Modernized Double-Boiling Process. Many millers in the Haryana and Punjab
states of India have adopted an ingenious system for producing parboiled basmati
rice for export (Fig. 9, 1b).
The plant consists of three tiers. In the original process, paddy first was steamed
in repeated batches in several small steaming tanks (holding 300–400 kg of grain)
located in the top tier. Next, the batches of steamed paddy were dumped
continuously, one after another, into a large, covered soaking tank (holding 20 t of
grain or more) located below in the second tier. When this lower tank was full,
ambient water was added and, thereby, heated, and the paddy was soaked for 10–15
hr. Then the water was discharged and the soaked paddy flowed into another bank
of steaming tanks located in the bottom tier, where it was steamed and then
conveyed into an LSU-type dryer (or to the drying yard in small mills). In recent
times, the process has been further improved by many millers. Now one initial
steaming, one soaking (capacity 8–10 t), and one final steaming tank each are set up
as a subunit, and approximately four such subunits constitute one processing unit.
The soaking tank is fully enclosed; therefore, 6–8 hr is sufficient for soaking. A
photograph of one of these installations is shown in Figure 11. The rice may or may
not have a malodor depending on the type of plant, management, hygiene, and
weather.
Modern Processes. Several elegant processes for parboiling based on the system
1c in Figure 9 were developed in Europe and the United States from approximately
the time of World War II and have been continuously improved since then.
Currently, parboiled rice is produced by some variant of these processes in the
United States, Europe, Thailand, and Central and South America.
Parboiling / 353
All of these modern plants use very sophisticated techniques. All are continuous
systems with very high capacities (5–30 t of paddy per hour), working year round,
where hundreds of tonnes of food (paddy) are soaked, cooked, and moved around in
a continuous stream. This makes it imperative for these processes to have a high
degree of mechanization and automatic control to ensure strict hygiene and freedom
from hazard.
The details of these processes are confidential. However, their patents expired
long ago. From experience gained over the decades, the author can say that two
processes appear to be in general use.
One is the basic system of the original Uncle Ben’s Conversion process (Kik and
Williams, 1945) and Gariboldi’s Crystallo process. Paddy first is deaerated by
applying vacuum, then soaked at 55–70°C under pneumatic pressure. The rice is
steamed in a closed rotating autoclave, followed by drying, first under vacuum and
heat in the same autoclave and then in a continuous-flow dryer. This batch system
probably now operates in only a few plants in Europe and South America.
The second system, more widely used, perhaps by all companies in the United
States and several others in Latin America, Thailand, and Europe, can be called by
the generic name of the “American system” or the “continuous system.” It is
Fig. 11. Photograph of an installation of a three-tier “modernized double boiling” system of parboiling.
(Courtesy Induss Food Products and Equipments Ltd., Kolkata)
essentially an evolution of the original Malek process developed a half-century ago.

The following gives a fair idea of this system, as popularly understood (Maung
Maung, 1977; Velupillai, 1994). Paddy is soaked with or without initial deaeration
and later pneumatic pressure in a series of soaking tanks at ~70°C. The soaked
paddy is removed from the tanks in sequence and steamed continuously in a vertical
autoclave. It then passes into a rotary (or a fluidized-bed) dryer, where, in the first
pass, it is dried with hot air at a temperature of 150–260°C down to ~20–25%
moisture. The paddy then is dried in a series of continuous-flow dryers to gradually
reduce the moisture content to 12–14% with or without tempering between each
stage.
Detailed descriptions, as known or surmised, of these two and other, earlier
processes (e.g., “Avorio,” Fernandez, Arlesiene, and so on, now mostly abandoned)
are available in Gariboldi (1984).
No details of the rotating or vertical autoclave described above for steaming are
available in public. Adhikaritanayake and Noomhorm (1998) have described a
continuous-flow vertical steaming unit consisting of two coaxial pipes. Steam is
passed through the inner (perforated) pipe; soaked paddy flows down the outer,
annular space, thereby being steamed. The equipment is so arranged that the
condensate flows out without disturbing the paddy.
Other Developments. The soaked paddy may be heated by other means, as shown
in Figure 9, system 2. These methods have, so far, been tested only in laboratory
scale. Among these, the method having the greatest potential is microwave-heating
parboiling. This, too, is still in the developmental stage, but it has considerable
potential and is discussed in detail by Velupillai (1994). Among its claimed
advantages are 1) all three steps of soaking, steaming, and drying can be done
sequentially in the same equipment by adjusting the microwave power and
incumbent pressure, and 2) all three steps can be accomplished in much shorter
times than in the conventional processes.
Another modification claimed is shown in Figure 9, process 1d or 2d. In this
system, paddy is not soaked to saturation (~30% moisture) but only partially (~24%
moisture), then tempered to even out the moisture within the grain (Pillaiyar et al,
1993). The authors call this process the “short-soaking-tempering” system of
parboiling. The rationale—also noted by Ali and Bhattacharya (1982) and
Velupillai (1994)—is that ~24% moisture evenly distributed within the grain is
enough to gelatinize it by atmospheric steaming. The low grain moisture should
provide considerable time and cost saving for drying. Pillaiyar (personal
communication) states that the process has been field tested in several rice mills in
India with favorable results.
PARBOILING BY SOAKING ALONE

Indian scientists always have felt a peculiar fascination for the challenge of
devising a modernized version of traditional home-scale parboiling (viz., parboiling
by hot water alone). Ali and Ojha (1977) at the Rice Process Engineering Centre
(RPEC), Kharagpur, India, suggested that soaking paddy at 70–75°C for a suffi-
ciently long time was enough to parboil the rice; they called this the RPEC process.
Bandyopadhyay and Roy (1977) made a similar observation. Kulkarni and Bal
(1986) suggested that their system of using vacuum before soaking at 70°C could
yield parboiled rice without further steaming.
Parboiling / 355
The above authors took grain translucence alone as a benchmark for parboiling.
A caution is due here. Unnikrishnan et al (1982) noted that soaking at 50–70°C
often gave grain translucence without measurable gelatinization. On the other hand,
basing their work on the soaking and product-property data in the literature, these
authors have shown how best to achieve the desired end. Paddy is soaked in water at
80–85°C for ~1.5 hr, at which time the moisture absorbed is sufficient to eventually
reach and gelatinize the grain center when equilibrated but is not enough to split the
husk. The water is drained out, and the grain is tempered for 1–1.5 hr in the same
container, which is insulated to retain the heat; the moisture penetrates the center
and the heat gelatinizes it (Fig. 9, process 3a). This system obviously would yield
only very lightly parboiled rice.
The German company Schule used to provide the following process (Hans
Keesenberg, personal communication). Paddy is soaked in medium-temperature
water for 2–3 hr under hydrostatic pressure. The pressure is released, and water at
~90°C is admitted to cook the rice (Fig. 9, process 3b). It is predried in a vibratory
dryer in a semifluidized condition, then dried in three passes in columnar dryers.
The process apparently never became popular, probably because the high grain
moisture inevitable in this process would cause a number of problems.
PRESSURE -PARBOILING
An interesting development came from the PPRC, Tiruvarur, India, in the 1970s
(Iengar et al 1972, 1974a,b). In this method, paddy is not saturated with water but is
only briefly soaked or wetted. Then, it is steamed under high pressure and becomes
fully parboiled (Fig. 9, process 4). The process was called “pressure-parboiling,”
which is actually a misnomer, because the distinctive feature of the process is not
pressure steaming (which also can be used in processes 1 and 2 in Figure 9), but low
grain moisture and pressure steaming.
The method has two great advantages: 1) the whole process is completed in a
very short time and 2) the parboiled grain has a low (<25%) final moisture, reduc-
ing the cost and time of drying. New millers in the Punjab state of India (a nontra-
ditional region where large-scale rice farming started only after the “green revolu-
tion”) took to it avidly in the 1970s and 1980s. However, serious shortcomings of
the product were soon revealed, along with great consumer resistance. The rice was
too discolored and also so hard that it was difficult to cook. In fact, it earned the
nickname “iron rice.” The process was abandoned.
A study was conducted (Ali and Bhattacharya, 1982) to discover the reasons for
the cooking resistance. Details of the results are discussed in the section on physio-
chemical properties. For the present, this study showed that discoloration and
resistance to cooking can be reduced only by adopting processing conditions that
leave a small “white belly” in the center of the grain. Such conditions consist of
washing the paddy in water and steaming it under 1–2 kg/cm2 gauge pressure for
10–20 min. This modification was successfully tested in commercial scale. An
alternative approach successfully tested later is to use low-amylose varieties for the
process (Unnikrishnan and Bhattacharya, 1987b, 1989); the resulting product is not
too hard after cooking, although the high grain discoloration remains (Table 5). The
pressure-parboiling process, otherwise so attractive, could be revived with either of
these modifications.
Two outstanding properties of pressure-parboiled rice is that it is excellent for
making puffed rice (Chinnaswamy and Bhattacharya, 1986) and canned rice
TABLE 5
Cooking and Textural Properties of Conventional and Pressure-Parboiled Ricea
Cooked-Rice Textured
Amylose Water Cooking Elastic
Content Uptake Time Firmness Recovery
Variety Sampleb,c (%, db) (g/g) (min) (%) (%)
IR20 Raw 27.8 4.09 17 40.9 30.1
30-0-10 3.52 28 47.3 42.8
30-0-60 2.83 42 56.1 61.7
20-2-20 2.51 75 62.3 67.8
20-3-10 2.55 70 61.9 67.4
Indrasan Raw 24.7 3.84 19 29.8 14.5

30-0-10 3.51 25 34.1 17.1
30-0-60 3.15 30 38.7 23.2
20-2-20 2.99 40 45.3 34.1
20-3-10 2.97 40 49.0 38.3
PR103 Raw 16.6 4.06 17 15.4 4.3

30-0-10 3.61 22 15.5 3.9
30-0-60 3.49 25 18.5 4.2
20-2-20 2.92 35 21.6 4.7
20-3-10 3.06 35 20.3 3.6
a Source: Unnikrishnan and Bhattacharya (1989); used with permission.
b Parboiled samples are often identified by a three-number code: a) the first number indicates the
nominal grain moisture (%), b) the second, the steaming pressure (kg/cm2 gauge = kpscg), and c) the
third, the steaming time (min). Thus 30-0-10 would mean 30% moisture paddy steamed at
atmospheric pressure for 10 min, 20-2-20 would mean 20% moisture paddy steamed under 2 kpscg
for 20 min, etc.
c The 30% moisture pair (30-0-10, 30-0-60) represent conventional parboiling; while the 20% moisture
pair (20-2-20, 20-3-10) represent pressure-parboiled rice.

d Chopin-INRA viscoelastograph.
(Unnikrishnan and Bhattacharya, 1987b, 1989) (see section on products from par-
boiled rice).
SIMULTANEOUS DRYING AND PARBOILING

(DRY-HEAT PARBOILING)
In this approach to parboiling, the soaked paddy, instead of being steamed, is
subjected to high-temperature short-time (HTST) treatment. The rice not only dries,
but its starch also is gelatinized by the intense heat (Fig. 9, process 5). It is amazing
that this process has, in fact, been in use in different forms for a very long time, for
rather small-scale production, in different corners of the world but was unknown
outside these pockets. Its prime attraction apparently lay in the fact that it did not
require one to install and maintain a boiler—which is both expensive and
hazardous.
First, as mentioned earlier, the process had been prevalent for a long time in and
around the Punjab region of undivided India (the sella process in India and the
chatti process in Pakistan, both now abandoned; Gariboldi, 1984; Bhattacharya,
1985). Soaked paddy was manually HTST-heated with or without sand in small
portions on iron pans over a fire. Second, the same process has been in use in
smaller scale all over the Indian subcontinent for a long time for the preparation of
Parboiling / 357
Fig. 12. Schematic diagram of “gram” roaster. (Reprinted, with permission, from Ali and Bhattacharya,
1980a)
an intermediate product during the making of flaked rice and puffed rice (discussed
in the section on products). However, until a study by Ali and Bhattacharya (1976a),
it was not realized that the process was essentially a case of parboiling. Third, a
process called the “stove process” has been in operation in the distant Santa
Catarina province of Brazil for a long time, unknown to the rest of the world. There,
soaked paddy is passed continuously through a rotating drum that is heated by a fire
from below (hence the name “stove process”). It is said that this process was
discovered quite accidentally long ago. Apparently, this hot drum system had been
in use for drying of harvested paddy. Once, it is said, a batch of paddy got wet and
when this paddy was dried by the same system, it gave better milling yield and the
rice looked yellowish and translucent. Since 1988, it gradually has become an
established process of parboiling in the state. As now practiced, the process involves
soaking paddy in upright cylindrical tanks in hot water with circulation, after which
the soaked paddy is gelatinized as well as dried by several rotary cylinders over
stoves in a continuous process (Amato and Silveira Filho, 1991).
Several de novo research efforts in the last three decades generated new interest in
this dry-heat principle. Iengar et al (1971) at the PPRC found that heating soaked
paddy with sand could parboil it. Khan et al (1974) at the International Rice Research
Institute, Los Baños, Philippines, saw that freshly harvested wet paddy was parboiled
when heated with sand or exposed to hot fumes and flames. Following the lead of the
flaked rice process, Ali and Bhattacharya (1976a,b) at CFTRI heated soaked paddy
with sand and found that the rice was not only parboiled but also had some special
properties. The same authors (1980a) then explored the process (using hot sand in both
cases), first by using a small rotating, electrically heated roaster and later a bigger,
continuous “gram roaster,” heated by flame from below (Fig. 12).
The matter then was studied in detail by a team in the CFTRI in the 1980s using
the same gram roaster. Finally, a process was standardized and then released by the
Institute for a fee (CFTRI, 1992). Pillaiyar et al (1994) at PPRC also made produc-
tion trials using the same gram roaster, a rice-sand ratio of 1:4, and a mixture tem-
perature of 125°C. Roasting the paddy down to 19–20% moisture gave a grain
temperature of ~90°C and good parboiling. The paddy later was dried to safe-
storage levels.
Alternately, Ali and Bhattacharya (1980a) showed the feasibility of obtaining

parboiled rice by briefly exposing soaked paddy to very hot air in the laboratory.
Vasan et al (1980), again at PPRC, dried soaked paddy in a dryer in three passes
with ~200°C air and saw that the paddy was parboiled.
Inasmuch as the paddy simultaneously is dried by this method and possesses the
unique property of quicker initial hydration (see the section on physiochemical
properties), this method holds some promise for the future, at least for producing a
niche product.
BROWN-RICE PARBOILING
In the second edition of this monograph, under “Future Developments”, the
author wrote that, if dehusked brown rice could be used as the feedstock for par-
boiling, that process would have a tremendous advantage. First, brown rice
hydrates very fast; in fact, in a few hours in ambient water (Fig. 6). Second,
brown rice weighs a little more than three-fourths as much as paddy and has two-
thirds its volume. Therefore, the process would be faster, simpler, and cheaper. A
surprising amount of progress in this direction has been made during the
intervening years.
Nagashima et al (1983) soaked brown rice at 45°C for 2 hr to 32% moisture,
steamed it at 100°C for 60 min, and dried it at 100°C for 1 hr. How the caking after
steaming and cracking during drying were avoided could not be ascertained. Itoh
and Kawamura (1985) used a more or less similar system. However, their objective
apparently was to obtain a quick-cooking parboiled brown rice; therefore, fissures
in the grain were not a handicap but actually were desired. Subsequently, Y. M.
Indudhara Swamy in the author’s laboratory (CFTRI, 1987) found that soaked
brown rice could be dry-heat parboiled with hot sand (Fig. 9, process 6a) without a
serious handling, caking, and cracking problem. The process then was tested in a
pilot scale at the Institute using the gram roaster (CFTRI, 1989). Subsequently,
Narasimha et al (2000) developed a full-fledged pilot-scale continuous system using
the same process, after modifying the roaster by reversing the sand and sand-grain
mixture paths, to provide more uniform heating and thereby prevent grain-to-grain
variation. The temperature of the rice-sand mixture was maintained at 120–130°C
with a residence time of 30 sec. The rice attained a grain temperature of 95–105°C
and a moisture of 20–22% at discharge. It was dried by simple aeration after
tempering. Milling breakage was less than 10%. Approximately 1% sand adhered to
the rice, but most could be removed by using a simple nonscouring polisher, and it
completely disappeared after whitening.
Guha and Ali (1998) examined the above process with different varieties. They
found that it worked well with relatively fine varieties but not with coarse varieties;
the extra heat needed apparently dried out the coarse grain too much for it to be
parboiled by the time it attained the required temperature. These products contained
too many chalky grains to be acceptable. Therefore, they devised an alternative
system. Brown rice was soaked in ambient water for 1 hr, to a moisture content of
~26%, followed by tempering for 5 hr to even out the moisture. The soaked brown
rice was steamed at 0.5 kg/cm2 for ~30 min. This system resulted in good parboiled
rice with normal properties with both fine and coarse varieties. Thus far, this
process (Fig. 9, 6b) has been tried only in the laboratory and awaits pilot-scale trial.
Kar et al (1999) soaked brown rice at 70–100°C, then cooled it at room temperature
for 2 hr, followed by open steaming for 20 min.
Parboiling / 359
Two relatively recent developments suggest that the seeming operational prob-
lems of the process have been largely overcome. The Uncle Ben’s company in the
United States took out a patent on a process (Kratochvil et al, 1994). Brown rice is
first soaked in ambient or warm water to saturation, then exposed to either hot air or
steam under high pressure for 15–30 sec. It is then tempered and finally dried. The
company is believed to be operating this system in their Greenville (Mississippi)
plant. The F. H. Schule company in Germany (Schule, 2001) offers a process using
a horizontal belt cooker for gelatinization of soaked brown rice and a belt dryer for
its drying. They claim that parboiling of brown rice is practiced on a small scale in
many places in Europe.
With so much development in less than two decades, this process seems to have
excellent future potential.
CHANGES DURING PARBOILING
Parboiling greatly changes the properties of the rice grain. These changes must
have their roots in the changes brought about in the grain constituents during the
parboiling process. However, more has been known about the properties of par-
boiled rice than about how they occurred. Only recently have scientists been in a
position to link the two.
An overview of the changes in the grain constituents (this section) and the prop-
erties of the processed grain (the next two sections) is presented below. The subjects
have been covered extensively in a review by Bhattacharya and Ali (1985), to
which the reader is referred for more detailed information and additional references.
Changes During Soaking
Inasmuch as rice is a seed and soaking for parboiling is akin to the initial phase
of germination, a large amount of enzymatic activity is to be expected. Ali and
Bhattacharya (1980b) noted appreciable enzymatic conversion of sucrose into
reducing sugars and de novo production of sugars and amino acids (Fig. 13). The
Fig. 13. Changes in reducing sugars (RS), sucrose (Suc), and free amino acids during incubation of
brown rice in water at different temperatures (°C). The total incubation mixture was analyzed.
(Reprinted, with permission, from Ali and Bhattacharya, 1980b)
well-known discoloration of parboiled rice is no doubt partly related to these pro-

ductions. Other reports of increase in reducing sugars (Anthoni Raj and
Singaravadivel, 1980) and in phenolic compounds (Singaravadivel and Anthoni Raj,
1979), of activities of diverse enzymes (Xavier and Anthoni Raj, 1996), and of
excretion of chloride during soaking (Anthoni Raj et al, 1996) have appeared.
In view of natural contamination, much microbial action during soaking may be
expected. However, what part it plays, if any, in the process or the product is not
known. Only the undesirable fermentation during the traditional parboiling has been
noted. The fermented odor is eliminated if the soaking occurs in hot water. Whether
any residual microbial activity persists and what effect it has are not known.
Ramalingam and Anthoni Raj (1996) have studied the microbial population and
concentration of various organic compounds in the soak-water effluent.
Changes in Organized Constituents
Steaming is the key step in which the unique properties of parboiled rice pri-
marily originate. Steaming disrupts all three organized constituents of rice—starch,
protein, and fat—in addition to causing other changes.
STARCH
Rice starch granules clearly are gelatinized during parboiling, because the onset
of all characteristic properties of parboiled rice is associated with steaming or mere
soaking at 70°C (Table 4). Indeed, birefringence of starch is lost, and the A-type
X-ray diffraction of raw-rice starch is destroyed (Raghavendra Rao and Juliano,
1970). A part of the gelatinized starch then reassociates or forms complexes with
lipid. These starch changes, the crux of the shift from raw to parboiled rice and the
basis of their respective properties, are discussed later in the section on the state of
starch in parboiled rice.
Many reports in the past suggested certain quantitative changes in starch (see
references in Bhattacharya and Ali, 1985), but these remain to be confirmed. The
total amylose content remains essentially unchanged (Raghavendra Rao and Juliano,
1970; Ali and Bhattacharya, 1972a). However, Mahanta and Bhattacharya (1989)
obtained evidence of thermal breakdown of starch. When parboiled-rice starch was
fractionated by gel-permeation chromatography, the proportion of the high-
molecular weight (MW) major fraction I progressively decreased and that of low-
MW minor fraction II increased as the steam pressure (i.e., temperature) increased
(Table 6). The lmax of the iodine complex of the two fractions also changed
accordingly. Substantial degradation of starch has been reported after extrusion
cooking of starchy foods (Colonna and Mercier, 1983; Schweizer et al, 1986; Lii et
al, 1992) as well as after popping of rice (Murugesan and Bhattacharya, 1989).
PROTEIN
The protein bodies in rice are ruptured and the solubility of the protein reduced
after parboiling (Raghavendra Rao and Juliano, 1970). Dimopoulos and Muller’s
(1972) data on protein solubility are shown in Table 7. The effect of this change in
protein on rice quality is not known. Despite earlier claims, the protein content of
brown or milled rice as well as the amino acid composition of the protein seem to
be unaffected by parboiling (Bhattacharya and Ali, 1985).
Parboiling / 361
TABLE 6
Fractionation of Parboiled Rice Flour on Sepharose Cl-2B Columna
Parboiling Condition
Steam
Carbohydrate
Paddy Moisture Pressure Time in GPCb Fr I lmax of Iodine Complex (nm)
(% wb) (kpscg) (min) (% of total) Fr I Fr II
Raw rice 69.5 571 634
30 0 10 68.5 571 638
1 10 53.8 569 647
3 20 33.0 581 622
22 1 10 47.4 577 644
3 20 16.1 588 620
17 1 10 55.8 564 649
3 20 20.9 581 621
12 3 20 38.4 587 621
a Source: Mahanta and Bhattacharya (1989); used with permission.
b Gel permeation chromatography. Fr = fraction.
TABLE 7
Effect of Processing Conditions on Solubility of Proteins
in Parboiled Ricea
Total Protein Extracted
Soaking Timeb Steaming Pressurec by Detergent Solutiond
(hr) (kg/cm2) (%)
Untreated 73.4
3.50 0.35 31.1
0.70 25.9
1.05 20.1
5.58 0.70 24.5
7.00 0.70 22.2
a Source: Dimopoulos and Muller (1972); used with permission.
b At 64°C.
c Steamed for 5 min at gauge pressure indicated.
d Sodium alkyl benzene sulfonate, 3% solution.
FAT
In their classic work, Subrahmanyan et al (1938) showed that milled parboiled
rice contains less fat than milled raw rice after equivalent degrees of milling, from
which they made the remarkable postulation that, as water solubles move into the
grain during parboiling, fat moves out. Whatever the truth of the former hypothesis,
a large body of observations has confirmed the latter (Bhattacharya et al, 1972;
Padua and Juliano, 1974; Benedito de Barber et al, 1977; Sondi et al, 1980) (Fig.
14). The total fat in brown rice remains more or less unchanged. Correspondingly,
the oil content of parboiled-rice bran has been repeatedly shown to be more than
that of raw-rice bran (Fig. 15). In fact, the location of the peak oil layer in the grain
moves slightly outward (Fig. 15). The amount of oil on the grain surface after vari-
ous degrees of milling also is higher after parboiling, and the peak layer shifts
slightly outward (Fig. 14). Mahadevappa and Desikachar (1968b) showed that, in
native brown rice, the oil occurs mostly in distinct bodies in the aleurone and germ
Fig. 14. Changes in the amount of fat in the total grain (t) and on the grain surface (s) in raw (R) and
parboiled (PB) rice with progressive milling with a McGill miller. (Reprinted from Bhattacharya et al,
1972; © Society of Chemical Industry. Permission is granted by John Wiley & Sons on behalf of the
SCI.))
Fig. 15. Oil content of raw and parboiled rice bran after various degrees of milling. Lower curves, oil
content in composite bran; upper curves, calculated mean oil content in successive bran layers.
Different symbols in the parboiled-rice curve represent different conditions of soaking (room
temperature to 80°C) and steaming (0.0–1.4 kg/cm2 gauge, 10–60 min). Milling was with the McGill
miller. (Adapted from Sondi et al, 1980; © Elsevier Science, used by permission)
Parboiling / 363
Fig. 16. Transverse sections of raw (left) and parboiled (right) brown rice, showing distinct bodies
containing oil in the aleurone layers of the former and their disruption in the latter. (Reprinted, with
permission, from Mahadevappa and Desikachar, 1968b)
but that these bodies are disrupted and converted into a band after parboiling (Fig.
16). The outward migration of fat must be related to this rupture of oil globules.
Possibly, the released oil seeps into the soft bran tissue but cannot penetrate the
harder endosperm.
Researchers have claimed that the outward movement of oil can be accelerated
further with an increase in the temperature (Chakravarty and Ghose, 1966) and
duration (Vasan et al, 1971) of soaking, as well as with an increase in the pressure
of steaming (Padua and Juliano, 1974). However, Sondi et al (1980) could not con-
firm it, because the oil content of rice, as well as its calculated fraction in bran at
various degrees of milling, remained constant irrespective of processing conditions
(Fig. 15). Some varietal difference exists in the extent to which bran is enriched in
oil after parboiling (Mukherjee and Bhattacharjee, 1978), suggesting differences in
the location and structure of oil globules.
The above changes in disposition of the grain fat have a significant effect on the
flow, packing, and milling properties and storage stability of parboiled rice (see the
section on product characteristics).
Changes in Other Constituents
VITAMINS
The greater amounts of thiamin and nicotinic acid in milled parboiled rice
compared with milled raw rice is a matter of history. However, this is a complex
change. The total thiamin in brown rice actually is reduced after parboiling (Done,
1949; Subba Rao and Bhattacharya, 1966; Padua and Juliano, 1974). Losses occur
both during soaking, being considerable if the husk splits open, and during steam-
ing, especially if done under pressure (Table 8). Some loss probably occurs in
nicotinic acid as well. In other words, the greater vitamin content in milled par-
boiled grain arises from the reduced loss of vitamins during the milling of brown
rice (Table 8). The classic data of Aykroyd et al (1940) are shown in Figure 17. The
milled grain apparently is similarly enriched with other B vitamins (Kik, 1955), but
the riboflavin content seems to be unaffected (Bolling and El-Baya, 1975; Ocker et
al, 1976).
TABLE 8
Loss of Thiamin During Soaking, Steaming, and Milling of Paddya
Percent Lossc of Thiamin
Soaking Steaming During Milling
During Parboiling
Temperature Timeb Pressure Time Before After
(°C) (min) (kpscg) (min) Soaking Steaming Steaming Steaming
Not soaked 0 10 10 90 64
Room temperature a 0 10 0 … 93 35
b 0 10 0 … 93 34
c 0 10 1 … 93 32
50 a 0 10 0 10 85 30
b 0 10 1 … 80 25
c 0 10 3 8 76 23
60 a 0 10 0 … 82 23
b 0 10 6 … 69 19
c 0 10 16 … 68 24
70 a 0 10 3 … 60 23
b 0 2 6 3 46 26
5 3 25
10 4 24
40 14 …
60 19 25
0.35 10 11 …
0.70 10 15 25
1.40 10 28 29
c 0 10 37 … 31 27
80 a 0 10 13 … 33 29
b 0 10 17 2 34 30
c 0 10 31 … 32 30
a Adapted from Subba Rao and Bhattacharya (1966); © American Chemical Society. Used with
permission.
b Soaking time: b = time required for optimal soaking at the indicated temperature; a and c =
undersoaking and oversoaking, respectively.

c Loss expressed as percent of B present in original paddy (for soaking), soaked paddy (for steaming),
1
and soaked and steamed paddy (for milling).
Fig. 17. Effect of milling on thiamin (left) and nicotinic acid (right) in raw and parboiled rice.
(Reprinted, with permission, from Aykroyd et al, 1940)
Parboiling / 365
SUGARS AND AMINO ACIDS

Milled parboiled rice contains a greater amount of sugar than milled raw rice
(Williams and Bevenue, 1953). Here again, the final sugar content is a dynamic
end-product (Ali and Bhattacharya, 1980b). Considerable enzymatic conversion of
nonreducing to reducing sugar occurs (Fig. 13), a large part of the sugar is leached
out into the soaking medium, and a small drop occurs during heating due to the
Maillard reaction. However, once again, the major change is a greatly reduced loss
of sugar (mainly of sucrose, which resides largely in the outer layers) during milling
compared with that in raw rice (Fig. 18; note the striking similarity to Fig. 17). The
extent of enzymatic production and leaching varies according to the processing
conditions; therefore, actual final sugar content also varies widely (sucrose is 0.02%
in raw milled rice and 0.45–0.68% in parboiled milled rice; reducing sugars are
0.06% and 0.06–0.18%, respectively).
Enzymatic production of amino acids (Fig. 13) and their reduced loss during
milling occur similarly. Again, samples vary widely, depending on the processing
conditions (0.23 mmol/100 g in raw rice, 0.52–0.86 mmol/100 g in parboiled rice).
MINERALS
Data on the content of ash and various minerals in brown and milled rice before
and after parboiling as reported by many authors have been collated by
Bhattacharya and Ali (1985). These show that: 1) little or no change occurs in the
Fig. 18. Changes in reducing sugar (RS) and sucrose (Suc) contents in raw and parboiled brown rice
during progressive milling with a McGill miller. (Reprinted, with permission, from Ali and
Bhattacharya, 1980b)
content of any constituent in brown rice after parboiling; 2) the content of total ash,
phosphorus, calcium, iron, manganese, molybdenum, and chromium is definitely
greater in parboiled than in raw rice after milling (in other words, the losses of these
minerals during milling are reduced, just like those of vitamins and sugars); and 3)
magnesium, zinc, and copper contents after milling are unaffected. Yang and Cho
(1995) reexamined the effect of parboiling and milling on chemical composition.
MECHANISM OF REDUCED LOSS OF CONSTITUENTS DURING MILLING

Two theories have been offered to explain the reduced loss of water-soluble
constituents (B vitamins, sugars, amino acids, and most minerals) during milling of
brown rice after parboiling. The case of thiamin has received the greatest attention
(see Bhattacharya and Ali, 1985, for the numerous earlier references).
The most widely propounded theory is that the water-soluble constituents diffuse
into the endosperm during parboiling. Subba Rao and Bhattacharya (1966) reviewed
these data and felt that the evidence was not conclusive. They observed that the loss
of thiamin during milling remained high after soaking of paddy without gelatini-
zation (i.e., soaking below 70°C), no matter how long the paddy was soaked, but
that the loss dropped sharply following gelatinization, whether that was achieved by
soaking alone or by soaking and steaming (Table 8). The case was identical with
respect to sugars and amino acids (Ali and Bhattacharya, 1980b). The authors
concluded that the reduction in milling loss is not caused by inward migration of the
constituents but by their adhesion to the endosperm caused by gelatinization.
Padua and Juliano (1974), however, found that removal of three successive 10%
outer portions from brown rice removed virtually all the thiamin from raw rice but
only about half from parboiled rice (Table 9). These data strongly suggest that B1
actually diffuses into the endosperm. Benedito de Barber et al (1977) similarly
removed three successive 5% portions from brown rice and observed similar
results.
These two sets of data are difficult to reconcile. If the latter data prove inward
migration, the migration must occur, on the basis of the data in Table 8, during the
brief period of steaming rather than during the long period of soaking. It is not clear
how and why this occurs. Further research in the area is necessary. In addition, the
TABLE 9
Distribution of Thiamin in Successive Milling Fractions of Raw and Parboiled Ricea,b
Raw Rice Parboiled Ricec
Weight Weight
Fraction Thiamin Fraction Thiamin
Milling (percent of Amount Percent of (percent of Amount Percent of
Fraction brown rice) (mg/g) Totald brown rice) (mg/g) Totald
First 11 26.9 82 12 4.6 21
Second 9 6.0 15 9 4.4 15
Third 10 0.60 2 7 3.8 10
Residual rice 70 0.05 1 72 2.0 54
a Source: Padua and Juliano (1974); © Society of Chemical Industry. Permission is granted by John
Wiley & Sons on behalf of the SCI.
b Milling with Satake Laboratory emery mill.
c Soaked at 60°C for 6 hr, steamed at 121°C for 10 min.
d Thiamin content of brown rice: raw, 3.58 µg/g; parboiled, 3.10 µg/g.
Parboiling / 367
theory, whatever it may be, must explain why reduction in milling loss of water
solubles after parboiling is so high in some case (B1, niacin, sugars), moderate in
others (most minerals), and nil in some (riboflavin, magnesium, zinc, and
copper).
Other Changes
The heat treatment has several other effects. First, enzymes present in the grain
are largely inactivated (Barber et al, 1983). Second, grain discoloration occurs due
to Maillard reaction. Third, antioxidants in rice are inactivated (Sowbhagya and
Bhattacharya, 1976), making the rice relatively prone to autoxidation. Fourth, heat-
labile B vitamins are partly destroyed. The effects of these changes are discussed
below.
THE PRODUCT AND ITS CHARACTERISTICS
The above changes in the grain constituents profoundly affect the properties of
rice, so much so that parboiled rice virtually becomes a different grain altogether.
The major effects can be summarized as: 1) the relatively opaque, white, raw rice
grains become rather glassy, translucent, and hard, with a color of light amber; 2)
the flow and packing properties of the rice are modified; 3) the head-rice yield
greatly increases; 4) the B-vitamin content of the rice after milling improves dra-
matically; 5) the oil content of the bran increases considerably; 6) the hydration,
cooking behavior, and eating quality change; and 7) the product-making quality of
the rice is profoundly altered. The extent of these changes, moreover, is related to
the type and severity of the treatment.
Physical Properties
GRAIN CHARACTERISTICS
Milled parboiled rice is slightly (~5%) shorter and broader than milled raw rice
(Bhattacharya and Ali, 1985; Lu and Lii, 1996), possibly due to some realignment
of the cooked grain. Recently, Sowbhagya et al (1993) found that this is true of
regular steam-parboiled rice but not so of dry-heat- and pressure-parboiled rice,
which are slightly longer and thinner, with a higher length-breadth ratio. In addition,
the natural ridges on brown rice are evened out in the steam-parboiling process but
accentuated in the other two processes (due to the quick-dried grain shrinking and
then being frozen in that state). Interestingly, if the dry-heat-parboiled rice is
moistened and held for a while, the grain rearranges itself and the pronounced
ridges disappear (Ali and Bhattacharya, 1980a).
Rice grains, which are rather opaque, become somewhat glassy and translucent
upon parboiling. Any chalky areas in the original rice also become translucent.
Probably, the gelatinized starch granules and disrupted protein bodies adhere to
each other to form a compact mass, reducing light scattering at the boundaries of
the granules (Fig. 19). Hardening of the rice grain is another well-known effect of
parboiling (Raghavendra Rao and Juliano, 1970). The hardness is proportional to
the severity of the heat treatment during processing (Kimura et al, 1976; Pillaiyar
and Mohandoss, 1981a; Islam et al, 2001).
Fig. 19. Photomicrograph of an incompletely parboiled rice endosperm in phase-contrast illumination,

showing a clear difference between the gelatinized outer layer (A) and the ungelatinized inner portion
(i.e., “white belly” area) (B). (Reprinted, with permission, from Raghavendra Rao and Juliano, 1970; ©
American Chemical Society)
Rice is slightly discolored by parboiling, becoming a light yellow to amber. The

discoloration seems to be caused mainly by nonenzymatic browning of the Maillard
type, because it is inhibited by bisulfite (Charlton, 1923; Houston et al, 1956;
Mecham et al, 1961; Jayanarayanan, 1964). The causative factors seem to be the
relatively high level of reducing sugars and amino acids in parboiled rice and the
heat treatment. However, the husk pigment may partly contribute.
The intensity of discoloration is directly proportional to the total heat treat-
ment during soaking and steaming (see Bhattacharya and Ali, 1985, for earlier
references; Kimura et al, 1976, 1993; Lu and Lii, 1996; S. Bhattacharya, 1996).
Soaking at high temperatures seems to show a relatively large effect, perhaps
partly caused by the splitting of the husk and absorption of husk pigment (Fig. 20,
left). Steaming at high pressures also causes severe discoloration (Table 10; Fig.
20, right), all the more so at low grain moisture (see section “pressure-
parboiling”; Ali and Bhattacharya, 1982; Chinnaswamy and Bhattacharya, 1986).
Drying at too high temperatures may intensify the color (Mohandoss and
Pillaiyar, 1982; Velupillai and Verma, 1982). The pH of soaking water has been
reported to affect the grain color (tested in brown rice without milling), the
discoloration being minimal at pH 5 (Jayanarayanan, 1964). This might be partly
because the bran pigment is an acid-base indicator and turns yellow in alkaline
pH (Bhattacharya and Subba Rao, 1966b), whereas an acid pH may cause some
starch hydrolysis and, hence, enhance the Maillard reaction. However, Maillard
reaction is known to progressively intensify at pH 7 and above (Renn and Sathe,
1997; Ajandouz and Puigserver, 1999).
Parboiling / 369
Fig. 20. Left: Effect of temperature and time of soaking of paddy on color of resultant milled parboiled
rice (open-steamed for 10 min in each case). RT = room temperature. (Reprinted, with permission,
from Bhattacharya and Subba Rao, 1966b; ¸ American Chemical Society) Right: Response surface of
color difference of parboiled rice against different pressures and times of steaming. (Reprinted from S.
Bhattacharya, 1996; © Elsevier Science, used with permission)
TABLE 10
Effect of Pressure and Time of Steaming
on Color of Milled Parboiled Ricea
Steamingb
Pressure Time Color Difference
(kpscg)c (min) (DE)
0 2 8.8
5 9.2
10 10.0
20 12.1
40 11.4
60 12.1
0.35 10 12.8
0.70 10 11.9
1.05 10 16.6
a Source: Bhattacharya and Subba Rao (1966b); © American Chemical
Society. Used with permission.
b Paddy soaked at 70°C for 2.5 hr, then steamed at indicated pressure.
c kg/cm2 gauge.
Bhattacharya and Subba Rao (1966b) studied the characteristics of the color in
two cultivars. They found that the “dominant wavelength” of the color surprisingly
remains practically unchanged at ~578 nm in both raw and parboiled rice. But its
“excitation purity” increases and “luminance” (Y) decreases with increasing heat
treatment during soaking and steaming. In other words, although the color becomes
deeper and darker with increasing discoloration, its “hue” remains largely
unchanged. Ali and Bhattacharya (1982) also noted that even the pronounced dis-
coloration of pressure-parboiling, as perceived by the eye, involved mainly a
“darkening” (reduced Y) of the color and not so much a change in its hue. S.
Bhattacharya (1996) recently studied the kinetics of the color development under
different pressures (0–2 kg/cm2) and times (0–60 min) of steaming. She found that
Hunter L (lightness = luminance, Y) decreased more with time than with pressure of
steaming. Hunter b (yellowness), chroma, and DE were affected by both pressure
and time (Fig. 20, right) and all followed a zero order kinetics.
MILLING QUALITY
After vitamin content, the remarkable milling quality of parboiled rice is its sec-
ond most widely observed property. Considerable “reduction” in grain breakage
during milling after parboiling has been repeatedly mentioned by countless authors.
However, what caused the improvement and why were little understood. The situa-
tion was clarified by Bhattacharya (1969, 1980). After rice is parboiled and air dried
in the shade, grain breakage during milling of any sample is close to zero,
irrespective of its breakage before processing. In fact, even paddy deliberately
damaged to give nearly 100% breakage, and separated immature grains that shatter
almost completely during milling, all yield nearly 100% whole grains after par-
boiling (Table 11). Examination in transmitted light shows no trace of cracks,
chalkiness, or cloudiness (characteristic of immature grains) after parboiling, no
matter what their content was before treatment. Clearly, the swelling of the starchy
endosperm during gelatinization completely heals the preexisting defects. As a
result, grain breakage is not reduced but virtually eliminated.
However, parboiled paddy, like raw rice, can itself crack if improperly dried and
will correspondingly break when milled. Incomplete parboiling (“white belly”) also
may cause some breakage (Subrahmanyan et al, 1955; Bhattacharya and Subba Rao,
1966a). In other words, parboiled rice should always give nearly 100% whole grains
irrespective of the variety or preparboiling history, except when the processing,
especially drying, is defective. This may explain why different workers got different
extents of “improvement” in milling after parboiling. Whether hardening of the
grain after parboiling contributes to the improved milling quality, as is generally
claimed, is yet to be experimentally demonstrated.
Some additional changes occur in the milling behavior of parboiled rice. The husk
slightly splits after parboiling; shelling therefore becomes easier, consuming less
energy. This is an advantage. However, the two following changes are disad-
vantageous. Parboiled rice, being harder, requires greater time or force and, therefore,
TABLE 11
Effect of Parboiling on Milling Quality
of Damaged Raw and Parboiled Paddya,b
Breakage (%) on Milling
Paddy Treatment As Is After Parboiling
Raw Nil 35.9 0.7
Oven-driedc 100.0 0.9
Wettedc 71.0 0.8
Parboiled Nil 0.6 …
Oven-dried 99.0 0.4d
Wetted 10.5 0.7d
a Source: Bhattacharya (1969), used with permission.
b After parboiling, paddy was air-dried and milled in a McGill miller to
8–10% degree of milling.
c After oven-drying or wetting, samples were exposed to ambient air to
equalize moisture.
d Indicates reparboiling.
Parboiling / 371
more energy, for whitening (Fig. 21). In addition, bran from parboiled rice contains
more oil and is flakier than raw-rice bran. Hence, the mill screen tends to get clogged.
Also, the milled rice appears oily and sticky, especially if undermilled in metal
pearlers (Raghavendra Rao et al, 1967; Halim and Bhattacharya, 1978). Millers
usually overcome these problems by adding a little husk or chalk to the brown rice
during whitening or by maintaining a high pressure in the cone (Gariboldi, 1984).
OTHER PROPERTIES
One of the effects of the greater oil content on the milled-rice surface (Fig. 14) is
that the flow (angle of repose) and packing (bulk density) properties of rice, par-
ticularly in the range of intermediate milling, are greatly affected by parboiling
(Fig. 22). The poor flow properties of medium-milled parboiled rice is in agreement
with its observed “stickiness,” as mentioned above.
Fig. 21. Increase in degree of milling of raw and parboiled rice with increasing time (left, 0.9-kg weight
used) and weight (right, milled for 5 sec) applied during milling with the McGill miller no. 3.
(Reprinted, with permission, from Raghavendra Rao et al, 1967)
Fig. 22. Physical properties of raw and parboiled rice as affected by degree of milling. D = density, BD =
bulk density, P = porosity, AR = angle of repose. Milling with the McGill miller. (Reprinted, with
permission, from Bhattacharya et al, 1972; © Society of Chemical Industry. Permission is granted by
John Wiley & Sons on behalf of the SCI)
Hygroscopic equilibria of parboiled rice have not been properly studied. Houston
and Kester (1954) found that its equilibrium moisture content was slightly higher
than the literature values for raw rice up to 50% rh, after which the trend reversed.
Simultaneous determination with the two rice types is necessary for proper
comparison.
Qualities of Parboiled Rice Relevant to Consumers
COOKING AND EATING QUALITIES

The third most widely studied property of parboiled rice is its altered cooking
quality. Numerous workers have shown that parboiled rice needs a longer time to
cook than raw rice; however, it retains better shape, is fluffier and less sticky, and
loses less solids into the cooking water (Bhattacharya and Ali, 1985; Islam et al,
2001, 2003). Soaking alone of paddy at temperatures up to 60°C has little influence
on the water uptake of the resultant milled rice during its subsequent cooking, but
the value decreases if the paddy was soaked at 70°C or more, even without
steaming (Table 4). Steaming or heating of soaked paddy, of course, causes a drop
in water uptake, a progressive decline with increasing severity of heating, and a
simultaneous decrease in loss of solids during cooking. Thus, gelatinization is the
original cause; then, as with other properties, the combined heat treatment during
soaking and steaming (also drying: Mohandoss and Pillaiyar, 1982) determines to
what extent the cooking quality is affected.
In parallel with its slower cooking, cooked parboiled rice has been known all
along to consumers to be firmer in texture than cooked raw rice. This has been
experimentally confirmed since the 1970s in numerous texture measurements.
Although definitely slightly colored after being parboiled, the rice becomes virtually
as white as raw rice after cooking.
One peculiarity of parboiled rice is that it expands less in length but more in
breadth during cooking compared with raw rice (Kurien et al, 1964; Mahadevappa
and Desikachar, 1968a; Raghavendra Rao and Juliano, 1970; Park and Cho, 1995).
For this reason, parboiled rice appears more stocky and round after cooking (Fig.
23). In other words, the differences in grain dimensions that already exist in the
uncooked state are further accentuated after cooking. The reasons for this change
are not known.
KEEPING QUALITY
Parboiled rice is resistant to fat hydrolysis because of the inactivation of its
lipase, but it is rather prone to oxidative rancidity (Houston et al, 1954). In a com-
parative study, Sowbhagya and Bhattacharya (1976) noted that the oil in raw rice is
highly stable, while that in parboiled rice is susceptible to oxidation (Fig. 24). Low
grain moistures, exposure to light, increasing degree of milling, high temperatures,
and grinding enhance fat oxidation but, curiously, more so in raw than in parboiled
rice. The destruction of antioxidants as well as the redistribution of fat in the grain
after parboiling were thought to be responsible for its rapid lipid autoxidation.
Tsuzuki and Tanaka (1981) and Kato et al (1983) showed changes in carbonyl
compounds, head space volatiles, and free fatty acids after parboiling. However, the
quicker rancidity does not normally appear to pose a problem to consumers of par-
boiled rice. If rice is stored in woven bags, the darkness delays oxidation and any
odor is quickly dissipated. When the rice is stored in closed containers, particularly
Parboiling / 373
Fig. 23. Raw (R) and mildly (M) and severely (S) parboiled (PB) rice cooked with 2.5 times its weight
of water. The variety was N10B, a scented rice.
Fig. 24. Development of peroxide and carbonyl compounds in raw and parboiled milled rice during
open storage at 60°C in the dark. (Reprinted, with permission, from Sowbhagya and Bhattacharya,
1976; © Institute of Food Technologists)
in transparent packages, the rancidity may become apparent. In modern processes in

the United States and Europe, suitable antioxidants are said to be added sometimes
to the final rice (Gariboldi, 1974).
Parboiled rice is less susceptible to insect infestation than raw rice. Vinacke et al
(1950) found that the number of moths and beetles (of two and six species, respec-
tively) found after storage for four months in a warehouse plus two months in a
laboratory was less in parboiled than in raw rice (2.1 instead of 3.3 adults per 100 g
of rice). Laboratory results were more striking; parboiled rice was virtually immune
to progeny development of several common species of insects, whereas raw rice
was highly susceptible (McGaughey, 1974; Sing, 1980).
Unnikrishnan and Bhattacharya (1995) found that milled parboiled rice ages
virtually just like raw rice. This phenomenon disproves the notion that parboiling
itself is a kind of artificial aging of rice (Sanjiva Rao, 1948). On the other hand,
the fact that rice ages even after parboiling tends to disprove the involvement
of enzymes with the phenomenon of rice aging (Sreenivasan, 1939; Chrastil,
1994).
NUTRITIVE VALUE
The better nutritive value of milled parboiled compared with milled raw rice is
by now a legend. Parboiled rice after milling contains more B vitamins than raw
rice (as described above; Yang and Cho, 1995). The loss of vitamins during storage
of rice (Cailleau et al, 1945; Vinacke et al, 1950; Narayana Rao et al, 1954) and
also during its washing (Subrahmanyan et al, 1938; Vinacke et al, 1950;
Swaminathan, 1941, 1942) also is reduced. Minerals, especially calcium, phospho-
rus, and iron, also are present in greater amounts.
The protein content remains on the whole unchanged, as mentioned before. Kik
(1955, 1965) mentions that the nutritive value of protein is improved, but Eggum
(1979) showed that the digestibility of the protein is reduced and the biological
value is increased, with the result that the total utilizable protein remains
unchanged. On the other hand, parboiling seems to result in a small (lysine) to con-
siderable (tryptophan and methionine) drop in available amino acids (Benedito de
Barber et al, 1977). Tetens et al (1997) suggest that parboiled rice has more resistant
starch.
BRAN QUALITY
Parboiled-rice bran contains more oil than raw-rice bran, which is a clear
advantage. It is also generally stable to development of free fatty acids (FFA)
(Viraktamath and Desikachar, 1971) (Fig. 25), evidently because of the inactivation
of lipase (Bolling and E1-Baya, 1975; Shaheen et al, 1975; Desikachar, 1977).
However, the amount of FFA usually rises somewhat in parboiled bran in the
industry in India, which seems to be caused by the unfortunate practice of under-
drying parboiled paddy, with consequent insect and fungal infestation of the rela-
tively moist bran. Proper drying and fumigation prevent FFA development (C. S.
Viraktamath, CFTRI, Mysore, India, personal communication). Kumaravel et al
(1980) found that parboiled-rice bran stored in high-density polyethylene bags
remained stable, whereas that in jute bags showed insects and high FFA. The origi-
nal FFA in the bran after milling also is lowered by parboiling (Anthoni Raj and
Singaravadivel, 1982).
Parboiling / 375
Fig. 25. Development of free fatty acids (FFA) in bran from raw, soaked, and parboiled rice (soaked at
70°C for 3 hr, open-steamed for 10 min). (Reprinted, with permission, from Viraktamath and
Desikachar, 1971)
ADVANTAGES AND DISADVANTAGES OF PARBOILING

Summarizing, one can say that the following are the advantages and disadvan-
tages of parboiling.
The advantages are that 1) shelling is easier; 2) grain breakage during milling is
dramatically reduced (in fact, parboiling can salvage rain- or drying-damaged
paddy); 3) milled rice contains greater amounts of B vitamins and minerals; 4) loss
of nutrients during washing is reduced; 5) grains remain integral and do not mash
after cooking, and the loss of solids in cooking water is reduced; 6) insect infesta-
tion and loss of nutrients during storage of rice are reduced; 7) the bran contains
more oil, which is relatively stable to FFA development; and 8) parboiled rice is
suitable for making three rice products (canned, expanded, and flaked rice) for
which raw rice is not suitable (see the section on products).
However, parboiled rice has certain disadvantages: 1) brown rice needs more
time and energy to whiten; 2) the mill screen tends to get choked during its whit-
ening; 3) rice undermilled in metal pearlers looks oily and unattractive; 4) the flow
and packing properties of undermilled rice are poorer; 5) the rice needs more time
and, hence, more energy to cook; 6) its harder texture after cooking is not liked by
many consumers; and 7) it is more prone to oxidative rancidity.
FUNDAMENTAL NATURE OF PARBOILED RICE
Parboiled rice has many unusual physicochemical properties. Moreover, these

properties differ somewhat from steam- to dry-heat- to pressure-parboiled rice.
These properties initially came to light during graduate work conducted at CFTRI in
Mysore, India, during the 1970s (Ali, 1979). It is these properties that initially
raised the question about the fundamental nature of parboiled rice as an archetype of
precooked cereal foods.
Physicochemical Properties
The most striking, and perhaps symbolic, of all its unusual properties is the
double paradox that parboiled rice (especially conventional steam-parboiled rice),
while being precooked, hydrates at a lower rate than raw rice during cooking at high
temperatures, and yet hydrates faster and to a greater extent than the latter at low
temperatures. Indeed, this contrariness increases with increasing severity of heat
treatment during parboiling (Fig. 26). Similarly, the equilibrium moisture content
attained by rice when soaked in water at room temperature (EMCS) is greater in
parboiled than in raw rice (Fig. 27). Solubility of amylose from rice flour (Ali and
Bhattacharya, 1972a) and swelling and solubility of rice flour (Schoch, 1964;
Unnikrishnan and Bhattacharya, 1981) also show the same trend. In keeping with
these contrary behaviors, a suspension of parboiled rice flour in water shows a
Fig. 26. Hydration of raw (R) and parboiled (PB) rice in water at different temperatures for different
times. M = mildly and S = severely parboiled. (Reprinted, with permission, from Ali and Bhattacharya,
1972a)
Parboiling / 377
higher viscosity at room temperature (Unnikrishnan and Bhattacharya, 1983)—as

well as a higher sediment volume (Bhattacharya and Ali, 1976)—but a lower paste
viscosity in the Brabender viscograph (Raghavendra Rao and Juliano, 1970; Arai et
al, 1975; Priestley, 1976a; Ali and Bhattacharya, 1980c; Islam et al, 2001) (Fig. 28)
than raw rice flour. Dispersion of flour in dilute alkali or dimethyl sulfoxide simi-
Fig. 27. Hydration of raw and mildly parboiled rice in water at room temperature. (Reprinted, with
permission, from Indudhara Swamy et al, 1971)
Fig. 28. Brabender viscograms of raw and parboiled (PB) rice (9.6% slurry, dry basis). VM = very
mildly, M = mildly, S = severely parboiled. (Adapted from Ali and Bhattacharya, 1980c; used by
permission)
larly shows a greater gel volume for parboiled than raw rice at ambient tempera-
tures (Pillaiyar, 1984a,b, 1985). Alkali also has a similar contrary effect. Parboiled
rice grains are attacked by alkali so dilute that it does not affect the raw grains at all.
However, in moderate and concentrated alkali, parboiled rice grains seem to be
slightly less affected than the raw grains (Fig. 29). Unnikrishnan and Bhattacharya
(1988) found that parboiled rice flour shows a greater alkaline gel mobility than raw
rice flour. All these contrary properties, it should be reiterated, are accentuated by
increasing severity of processing. A few typical examples of some of these
paradoxical properties are shown in Table 12.
The properties of dry-heat-parboiled rice (prepared by conduction heating of
soaked paddy with hot sand rather than by steaming: Fig. 9, process 5), are some-
what different. It shows a higher EMCS and slurry viscosity than steam-parboiled
rice (Ali and Bhattacharya, 1976a; Unnikrishnan and Bhattacharya, 1981, 1983).
However, if this dry-heat-parboiled rice is moistened and tempered, the properties
tend to revert toward those of conventional parboiled rice (Table 12).
Properties of pressure-parboiled rice, as mentioned before (in the section on
production systems), are again very different. In an effort to understand its strong
cooking resistance, Ali and Bhattacharya (1982) made a detailed study of the effect
of a whole range of grain moistures and times and temperatures of steaming. The
data provided a mine of information (Fig. 30). Grain translucence and EMCS
determined immediately after processing (I-EMCS), both indexes of gelatinization,
Fig. 29. Degradation of raw (R) and very mildly (V), mildly (M), and severely (S) parboiled rice grains
in very dilute (1.0%), dilute (1.5%), and moderate (2.0%) aqueous KOH. (Reprinted, with permission,
from Ali and Bhattacharya, 1972b)
Parboiling / 379
Fig. 30. Effect of paddy moisture, steam pressure, and steaming time on properties of resulting parboiled
rice. Steaming time: a, b, c, and d = progressively increasing time. The symbols, depending on the degree
of shading, represent the lowest degree (open square) to the highest degree (closed square) of each property
within the range indicated. EMCS = equilibrium moisture content attained by rice when soaked in water at
room temperature; I = immediate and F = final. W96 = water uptake when cooked at 96°C for 1 hr.
(Adapted from Ali and Bhattacharya, 1982; Chinnaswamy and Bhattacharya, 1986. Used with permission)
increased not only with increasing moisture but also with increasing steam pressure
even at low grain moisture—which explained the rationale of the pressure-parboil-
ing process. Retrogradation (i.e., a drop in EMCS determined in final dried and
milled rice [F-EMCS] compared with I-EMCS) was high in high-moisture and low
in low-moisture samples, as expected from the properties of steam- and dry-heat-
parboiled rice discussed above. Surprisingly, reduction in high-temperature hydra-
tion (i.e., resistance to cooking) was not necessarily related to retrogradation.
Indeed, it was maximum in intermediate-moisture samples (Fig. 30), even though
these had rather high F-EMCS, indicating low retrogradation (i.e., precisely in the
optimum pressure-parboiled samples, which explained their “iron-rice” appella-
tion). These results were confirmed later in a related work (Chinnaswamy and
Bhattacharya, 1986).
State of Starch in Parboiled Rice
The peculiar physicochemical properties of various kinds of parboiled rice dis-

cussed above suggest two changes in the state of starch as a result of processing: 1)
gelatinization and 2) some form or forms of reassociation and complexation.
Parboiling / 381
Fig. 31. Fall in equilibrium moisture content attained by rice when soaked in water at room
temperature (EMCS) of just-steamed parboiled (PB) paddy with time of air-drying (left) and with
holding at various temperatures (right). VM = very mildly, M = mildly, S = severely parboiled, RT =
room temperature. (Reprinted, with permission, from Ali and Bhattacharya, 1976b)
Starch gelatinization is evident from the fact that all characteristic properties of
parboiled rice are associated with steaming or heating of soaked paddy or else with
its mere soaking if carried out at a sufficiently high temperature (Table 4). This was
confirmed by Raghavendra Rao and Juliano (1970) from the loss of A-type X-ray
diffraction and of granule birefringence after parboiling. The low-temperature
hydration and solubility properties, alkali solubility, slurry viscosity, and sediment
volume of parboiled rice and flour discussed above follow from this starch gelatini-
zation.
On the other hand, the converse properties of reduced high-temperature hydra-
tion and solubility and the firmer texture of cooked rice are indications of certain
starch reassociation or complexation. Ali and Bhattacharya (1972a,b, 1976a,b), who
first observed these contrary properties of parboiled rice, assumed that starch was
retrograded after cooking during its relatively slow drying, and that this caused the
cooking resistance. Retrogradation was clearly demonstrated when these authors
found that the hydration power of normal steam-parboiled rice as expressed in
EMCS was unusually high immediately after steaming (I-EMCS), and then
gradually fell during air drying or holding to the normally encountered final values
(F-EMCS) (Fig. 31). Ali and Bhattacharya (1976b) found that the extent of reasso-
ciation depended on both the moisture content and the storage temperature. Retro-
gradation was maximum at a moisture content of ~25% at room temperature, at
over 30% moisture at 8°C, and at 20% moisture or less at 50°C or more.
Doubt concerning the above hypothesis arose from X-ray and related studies.
Charbonniere (1975), who had studied the X-ray diffraction of rice products, observed
that, although raw rice gave a clear A-type X-ray pattern, parboiled rice gave a V-type
pattern of helical amylose (i.e., lipid-amylose complexation). Canned rice, but not par-
boiled rice, gave a clear B-type pattern of retrograded starch, which disappeared upon
heating (Fig. 32). Priestley (1976a,b) also observed a helical lipid-amylose type V-
pattern in parboiled rice. Mahanta et al (1989) examined this matter in a wide range of
samples with both X-ray and differential scanning calorimetry (DSC) methods. They,
too, detected a lipid-amylose complex, but its diffraction intensity as well as enthalpy
were low. The presence of a B X-ray pattern could neither be confirmed nor refuted
from their data. In any case, the overall degree of crystallinity seemed to be too low.
Further doubts arose because starch retrogradation alone could not explain the
cooking resistance of parboiled rice or its firmer texture. When a portion of con-
ventional steam-parboiled rice was rapidly dried with very hot air and another por-
tion was air dried in the normal way, rice milled from the former had a higher
Fig. 32. X-ray diffractograms of raw rice, parboiled rice, amylose-butanol complex, stored canned rice,
and potato starch. (Reprinted, with permission, from Charbonniere, 1975)
Parboiling / 383
EMCS (lower retrogradation) than that from the latter, as expected. Yet the two
samples showed no difference in their cooking rate in boiling water (Table 13).
Similarly, the high EMCS of dry-heat-parboiled rice dropped appreciably when the
rice was moistened and tempered to promote starch retrogradation (Table 12). Yet
this induced retrogradation reduced its rate of water absorption in boiling water only
marginally (Mahanta, 1988). The same phenomenon is seen in pressure-parboiled
rice (Fig. 30): samples showing the highest retrogradation index (drop from I-
EMCS to F-EMCS) did not necessarily have the highest cooking resistance index,
and vice versa. Clearly, the low- and high-temperature hydration properties of par-
boiled rice seem to be governed independently.
The confusion in this matter now has been substantially resolved by recent
studies of starch polymorphism occurring in starchy foods under different condi-
tions of heat treatment (Biliaderis et al, 1986; Mesters et al, 1988; Biliaderis and
Galloway, 1989; Biliaderis, 1992).
Several points have emerged.
Moisture. The moisture content of starch has a pronounced effect on its DSC
endotherms; sharp endotherms are observed or formed only at 60% water content
by volume (~50% moisture or more by weight); the endotherms are broadened as
well as pushed to higher temperatures at lower moistures as a result of the compos-
ite effect of partial melting, annealing, and crystallization.
Retrograded Starch. Retrograded starch, especially retrograded amylopectin (R-
Ap), does not necessarily show a B X-ray pattern, especially at low moistures.
Type of Complex. Lipid-amylose complexes (L-Am) are of two types. One (L-Am I)
is formed at relatively low temperatures and the other (L-Am II) at high temperatures.
L-Am may have a low degree of crystallization intrinsically, yet have a marked effect.
Type of Endotherm. Six types of endotherms involving starch potentially exist: 1)
normal-gelatinization endotherm (melting temperature, Tm = 65–80°C) in uncooked
starch, 2) annealed starch endotherm formed upon heating at low moistures (Tm =
80–85°C), 3) R-Ap endotherm (Tm = 50–55°C), 4) L-Am I endotherm (Tm =
~100°C), 5) L-Am II endotherm (Tm = 110–120°C), and 6) retrograded amylose
endotherm (R-Am) (Tm = 140–150°C).
Biliaderis et al (1993) recently studied the status of starch in seven rice varieties
(one waxy and six nonwaxy) having different amylose and GT combinations and
parboiled at 100, 120, and 127°C. All parboiled samples showed the presence of R-
TABLE 13
Effect of Slow (Shade) and Quick (Very Hot-Air) Drying
on Hydration Properties of Parboiled Ricea,b
Steaming EMCSd (% wb) After Water Uptake (g/g) After
Pressure Time Shade Hot-Air Shade Hot-Air
(kpscg)c (min) Drying Drying Drying Drying
1.0 20 39.0 45.6 3.89 3.93
1.5 20 48.4 53.3 3.41 3.43
2.0 10 50.0 55.8 3.61 3.62
15 57.3 58.5 3.41 3.40
a Source: Ali and Bhattacharya (1982); used with permission.
b Paddy soaked in ambient water for 40 min and steamed as indicated.
c kg/cm2 gauge.
d Equilibrium moisture content attained by rice when soaked in water at room temperature.
Ap. In addition, low-GT samples gave L-Am I in 100°C-parboiled samples but L-

Am II when parboiled at 120–127°C. High-GT samples, on the other hand, gave L-
Am II and also some annealed starch and R-Am depending on the circumstances.
Ong and Blanshard (1994, 1995) did similar studies of several samples and showed
somewhat similar complexes. However, their data are rather difficult to interpret
because the samples and the processing conditions are not identified.
These results provide a fair explanation of most cooking and related properties of
parboiled rice. For example, conventional parboiled rice produced by atmospheric-
pressure steaming would likely have R-Ap and L-Am I. As a result it would have a
relatively low EMCS because of the staled amylopectin (R-Ap), but also a moderate
but not very high resistance to cooking because the L-Am I would be amenable to
slow melting at boiling temperature. On the other hand, parboiled rice produced by
high steam pressures (especially the pressure-parboiled rice, with its low moisture and
high-temperature processing) would likely have annealed starch as well as L-Am II
with or without R-Ap (depending on the grain moisture). L-Am II would not melt at
boiling temperature; therefore, the difficulty of its cooking would be explained. The
same data also would explain why certain samples, even though showing high EMCS
(caused by the absence of R-Ap due to low end moisture; for example, in dry-heat- or
pressure-parboiled rice), may be relatively easy or hard to cook, depending on whether
the conditions favor L-Am I or L-Am II. Similarly, although the method of moistening
and tempering causes a precipitate drop in EMCS in many samples (caused by de
novo formation of R-Ap), it may have hardly any effect on the water uptake at boiling
temperature, which is controlled by the presence or absence of and type of L-Am.
Finally, the remaining factor to consider is the effect of thermal breakdown of
starch during processing. As mentioned in the section on changes in constituents,
Mahanta and Bhattacharya (1989) observed clear evidence of thermal degradation of
starch under high steam pressures. However, its effect on rice properties is not clear.
Unnikrishnan and Bhattacharya (1987b, 1989) noted that, when parboiled rice was
prepared under increasing steam pressure, there was a hint that the initial trend of
decreasing parboil-canning solids loss with increasing steam pressure tended to be
reversed at very high pressures (Table 14). Mahanta (1988), working with an
intermediate-amylose variety (Intan), confirmed this. She observed not simply a
tendency but a clear reversal of high-temperature water uptake, cooked-rice firmness,
and canning solids loss after treatment at high steam pressures (2–3 kg/cm2) (Fig. 33).
She speculated that this peculiar reversal might be related to the thermal breakdown of
starch at the high steam pressures observed (Mahanta and Bhattacharya, 1989); in
other words, that starch breakdown might increase water uptake and solids loss during
cooking. From a consideration of these two results, there is a hint that thermal
degradation of starch at high steam pressures during parboiling might be inversely
related to the amylose content of the variety. If that is the case, the thermal breakdown
of starch may play a big role in the cooking behavior of parboiled rice, especially in
low-amylose varieties after high-pressure steaming.
Effect of Rice Variety on Properties of Parboiled Rice
Varietal differences in the starting rice affect parboiled-rice quality. This aspect
has been brought to light by relatively recent studies. The earlier work in this
respect was in relation to suitability for making canned rice, discussed in the section
on products below.
Parboiling / 385
Early workers (see Bhattacharya and Ali, 1985, for references) noted certain dif-
ferences in textural and other properties of parboiled rice made from different
varieties. Unnikrishnan and Bhattacharya (1987a) made a careful study of this
aspect with two waxy and 11 nonwaxy varieties of a wide range of quality types.
They observed that all varieties showed the usual phenomena of contrary low- and
high-temperature hydration, solubility, and solids loss as a result of parboiling.
What is more, the extent of the above changes compared with the respective raw
TABLE 14
Cooked-Rice Firmness and Parboil-Canning Solids Loss
from Conventional and Pressure-Parboiled Ricea,b
Rice
Quality Amylose (% db)

Typec Total Insoluble MPB SPB PPB1 PPB2 PPB3
Firmness (%)
I 29.2 18.0 64.0 69.2 72.1 72.4 74.7
III 29.9 11.1 42.1 52.3 65.2 66.0 68.2
V 24.3 9.6 37.4 46.5 57.0 58.8 61.4
VI 24.6 8.4 32.5 39.4 51.4 55.8 59.5
VII 19.4 6.6 26.8 29.7 28.2 29.8 39.1
Canning Solids Loss (%, db)

I 29.2 18.0 17.2 12.2 10.2 11.3 15.1
III 29.9 11.1 25.4 17.4 13.4 15.2 18.0
V 24.3 9.6 30.3 27.1 15.4 17.0 21.9
VI 24.6 8.4 33.4 28.3 17.6 18.5 24.3
VII 19.4 6.6 36.4 33.7 26.1 27.5 32.9
a Source: Unnikrishnan and Bhattacharya (1987b); used with permission.
b Parboiled samples are often identified by a three-number code: a) the first number indicates the
nominal grain moisture (% wb), b) the second, the steaming pressure (kg/cm2 gauge = kpscg), and c)
the third, the steaming time (min). Mildly parboiled (MPB) = 30:0:10, severely parboiled (SPB) =
30:1:10, pressure-parboiled (PPB1) = 17-2.5-15, PPB2 = 12:3:15, PPB3 = 12:3:30.
c Rice quality type is based primarily on total and insoluble amylose contents (see Chapter 15 in this
volume).
Fig. 33. Effect of steam pressure during parboiling (steamed for 20 min each at the gauge pressure
indicated) on 96°C water uptake (A), cooked-rice Viscoelastograph firmness (B), and parboiled-
canning solids loss (C). The numerals along the curves indicate the initial paddy moisture. Symbols
enclosed in circles indicate white-bellied samples. (Courtesy: Mahanta, 1988)
rice remained largely constant in all varieties (Table 15). This applied to rice-flour
slurry viscosity and pasting behavior as well. In other words, the differences in
properties existing among rice varieties were carried over intact after parboiling.
The situation is the same with respect to the texture of cooked rice. Arai et al
(1975) and Kimura (1983) observed that the stickiness and tenderness of cooked
rice decreased after parboiling; but japonica varieties continued to remain stickier
than Indica varieties even in the processed form. Unnikrishnan and Bhattacharya
(1987a,b) noted that, in their 11 types of nonwaxy rice, sensory stickiness, tender-
ness, and moistness scores decreased and viscoelastograph firmness and elastic
recovery values increased from raw to mild to severely parboiled rice. However, the
intervarietal gradation among the original rice types again were maintained (Table
16, firmness data). This also was true of pressure-parboiled rice. The sensory and
instrumental textural values were highly correlated with both total and insoluble
amylose contents as much before as after parboiling.
Biswas and Juliano (1988) studied 12 varieties (one waxy and several pairs of
different amylose-GT combinations) after steam-parboiling at 100–131°C. Low-
GT varieties of paddy imbibed a greater amount of moisture after soaking (this is
true if the temperature is not adjusted down) and, accordingly, had a higher
degree of gelatinization upon steaming at 100°C. Otherwise, the amylose content
was the major factor in EMCS, gel mobility, and cooked-rice hardness and
stickiness.
Clearly, the varietal effect on properties of parboiled rice more or less mirrors
that in raw rice (discussed in a separate chapter in this volume). This is fortunate,
because it implies that 1) one can predict the quality of the processed rice by
knowing that of the feedstock, 2) one can choose the variety to be processed
depending on the desired end-product, and 3) no special test other than normal
quality tests are necessary for screening rice varieties for suitability for parboiling.
Two additional conclusions may be that, if a low-amylose rice is used for process-
ing, 1) the resulting parboiled rice may become acceptable to those who use high-
amylose raw rice (as in South Asia), and 2) pressure-parboiled rice could be
acceptable and the process could be revived in view of its several other advantages
(Tables 5 and 14).
Tests for Parboiled Rice
Originally, no criteria other than grain translucence and color identified or

characterized parboiled rice. Grain translucence differentiates parboiled rice from
raw rice except when paddy is merely soaked in water at 50–70°C, when
translucence appears without gelatinization (Unnikrishnan et al, 1982). Grain
color gives a fair idea of the degree of parboiling, because discoloration increases
with increasing severity of heat treatment. Clarification of its physicochemical
properties during the last decades now has provided several elegant tests of
parboiled rice. The possible tests, mainly applicable to conventional steam-
parboiled rice, are summarized in Table 17. Among these, the alkali reaction test,
although only semiquantitative, is the simplest and possibly the most elegant. The
alkaline gel test is probably the simplest quantitative test. EMCS, except for its
time requirement, is another exceedingly simple quantitative test. The latest test
being explored is the one using near-infrared radiation—but it has a long way to
go yet.
Parboiling / 387
TABLE 16
Texture and Canning Stability of Cooked Raw and Parboiled Ricea
Canning Solids Loss (%)
Firmnessc (%) MPB/Raw SPB/Raw
Varietyb Ratio Ratio
No. Raw MPBd SPBd Raw MPB SPB (%) (%)
11 54.2 60.1 65.0 22.3 16.3 10.6 73.1 47.5
12 62.0 65.3 70.0 20.2 14.3 10.1 70.8 50.0
21 49.4 55.4 60.2 22.6 17.1 11.6 75.7 51.3

22 45.5 52.6 59.4 26.2 19.6 14.0 74.8 53.4
31 38.3 45.8 54.9 32.6 24.8 18.5 76.1 56.7

32 36.6 44.2 54.2 30.9 23.0 16.5 74.4 53.4
41 29.1 39.2 50.4 36.4 28.4 23.4 78.0 64.3

42 27.7 42.5 52.7 35.1 28.2 22.7 80.3 64.7
51 24.0 36.5 47.1 34.6 28.8 25.9 83.2 74.8
71 22.2 28.3 37.2 38.5 34.5 33.0 89.6 85.7

72 20.3 24.6 35.6 42.3 36.9 36.1 87.2 85.3
81 14.1 14.8 … 55.0 44.8 … 81.5 …

82 14.0 15.2 … 51.9 45.3 … 87.3 …
a Source: Unnikrishnan and Bhattacharya (1987a); used with permission.
b The first digit of the variety code identifies the quality type of the rice. Thus, 21 indicates that the
variety belongs to quality type II, 72 = type VII and so on. The subsequent digit stands for the serial
no. within the type. The rice quality types are characterized by the amounts of total and insoluble
amylose contents, which are highest in type I and least in type VIII (waxy). (See Chapter 15 in this
volume.)
c Chopin-INRA viscoelastograph.
d Parboiled samples are often identified by a three-number code: a) the first number indicates the
nominal grain moisture (% wb), b) the second, the steaming pressure (kg/cm2 gauge = kpscg), and c)
the third, the steaming time (min). Mildly parboiled (MPB) = 30:0:10, severely parboiled (SPB) =
30:1:10.
Tests also are needed to identify 1) the temperature of soaking, 2) the type of
parboiling (conventional versus dry-heat versus pressure-parboiling), 3) the degree
of gelatinization, 4) the extent of amylopectin retrogradation, 5) type and extent of
lipid-amylose complexation, and 6) the degree of starch breakdown. Some progress
in these objectives has been made.
TEMPERATURE OF SOAKING
The temperature at which the paddy was soaked for parboiling can be
determined roughly on the basis of an unpublished work of the author carried out
decades ago. When milled parboiled rice was placed in a corked test tube and
heated in an oven, the degree of induced discoloration (light to extremely dark
amber) was related, apart from the temperature of heating, to the temperature at
which the paddy had been soaked during parboiling (Table 18). The heat-induced
discoloration was negligible after ambient soaking, highest after soaking at ~60°C,
and again negligible after 80°C soaking. Evidently, this was related to the enzymic
release of reducing sugar during soaking discussed earlier (Fig. 13).
Parboiling / 389
TABLE 17
Tests for Parboiled Rice
Property Principle of Test Referencesa
Water-uptake ratio Rice is cooked in water at a)50–60Cb Cand b) boiling (1), (2)
temperature for a definite time. The ratio of the two water-
uptake values, expressed as a percent, goes up from about
5% for raw rice to 50% or more for severely parboiled rice.
EMCSc Rice is soaked in ambient water overnight. The equilibrium (1), (2)
moisture content attained goes up from about 28% (wb) for
raw rice to 50% or more for severely parboiled rice.
Soluble-amylose ratio Rice flour is extracted with water at a) 40–50Cb and b) boiling (1), (2)
temperature. The ratio of the dissolved amylose, expressed
as a percent, goes up from about 2% for raw to about 15%
or more for severely parboiled rice.
Alkali degradation Rice kernels are put in 0.5–1.0% KOH. Extent of kernel (3), (4)
degradation increases with increasing degree of parboiling.
Slurry viscosity Rice flour is made into a slurry with water (20%, db). (2), (5)
Viscosity increases with increasing degree of parboiling.
Sediment volume A 2% slurry of rice flour in 0.05N HCl is allowed to stand in a (6)
measuring cylinder. Sediment volume increases with
parboiling.
Alkaline gel length 100 mg of flour is shaken with 4 mL of 1.25% KOH and (7)
centrifuged (or allowed to stand for 30 min). The gel length
increases with increasing severity of parboiling.
DMSOd gel length Similar to alkaline gel test. (8)
Gel mobility 90–130 mg of rice flour is dispersed in 2.0 mL of 0.2N KOH (2), (9)
in a 13- × 150-mm test tube. The length (mm) of gel flow
after 1 hr, when the tube is laid horizontally, increases with
severity of parboiling.
Canning stability 5 g of rice is cooked under pressure in excess water and (2), (10)
strained. Loss of solids decreases with severity of
parboiling.
NIRe The near-infrared spectrum is obtained and examined for its (11), (12)
defining characteristics.
a References: (1) Ali and Bhattacharya (1972a), (2) Unnikrishnan and Bhattacharya (1987a), (3) Ali
and Bhattacharya (1972b), (4) Mohandas and Pillaiyar (1980), (5) Unnikrishnan and Bhattacharya
(1983), (6) Bhattacharya and Ali (1976), (7) Pillaiyar (1984a, 1984b), (8) Pillaiyar (1985), (9)
Unnikrishnan and Bhattacharya (1988), (10) Webb and Adair (1970), (11) Kimura et al (1995), (12)
Shimizu et al (2001).
b The temperature is adjusted depending on the gelatinization temperature of the variety.
c Equilibrium moisture content attained by rice when soaked in ambient water.
d Dimethyl sulfoxide.
e Near-infrared.
TYPE OF PARBOILING
As for identifying the type of parboiling, Mahanta and Bhattacharya (1995)
made an interesting observation while doing the alkali test. Conventional
parboiled rice produced with atmospheric steam pressure, dry-heat-parboiled rice,
and pressure-parboiled rice that had a chalky center all retained a chalky central
grain mass after alkali digestion. On the other hand, fully translucent pressure-
parboiled rice as well as conventional parboiled rice produced by steaming at 1
kg/cm2 gauge pressure gave a virtually transparent central mass (Fig. 34). This
result coupled with the grain color (pressure-parboiled rice is highly discolored)
would easily enable the researcher to distinguish pressure-parboiled rices with
and without a chalky center as well as to distinguish conventional parboiled rice
TABLE 18
Effect of the Temperature of Soaking of Paddy on the Discoloration
of Resulting Milled Parboiled Rice Upon Reheatinga,b
Approximate Colorc of Parboiled Rice
Soaking Temperature
(°C) Before Heating After Heating
RTd ± +
50 ± 3+
60 ± 5+
70 ½+ 3+
80 + 1½+
a K. R. Bhattacharya, unpublished.
b Milled rice placed in a closed tube and heated in an oven at 90°C for 18 hr.
c The number of plus signs shows the approximate relative color. ± =
faint.
d Room temperature (23–27°C).
produced under atmospheric pressure from that produced under high steam
pressure. Finally, dry-heat-parboiled grains also stood out by being invariably
longitudinally split after alkali treatment, with the two split halves moving
substantially apart. In a similar test, Pillaiyar et al (1997) observed that, when a
few grains of parboiled rice are heated with 1 mL of 5% KOH in a boiling water
bath in a tube for 4 min, pressure-parboiled grains are converted into a lumpy gel,
whereas all other grains remain chalky. Another possible test for dry-heat- and
pressure-parboiled rice is their EMCS values before and after treatment with
moisture. These products have a high EMCS as produced, but their EMCS value
drops precipitately upon moistening and tempering due to induced amylopectin
staling (Fig. 30, Table 12).
GELATINIZATION
Birch and Priestley (1973) suggested that the degree of gelatinization of starch
could be determined from the ratio of amylose dissolved in 0.2N KOH to that dis-
solved in 0.5N KOH. Unnikrishnan (1986) and Mahanta (1988), working in the
author’s laboratory, confirmed that the ratio increased with increasing severity of
parboiling irrespective of the type of processing. What is more, amylopectin staling
(R-Ap) induced in the samples by moistening and tempering had little or no effect
on the property (Bhattacharya and Ali, 1985). Therefore, this ratio is a good index
of the degree of gelatinization regardless of any amylopectin staling and, pre-
sumably, any lipid-amylose complexation.
RETROGRADATION
Mahanta (1988) also noted that the I-EMCS of a sample showed a very high cor-
relation with the above alkali-soluble amylose ratio (r = 0.933, n = 34). In other
words, the I-EMCS also is an excellent index of the degree of gelatinization. Now,
if the F-EMCS of the sample as determined after drying and milling is compared
with the I-EMCS, then the percentage drop in F-EMCS of the finished product
compared with its I-EMCS before drying gives a clear indication of its amylopectin
staling. What is more, if the milled sample is now moistened and tempered, the
further fall in EMCS would provide additional information about the state of amylo-
pectin staling in the sample.
Parboiling / 391
Fig. 34. Degradation of various parboiled rice kernels in 0.9% KOH after 24 hr. Top, steam-parboiled
rice. Code: first number = nominal grain moisture (%), second number = steaming pressure (kg/cm2
gauge = kpscg), and third number = steaming time (min). Bottom, dry-heat-parboiled rice. Code: first
number = roasting temp (°C), second number = time (min), and third number = final grain moisture
(%). The rice in the bottom left dish (30-0-10) was steam-parboiled and included for comparison.
(Reprinted, with permission, from Mahanta and Bhattacharya, 1995)
STARCH COMPLEXATION
Unfortunately, as of now, the degree of resistance to cooking in boiling water is
the only approximate index of the extent and nature of lipid-amylose complexation
(apart from DSC).
STARCH BREAKDOWN
As for the thermal breakdown of starch, no simple test exists at the moment
(only gel chromatography). However, starch breakdown may cause enhanced water
absorption and canning solids loss during cooking (Table 14, Fig. 33), which may
provide some indirect indication.
PRODUCTS FROM PARBOILED RICE
Parboiled rice yields at least three products for which raw rice is not suitable or
not optimal.
Canned Rice
Canned rice produced before the advent of parboiling was largely unacceptable.
Parboiling provides a high degree of grain stability during heat processing, which is
essential in this product (Burns and Gerdes, 1985). Solids loss during processing
also is reduced.
The variety has a strong influence on the quality of canned rice. The earliest
work in this area was that of Webb et al (1968), who screened 4,000 rice varieties in
the United States Department of Agriculture’s world collection. Parboil-canning
stability, expressed as loss of solids when rice is heated with excess water under
pressure (Webb and Adair, 1970), correlated very well with amylose content,
starch-iodine-blue value, alkali reaction value, kernel width, and protein content in
2,322 nonwaxy varieties. Feillet and Alary (1975) and Alary et al (1977) found with
49 French varieties that the firmness of canned rice was proportional to amylose
content. Unnikrishnan and Bhattacharya (1987a,b) showed that variety as well as
the processing type and conditions affect canned-rice quality (Tables 14 and 16);
pressure-parboiling, especially of high-amylose rice, yielded the best base for can-
ning.
Puffed Rice
Puffed rice is produced by HTST heating of milled parboiled rice in sand, oil, or
air. It is a traditional product widely consumed in many Asian countries from
ancient times. The degree of puffing is related to the type and degree of parboiling
as well as to the variety.
Roberts et al (1954) found that the puffing expansion of a U.S. long-grain rice
increased with increase in the temperature of steaming during its parboiling up to
130°C. This was confirmed by Chinnaswamy and Bhattacharya (1983b) with an
intermediate-amylose variety. However, they found that dry-heat-parboiled rice
expands more than steam-parboiled rice, but the expansion drops when the rice is
moistened to induce retrogradation before puffing. Subsequently, the same authors
(1986) found that certain pressure-parboiled rice, but not all, gives still higher
expansion (Fig. 30). From this, they devised a new processing system for preparing
the base material for making puffed rice.
Variety (amylose content) also has a profound effect on the degree of puffing.
Antonio and Juliano (1973) thought that the extent of expansion was inversely
related to the amylose content, which is highest in waxy rice. Villarreal and Juliano
(1987) again found that waxy rice puffed best, but intermediate-amylose rice was as
Parboiling / 393
good. However, in the latter work, paddy was soaked at 45–50°C for 6 hr, which
would be insufficient for hydration.
Chinnaswamy and Bhattacharya (1983a), studying 35 varieties with the full
range of amylose contents, and after parboiling by steaming at 1.5 kg/cm2 gauge
pressure, found that the highest puffing is given by varieties having 27.5% total
amylose and 13.5% hot-water-insoluble amylose (Fig. 35). Chandrasekhar and
Chattopadhyay (1991) more or less confirmed this finding; they also found a small
positive correlation with length-breadth ratio and a negative one with protein.
Chinnaswamy and Bhattacharya (1984) later found that the processing and varietal
variables in seven carefully selected varieties are well interrelated. An optimum
steam pressure for parboiling exists for each amylose class of rice. These conditions
are 0 kg/cm2 gauge pressure of steaming for 0% amylose (waxy rice), 0.5 kg for
21% amylose, 1 kg for 25% amylose, 2 kg for 28% amylose, and 3 kg for 29%
amylose. However, overall, the 27.5% total and 13.5% insoluble-amylose variety
gives the highest expansion (Fig. 36).
Chandrasekhar and Chattopadhyay (1990) noted, by scanning electron micros-
copy, that milled parboiled rice has a ~0.1-mm-thick peripheral layer in which
swollen starch granules are embedded in a protein matrix (Fig. 37). They surmised
that this layer holds the pressure that allows the grain to puff by HTST heating.
The puffing ability of parboiled rice is clearly a consequence of its gelatinization,
because raw rice expands but little when similarly heated. (However, raw rice
expands very well by gun puffing—indeed much more than parboiled rice does by
sand heating [Villareal and Juliano, 1987]. The underlying basis is clearly differ-
ent.) The observations that dry-heat-parboiled rice puffs better than steam-parboiled
rice, that this superiority is negated by induced retrogradation, and that pressure-
parboiled rice puffs best would suggest that amylopectin staling hampers puffing
but that lipid-amylose complex type II perhaps helps it. The fact that different
amylose classes of rice have different steam requirements might suggest some
negative influence of thermal breakdown of starch on puffing.
Fig. 35. Relation between amylose content (total and hot-water insoluble) and expansion ratio of
parboiled rice (paddy parboiled by steaming at 1.5 kg/cm2 gauge) puffed by high-temperature short-
time heating. (Adapted from Chinnaswamy and Bhattacharya, 1983a; used by permission)
Fig. 36. Interdependence of expansion during high-temperature short-time puffing of parboiled rice,
amylose content of rice, and steam pressures (kg/cm2 gauge, indicated) used for parboiling. (Reprinted,
with permission, from Chinnaswamy and Bhattacharya, 1984)
Flaked Rice
Flaked rice is another product, also very popular on the Indian subcontinent,
made from parboiled rice. Again, dry-heat-parboiled rice and not steam-parboiled
rice is best for this purpose (Ali and Bhattacharya, 1976a). Optimum conditions for
making flaked rice by the traditional process—batch sand roasting of soaked paddy
followed by flaking in an edge-runner (ER-Fl)—have been standardized
(Ananthachar et al, 1982). A modernized process—involving continuous roasting of
paddy in a gram roaster, followed by dehusking, partial whitening, and flaking in a
pair of flaking rolls (FR-Fl)—has been devised (Narasimha et al, 1982). There is a
clear gradation of physicochemical properties from raw to dry-heat-parboiled to
ER-Fl to FR-Fl rice (Ali and Bhattacharya, 1976b; Mujoo and Ali, 1998, 1999,
2000). Apart from cooking during the parboiling, some additional mechanical
damage of starch granules occurs during flaking, as well as some molecular
degradation.
The popular pinipig of the Philippines is simply flaked parboiled waxy rice
(Antonio and Juliano, 1974). Another traditional product is hurum in the Assam
state of India, a parboiled waxy rice that is first flaked and then dabbed with oil and
puffed by heating (Mishra et al, 2000).
THE FUTURE
Parboiled rice, as an exotic curiosity of remarkable quality, became known to the

world outside South Asia during the early part of the twentieth century. Work on
understanding the product and on improving the technology began then and
continued until a little beyond the middle of the century. Parboiling as a cereal
Parboiling / 395
Fig. 37. Scanning electron microscopy photomicrograph of transverse section of a milled parboiled
rice grain. Top, transverse section, showing the peripheral layer (PL) as distinct from the inner
endosperm (IE). Bottom, magnified view of PL. CW = cell wall, g = swollen granules, PM = protein
matrix. (Reprinted, with permission, from Chandrasekhar and Chattopadhyay, 1990)
science began to be studied in the second half of the century. It is now a mature
science and a mature technology. Today we have excellent technologies for making
high-quality parboiled rice. We also have a reasonably clear idea of the changes that
take place in the rice grain during parboiling, the properties of parboiled rice, and
the fundamental states of starch and fat in parboiled rice that account for its singular
properties. However, several unknown areas remain.
In the last edition of this monograph, the author wrote that there were signs of a
coming change in production technology. That view clearly was premature. No
doubt there is a clear path from primitive to improved to semimodern to modern
technology to be traversed, and it is being traversed. However, the current modern
parboiling system works so well and is so well entrenched that there is little chance
of a major change in the technology in the near future for the bulk of production.
Nonetheless, certain new processes have potential. Parboiling of brown rice
seems to have already made a place for itself. Reports suggest that its use is
spreading, although still limited. This is a process to be watched. The so-called

pressure-parboiling process has potential only if low-amylose varieties are used
(Table 5) and the use of bisulfite is permitted to reduce discoloration. However, the
process has excellent potential for making the base material for canned rice (Table
14) and puffed rice (Fig. 30). Dry-heat-parboiled rice has some good potential for
use as a niche product because of its superior low-temperature hydration.
Microwave parboiling is another technique with some potential.
A lot of work remains to be done on the science of parboiling. It is true that the
changes during parboiling, the properties of the product, and the effect of rice
variety on parboiled-rice quality have been fairly exhaustively explored. But much
work remains to be done on the state of starch in processed rice. Recent
developments in starch polymorphism have given us a glimpse of why parboiled
rice behaves the way it does. But this is mainly a beginning. These descriptions
haven’t provided all the answers—they have probably only enabled us to ask the
appropriate questions. Research is needed to know precisely what polymorph of
starch is formed under what precise conditions, what precise effects each
polymorph has on the properties of the product, and the combined effect of different
combinations of polymorphs on product quality.
At least three other questions remain to be answered. The first concerns the
property of puffing of parboiled rice upon HTST heating. Rice samples of different
amylose contents require different optimum steam pressures of parboiling in order
to puff (Fig. 36); dry-heat-parboiled rice puffs more than steam-parboiled rice, and
some pressure-parboiled rice puffs best, but others do not. These differences might
be related to different states and combinations of gelatinized starch, retrograded (R-
Ap) starch, lipid-amylose (L-Am I and L-Am II) complexes, annealed starch, and
the thermal breakdown of starch. Clarification of this matter will enhance the
theoretical understanding of starch modification during hydrothermal treatment and
its relation to puffing and also may provide a better technology for puffing.
A second question concerns the role that starch breakdown during high-pressure
parboiling plays in parboiled-rice properties. Thermal breakdown of starch under
certain conditions is a fact. But what precise effect it has on product quality is not
yet known with any degree of assurance. This needs to be clarified. Also important
to know is whether this phenomenon can be made use of in any way—for instance,
in preparing parboiled rice that is relatively easier to cook (Fig. 33).
The last question concerns the role of protein. Any idea that protein has no effect
on product quality is difficult to believe. That the protein bodies are disrupted and
the solubility of protein decreases upon parboiling are well known, but the effects of
these changes on product properties are not. One also does not know whether there
are protein-starch or protein-lipid interactions and how these may affect rice
property. Focused research in this area is necessary.
Any other process or treatment that can reverse the difficulty of cooking par-
boiled rice without disturbing its advantages would be a boon to the industry. This
may be what is promoting the progressive push toward the production of quick-
cooking parboiled rice, a process that is likely to gather momentum as time goes on.
ACKNOWLEDGMENTS
I am most grateful to my informants (mentioned in the footnote) for providing a wealth of information
on the present status of the parboiling industry in the world. Steven Danforth, Antonio Bianchi, Salvador
Barber, and Vichai Sriprasert, in particular, despite being ever burdened in their own enterprises, readily
Parboiling / 397
responded to my repeated questions for months. The result is the first-ever compilation of the topic in the
section on the present status of the industry. Credit for this compilation truly is due to these and other
informants. Any errors are the sole responsibility of the compiler. I am also thankful to my many young
colleagues in the Rice Research and Development Centre for helping me out in this work in many ways.
REFERENCES
Achaya, K. T. 1998a. Page 110 in: Indian Food: Ali, S. Z., and Bhattacharya, K. R. 1982. Studies
A Historical Companion. Oxford University on pressure parboiling of rice. J. Food Sci.
Press, Delhi, India. Technol. 19:236-242.
Achaya, K. T. 1998b. Page 212 in: A Historical Amato, G. W., and Silveira Filho, S. 1991. Par-
Dictionary of Indian Food. Oxford University boilização do Arroz no Brasil. Ciente, Porto
Press, Delhi, India. Alegre. (In Portuguese)
Acton, W. H., Ghosh, S., and Dutt, A. 1933. The Ananthachar, T. K., Narasimha, H. V., Shankara,
vitamin B content of different samples of R., Gopal, M. S., and Desikachar, H. S. R.
Indian rice by Spruyt’s colorimetric method. 1982. Improvement of the traditional process
II. Indian J. Med. Res. 21:103-107. for rice flakes. J. Food Sci. Technol. 19:47-50.
Adhikaritanayake, T. B., and Noomhorm, A. Anthoni Raj, S., and Singaravadivel, K. 1980.
1998. Effect of continuous steaming on par- Influence of soaking and steaming on the loss
boiled rice quality. J. Food Eng. 36:135-143. of simpler constituents in paddy. J. Food Sci.
Ajandouz, E. H., and Puigserver, A. 1999. Technol. 17:141-143.
Nonenzymatic browning reaction of essential Anthoni Raj, S., and Singaravadivel, K. 1982.
amino acids: Effect of pH on caramelization Reduction in free fatty acids due to parboiling
and Maillard reaction kinetics. J. Agric. Food of paddy. J. Food Sci. 47:692-693.
Chem. 47:1786-1793. Anthoni Raj, S., Singaravadivel, K., and
Alary, R., Laignelet, B., and Feillet, P. 1977. Subramaniyan, V. 1996. Excretion of
Effect of amylose content on some characteris- chlorides during soaking of paddy. J. Food
tics of parboiled rice. J. Agric. Food Chem. Sci. Technol. 33:57-59.
25:261-264. Antonio, A. A., and Juliano, B. O. 1973.
Ali, N., and Ojha, T. P. 1976. Determination of Amylose content and puffed volume of
optimum soaking conditions for parboiling of parboiled rice. J. Food Sci. 38:915-916.
paddy. Rice Process Eng. Cent. Rep. (Kharag- Antonio, A. A., and Juliano, B. O. 1974. Physico-
pur, India) 2(1):17-20. chemical properties of glutinous rices in relation
Ali, N., and Ojha, T. P. 1977. Cooking quality of to pinipig quality. Philipp. Agric. 58:17-23.
raw rice and parboiled rice. Rice Process Eng. Arai, K., Raghavendra Rao, S. N., and
Cent. Rep. (Kharagpur, India) 3(1):22-26. Desikachar, H. S. R. 1975. Studies on the
Ali, S. Z. 1979. Studies on parboiling of rice. effect of parboiling on japonica and indica
Ph.D. thesis. University of Mysore, Mysore, rice. Jpn. J. Trop. Agric. 19:7-14. (In Japan-
India. ese)
Ali, S. Z., and Bhattacharya, K. R. 1972a. Hydration Aykroyd, W. R. 1932. The effects of parboiling
and amylose-solubility behaviour of parboiled and milling on the antineuritic vitamin and
rice. Lebensm. Wiss. Technol. 5:207-212. phosphate content of rice. J. Hyg. 32:184-192.
Ali, S. Z., and Bhattacharya, K. R. 1972b. An Aykroyd, W. R., Krishnan, B. G., Passmore, R.,
alkali reaction test for parboiled rice. and Sundararajan, A. R. 1940. The rice prob-
Lebensm. Wiss. Technol. 5:216-218. lem in India. Indian Med. Res. Mem. No. 32.
Ali, S. Z., and Bhattacharya, K. R. 1976a. Com- Bakshi, A. S., and Singh, R. P. 1980. Kinetics of
parative properties of beaten rice and parboiled water diffusion and starch gelatinization
rice. Lebensm. Wiss. Technol. 9:11-13. during rice parboiling. J. Food Sci. 45:1387-
Ali, S. Z., and Bhattacharya, K. R. 1976b. Starch 1392.
retrogradation and starch damage in parboiled Bakshi, A. S., and Singh, R. P. 1982. Modelling
rice and flaked rice. Staerke 28:233-240. rice parboiling process. Lebensm. Wiss. Tech-
Ali, S. Z., and Bhattacharya, K. R. 1980a. High- nol. 15:89-92.
temperature drying-cum-parboiling of paddy. Bal, S., Maheshwari, R. C., Prasad, S., and
J. Food Process Eng. 4:123-136. Parker, R. E. 1976. Effect of drying air
Ali, S. Z., and Bhattacharya, K. R. 1980b. temperature on performance of a recirculating
Changes in sugars and amino acids during par- batch dryer for drying parboiled paddy. Rice
boiling of rice. J. Food Biochem. 4:169-179. Process Eng. Cent. Rep. (Kharagpur, India)
Ali, S. Z., and Bhattacharya, K. R. 1980c. Pasting 2(1):8-15.
behaviour of parboiled rice. J. Texture Stud. Bandopadhyay, B., and Ghose, T. K. 1965. Stud-
11:239-245. ies on the hydration of Indian paddy. I. A rate
equation for the soaking operation. Indian J. paddy for optimum milling quality. Cereal
Technol. 3:360-365. Chem. 44:592-600.
Bandyopadhyay, S., and Roy, N. C. 1975. Predic- Bhattacharya, K. R., and Subba Rao, P. V. 1966a.
tion of time of hydration during parboiling of Processing conditions and milling yields in
paddy from activation energy. J. Food Sci. parboiling of rice. J. Agric. Food Chem.
Technol. 12:197-199. 14:473-475.
Bandyopadhyay, S., and Roy, N. C. 1976. Kinet- Bhattacharya, K. R., and Subba Rao, P. V.
ics of absorption of liquid water by paddy 1966b. Effect of processing conditions on
grains during parboiling. Indian J. Technol. quality of parboiled rice. J. Agric. Food Chem.
14:27-30. 14:476-479.
Bandyopadhyay, S., and Roy, N. C. 1977. Bhattacharya, K. R., Ali, S. Z., and Indudhara
Studies on swelling and hydration of paddy by Swamy, Y. M. 1971. Commercial drying of
hot soaking. J. Food Sci. Technol. 14:95-98. parboiled paddy with LSU dryers. J. Food Sci.
Bandyopadhyay, S., and Roy, N. C. 1978. A Technol. 8:57-63.
semi-empirical correlation for prediction of Bhattacharya, K. R., Sowbhagya, C. M., and
hydration characteristics of paddy during Indudhara Swamy, Y. M. 1972. Some physical
parboiling. J. Food Technol. 13:91-98. properties of paddy and rice and their
Barber, S., Benedito De Barber, C., and Novo, interrelations. J. Sci. Food Agric. 23:171-186.
M. 1983. Inactivation of rice bran enzymes Bhattacharya, S. 1996. Kinetics on colour
during parboiling. Pages 85-86 in: Research in changes in rice due to parboiling. J. Food Eng.
Food Science and Nutrition. Vol. I, The 29:99-106.
Production, Preservation and Processing of Biliaderis, C. G. 1992. Structures and phase tran-
Food. J. V. McLoughlin and M. B. McKenna, sitions of starch in food systems. Food Tech-
Eds. Boole Press, Dublin. nol. 46:98-109, 145.
Bayram, M. 2000. Bulgur around the world. Biliaderis, C. G., and Galloway, G. 1989. Crys-
Cereal Foods World 45:80-82. tallisation behavior of amylose-V complexes:
Becker, H. A. 1960. On the absorption of liquid Structure-property relationships. Carbohydr.
water by the wheat kernel. Cereal Chem. Res. 189:31-48.
37:309-323. Biliaderis, C. G., Page, C. M., Maurice, T. J., and
Benedito de Barber, C., Martinez, J., and Barber, Juliano, B. O. 1986. Thermal characterization
S. 1977. Effects of parboiling processes on the of rice starches: A polymeric approach to
chemical composition and nutritional charac- phase transitions of granular starch. J. Agric.
teristics of rice bran. Pages 121-130 in: Rice Food Chem. 34:6-14.
By-Products Utilization, Vol. IV. S. Barber Biliaderis, C. G., Tonogai, J. R., Perez, C. M., and
and E. Tortosa, Eds. Inst. Agroquimica Juliano, B. O. 1993. Thermophysical properties
Technol. Alimentos, Valencia, Spain. of milled rice starch as influenced by variety and
Bhattacharya, K. R. 1969. Breakage of rice parboiling method. Cereal Chem. 70:512-516.
during milling, and effect of parboiling. Cereal Birch, G. G., and Priestley, R. J. 1973. Degree of
Chem. 46:478-485. gelatinization of cooked rice. Staerke 25:98-100.
Bhattacharya, K. R. 1980. Breakage of rice during Biswas, S. K., and Juliano, B. O. 1988. Labora-
milling: A review. Trop. Sci. 22:255-276. tory parboiling procedures and properties of
Bhattacharya, K. R. 1985. Parboiling of Rice. parboiled rice from varieties differing in starch
Pages 289-348 in: Rice Chemistry and Tech- properties. Cereal Chem. 65:417-423.
nology. 2nd ed. B. O. Juliano, Ed. Am. Assoc. Bolling, H., and El-Baya, A. W. 1975. Einfluss von
Cereal Chem., St. Paul, MN. Parboiling auf die physikalischen und chemi-
Bhattacharya, K. R., and Ali, S. Z. 1970. schen Eigenschaften des Reises (Effect of par-
Improvement in commercial sun drying of parboiling on the physical and chemical properties
boiled paddy for better milling quality. Rice J. of rice). Getreide Mehl Brot 29:230-233.
73(9):3,4,9,12-15. Braddon, W. L. 1907. The Cause and Prevention
Bhattacharya, K. R., and Ali, S. Z. 1976. A sedi- of Beri-beri. Rebman, London.
mentation test for pregelatinized rice products. Burns, E. E., and Gerdes, D. L. 1985. Canned
Lebensm. Wiss. Technol. 9:36-37. rice foods. Pages 557-567 in: Rice: Chem-
Bhattacharya, K. R., and Ali, S. Z. 1985. istry and Technology. 2nd ed. B. O. Juliano,
Changes in rice during parboiling, and Ed. Am. Assoc. Cereal Chem., St. Paul, MN.
properties of parboiled rice. Pages 105-167 in: Cailleau, R., Kidder, L. E., and Morgan, A. F.
Advances in Cereal Science and Technology, 1945. The thiamine content of raw and par-
Vol. VII. Y. Pomeranz, Ed. Am. Assoc. Cereal boiled rices. Cereal Chem. 22:50-60.
Chem., St. Paul, MN. CFTRI. 1969. Parboiling of paddy. Project Circ.
Bhattacharya, K. R., and Indudhara Swamy, Y. 7 (rev.). Cent. Food Technol. Res. Inst.,
M. 1967. Conditions of drying parboiled Mysore, India.
Parboiling / 399
CFTRI. 1987. Annual Report for the year 1986–87. Craufurd, R. Q. 1963. Sorption and desorption in
Cent. Food Technol. Res. Inst., Mysore, India. raw and parboiled paddy. J. Sci. Food Agric.
CFTRI. 1989. Annual Report for the year 1988–89. 14:744-750.
Cent. Food Technol. Res. Inst., Mysore, India. Desikachar, H. S. R. 1977. Preservation of by-
CFTRI. 1992. Process for the Dry Heat products of rice milling. Pages 1-32 in: Rice
Parboiling of Rice. Cent. Food Technol. Res. By-Products Utilization, Vol. II. S. Barber and
Inst., Mysore, India. E. Tortosa, Eds. Inst. Agroquimica Technol.
Chakravarty, H. B., and Ghose, T. K. 1966. Studies Alimentos, Valencia, Spain.
on the hydration of Indian paddy. IIIA: Effi- Desikachar, H. S. R., Majumder, S. K., Pingale,
ciency of elevated temperature on composition S. V., Swaminathan, M., and Subrahmanyan,
of oil in bran. Pages 21-25 in: Souvenir, Sem- V. 1955. Prevention of off-flavours due to fer-
inar on Modern Technology of Rice Milling. mentation in the commercial production of
East India Rice Mills Assoc., Calcutta, India. parboiled rice. Bull. Cent. Food Technol. Res.
Chandrasekhar, P. R., and Chattopadhyay, P. K. Inst. Mysore 5:50-53.
1990. Studies on microstructural changes of Desikachar, H. S. R., Lakshminarayana, S. K.,
parboiled and puffed rice. J. Food Process Pre- and Subrahmanyan, V. 1957. Improvements in
serv. 14:27-37. parboiling of paddy. Res. Ind. 2:150-153.
Chandrasekhar, P. R., and Chattopadhyay, P. K. Dimopoulos, J. S., and Muller, H. G. 1972. Effect
1991. Rice puffing in relation to its varietal of processing conditions on protein extraction
characteristics and processing conditions. J. and composition and on some other physico-
Food Process Eng. 14:261-277. chemical characteristics of parboiled rice.
Charbonniere, R. 1975. Structure physique de Cereal Chem. 49:54-62.
l’amidon du riz natif, étuvé, précuit, appertisé, Done, J. 1949. The effects of various treatments
retrogradé (Physical structure of native, par- of paddy rice on the aneurin and nicotinic acid
boiled, precooked, canned, and retrograded rice content of milled grain. Br. J. Nutr. 3:335-345.
starch). Bull. Inf. Rizicul. France 160:14-21. Eggum, B. O. 1979. The nutritional value of rice
Charlton, J. 1923. The prevention of nuisances in comparison with other cereals. Pages 91-
caused by the parboiling of paddy. Agric. Res. 111 in: Chemical Aspects of Rice Grain
Inst. Bull. 146. Pusa (Calcutta), India. Quality. Int. Rice Res. Inst. Los Baños,
Chinnaswamy, R., and Bhattacharya, K. R. Laguna, Philippines.
1983a. Studies on expanded rice. Physico- Feillet, P., and Alary, R. 1975. Influence des
chemical basis of varietal differences. J. Food caracteristiques variétales, de l’étuvage et de
Sci. 48:1600-1603. l’usinage sur l’aptitude des riz a l’appertisation
Chinnaswamy, R., and Bhattacharya, K. R. 1983b. (Influence of varietal characteristics, of par-
Studies on expanded rice. Optimum processing boiling, and of milling on the suitability of
conditions. J. Food Sci. 48:1604-1608. rice for canning). Bull. Inf. Rizicul. France
Chinnaswamy, R., and Bhattacharya, K. R. 1984. 158:15-18.
Relationship between amylose content and Fletcher, W. 1907. Rice and beri-beri: Pre-
expansion characteristics of parboiled rice. J. liminary report on an experiment conducted at
Cereal Sci. 2:273-279. the Kuala Lumpur lunatic asylum. Lancet
Chinnaswamy, R., and Bhattacharya, K. R. 1986. 1:1776-1779.
Pressure-parboiled rice: A new base for making Fraser, H., and Stanton, A. T. 1909. An enquiry
expanded rice. J. Food Sci. Technol. 23:14-19. concerning the etiology of beriberi. Lancet
Chrastil, J. 1994. Effect of storage on the 1:451.
physicochemical properties and quality factors Gariboldi, F. 1974. Rice Parboiling. FAO Agric.
of rice. Pages 49-81 in: Rice Science and Dev. Paper 97. Food Agric. Org. of the U.N.,
Technology. W. E. Marshall and J. I. Rome.
Wadsworth, Eds. Marcel Dekker, New York. Gariboldi, F. 1984. Rice Parboiling. FAO Agric.
Cnossen, A. G., and Siebenmorgen, T. J. 2000. Serv. Bull. 56. Food Agric. Org. of the U.N.,
The glass transition temperature concept in Rome.
rice drying and tempering: effect on milling Ghose, R. L. M., Ghatge, M. B., and
quality. Trans. ASAE 43:1661-1667. Subrahmanyan, V. 1960. Rice in India (rev.).
Colonna, P., and Mercier, C. 1983. Macromo- Indian Counc. Agric. Res., New Delhi, India.
lecular modifications of manioc starch compo- Guha, M., and Ali, S. Z. 1998. Parboiling of
nents by extrusion-cooking with and without brown rice: Effect of variety and parboiling
lipids. Carbohydr. Polym. 3:87-108. conditions. J. Food Sci. Technol. 35:504-508.
Craufurd, R. Q. 1962. Moisture changes in raw and Haley, W. L. 1960. Bulgor, an ancient wheat
parboiled paddy in West Africa and their influ- food. Cereal Sci. Today 5:203-207.
ence upon milling quality. II. Changes during Halim, A., and Bhattacharya, K. R. 1978. Effect
drying. Empire J. Exp. Agric. 30:321-329. of degree and pressure of milling on the sticki-
ness of milled parboiled rice. J. Food Qual. Jagannadha Rao, P. V. K., Bal, S., and
1:349-358. Chakraverty, A. 1997. Use of pneumatic pres-
Houston, D. F., and Kester, E. B. 1954. Hygro- sure in parboiling paddy. Agric. Macha. Asia,
scopic equilibria of whole-grain edible forms Africa Latin Am. 28:69-71.
of rice. Food Technol. 8:302-304. Jayanarayanan, E. K. 1964. Der Einfluss der ver-
Houston, D. F., Hunter, I. R., McComb, E. A., arbeitungsbedingungen auf das Braunwerden
and Kester, E. B. 1954. Deteriorative changes von “parboiled” Reis (Effect of operating con-
in the oil fraction of stored parboiled rice. J. ditions on the browning of parboiled rice).
Agric. Food Chem. 2:1185-1190. Nahrung 8:129-137.
Houston, D. F., Hunter, I. R., and Kester, E. B. Kar, N., Jain, R. K., and Srivastav, P. P. 1999.
1956. Storage changes in parboiled rice. J. Parboiling of dehusked rice. J. Food Eng.
Agric. Food Chem. 4:964-968. 39:17-22.
Iamsuri, K. 1977. Traditional and modern meth- Kato, H., Ohta, T., Tsugita, T., and Hosaka, Y.
ods of parboiling in Thailand. Pages 393-397 1983. Effect of parboiling on texture and
in: Working Papers, Action Oriented Field flavor components of cooked rice. J. Agric.
Workshop for Prevention of Post-Harvest Food Chem. 31:818-823.
Rice Losses, Alor Setar, Kedah, Malaysia, Khan, A. V., Amilhussin, A., Arboleda, J. R.,
1977. I. Shamsudin and F. Abdullah, Eds. Manalo, A. S., and Chancellor, W. J. 1974.
Food Agric. Org. of the U.N. and Govern- Accelerated drying of rice using heat-conduc-
ment of Malaysia. tion media. Trans. ASAE 17:949-955 (Food
Iengar, N. G. C., Bhaskar, R., and Dharmarajan, P. Sci. Technol. Abstr. 7:7M845, 1975).
1971. Studies on sand parboiling and drying of Kik, M. C. 1955. Influence of processing on
paddy. J. Agric. Eng. New Delhi 8(2):51-54. nutritive value of milled rice. J. Agric. Food
Iengar, N. G. C., Bhaskar, R., and Dharmarajan, Chem. 3:600-603.
P. 1972. Studies on methods of parboiling— Kik, M. C. 1965. Nutritional improvement of rice
Pressure parboiling. J. Agric. Eng. New Delhi diets and effect of rice on nutritive value of
9(4):40-51. other foodstuffs. Ark. Agric. Exp. Stn. Bull.
Iengar, N. G. C., Gangadharan, N. S., and 698.
Rajendran, G. 1974a. Recent developments in Kik, M. C., and Williams, R. R. 1945. The nutri-
pressure parboiling of paddy. J. Agric. Eng. tional improvement of white rice. Nat. Res.
New Delhi 11(1):31-33. Counc. Bull. 112. Washington, D.C.
Iengar, N. G. C., Gangadharan, N. S., and Kimura, T. 1983. Properties of parboiled rice
Rajendran, G. 1974b. Recent developments in produced from Japanese paddy. Agric. Mecha.
the pressure parboiling of paddy—Direct par- Asia, Africa Latin Am. 14:31-33.
boiling. J. Agric. Eng. New Delhi 11(5,6):28- Kimura, T. 1989. Improvement of rice parboiling
29. with batch tank. Agric. Eng. 4:2445-2453.
Igathinathane, C., and Chattopadhyay, P. K. Kimura, T. 2001. The Theory and Technology of
1999a. Moisture diffusion modeling of drying Rice Parboiling. Japan International Coopera-
in parboiled paddy components. Part I: tion Agency, Tsukuba International Centre,
Starchy endosperm. J. Food Eng. 41:79-88. University of Tsukuba, Japan.
Igathinathane, C., and Chattopadhyay, P. K. Kimura, T., Matsuda, J., Ikeuchi, Y., and
1999b. Moisture diffusion modelling of drying Yoshida, T. 1976. Basic studies on parboiled
in parboiled paddy components. Part II: Bran rice. Part III. Effect of processing conditions
and husk. J. Food Eng. 41:89-101. on the rate of gelatinization of parboiled rice.
Indudhara Swamy, Y. M., Ali, S. Z., and J. Jpn. Soc. Agric. Mach. 38:379-383. (In
Bhattacharya, K. R. 1971. Hydration of raw Japanese)
and parboiled rice and paddy at room tem- Kimura, T., Bhattacharya, K. R., and Ali, S. Z.
perature. J. Food Sci. Technol. 8:20-22. 1993. Discoloration characteristics of rice dur-
Islam, M. R., Shimizu, N. and Kimura, T. 2001. ing parboiling. I. Effect of processing condi-
Quality evaluation of parboiled rice with tions on the color intensity of parboiled rice. J.
physical properties. Food Sci. Technol. Res. Soc. Agric. Structures Jpn. 24:23-30.
7:57-63. Kimura, T., Shimizu, N., Shimohara, T., and
Islam, M. R., Shimizu, N., Furuichi, S., and Warashina, J. 1995. Trials of quality
Kimura, T. 2003. Effect of processing evaluation for parboiled and other rice by
conditions on cooking qualities of parboiled means of the near infrared spectroscopy and
rice. J. Jpn. Soc. Agric. Machinery 65:85-93. the rapid visco analyzer. J. Soc. Agric.
Itoh, K., and Kawamura, S. 1985. Studies on Structures Jpn. 25:175-182.
parboiled rice. I. Processing conditions of par- Kratochvil, C. P., Lin, Y. H. E., and Novak, T. J.
boiled rice and its quality. J. Jpn. Soc. Food 1994. Parboiled rice product and method of
Sci. Technol. 32:471-479. making same. U.S. patent 5316783.
Parboiling / 401
Kulkarni, S. D., and Bal, S. 1984. On the influ- Action Oriented Field Workshop for Preven-
ence of evacuation on liquid water absorption tion of Post-Harvest Rice Losses, Alor Setar,
by paddy grains. J. Food Sci. Technol. 21:318- Kedah, Malaysia, 1977. I. Shamsudin and F.
320. Abdullah, Eds. Food Agric. Org. of the U.N.
Kulkarni, S. D., and Bal, S. 1986. Studies on and Government of Malaysia.
effect of evacuation of paddy on quality char- Mazumder, A. C., Bose, A. N., Ganguli, N. C.,
acteristics of parboiled rice. J. Food Sci. Tech- and Guha, B. C. 1960a. Pilot plant studies on
nol. 23:33-36. parboiling of rice. I. Soaking and gelatiniza-
Kumaravel, S., Singaravadivel, K., Vasan, B. S., tion. J. Biochem. Microbiol. Technol. Eng.
and Anthoni Raj, S. 1980. Storage of parboiled 2:431-438.
bran. J. Oil Technol. Assoc. India 12:49-51. Mazumder, A. C., Bose, A. N., Ganguli, N. C.,
Kurien, P. P., Radhakrishna Murty, R., and Guha, B. C. 1960b. Pilot plant studies on
Desikachar, H. S. R., and Subrahmanyan, V. parboiling of rice. II. Dehydration. J. Biochem.
1964. Effect of parboiling on the swelling Microbiol. Technol. Eng. 2:439-445.
quality of rice. Cereal Chem. 41:16-22. McCarrison, R., and Norris, R. V. 1924. The
Lii, C. Y., Chang, Y. H., and Lai, S. S. 1992. relationship of rice to beri-beri in India. Indian
Degradation of rice starch in a single screw Med. Res. Mem. No. 2.
extruder. J. Chin. Agric. Chem. Soc. 30:282- McGaughey, W. H. 1974. Insect development in
290. milled rice: Effects of variety, degree of mill-
LSU. 1998. Ohmic heating of rice and rice by- ing, parboiling and broken kernels. J. Stored
products. Dept. of Biological and Agricultural Prod. Res. 10:81-82.
Engineering, Louisiana State University, Mecham, D. K., Kester, E. D., and Pence, J. W.
Baton Rouge. (Online: 1961. Parboiling characteristics of California
www.bae.lsu.edu/3276.htm) medium-grain rice. Food Technol. 15:475-
Lu, S., and Lii, C. Y. 1996. Effects of hydro- 479.
thermal treatment on the quality and Mestres, C., Colonna, P., and Buleon, A. 1988.
physicochemical properties of kernels of Characteristics of starch networks within rice
three rice varieties. J. Chin. Agric. Chem. flour noodles and mungbean starch vermicelli.
Soc. 34:1-12. J. Food Sci. 53:1809-1812.
Mahadevappa, M., and Desikachar, H. S. R. Mishra, A., Goswami, R. G., and Mahanta, C, L.
1968a. Expansion and swelling of raw and 2000. Identifying the essential steps and proc-
parboiled rice during cooking. J. Food Sci. ess variables to make Hurum. J. Food Sci.
Technol. 5:59-62. Technol. 37:455-458.
Mahadevappa, M., and Desikachar, H. S. R. Mohandoss, R., and Pillaiyar, P. 1980. Alkali
1968b. Some observations on histology of raw reaction test as an indicator of the degree of
and parboiled rice. J. Food Sci. Technol. 5:72- parboiling of rice. Oryza 17:158-160.
73. Mohandoss, R., and Pillaiyar, P. 1982. Influence of
Mahanta, C. L. 1988. Studies on nature of par- parboiling and drying methods on the quality of
boiled rice. Ph.D. thesis, University of parboiled rices. Madras Agric. J. 69:185-193.
Mysore, Mysore, India. Mujoo, R., and Ali, S. Z. 1998. Susceptibility of
Mahanta, C. L., and Bhattacharya, K. R. 1989. starch to in vitro enzyme hydrolysis in rice,
Thermal degradation of starch in parboiled rice flakes and intermediary products.
rice. Starch/Staerke 41:91-94. Lebensm. Wiss. Technol. 31:114-121.
Mahanta, C. L., and Bhattacharya, K. R. 1995. Mujoo, R., and Ali, S. Z. 1999. Molecular degra-
An alkali-reaction test to distinguish open- dation of rice starch during processing to
steam, pressure-steam and dry-heat parboiled flakes. J. Sci. Food Agric. 79:941-949.
rice. J. Food Sci. Technol. 32:400-402. Mujoo, R., and Ali, S. Z. 2000. Changes in
Mahanta, C. L., Ali, S. Z., Bhattacharya, K. R. physico-chemical and rheological properties
and Mukherjee, P. S. 1989. Nature of starch of rice during flaking. Int. J. Food Prop.
crystallinity in parboiled rice. Starch/Staerke 3:117-135.
41:171-176. Mukherjee, R. K., and Bhattacharjee, M. 1978.
Malkanthi, O. S., Kimura, T., and Shimizu, N. Distribution of oil in the bran layers of
2001. Present Status of Rice Post Harvest slender, medium and short grain varieties of
Technology in Sri Lanka. Presented at the rice, and effect of parboiling. J. Am. Oil
Annual Meeting of Kanto Branch, Japanese Chem. Soc. 55:463-464.
Society of Agricultural Machinery. Available Murugesan, G., and Bhattacharya, K. R. 1989.
from the Institute of Agricultural and Forest The nature of starch in popped rice.
Engineering, University of Tsukuba, Japan. Carbohydr. Polym. 10:215-225.
Maung Maung, M. 1977. Parboiling process in Nagashima, T., Takano, K., Suzuki, Y., Baba, T.,
Burma. Pages 398-402 in: Working Papers, and Kamoi, I. 1983. Study on parboiling of
brown rice (Japonica). J. Agric. Sci. 28:238- Malaysia, 1977. I. Shamsudin and F.
243. Abdullah, Eds. Food Agric. Org. of the U.N.
Narasimha, H. V., Ananthachar, T. K., Shankara, and Government of Malaysia.
R., Gopal, M. S., and Desikachar, H. S. R. Pillaiyar, P. 1984a. Applicability of the rapid gel
1982. Development of a continuous process test for indicating the texture of commercial
for making rice flakes. J. Food Sci. Technol. parboiled rices. Cereal Chem. 61:255-256.
19:233-235. Pillaiyar, P. 1984b. A rapid test to indicate the
Narasimha, H. V., Shankara, R., Sreedhara texture of parboiled rices without cooking. J.
Murthy, S., Indudhara Swamy, Y. M., Texture Stud. 15:263-273.
Sathyendra Rao, B. V., and Srinivas, A. 2000. Pillaiyar, P. 1985. A gel test to parboiled rice
A continuous process for parboiling brown using dimethyl sulphoxide. J. Food Sci. Tech-
rice. J. Food Sci. Technol. 37:323-326. nol. 22:1-3.
Narayana Rao, M., Viswanatha, T., Mathur, P. Pillaiyar, P. 1989. Effect of open and closed dry-
B., Swaminathan, M., and Subrahmanyan, V. ing on the quality of parboiled rice. J. Agric.
1954. Effect of storage on the chemical Eng. New Delhi 26:157-163.
composition of husked, undermilled and Pillaiyar, P., and Mohandoss, R. 1981a. Hardness
milled rice. J. Sci. Food Agric. 5:405-409. and colour in parboiled rices produced at low
Narayanswami, C. K. 1956. The Rice We Eat. and high temperatures. J. Food Sci. Technol.
All India Khadi and Village Industries Board, 18:7-9.
Bombay, India. Pillaiyar, P., and Mohandoss, R. 1981b. Cooking
Ocker, H. D., Bolling, H., and El-Baya, A. W. qualities of parboiled rices produced at low
1976. Effect of parboiling on some vitamins and high temperatures. J. Sci. Food Agric.
and minerals of rice: Thiamine, riboflavin, cal- 32:475-480.
cium, magnesium, manganese and phosphorus. Pillaiyar, P., Venkatesan, V., and Narayanasamy,
Riso 25:79-82. R. V. 1977. Parboiling of soaked paddy by
Ong, M. H., and Blanshard, J. M. V. 1994. The closed heating. J. Food Sci. Technol. 14:226-
significance of the amorphous-crystalline tran- 227.
sition in the parboiling process of rice and its Pillaiyar, P., Vasan, B. S., Rajendran, G., and
relation to the formation of the amylose-lipid Dharmarajan, V. 1980. Prevention of foul
complex and recrystallisation (retrogradation) odour and minimising soaking loss in conven-
of starch. Food Sci. Technol. Today 8:217- tional parboiling of paddy. J. Food Sci. Tech-
226. nol. 17:228-231.
Ong, M. H., and Blanshard, J. M. V. 1995. The Pillaiyar, P., Singaravadivel, K., Desikachar, H.
significance of starch polymorphism in com- S. R., and Subramaniyan, V. 1993. Low mois-
mercially produced parboiled rice. ture parboiling of paddy. J. Food Sci. Technol.
Starch/Staerke 47:7-13. India 30:97-99.
Padua, A. B., and Juliano, B. O. 1974. Effect of Pillaiyar, P., Singaravadivel, K., and Desikachar,
parboiling on thiamine, protein and fat of rice. H. S. R. 1994. Quality changes in HTST proc-
J. Sci. Food Agric. 25:697-701. essing of rice parboiling. J. Sci. Food Agric.
Palani Muthu, V., and Chattopadhyay, P. K. 65:229-231.
1992. Thin layer drying characteristics of Pillaiyar, P., Sabarathinam, P. L., Subramaniyan,
parboiled milled rice. J. Food Sci. Technol. V., and Sulochana, S. 1996. Parboiling of
29:275-277. paddy using thermic fluid. J. Food Eng.
Park, S. H., and Cho, E. J. 1995. Physical and 27:267-278.
cooking characteristic properties of parboiled Pillaiyar, P., Singaravadivel, K., and Desikachar,
rice. J. Korean Soc. Food Sci. 11:126-132. H. S. R. 1997. A rapid test to differentiate hard
Paul, H. 1954. The control of smell in parboiled pressure parboiled rice from other parboiled
rice by use of acetic acid. Farm J. Br. Guiana rices. J. Food Sci. Technol. 34:532-533.
17(1):10-11. (Cited in Kasasian, 1982) Pillaiyar, P., Sabarathinam, P. L., and Sulochana,
Perdon, A., Siebenmorgen, T. J., and S. 1998. Variations in hydration of paddy
Mauromoustakos, A. 2000. Glassy state grains within and among panicles of pure vari-
transition and rice drying: Development of a ety during parboiling. J. Food Sci. Technol.
brown rice state diagram. Cereal Chem. 35:274-275.
77:708-713. Priestley, R. J. 1976a. Studies on parboiled rice.
Pillaiyar, P. 1977. Improvements in conven- I. Comparison of the characteristics of raw and
tional parboiling of paddy: Elimination of parboiled rice. Food Chem. 1:5-14.
foul odour and increase in yield of rice. Priestley, R. J. 1976b. Studies on parboiled rice.
Pages 403-414 in: Working Papers, Action II. Quantitive study of the effects of steaming
Oriented Field Workshop for Prevention of on various properties of parboiled rice. Food
Post-Harvest Rice Losses, Alor Setar, Kedah, Chem. 1:139-148.
Parboiling / 403
Raghavendra Rao, S. N., and Juliano, B. O. 1970. rate of lipid hydrolysis and deterioration of rice
Effect of parboiling on some physicochemical bran. Cereal Chem. 52:1-8.
properties of rice. J. Agric. Food Chem. Shimizu, N., Yanagisawa, T., Okadome, H.,
18:289-294. Toyoshima, H., Andren, H., Kimura, T., and
Raghavendra Rao, S. N., Narayana, M. N., and Ohtsubo, K. 2001. Determination of maximum
Desikachar, H. S. R. 1967. Studies on some viscosity of milled rice flours using near-infra-
comparative milling properties of raw and par- red transmittance spectroscopy. Food Sci.
boiled rice. J. Food Sci. Technol. 4:150-155. Technol. Res. 7:104-109.
Rama Rao, V. V., and Bonde, R. S. 1977. Drying Sing, K. 1980. Influence of milled rice on insect
characteristics of parboiled paddy. Rice Proc- infestation. I. Oviposition and development of
ess Eng. Cent. Rep. (Kharagpur, India) post-harvest pests in different types of milled
3(1):13-18. rice. Angew. Entomol. 90:1-9 (Food Sci.
Ramalingam, N., and Anthoni Raj, S.1996. Stud- Technol. Abstr. 13:6M562, 1981).
ies on the soak water characteristics in various Singaravadivel, K., and Anthoni Raj, S. 1979.
paddy parboiling methods. Bioresour. Leaching of phenolic compounds during soak-
Technol. 55:259-261. ing of paddy. J. Food Sci. Technol. 16:77-78.
Ramiah, K. 1937. Rice in Madras. Government Sluyters, J. A. F. M. 1963. Milling studies on
Press, Madras, India. parboiled rice in Nigeria. Trop. Agric.
Renn, P. T., and Sathe, S. K. 1997. Effects of pH, Trinidad 40:153-158.
temperature, and reactant molar ratio on L-leu- Sondi, A. B., Reddy, I. M., and Bhattacharya, K.
cine and D-glucose Maillard browning R. 1980. Effect of processing conditions on
reaction in an aqueous system. J. Agric. Food the oil content of parboiled-rice bran. Food
Chem. 45:3782-3787. Chem. 5:277-282.
Roberts, R. L., Potter, A. L., Kester, E. B., and Sowbhagya, C. M., and Bhattacharya, K. R.
Keneaster, K. K. 1954. Effect of processing 1976. Lipid autoxidation in rice. J. Food Sci.
conditions on the expanded volume, color and 41:1018-1023.
soluble starch of parboiled rice. Cereal Chem. Sowbhagya, C. M., Ali, S. Z., and Ramesh, B. S.
31:121-129. 1993. Effect of parboiling on grain dimensions
Roy, P. 2001. Improvement of Energy Require- of rice. J. Food Sci. Technol. 30:459-461.
ment in Traditional Parboiling Process. A Sreenivasan, A. 1939. Studies on quality in rice.
Report on Field Study. Inst. Agric. Forest Eng, IV. Storage changes in rice after harvest.
University of Tsukuba, Japan. Indian J. Agric. Sci. 9:208-222.
Sanjiva Rao, B. 1948. Rice: Physicochemical Srinivasa Aiyangar, P. T. 1985. Page 59 in: Pre-
aspects of its curing to secure vitamin conser- Aryan Tamil Culture. Asian Educational Ser-
vation and improvement in storage and vices, New Delhi, India. (Originally published
cooking quality. Proc. 35th Indian Sci. Congr. 1930, reprinted 1985)
Part II:259-270. Subba Rao, P. V., and Bhattacharya, K. R. 1966.
Sanjiva Rao, B., and Guha, B. C. 1952. Report of Effect of parboiling on thiamine content of
the Rice Technology Committee. Ministry of rice. J. Agric. Food Chem. 14:479-482.
Food, Government of India, New Delhi. Subrahmanyan, V., Sreenivasan, A., and Das
Sarda, P. S. 1965. Improved method of parboiling Gupta, H. P. 1938. Studies on quality of rice.
in Sarda Rice and Oil Mills. J. Food Sci. Tech- I. Effect of milling on the chemical
nol. 2:105-107. composition and commercial qualities of raw
Schoch, T. J. 1964. Swelling power and and parboiled rice. Indian J. Agric. Sci. 8:459-
solubility of granular starches. Pages 106-108 486.
in: Methods in Carbohydrate Chemistry, Vol. Subrahmanyan, V., Desikachar, H. S. R., and
IV. R. L. Whistler, Ed. Academic Press, New Bhatia, D. S. 1955. Commercial methods of
York. parboiling paddy and improvement of the
Schule. 2001. Rice Parboiling System. F. H. quality of parboiled rice. J. Sci. Ind. Res.
Schule Mühlenbau GmbH, Reinbek, Germany. 14A:110-114.
(Online: www.schulerice.com) Sunnyland Mills. 2001. History of bulgur wheat.
Schweizer, T. F., Reimann, S., Solms, J., (Online: www.sunnylandmills.com)
Eliasson, A.-C., and Asp, N.-G. 1986. Influ- Swaminathan, M. 1941. The effect of washing
ence of drum-drying and twin-screw extrusion and cooking on the nicotinic acid content of
cooking on wheat carbohydrates. II. Effects of raw and parboiled rice. Indian J. Med. Res.
lipids on physical properties, degradation and 29:83-88.
complex formation of starch in wheat flour. J. Swaminathan, M. 1942. The effect of washing
Cereal Sci. 4:249-260. and cooking on the vitamin B1 content of raw
Shaheen, A. B., El-Dash, A. A., and El-Shirbeeny, and parboiled milled rice. Indian J. Med. Res.
A. E. 1975. Effect of parboiling of rice on the 30:409-416.
Tata, S. N. 1962. Parboiled rice. Rice News cold soaking process of paddy. J. Food Sci.
Teller 10(2/3):50-52. Technol. 19:31-34.
Taylor, J., Martin, C. De C., and Thant, U. 1928. Vasan, B. S., Iengar, N. G. C., Subramanyam, T.
Preliminary inquiry into beri-beri in Burma. V., Chandrasekaran, R., and Subrahmanyan,
Indian Med. Res. Mem. No. 8. (Cited in V. 1971. Effect of processing on the move-
Aykroyd et al, 1940) ment of oil in the rice kernel and its relation to
Tetens, I., Biswas, S. K., Glitso, L. V., Kabir, K. the oil content of paddy and the yield of oil
A., Thilsted, S. H., and Choudhury, N. H. from rice bran. Interregional Seminar on the
1997. Physico-chemical characteristics as indi- Industrial Processing of Rice. Food and Agri-
cators of starch availability from milled rice. J. culture Organization, ECAFE, and Govern-
Cereal Sci. 26:355-361. ment of India, Madras.
Tsuzuki, E., and Tanaka, K. 1981. Studies on the Vasan, B. S., Basheer, J., and Venkatesan, V.
characteristics of scented rice. VII. Effect of 1980. Use of hot air for parboiling and drying
parboiling treatment on some fatty acid of paddy. J. Food Sci. Technol. 17:246-247.
compositions and flavor components in rice Vedder, E. B. 1913. Beri-beri. William Wood,
grains. Bull. Fac. Agric. Miyazaki Univ. New York.
28:25-30. Velupillai, L. 1994. Parboiling rice with micro-
Unnikrishnan, K. R. 1986. Studies on parboiled wave energy. Pages 263-274 in: Rice Science
rice. Ph.D. thesis, University of Mysore, and Technology, W. E. Marshall and J. I.
Mysore, India. Wardsworth, Eds. Marcel Dekker, New York.
Unnikrishnan, K. R., and Bhattacharya, K. R. Velupillai, L., and Verma, L. R. 1982. Parboiled
1981. Swelling and solubility behaviour of rice quality as affected by the level and distri-
parboiled rice flour. J. Food Technol. 16:403- bution of moisture after the soaking process.
408. Trans. ASAE 25:1450-1456.
Unnikrishnan, K. R., and Bhattacharya, K. R. Velupillai, L., Verma, L. R., and Wadsworth, J. I.
1983. Cold-slurry viscosity of processed rice 1989. Quality aspects of microwave-vacuum-
flour. J. Texture Stud. 14:21-30. dried parboiled rice. Trans. ASAE 32:1759-
Unnikrishnan, K. R., and Bhattacharya, K. R. 1762.
1987a. Influence of varietal difference on Villareal, C. P., and Juliano, B. O. 1987. Varietal
properties of parboiled rice. Cereal Chem. differences in quality characteristics of puffed
64:315-321. rices. Cereal Chem. 64:337-342.
Unnikrishnan, K. R., and Bhattacharya, K. R. Vinacke, W. R., Hartzler, E., and Tanada, Y.
1987b. Properties of pressure-parboiled rice 1950. Processed rice in Hawaii. Tech. Bull.
as affected by variety. Cereal Chem. 64:321- 10. Univ. Hawaii Agric. Exp. Stn., Honolulu.
323. Viraktamath, C. S., and Desikachar, H. S. R. 1971.
Unnikrishnan, K. R., and Bhattacharya, K. R. Inactivation of lipase in rice bran in Indian rice
1988. Application of gel consistency test to mills. J. Food Sci. Technol. 8:70-74.
parboiled rice. J. Food Sci. Technol. 25:129- Wadsworth, J. I., Velupillai, L., and Verma, L. R.
132. 1990. Microwave-vacuum drying of parboiled
Unnikrishnan, K. R., and Bhattacharya, K. R. rice. Trans. Agric. 33:199-210.
1989. Reviving the pressure-parboiling proc- Webb, B. D., and Adair, C. R. 1970. Laboratory
ess by use of low-amylose varieties of paddy. parboiling apparatus and methods of eval-
Indian Food Ind. 8:25-28. uating parboil-canning stability of rice. Cereal
Unnikrishnan, K. R., and Bhattacharya, K. R. Chem. 47:708-714.
1995. Changes in properties of parboiled rice Webb, B. D., Bollich, C. N., Adair, C. R., and
during ageing. J. Food Sci. Technol. 32:17- Johnston, T. H. 1968. Characteristics of rice
21. varieties in the U.S. Department of Agriculture
Unnikrishnan, K. R., Viraktamath, C. S., Collection. Crop. Sci. 8:361-365.
Krishnamurthy, H., and Bhattacharya, K. R. Williams, K. T., and Bevenue, A. 1953. A note on
1982. Parboiling of paddy by simple soaking the sugars in rice. Cereal Chem. 30:267-269.
in hot water. J. Food Technol. 17:499-506. Xavier, I. J., and Anthoni Raj, S. 1996. Enzyme
Vasan, B. S., and Ganesan, G. 1981. Electrical changes in rough rice during parboiling. J.
heating of soaked paddy for parboiling. J. Food Biochem. 19:387-389.
Agric. Eng. 18:88-92. Yang, M. O., and Cho, E. J. 1995. The effect of
Vasan, B. S., and Kumaravel, S. 1982. The role milling on the nutrients of raw and parboiled
of dissolved oxygen in the improvement of rices. J. Korean Soc. Food Sci. 11:51-57.
CHAPTER 14
ENRICHMENT AND FORTIFICATION OF RICE
Diane Wright Hoffpauer

Light Heart, LLC
Crowley, Louisiana
The most significant change in the enrichment of rice in the past several years
is the addition of folic acid to the U.S. standard of identity for rice and other
cereal grains. As research reveals new insights into the treatment and prevention
of disease with nutrients, traditional suppliers of cereal grains have responded
accordingly, with various kinds of enrichment and fortification in items such as
breakfast cereals, food bars, and snack foods. The market for functional foods
using cereal grains as an ingredient has also increased in the last decade. Before
adding folic acid to the standard of identity for cereal grains to prevent neural
tube defects (NTDs), the Food and Drug Administration (FDA) conducted an
extensive three-year study on the feasibility of adding folic acid to grain products.
This study period included discussions by food and medical professionals,
manufacturers, and the public. The evidence of the positive benefits of folic acid
enrichment proved overwhelming. This is just one of many cases in which
enrichment and fortification of cereal and other foods made a significant impact
on the prevention of disease.
Enrichment generally refers to the restoration of vitamins and minerals lost
during processing. Fortification generally means adding vitamins and minerals to
foods in higher amounts than were present before processing. Before the standards
of identity were adopted to define enriched foods, the term fortification was used
when any vitamins and minerals were added to foods (Hoffpauer and Wright,
1994).
HISTORY OF REGULATIONS
The standard for enriching rice was published in the Federal Register and
became effective February 27, 1958, after various methods of rice enrichment had
been studied. In 1968, a joint policy statement was issued by the Council on Foods
and Nutrition of the American Medical Association and the Food and Nutrition
Board of the National Academy of Sciences-National Research Council endorsing
the continuation of enrichment programs, which included the enrichment of white
rice with thiamin, riboflavin, niacin, and iron. In addition, the Council on Foods and
Nutrition and the Food and Nutrition Board, in the same policy statement, continued
to endorse the addition of nutrients to foods in situations that include all of the
following circumstances:
405
1. The intake of the nutrient is below the desirable level in the diets of a signifi-
cant number of people.
2. The food used to supply the nutrient is likely to be consumed in quantities that
will make a significant contribution to the diet of the population in need.
3. The addition of the nutrient is not likely to create an imbalance of essential
nutrients.
4. The nutrient added is stable under proper conditions of storage and use.
5. The nutrient is physiologically available from the food.
6. There is reasonable assurance against excessive intake to a level of toxicity.
Based on improved techniques for evaluating the nutritional status of the popula-
tion, the board reiterated its adherence to this policy in 1973 (FNB/NRC, 1974). In
addition to standard enrichment, the board also gave guidelines for the fortification
of cereal grains. It noted that a number of significant changes had taken place in the
food consumption patterns of the population. Advances in food technology had led
to the introduction of many processed foods into the U.S. diet. Total energy
requirements had declined with improved transportation and mechanization. More
people had become weight conscious, in a trend that continues today. Because of
these conditions, consumption of the basic staples of the diet of the 1940s and
1950s, such as wheat, corn, oats, and rice, had declined significantly. In 1974, the
board established a recommended dietary allowance for 31 essential nutrients
either known to be essential for humans or presumed to be essential on the basis
of demonstrated need in animal diets (FNB/NRC, 1974). Nearly 30 years after
the initial assessment of dietary needs in the U.S. population, this review
prompted a recommendation for cereal-grain product fortification with vitamin
A, thiamin, riboflavin, niacin, pyridoxine, folic acid, iron, calcium, magnesium,
and zinc. A shift in the dietary needs of the population prompted a similar review
in the 1990s.
CURRENT REGULATIONS
The regulations published by the FDA on April 1, 2000, were preceded by three
years of intense discussion by the FDA, food and medical professionals, and manu-
facturers (FDA, 2000). These discussions focused mainly on scientific evidence
supporting the role of folate in preventing NTDs. An NTD usually takes the form of
a condition known as spina bifida, in which lower-back vertebrates fuse and the
spinal cord is exposed, or anacephaly, which is characterized by a lesion in the
upper neural tube and little or no brain development. The failure of the neural-tube
fusion occurs in the period between the 24th and 27th day of pregnancy, usually
before a woman is aware that she is pregnant (Hoffpauer and Bonnette, 1998).
Therefore, folic acid consumption must occur early in the pregnancy or preferably
before conception. This information led the Public Health Service in 1992 to
recommend that all women of childbearing age consume at least 400 mcg of folate
per day (Stevenson et al, 1995).
Some of the most important evidence in support of folate fortification came from
a study conducted by the British Medical Research Council. The study, involving 33
research centers in seven countries, examined women who had a previous NTD
pregnancy. The women were randomly assigned in a double-blind protocol to one
of four groups: folic acid, folic acid and multivitamin, multivitamin alone, and pla-
cebo. The groups consuming folic acid alone or folic acid and a multivitamin
Enrichment and Fortification / 407
experienced a 72% reduction in recurrence of NTDs, compared to the placebo

group or the group consuming a multivitamin with folic acid (MRC, 1991).
Megaloblastic anemia was the first condition identified that responded to
folic acid therapy. The condition is characterized by large, malformed, multi-
segmented blood cells. Folate deficiency is thought to cause this condition by
preventing proper cell division due to a lack of nucleic acids for DNA synthesis
during cell division. Because folate is required for nucleic acid synthesis and
nucleic acids are required for DNA synthesis, treatment with folic acid relieves
the condition by remedying the folate deficiency and allowing proper nucleic
acid synthesis. Megaloblastic anemia is common during pregnancy, especially
in impoverished, malnourished populations. However, pernicious anemia
(vitamin B-12 deficiency), a very rare condition during pregnancy and in the
nonelderly, can also result in megaloblastic anemia (Hoffpauer and Bonnette,
1998).
One negative aspect of folic acid fortification identified during the review period
concerned pernicious anemia. Considerable discussion was conducted on the effects
of folate consumption at a level above 1.0 mg. At this level, folate masks the
symptoms of pernicious anemia. In particular, elderly individuals, who are most
likely to develop pernicious anemia, could be harmed, as delay of diagnosis could
result in permanent neurological damage. In the final FDA rule, the fortification
level of 1.0 mg of folate per day was chosen as the safe upper limit. It was
concluded that the 1.0-mg/day limit would be most likely to affect all population
groups without increasing the risk to persons deficient in vitamin B-12 (FDA,
2000). There has been extensive debate on whether large doses of folic acid are
harmful, but the general consensus has been that the benefits in disease prevention
require immediate action for the health of the public. Diagnostic methods for
pernicious anemia are specific and accurate. Monitoring the groups at risk and using
the established diagnostic methods should minimize the risk of folate-masking
(Hoffpauer and Bonnette, 1998).
The current standard of identity for enriched rice as set forth in Title 21 of the
Code of Federal Regulations (CFR) covers forms of milled rice (except rice coated
with talc and glucose and known as coated rice) to which nutrients have been added
so that each pound of rice contains the amounts shown in Table 1.
The forms of milled rice included in the standard of identity are white rice,
instant white rice, parboiled rice, and instant parboiled rice. Provisions are included
in the standard of identity for the addition of butylated hydroxytoluene as an
optional ingredient in an amount not to exceed 0.0033% by weight of the finished
food (FDA, 2000). This optional antioxidant ingredient is allowed because of the
high oil content of parboiled rice.
The current regulations also require the following statement for powder enrich-
ment: “To retain vitamins do not rinse before or drain after cooking.” Rice that has
been fortified with nutrients in a form that is rinse resistant (i.e., other than powder)
do not require this statement on the package.
Differences exist between federal, state, and international regulations. The U.S.
federal standard of identity for milled rice is voluntary, unless the manufacturer puts
“enriched” on the label. In such cases, the vitamins and minerals listed in the stan-
dard of identity for enriched rice must be included at the levels shown. State laws
vary from state to state. In some states, enriched rice is required, and all interstate
sales of rice to those states must be enriched rice. For international sales of rice, the
TABLE 1
Standard of Identity for Enriched Ricea
Minimum Level Maximum Level
Nutrient (mg/lb) (mg/lb)
Thiamin 2.0 4.0
Riboflavinb 1.2 2.4
Niacin or niacinamide 16.0 32.0
Folic acid 0.7 1.4
Iron (Fe) 13.0 26.0
a Source: FDA (2000).
b Note that as of February 25, 1958, by the order of the Commissioner of Food and Drugs appearing at
23 FR 1170, the requirement for riboflavin has been stayed until final action is taken disposing of the
objections, after public hearing thereon, and vitamin D and calcium enrichment remain optional.
laws vary from country to country. It is best to consult an expert on international

law where sales of enriched rice are concerned.
IMPACT OF FOLIC ACID FORTIFICATION
On June 20, 2001, the Journal of the American Medical Association published
the results of a study to determine the impact of folic acid fortification on the U.S.
food supply (Honein et al, 2001). After the FDA authorized the addition of folic
acid to enriched grain products in March 1996 (FDA, 1996), it was expected that
the fortification would add 100 mcg to the daily diet of the average person and
result in approximately 50% of all reproductive-aged women receiving 400 mcg of
folate from all sources. Using birth certificate information provided by state vital
statistic offices and compiled by the Centers for Disease Control and Prevention
(CDC) and the National Center for Health Statistics, an analysis of the incidence of
folate-preventable NTDs was conducted. The baseline mean and standard deviation
were based on 1990 and 1996 data. Data collected from U.S. birth certificates for
births after mandatory folic acid fortification (October 1998 through December
1999) showed a 19% decline in the birth prevalence of NTDs and a 23% decline in
spina bifida prevalence. Because of the public health importance of the NTD
decline, an announcement was published on the CDC’s NCHS website in December
2000 (Honein et al, 2001). Further studies will show the full impact of this
fortification.
RICE ENRICHMENT PROCESSES
Powder Enrichment
Currently two forms of enrichment are used commercially in the United States.
The first form is a preblended powder mixture of thiamin, riboflavin (if requested),
niacin or niacinamide, folic acid, and either ferric orthophosphate (white iron),
ferrous sulfate (yellow iron), or reduced iron. Riboflavin and ferrous sulfate give the
powder and the rice a slight off-white to yellow color, which is undesirable to some
consumers. Reduced iron has the potential to turn the rice gray to black. Ferric
orthophosphate is the most requested form of iron used in the industry due to its
white color and ability to blend with the white rice. However, when ferric ortho-
phosphate is oxidized or contains excessive moisture, it can turn tan, yellow, purple,
TABLE 2
Forms of Enrichment Showing Preblended Levels
Vitamin or Level Label
Forms of Enrichment Mineral (g/lb) (mg/g) Addition Ratea
Vitamin/mineral powder Thiamin 3.20 7.05 1 oz/100 lb riceb
Niacin 25.60 56.44
Folic acid 1.12 2.47
Iron 33.60 74.07
Filler 390.08 859.96
Vitamin/mineral coated kernel Thiamin 0.40 0.88 1 lb/199 lb ricec
Niacin 3.20 7.05
Folic acid 0.14 0.31
Iron 2.60 5.73
Filler 447.26 986.02
a When vitamin/mineral enrichment is added at this rate, the final product will contain the federal
standard of identity for thiamin, niacin, folic acid, and iron.
b Or 0.62 mg/g (0.06%).
c Or 5.03 mg/g (0.5%).
and/or black. When powder enrichment is used in the packaged product, the state-
ment, “To retain vitamins do not rinse before or drain after cooking” is required
according to 21 CFR 137.350 4(c). One of the disadvantages of powder enrichment
in rice is that 20-100% of the enrichment washes off the rice, depending on the
amount of water used in rinsing and the application time. Other disadvantages of
powder enrichment are that the vitamins and minerals are less stable during storage
and after application; the uniformity of application can be a problem and may cause
assay problems; and the vitamins and minerals can react easily with other food
components. The principal advantage of powder enrichment is that the blend is less
expensive than other forms of enrichment.
Coated-Kernel Enrichment
The second form of enrichment available commercially is a premixed and treated

kernel-type enrichment. This “premix,” as it is known in the rice industry, is a pow-
der-blended enrichment that is applied to the milled rice grains and coated with a
water-insoluble surface to retain the vitamins and minerals during rinsing. The
enrichment of the rice grains is concentrated so that, when added to the milled rice
at 0.5%, all of the enrichment required per pound of rice is provided according to
the standard of identity.
One unique aspect of kernel-type enrichment is that the rice can be rinsed with-
out losing the enrichment before cooking. The insoluble food-grade coating is bro-
ken down when it reaches the acid environment of the stomach, thereby releasing
the vitamins and minerals. In addition, the vitamins and minerals are stable and do
not react with other food components, and the kernels are easy to detect and assay
in the final product (Hoffpauer and Wright, 1994). Kernel-type enrichment is
slightly more expensive than powder-type enrichment because it is not as concen-
trated as powder enrichment (Table 2).
Many of the suppliers of instant white rice package the rice in “boil-in-the-bag”
containers. This rice must be enriched with a rinse-resistant premix such as the
coated kernel-type enrichment currently available.
Extruded Kernels
The extruded-kernel premix is prepared by mixing rice starch or flour with the
vitamin and mineral powder and extruding it into simulated grains. The patented
extruded enrichment product on the market, named Ultra Rice, is produced
primarily for the international market. Both the coated and extruded methods allow
the packaged “boil-in-the-bag” rice to be cooked in excess water while retaining at
least 85% of the enrichment, as required by the standard of identity (FDA, 2000).
Application of Folic Acid
Following the comment period allowed by the FDA and the final ruling in 1996,
manufacturers of enriched grain products were required to add folic acid to the
enrichment products. The two main concerns were color and stability. Folic acid is
yellow in its pure form. It is also unstable in light and can absorb up to 10% water
by weight. These characteristics had to be analyzed in order to have a smooth tran-
sition to folic-enriched products.
The yellow color of the nutrient in powder enrichment preparations is not easily
detectable at the levels required for enrichment at 1.4 mg/lb. The enrichment pow-
ders for flour, cornmeal or grits, pasta, farina, and bread that already contained ribo-
flavin did not change considerably in color. Riboflavin is also yellow and is similar
in density to folic acid. These powders are also normally enriched with reduced
iron, which is gray to black in color. The density and color of these mixes were
virtually the same as before the folic acid addition.
The enrichment powders for rice, however, did have a slight color change, from
grayish white to a cream color. Once added to the rice, the change was negligible
and not noticeable by the consumer. The stability of riboflavin and folic acid con-
tinues to be a problem in an environment unprotected from heat and light. Stability
studies may need to be undertaken in the future to evaluate the effects of this insta-
bility. Before such studies are performed, overages in these nutrients may be neces-
sary to achieve proper levels after application.
For enrichment of rice by coated-grain premix, the color was a problem at the
beginning of the transition. Adjustments had to be made to the coating process, and
masking of the color was necessary in some cases. With the addition of a concen-
trated grain premix, where color variation may be perceived by the consumer as
“defective” grains, it is important that the premix blend with the final rice product.
Additional steps have been taken in the coated-grain process to further stabilize the
folic acid so that it is not affected by light. This should ensure that the levels remain
constant in the final product. Like the coated-grain process preceding folic enrich-
ment, the nutrients are not rinsed off, as they are in powder enrichment (Hoffpauer
and Bonnette, 1998).
TESTING ENRICHMENT AND FORTIFICATION LEVELS
Testing methods for determining rinse resistance of “not less than 85 percent of
the minimum quantities of the substances named…” are included in CFR 2, section
137.350, part 4 (e) (FDA, 2000). This method details the process for diluting and
rinsing the enriched rice, and cross-references the methods of the Association of
Official Analytical Chemists for testing the remaining vitamins and minerals.
Methods are cited for determination of thiamin, riboflavin, niacin, iron, and cal-
cium. However, these methods, for the rinse test for enriched rice, are usually not
adequate for testing the enrichment before it is added to the rice. The concentration
is sufficient so that high-performance liquid chromatography (HPLC) methods are
more suitable for determining accurate results. For the minerals, accurate results
have been determined using atomic absorption (AA), with the appropriate dilution.
For the final nutrient levels in enriched rice after a premix has been added,
HPLC and AA may be used, but sampling and extraction are important. At least 1 lb
of the sample should be collected per test. The entire 1-lb sample should be ground
using a coffee grinder or other bench-top mill. Approximately 100 g of the ground
sample should be diluted with the mobile phase used for the HPLC method. After
dilution, the sample should be shaken or sonicated for at least 10 min before filtra-
tion. The appropriate HPLC methods for thiamin, riboflavin, niacin or niacinamide,
and folic acid may then be used to determine concentrations. Approximately 20 g of
the ground sample is usually sufficient for mineral determination. After weighing,
the dry sample should be placed in a crucible and partially ashed on a hot plate to
prevent splattering in the furnace. The sample should then be completely ashed in a
furnace overnight. After ashing, the sample should be diluted with concentrated
hydrochloric acid and distilled water for reading on the AA. These methods,
although not official, have provided reliable results for manufacturers.
GLOBAL RICE FORTIFICATION
Beginning with the benchmark study on rice fortification from 1947 to 1950 in
Bataan, Philippines, numerous models for rice fortification have been formulated.
The Bataan study was an effort to reduce the incidence of beriberi in the population
of a control group by fortifying white milled rice with thiamin. In this well-designed
and documented study, rice premix enriched with thiamin (2 mg/lb), niacin (16
mg/lb), and iron (13 mg/lb), according to the U.S. federal standard of identity, was
delivered free of cost to mills in Bataan in seven experimental regions and two
control regions. Inspectors were employed to verify compliance at the mills,
markets, and occasionally in homes. Surveys that took place in 1947 and 1948
showed that 90% of the clinical beriberi had disappeared in the experimental area as
compared to the control area (Salcedo et al, 1950). The results of the experiment
were even more pronounced when a reduction occurred in beriberi deaths of infants
who were nursed and did not eat the rice itself. The study also suggested that
enrichment of rice could decrease the mortality of populations from other diseases
such as tuberculosis, which is improved by good nutrition.
Bhattacharya (1964) reported research on enrichment of rice for India. The
method included enrichment with thiamin, nicotinic acid, and calcium by mixing
the nutrients directly with the rice, heating the mixture in a closed chamber to
90-95°C, and then cooling under insulation. Although the method was unsatisfac-
tory for riboflavin alone, it was effective for the combination of riboflavin 5-
phosphate and calcium chloride, giving an 85% retention after washing.
Another study on fortification of rice with lysine, threonine, thiamin, riboflavin,
vitamin A, and iron for preschool children in Chiang Mai, Thailand, from 1971 to
1975, showed less favorable results than previous studies (Gershoff et al, 1977).
The primary objective of the study was to determine the effects of limiting amino
acids in the diet of Thai children. No significant changes were found in the
anthropometric measurements or hemoglobin and hematocrit levels with fortified

rice use. Also, no significant change in morbidity occurred. It is not known whether
the incidence of infectious disease affected the study, but diseases such as
respiratory illness, mumps, measles, and whooping cough were prevalent. The
authors concluded that the retarded growth in the experimental group was not
caused by illness and could not be prevented by the nutrients provided in this study.
They suggested that future studies contain calcium, as this nutrient was clearly low
in the diets of the people studied.
Studies on the effect of iron, vitamin A, and calcium fortification of rice have
taken various routes. The technical difficulties of fortifying with iron that is ade-
quately absorbed, does not alter the appearance of the rice, and is bioavailable are
addressed in a study by Cook and Reusser (1983). They identify iron compounds
used for fortification and the relative cost of each, and they noted that fortification
with ascorbic acid would improve iron absorption but would be cost prohibitive.
The most soluble and bioavailable form of iron, ferrous sulfate, continues to provide
challenges for rice fortification because of its reactivity and color. Attempts to for-
tify grains with ferrous sulfate and seal them with talc as a whitening agent pro-
duced absorption of less than 50% as compared to the absorption of iron in other
foods eaten simultaneously. It was assumed that the iron was not completely
released from the coating.
The most prominent studies on fortifying rice with vitamin A have been with
retinyl palmitate (RP) in extruded synthetic rice made from rice flour. The product,
named Ultra Rice, had a cooking retention of 46-94% with a combination of satu-
rated oils, tocopherols, and ascorbate. This combination would be necessary for
stability in tropical environments such as that of the Philippines (Murphy et al,
1992). The bioavailability of Ultra Rice was tested using serum retinol values. The
values from pre- and postchallenge doses were compared to show the status of the
liver reserves of the vitamin in 83 children aged 11-77 months. The relative dose
response test was positive in 48 of the 83 subjects. The loss of vitamin A under
normal rice cooking conditions in the study was found to be 25.9 ± 9.1% (n = 70)
(Flores et al, 1994). In additional cooking and storage stability tests on Ultra
Rice, tests of RP showed that it was affected more by temperature than by relative
humidity. When the rice was stored at 23°C for six months, 85% of the RP was
retained. Storage at 35°C showed extensive losses in RP. The researchers
concluded that, if the storage temperature of the premix could not be controlled,
more rapid turnover of the premix and RP overages would be required (Lee et al,
2000).
Studies of the qualities of calcium-fortified rice have been performed in the last
several years (Hettiarachchy et al, 1995; Lee et al, 1995). Solutions of calcium lac-
tate, selected for its nonmetallic taste and pleasant mouthfeel, were soaked into the
rice grains, which were then autoclaved. Color, whiteness, translucency, and texture
of the final cooked rice were then evaluated. The results showed that its whiteness
did not differ from that of the control, but the calcium-fortified rice was
significantly more translucent than the control. No undesirable color was associated
with the product. The firmness values of the fortified rice increased; however, this
was not perceived to be negative, as American consumers prefer firm, dry, and
fluffy cooked rice. A study of the effects of washing on the calcium-fortified rice
samples found that the reduction was only about 5%, well within the U.S. standard
for rinse resistance (Hettiarachchy et al, 1996).
A report for two federal programs, Opportunities for Micronutrient Intervention

and U.S. AID, (P. B. Dexter, unpublished data) identified common micronutrient
deficiencies in global populations. Low energy and protein intakes continue to be
problem areas for rice-consuming countries. In addition, vitamin A, iron, and iodine
continue to be lacking in these diets. Other microdeficiencies noted include thiamin,
riboflavin, calcium, vitamin C, selenium, magnesium, and zinc, but these have not
been well documented. With increased technology in the area of rice fortification,
the potential of rice as a carrier for nutrient intervention in populations remains
strong.
CONCLUSION
The basic commodities of rice, corn, oats, and wheat have fed millions of people
over the years. However, the forms of these commodities are constantly changing. It
is important that, as diets and preparation of these grains change in the population,
we use them to their fullest potential. Using grains as carriers for supplemental
nutrition is as important as ever in the United States and around the world. Enrich-
ment and fortification play a vital role in maintaining nutrition and in the prevention
of disease. In the future, using whole grains such as rice in functional foods will
create a new dimension to our food systems.
REFERENCES
Bhattacharya, K. R. 1964. Enrichment of rice iron on preschool children. Am. J. Clin. Nutr.
with riboflavin. J. Food Sci. 1:45. 30:1185-1195.
Cook, J. D., and Reusser, M. E. 1983. Iron forti- Hettiarachchy, N. S., Gnanasambandam, R., and
fication: An update. Am. J. Clin. Nutr. 38:648- Lee, M. H. 1996. Calcium fortification of rice:
659. Distribution and retention. J. Food Sci.
FDA. 1996. Title 21 Code of Federal Regula- 61:195-197.
tions, Chapter 1, Parts 136, 137 and 139, Food Hoffpauer, D. W., and Bonnette, R. E., III. 1998.
standards: Amendment of standards of identity Enrichment update on folic acid. Cereal Foods
for enriched grain products to require addition World 43:365-367
of folic acid. Food and Drug Administration, Hoffpauer, D., and Wright, S. L., III. 1994.
Washington, DC. Enrichment of rice. Pages 195-204 in: Rice
FDA. 2000. Title 21 Code of Federal Regula- Science and Technology. W. E. Marshall and
tions, Chapter 1, Part 137, Cereal flours and J. I. Wadsworth, Eds. Marcel Dekker, Inc.,
related products, Section 137.350 Enriched New York.
rice, revised as of April 1, 2000. Food and Honein, M. A., Paulozzi, L. J., Mathews, T. J.,
Drug Administration, Washington, DC. Erickson, D. J., and Wong, L. C. 2001. Impact
Flores, H., Guerra, N. B., Cavalcanti, A. A., of folic acid fortification of the U.S. food sup-
Campos, F. A. C. S., Azevedo, M. C. N. A., ply on the occurrence of neural tube defects. J.
and Silva, M. B. M. 1994. Bioavailability of Am. Med. Assoc. 285:2981-2986.
vitamin A in a synthetic rice premix. J. Food Lee, J., Hamer, M. L., and Eitenmiller, R. R.
Sci. 59:371-372. 2000. Stability of retinyl palmitate during
FNB/NRC. 1974. Proposed Fortification Policy for cooking and storage in rice fortified with Ultra
Cereal-Grain Products. Food and Nutrition RiceTM fortification technology. J. Food Sci.
Board/National Research Council, National Acad- 65:915-919.
emy of Sciences, Washington, DC. pp. 8-9; 10-11. Lee, M. H., Hettiarachchy, N. S., McNew, R. W.,
Gershoff, S. N., McGandy, R. B., Suttapreyasri, and Gnanasambandam, R. 1995. Physico-
D., Promkutkao, C., Nondasuta, A., Pisolya- chemical properties of calcium-fortified rice.
butra, U., Tantiwongse, P., and Viravaidhaya, Cereal Chem. 72:352-355.
V. 1977. Nutrition studies in Thailand. II. Marshall, W. E., and Wadsworth, J. I., Eds. 1994.
Effects of fortification of rice with lysine, Rice Science and Technology. Marcel Dekker,
threonine, thiamin, riboflavin, Vitamin A, and Inc., New York.
MRC. 1991. Prevention of neural tube defects: F., Oliveros, S. B., Pascual, R., Santiago, L.
Results of the Medical Research Council C., and Valenzuela, R. C. 1950. Artificial
Vitamin Study. Lancet 338:131-137. enrichment of white rice as a solution to
Murphy, P. A., Smith, B., Hauck, C., and O’Connor, endemic beriberi. J. Nutr. 42:501-523.
K. 1992. Stabilization of vitamin A in a Stevenson, R. E., Dean, J. H., Allen, W. P., and
synthetic rice premix. J. Food Sci. 57:437-439. Kelly, M. 1995. Prevention program for
Salcedo, J., Bamba, M. D., Carrasco, E. O., Chan, reducing risk for neural tube defects. Morb.
G. S., Concepcion, I., Jose, F. R., DeLeon, J. Mort. Wkly. Rep. 44:141-142.
CHAPTER 15
RICE END-USE QUALITY ANALYSIS
C. J. Bergman
Rice Research Unit
Beaumont, Texas
K. R. Bhattacharya
Rice Research and Development Centre
Mysore, India
K. Ohtsubo
National Food Research Institute
Tsukuba Science City
Ibaraki, Japan
The many rice cultivars grown in the world vary greatly in their cooking, sen-
sory, and processing quality. Rice desirable to one group of people or suitable for
one product may not be so for another. Chemists began looking into these cultivar
differences in rice end-use quality early in the twentieth century. The cultivar dif-
ference in rice kernel response to dilute alkali observed by Warth and Darabsett
(1914) is probably the earliest recorded research on the subject. Further research
was sporadic and was focused by and large on the “cooking” (i.e., hydration prop-
erty) of rice in India. But the scope of this inquiry was widened to a greater degree
in the 1950s. At that time, rice in the United States was categorized into three grain
types (long, medium and short), each with precise dimensional and cooking-
processing characteristics (Adair et al, 1973). The inadvertent release in the early
1950s of the cultivar Century Patna 231, which had the dimensions of a typical U.S.
long-grain rice but different cooking and processing characteristics, came as a sur-
prise and was a financial disaster for the rice industry. This event brought forward
the need to understand the basis of rice end-use quality and to devise simple tests
for it. What followed was a research effort, coordinated by the United States
Department of Agriculture (USDA), that clarified the essentials of rice quality,
which remain substantially valid even today. Beachell and Halick (1956, 1957)
recorded the first steps of this effort and Beachell and Stansel (1963) its final out-
come. Since the 1950s, much research in Germany (Federal Research Center for
Cereal and Potato Processing), Spain (Institute of Agronomy and Food Technol-
ogy), Japan (National Food Research Institute and Niigata University), and espe-
cially India (Central Food Technological Research Institute) and the Philippines
415
(International Rice Research Institute) has taken our understanding of rice end-use
quality to new levels. And today, research efforts in this arena exist on nearly every
continent.
Across the world, rice exists in many combinations of length-width ratio, grain
weight, bran color, and endosperm chemical characteristics. The physical form of
rice kernels is not associated with their cooking and processing qualities. End-use
quality characteristics desired by one region of the world may be completely unac-
ceptable to another. Most rice does not enter the export market but is consumed
close to where it was produced. Consequently, consumers are often exposed only to
local cultivars and are unaware of the great variation in the physical and chemical
characteristics of the world’s rice germ plasm. Rice that enters the export market is
typically limited to a select few combinations of kernel physical and chemical traits.
This facilitates trade by simplifying the language required to identify rice quality-
types and is designed to meet the quality requirements of rice-importing nations.
For example, rice in the export market that has a long, thin grain typically has a
firm, nonsticky cooked texture and is called “long-grain rice.” But, within all of the
world’s rice germ plasm, that which has a long grain is not always firm cooking and
likewise firm-cooking rice does not always exist as a long grain.
Rice produced in the United States enters both the export and domestic markets.
The three market classes in the United States are based on grain shape, amylose
content, and gelatinization temperature (Fig. 1). Conventional long-grain rice is at
least three times longer than it is wide; it has an intermediate gelatinization tem-
perature and intermediate amylose content. After cooking, it is firm and fluffy.
Consumers in areas of the world such as North and South America, Southern China,
and Europe prefer this type of rice. In the U.S. and export markets, most parboiled
and quick-cooking rice begins as conventional long grains. Conventional medium
grains are from 2.1 to 2.9 times longer than they are wide. After cooking, they are
soft, moist, and sticky in texture. This type of rice is preferred by people from
Fig. 1. Traits commonly used to define rice quality types. Alkali spreading value corresponds to the
following gelatinization temperature ranges: low = <70°C, intermediate = 70–74°C, and high = >70ºC.
Other traits commonly used for rice quality categorization are grain dimensions, cooked kernel
elongation, firmness and stickiness, pasting properties, protein, and 2-acetyl-1-pyrroline content.
End-Use Quality Analysis / 417
Japan, Northern China, and North and South Korea. Many breakfast cereals are
produced using conventional medium-grain rice. This class of rice is low in amylose
and has lower gelatinization temperature compared to long-grain rice. Conventional
short-grain rice is less than two times longer than it is wide, while its other proper-
ties are similar to those of medium-grain rice. Because this type of rice is com-
monly used in sushi, some call it sushi rice.
In the United States, rice cultivars that are not conventional long, medium, or
short grains are known as specialty types. They have distinguishing traits that must
be determined using analytical methods besides those for amylose content and
gelatinization temperature. For example, Figure 1 shows that basmati-type rice
cultivars have amylose content and gelatinization temperature similar to those of
conventional long-grain rice types. They also contain high levels of a compound (2-
acetyl-1-pyrroline) that confers a “buttery” popcornlike smell, and after cooking,
the grains elongate into very long, thin kernels. This style of rice is preferred by
many people of Indian and Pakistani descent. Jasmine-style rice consists of long
grains that are low in amylose content and have low gelatinization temperature.
They also contain high levels of 2-acetyl-1-pyrroline and are soft and sticky after
cooking. This style of rice is increasing in popularity across the world but is most
desired by people from Thailand. Japanese premium-quality types (e.g.,
Koshihikari) are similar to conventional short grains in terms of grain type, amylose
content, and gelatinization temperature. However, this style of rice has other prop-
erties that are preferred by many people of Japanese and Korean descent. These
properties are said to include high degree of glossiness, lack of off-flavor, distinct
aroma, sticky but smooth texture, and a soft texture that is maintained after cooling.
Another type also preferred in Japan and currently under development in the United
States is the low-amylose type (e.g., Milky Queen and Snow Pearl), which is used
in chilled and frozen products due to its reduced susceptibility to retrogradation.
Superior-processing types are those long grains that have high amylose and inter-
mediate gelatinization temperature. They are differentiated from other high-amylose
types by their high setback, as determined using a Rapid Visco Analyser (RVA).
The arborio style of rice, classified as a medium-grain, traditionally is used for
making Italian risotto. It has a firm internal texture and a creamy exterior. Arborio
types often have a white core (e.g., chalky center) that is thought to be responsible
for their ability to take up the flavor of a cooking stock or sauce. Waxy rice (known
also as sticky or glutinous rice) has little to no amylose, appears completely opaque,
and is very soft and sticky after cooking. In addition to its use as an intact cooked
grain in foods such as desserts and steamed, savory rice balls, flour and starch pro-
duced from this type of rice are used by the food industry as thickeners, fat replac-
ers, etc.
Rice breeders in the United States, in collaboration with chemists, select culti-
vars for release that have characteristics falling within a market class or specialty
type. Therefore, the only quality types the food industry sees are those with a
defined set of quality attributes, and the total amount of variation in end-use quality
that exists in the world’s rice germ plasm can thus be underappreciated. Nearly
420,500 samples of rice and its related species are maintained in germ-plasm col-
lections across the world. Within these valuable genetic resources, rice breeders
find opportunities for improving cultivar yield potential and disease resistance.
Unique traits within this germ plasm also offer the potential to develop cultivars
with new industrial applications. For example, a new U.S. specialty type (e.g., the
cultivar Cadet) has been developed that, when combined with a patented processing
method, results in a new convenience food product marketed in Europe (Fig. 1).
Creating rice cultivars with novel combinations of grain traits often requires the
use of germ plasm from another country, and such new cultivars may not perform
well under local growing conditions. Using this unadapted breeding material and
targeting unique grain characteristics makes the breeding process very challenging.
As a result, some rice breeders are putting together teams of food scientists, chem-
ists, and molecular geneticists to define targeted quality characteristics and to assist
in the evaluation of breeding progeny. For example, food scientists are helping to
define the rice sensory attributes that today’s discriminating consumers demand
(Champagne et al, 1998). Chemists are defining the chemical composition of rice
that results in specific sensory and processing quality (Buttery et al, 1983). And new
selection tools such as gas chromatography and genetic markers are being devel-
oped by chemists and geneticists and incorporated into breeding programs to
enhance cultivar development (Bergman et al, 2000, 2001). In the future, there is
hope that these teams will include more professionals from the rice milling and food
processing industries. Such relationships can help ensure that these industries have
access to rice with quality attributes that perform well in their processes and meet
their consumers’ requirements.
End-use quality methodology can be categorized as those procedures that deter-
mine rice physical, functional, chemical, or sensory properties. This chapter focuses
on the analytical methods and instruments available to measure the first three of
these properties of rice. Aspects of these topics have previously been reviewed by
Juliano (1985, 1998), Kohlwey (1994), and Ohtsubo et al (1998).
PHYSICAL PROPERTIES
Milling Quality
Milling quality is composed of several factors that directly effect the value of
rough rice. It encompasses the total amount of milled rice recovered after milling
(total milled-rice yield) and the total amount of whole kernels recovered after mill-
ing (head-rice yield or whole-rice yield). The purity of the rough-rice samples is
also a component of milling quality. Specifically, the value of rough rice is nega-
tively affected by the presence of cracked kernels, red rice, discolored kernels, or
immature kernels.
Laboratory-sized mills are used to determine milling quality. They serve as
important tools used by rice breeders to assess the milling quality of breeding prog-
eny, by scientists to prepare samples for study, by government agencies to deter-
mine the grade of rice, and by processors to set the price of rice. Differences
between rice laboratory mills include the principle behind the milling process, i.e.,
friction versus abrasion. Also, the mills differ in the sample size required and
whether dehulling is first required. The McGill no. 3 miller is used by the USDA to
determine the grade designation of milled rice (USDA, 1995). The McGill no. 2
miller is preferred by some because it requires a smaller sample size and reportedly
is able to achieve milling yields equivalent to those of the no. 3 miller (Andrews et
al, 1992). McClung and Castro (1994) have demonstrated that a smaller version of
the no. 2 miller can be used to mill 50 g of rough rice. The milling yields obtained
were reported to be similar to those achieved when starting with brown rice in the
no. 2 miller. Other laboratory mills, such as the Test Rice Whitening Machine
(Yamamoto Corp.), Rice Miller Tester no. 60 (Grain Machinery Mfg. Corp.), and
Rice Test Mill (Satake Corp.), are also available.
Brown rice yield is determined by expressing dehulled rice as a weight percent-
age of rough rice. This measurement also indicates the amount of hull in a given
sample. Total milled-rice yield is determined by expressing combined broken- and
whole-kernel yield as a weight percentage of rough rice. Head-rice yield is the yield
of milled rice that is three quarters or more of normal kernel length, expressed as a
percentage of rough rice or total milled rice. The most common method for deter-
mining whole vs. broken kernels is to place a sample on a shaker table, which con-
sists of two inclined indent plates that vibrate. Various other types of sieves, indent
cylinders, and graders are also in use. Because these are subjective methods, some
effort has been made to develop instrumentation able to remove the subjectivity
from the separation of whole kernels from brokens. An instrument that is receiving
some attention is an automated grain inspection system called the GrainCheck
(Lloyd et al, 2001). This system utilizes image-analysis technology to measure ker-
nel color, length, and width. Another image-analysis system is the Single-Grain
Rice Inspector, developed by Kett Laboratory and in use in Japan. The manufactur-
ers of this instrument advertise that grain color and shape and the percentage of
whole grain, immature grain, and discolored kernels can be determined.
Kernel weakness that results in breakage during milling is reported to be related
to fissuring (cracking), chalkiness, and kernel dimensions (Bhattacharya, 1980).
Measurement of the latter two factors is discussed in sections below. Both environ-
ment and genetics are known to have an impact on fissure resistance (Jodari and
Linscombe, 1996). The environmental conditions that promote fissuring are rain
and changes in humidity just before harvest (Banaszek and Siebenmorgen, 1990;
Lan and Kunze, 1996a,b). Research suggests that grain shape and size, hull and bran
diffusivity, and the endosperm chemical makeup may control fissure resistance
(Srinivas et al, 1977; Steffe and Singh, 1980; Patil, 1988; Lu and Siebenmorgen,
1992; Juliano et al, 1993; Sarker et al, 1996).
The measurement of rice kernel fissures was reviewed by Bhattacharya (1980).
The methods described were of varying levels of sophistication, but all consisted of
light being transmitted from below a sample and manual counting of the number of
kernels with fissures or particular types of fissures. More recently, high-speed
microscopy imaging has been used to study the association of internal stress with
kernel fissuring (Jia et al, 2002). Also, an image-analysis (machine-vision) system
was developed that was able to quantify 94% of the fissures in a medium-grain
sample and 100% of the fissures in long-grain rice (Lan et al, 2002).
Another aspect of milling quality is the degree to which a given sample has been
milled. Degree of milling is a quantification of the amount of bran removed from
kernels during the milling process. The majority of consumers around the world
prefer well-milled rice that has little to no bran remaining on the kernels. This is
ironic considering that unmilled (brown), rice compared to milled rice, contains
more protein, lipids, vitamins, minerals, and phytochemicals with potential health
benefits. While nutritionally inferior, milled rice does have a longer shelf life than
unmilled rice.
Samples milled to different degrees reportedly have varying functionality and
sensory properties. For example, Champagne et al (1990) and Marshall (1992)
reported that a greater degree of milling was associated with decreased onset and
peak-gelatinization temperatures. Reduced cooking time and higher water-binding

capacity, swelling power, and peak viscosity were found for rice milled to a higher
degree (Champagne et al, 1990; Kim and Jeon, 1996; Perdon et al, 2001). Sensory
panel research revealed that flavor intensities of cooked rice milled to different
degrees were dependant on cultivar, growing location, and moisture content
(Champagne et al, 1997). Cooked rice with a higher degree of milling was found to
have lower corn flavor, raw-rice flavor, wet-cardboard flavor, haylike flavor, and
bitterness (Park et al, 2001). This sensory report also found greater degree of mill-
ing to be associated with increased agglomeration, adhesiveness, cohesiveness of
mass, inner moisture, and tooth packing, while hardness and chewiness decreased.
Therefore, it is important that only samples that have been milled to the same
degree be compared in studies related to end-use quality. This, however, is seldom
documented in published research.
The numerous methods developed to determine degree of milling have been
reviewed (Barber and Benedito de Barber, 1979). These methods can be categorized
into two groups: those that assess the amount of bran remaining on milled rice and
those that measure chemical components in the rice kernel’s outer layers. Visual
estimation of the bran remaining after milling is performed by comparing milled
samples to established rice standards, or by doing so after staining kernels with
various bran-specific dyes (Tani et al, 1952; USDA, 1995). Although, these
visualization methods are still in use, they rely on a person’s subjective judgment
and are thus susceptible to a high level of method error. Colorimetric determination
of extracted bran pigments has also been reported to be useful for determining the
amount of bran left after milling rice (Bhattacharya and Sowbhagya, 1972a, 1976).
Chemical components that decrease with greater degree of milling and have been
used to quantify this grain characteristic include lipids, ash, silica, protein, thiamin,
and fiber (Barber and Benedito de Barber, 1979). These methods suffer from error
because cultivars differ in the concentration gradient of these compounds (Shams-
Ud-Din and Bhattacharya, 1978). They also are not practical because most of them
are technically demanding and time-consuming. Nevertheless, in the last few
decades, extracting the amount of lipid in the outer layers of 10 g of milled rice
using a Goldfisch apparatus (i.e., determining surface lipids) has commonly been
used to determine degree of milling of rice (Hogan and Deobald, 1961). According
to Shams-Ud-Din and Bhattacharya (1978), this technique is suitable only for rice
that has been milled beyond the stage of being undermilled. Some have sought to
shorten the time of analysis for surface-lipid measurements. For example, a rapid
gas-chromatographic measurement of surface lipids that requires only 500 mg of
sample has been developed (C. Bergman, unpublished data). Also, Lam and Proctor
(2001) shortened the time of analysis for the standard Goldfish surface-lipid
measurement by changing the extraction solvent from the standard solvent,
petroleum ether, to the more polar isopropanol.
Others have reported methods to predict surface-lipid content rather than actually
measuring this characteristic. For example, Chen et al (1997) developed near-
infrared (NIR) reflectance calibration equations to predict surface-lipid content of
milled rice. Studying three cultivars, they reported the best equation utilizing visible
and NIR wavelengths, modified partial least squares, and pretreatments of standard
normal variate and first derivative. Liu et al (1998) developed a digital image-
analysis system that was able to predict (R2 = 0.95) the surface lipid content of one
cultivar milled to different degrees of milling. Gangidi et al (2002) reported the use
of diffuse-reflectance Fourier-transformed infrared spectroscopy to predict degree

of milling, as measured by solvent extraction of surface lipids. Analysis showed a
high degree of correlation (R2 = 0.96) between surface-lipid content and the spectra
in the 4000–400 cm (–1) range for one long- and one medium-grain cultivar. Where
most of these predictive method reports fall short is in the very small number of
cultivars that each studied. Rice germ plasm is so variable in terms of end-use qual-
ity characteristics that, to validate any predictive method, many cultivars with dif-
fering genetic backgrounds need to be evaluated.
Other recent developments in the measurement of degree of milling include the
use of image analysis to quantify the whiteness of a sample (Yadav and Jindal,
2001). Optical instruments, such as the Satake Milling Meter, quantify the
whiteness and translucency of a sample and compute a degree-of-milling value. A
whiteness score of 40 is considered by many in the U.S. rice industry to be a well-
milled sample. These instruments are very rapid, but their utility is hindered if
chalkiness is present or if the sample is inherently dark in color. The percentage of
bran by weight lost after milling is currently being studied by the American
Association of Cereal Chemists (AACC) to determine what percentage loss of
weight is suitable for describing over-, well-, and undermilled rice (C. Earp,
personal communication).
Kernel Dimensions
Consumers in some countries associate sensory quality type with kernel dimen-
sions, thus making this trait important to the entire rice industry in these countries.
Different segments of these rice industries use kernel-length and -width data
calculated from either rough, brown, or milled rice. Breeding programs typically
focus on brown rice during cultivar development to avoid the variation that can be
introduced into the data from the use of different mills or milling procedures. Table
1 shows the definitions for brown-rice kernel length, length-width ratio, and weight
that are used by U.S. rice breeders during cultivar development (Adair et al, 1973).
Food companies generally use milled-rice kernel dimensions in their purchasing
specifications, since milled rice is what will actually be used in their processing.
The USDA Grain Inspection, Packers and Stockyard Administration-Federal Grain
Inspection Service uses rough-, brown-, and milled-rice kernel length-width ratios
for grading purposes (Table 2; USDA, 1994).
Kernel dimension data can be collected by hand using a ruler or calipers and also
using various forms of instrumentation. In terms of instrumentation, digital image-
analysis systems are primarily in use. These systems consist of uniform diffused
TABLE 1
Size and Shape Classes Used as Guides in U.S. Rice Breeding Programsa
Brown Rice Grain
Length Length-Width Weight
Grain Type (mm) Ratio (mg)
Long 6.61–7.5 3.1 and more 15–20
Medium 5.51–6.6 2.1–3 17–24
Short up to 5.5 2.0 and less 20–24
a Data from Adair et al (1973).
TABLE 2
Length-Width Ratios Used to Determine Grain Typea
Length-Width Ratio of
Rice Form Long-Grain Medium-Grain Short-Grain
Rough 3.4 and more 2.3–3.3 2.2 and less
Brown 3.1 and more 2.1–3.0 2.0 and less
Milled 3.0 and more 2.0–2.9 1.9 and less
a Data from USDA (1995).
light focused on a sample; an image is taken with a camera and converted from an
analog to a digital signal, and a computer is used to calculate the desired dimen-
sions. Static and automated systems have both been reported (Paige et al, 1991;
Lloyd et al, 2001). These systems are not designated specifically for rice; thus,
some application development time is generally needed.
The world’s rice germ plasm exists in a wide array of sizes and shapes. Classi-
fying rice as long, medium, or short grain is a useful practice when developing cul-
tivars to fit defined market classes such as in the United States. However, these
definitions are not meant to, nor are they able to, describe all possible size and
shape combinations that exist across the world. The surface area per unit weight of
a rice sample (S, cm2/g) is able to capture the diversity in kernel morphology. In
addition to being a means of categorizing rice, surface area is an important physical
property of rice because it affects water uptake and cooking time, as discussed later.
A sphere has the least area for any given weight. As the roundness decreases and
the shape becomes elongated, flattened, or irregular, the value of S increases. In
addition, the surface area per unit weight of a particle increases as its particle
weight decreases (compare a marble, a tennis ball, and a soccer ball). A method to
classify rice into classes such as superfine, fine, medium, and coarse by its S value
was reported by Bhattacharya et al (1982b). This parameter can be approximated
using the method of Bhattacharya and Sowbhagya (1971), which was based on a
suggestion made by Husain et al (1968). They assumed that the rice grain was an
approximate cono-ellipsoid, i.e., a uniform ellipsoid (half the length of the rice)
with two cones on its two ends (each a quarter grain length). Therefore surface area
of one rice grain (s) is:
/ Ë /Û
V= 3 + Ì 3 Ü
Í Ý

= /3

% + 7
= /p PP

in which L, B, and T = length, breadth, and thickness of the grain in millimeters and
P = perimeter in millimeters.
Then the surface area per gram (S) is
6 = V 1 FP

J

in which N = number of kernels per gram.

This parameter requires the measurement of L, B, T, and grain weight. A simpler

but nearly as effective parameter for classification of rice is “normalized grain
weight” (10 W/L, mg/cm2), in which W is the mean grain weight in milligrams. It
expresses what the grain weight would have been had the grain, in its existing
shape, been exactly 1 cm long (Sowbhagya et al, 1984).
Grain Color
Rice chalkiness appears as a white (opaque) portion of a kernel or can make up

the entire kernel. This characteristic occurs in all cereal grains (Raju et al, 1991).
Rice kernels free of chalk and very translucent are desired by most segments of the
rice industry. The major exceptions are found in those that utilize waxy rice, which
is entirely opaque, and arborio types, which have a chalky core. Chalky kernels
affect the physical and functional properties of a rice sample (Lisle et al, 2000).
Both genetics and the environment play a role in controlling rice kernel chalk con-
tent, and thus breeders select against this trait throughout cultivar development
(Tashiro and Ebata, 1975; Tashiro and Wardlaw, 1991).
Chalk is expressed in a number of different ways. Some methods measure the
percentage of kernels containing any amount of chalk, while others report the per-
centage having a specific amount of chalk. Still others differentiate the percentage
of kernels containing chalk in varying places within the kernel: “white belly” indi-
cates that the ventral side contains chalk; “white core” or “white center” extends to
the edge of the ventral side and toward the center; and a streak along the dorsal side
is called “white back” (Ikehashi and Khush, 1979). Chalk measurements are most
commonly made using subjective visualization of kernels placed on a light box;
however, some breeding programs now rely on digital image analysis to quantify
chalk levels. For example, Reece and Blakeney (1993) reported a digital image-
analysis method that expresses chalk as the proportion of opaque to translucent
areas in a single layer of milled rice kernels.
Consumers also prefer kernels that are very white rather than inherently gray or
discolored from environmental effects such as stink bug infestation or kernel smut.
Whiteness can be measured using light meters such as the Satake Milling Meter and
the Kett Whiteness Meter, as well as digital image-analysis systems (Yadav and
Jindal, 2001). Inherent kernel whiteness should not be confused with degree of
milling. That is, differences in degree of milling affect kernel whiteness, but when
various samples are milled to the same degree of milling they can still show varia-
tion in whiteness. For example, the U.S. hybrids XL 6, 7, and 8 are inherently
whiter than many conventional U.S. long grains grown in the same location and
milled to the same degree (C. Earp, personal communication).
FUNCTIONAL PROPERTIES
Hydration During Cooking
Since rice is cooked in water before consumption, it is not surprising that studies
of rice quality have often begun with examination of its hydration properties. The
amount of water absorbed by a known quantity of rice when cooked in water in a
boiling bath for a defined time (usually 15–20 min) has been expressed as “swelling
number,” “water uptake,” or “water absorption ratio” (grams of water absorbed per
gram or 100 g of rice), “swelling ratio” (grams of cooked rice per gram or 100 g of
uncooked rice), or “volume expansion ratio” (volume of cooked rice per gram or
100 g of uncooked rice). The physical and chemical properties controlling the
amount of water absorbed by rice during the cooking process have been studied
extensively.
The earliest studies of rice hydration during cooking were by Sreenivasan (1938,
1939) and Sanjiva Rao (1938, 1948). Sanjiva Rao et al (1952) concluded that high-
quality cultivars had both a high amylose content and a high swelling number. This
conclusion can be understood in retrospect, since people from India prefer firm and
fluffy cooked rice (high amylose content) and small, slender grains (high water
uptake). This conclusion unwittingly established a paradigm of rice quality that
influenced subsequent thinking for decades. Water absorption began to be routinely
tested and reported upon in most research reports and breeding program
evaluations. Many understood water uptake literally as implying hydration ability, a
consequence of some intrinsic chemical property. The association originated in
observations that slender varieties (generally indica and high-amylose) showed
higher water uptake than shorter and rounder varieties (generally japonica and low-
amylose). This association then tended to suggest that a high-amylose rice absorbs,
or needs to absorb, more water to cook properly.
Parthasarathi and Nath (1953) cooked rice for different periods and suggested
that when 2 g of rice had absorbed 4.5 g of water (which means a water uptake
value of 2.25), the rice was optimally cooked, i.e., it had no hard center when
pressed between the fingers. Batcher et al (1956, 1957) and Halick and Keneaster
(1956) observed that the water uptake after cooking rice in excess boiling water for
a predetermined time was greater in long- than in short- and medium-grain U.S. rice
types. Interestingly, the value was practically constant within a grain type, regard-
less of differences in other properties among the cultivars. Hogan and Planck (1958)
and Halick and Kelly (1959) noted that water uptake at 70–80°C was in the reverse
order, being more in short and medium grains than in long grains. Also, the 70–
80°C value, unlike the 100°C value, differed among cultivars within a grain type in
accordance with their other properties. Halick and Kelly (1959) commented that
while the 70–80°C value reflected the compositional characteristic, “it is suggested
that at high temperatures water uptake may become a function of size and shape of
the endosperm, and cultivar differences in composition are then masked.”
Bhattacharya and Sowbhagya (1971) studied the hydration phenomenon in 65
nonwaxy samples. They started with the premise that, since rice was never cooked
(in excess water) in Indian homes for a given time but only to a given end-point
(disappearance of a hard core), cooking rice for a predetermined time—as may
appear mandatory in “scientific studies” in a laboratory—was misleading. They
argued that water uptake values should be compared after cooking each cultivar for
its own “optimal cooking time,” as defined by the Desikachar and Subrahmanyan
(1961) test, viz., the time at which the opaque central core disappears when the
grain being cooked is pressed between two glass slides. The authors confirmed that
water uptake after cooking for a fixed period of 20 min (W20) indeed differed widely
among cultivars. The data indicated that W20 was inversely related to the grain
thickness of the cultivar (r = –0.832***, n = 20). Accordingly, the “optimum cook-
ing time” was directly proportional to the grain thickness (r = 0.877***), and W20
was inversely related to the optimum cooking time (r = –0.843***). On the other
hand, when the samples were cooked for their individual optimum cooking times,
the resulting “optimal-time water uptake” (Wot) was nearly constant at around 2.5 g
of water absorbed per gram of rice, regardless of the large differences in their W20
values, size, shape, thickness, or chemical properties. All the cooked samples
attained a nearly constant moisture content of 73–74%. This last conclusion about
nearly identical rice moisture content, at optimal cooking time, among cultivars,
irrespective of their amylose, gelatinization temperature (GT) and gel consistency
was later confirmed by Juliano et al (1981a). Clearly the “water uptake” parameter,
as defined (i.e., the value after a given amount of time), had a restricted meaning. It
was, in reality, an expression of the hydration at the indicated time, i.e., of the rate
of hydration of the sample, not of its hydration ability, nor of its affinity for water.
Bhattacharya and Sowbhagya (1971) then suspected that perhaps the surface area
per unit weight of the rice was the determining factor that would explain why the
value varied inversely with the grain size, including grain thickness. The authors
calculated the S of the cultivars by using the approximate formula of Husain et al
(1968). Indeed, W20 turned out to be highly correlated with S (r = 0.747***; n = 45).
This explained why small and slender grains had a greater water uptake value (at a
given time) than large and round ones. In agreement, broken rice absorbed water
much faster than whole grains. Cracked and chalky grains also hydrated faster. The
authors further noted that the W20 values of their samples were related neither to the
amylose (17–29%, dry basis) nor protein (6.0–11.5%) contents nor to the alkali
score (0–8: an inverse index of the GT) of the samples. But the W-S ratio (weight of
water absorbed per unit of surface area) showed a small range (coefficient of
variation, 8.2%) and a low correlation with the protein content (r = –0.421**) and
the GT (r = –0.353*; n = 45). The authors concluded that water absorption by rice
during cooking in excess boiling water for a given time was a function
predominantly of the physical size and shape of the grain (as well as cracks and
chalkiness).
The authors also studied hydration at 70–80°C. The results were strongly
affected (inversely) by the GT of the samples. In fact this value could be used to
index the GT. However, they argued that since water absorption was also affected
by the grain surface area, it would be best for a correct estimation of GT to express
the 80°C value as a percent of that at 100°C (whereby the effect of surface area
would be cancelled). Bhattacharya et al (1972, 1982a) confirmed that the above
ratio (W80°/W100°) was highly correlated with the GT (r = –0.791***, n = 144) as
well as with the alkali digestion score (r = 0.924***, n = 40).
The results of Hogan and Planck (1958) and Halick and Kelly (1959) can be
understood in the light of the results described above. Because of the lower GT of
the medium- and short-grain cultivars than the long-grain ones, the former types
absorbed more water at 70–80°C than the latter. The 70–80°C value also brought
out the atypical cultivars within each grain type because of the differences in their
GT values. But the 100°C value, being dependent mainly on the surface area and
largely independent of GT or amylose content, was higher in the long grains than in
the other two types and also failed to differentiate the atypical cultivars within a
type, which had very different GT and/or amylose values.
Juliano et al (1965), after cooking rice in an automatic electric cooker, concluded
that water uptake of rice was inversely related to the protein content as well as to
the GT. As a result, high-protein or high-GT rice was believed to need a longer
cooking time and more water to cook satisfactorily. The authors mentioned that
amylose, although not a dominant factor in cooking, did influence the cooking time
positively if calculated independent of protein (partial correlation analysis). Later,

Juliano (1971a, 1972) mentioned that the water absorption and volume expansion
by rice during cooking were positively related to the amylose content and that, for
optimum texture (not defined), more water was needed to cook high-amylose than
low-amylose rice (see also del Mundo, 1979; Juliano, 1979; Perez and Juliano,
1979).
Juliano et al (1969) reported that, after cooking rice in excess water in a boiling
bath, the time needed for cooking was longer by at least 1 min for high-GT rice than
for low-GT rice. At the same time, GT was found to be unrelated to water uptake or
volume expansion ratio. After cooking 10 samples of rice in excess water for their
individual optimal cooking times (i.e., until the opaque core disappeared), Juliano
and Perez (1983) concluded that the cooking time was positively related to the GT
(r = 0.75*) but was not related to the surface area (r = –0.49NS). It would be difficult
to explain the results in Bhattacharya and Sowbhagya (1971) based on these consid-
erations. For instance, cv. Century Patna 231 has low amylose content and high GT;
Toro also has low amylose content but low GT; conventional U.S. long grains have
intermediate amylose content and GT. Yet all of them have more or less equal
water uptake. The same applies to Early Prolific, which (among medium grains) has
a high GT. These results stand explained on the primary basis of size and shape, i.e.,
surface area per unit weight.
The above differences in conclusions can be partly explained. The results of
Bhattacharya and Sowbhagya (1971) pertain to the cooking of rice by putting it
directly into excess boiling water. In this situation, water uptake (at a given time) is
related primarily to the grain size and shape. But when cooked in a cooker, the rice
is exposed for the initial several minutes to a temperature much below that of boil-
ing. Room-temperature hydration of rice is inversely related to amylose content and
GT, and a low-GT rice would start rapid hydration as the temperature passed
through the 60–80°C range (Bhattacharya et al, 1979). A high-amylose and/or high-
GT rice would therefore need a longer time to cook (or would absorb less water in a
given time) than a lower-amylose or low-GT rice when put in ambient water and
gradually heated, and vice versa. These issues may partly explain the persistent
belief that high amylose content or high GT lowers the water uptake of rice or
increases its cooking time, and vice versa.
Desikachar and Subrahmanyan (1961) found that rice grains developed cracks
when soaked in water and that such rice absorbed water and cooked faster than rice
directly cooked without soaking. The presoaked grains also elongated more than
those that were directly cooked. These facts were confirmed by Sowbhagya and Ali
(1991). Desikachar et al (1965) showed that the bran layer in brown rice was a bar-
rier to water absorption during cooking and that a certain minimum degree of mill-
ing was necessary for rice to absorb water unhindered.
Water uptake (W, grams of water absorbed per gram of rice at a given time) can
be determined by the method adopted by Batcher et al (1956, 1957) or by that of
Bhattacharya and Sowbhagya (1971). In the latter case, rice is put directly into
water in a preheated tube and cooked for the desired time. The rice is then strained,
surface-dried over filter paper, and weighed. The ratio of the weight of cooked rice
to that of the original rice gives the apparent swelling ratio (SR). The apparent
water uptake (W) is the apparent swelling ratio minus one (SR–1). Determination
of the true water uptake (W, i.e., corrected for solids lost during cooking) requires
the estimation of the moisture contents of the cooked as well as the original rice:
Fig. 2. Photograph of rice samples cooked in graduated tubes, illustrating the test for volume ex-
pansion. Tube 1 contains freshly harvested rice; tube 2 contains aged rice of the same variety; and tube
3 contains rice of tube 1 treated by an accelerated aging process described by Bhattacharya et al,
(1964). (Courtesy K. R. Bhattacharya)
: = 0 +-00
F R
in which Mc and Mo are the moisture contents (dry basis, g/g) of the cooked and
original rice, respectively.
Volume expansion during cooking has not been used to study cultivar differ-
ences; however, it has been reported to be a useful tool for studying rice aging
(Desikachar, 1956). The test is conducted as follows. Rice and distilled water are
placed into a round-bottom graduated boiling tube. The tubes are placed in an auto-
clave in which the water is vigorously boiling. The autoclave is then covered, but
not closed, allowing the rice to be cooked in steam, at atmospheric pressure, for 45
min, and the volume is then read (Fig. 2).
Cooking time can be determined by the Desikachar and Subramanyan (1961)
method. A small quantity of rice is put directly into water preheated in a test tube
kept immersed in a vigorously boiling water bath. A few grains are removed at
intervals and pressed between two glass slides. The time at which the opaque cen-
tral core just disappears is the cooking time. In the literature, this test is often
referred to as the Ranghino (1966) test.
Loss of Solids During Cooking
Another factor of importance during the cooking of rice is the amount of dis-
solved and undissolved solids lost into the cooking or processing water. This prop-
erty has been examined by several groups (Batcher et al, 1956, 1957; Chikubu et al,
1960; Borasio et al, 1964; Hampel 1965, 1967; Juliano et al, 1969), but no definite
agreement exists among their conclusions. However, it does seem that the loss of
solids during cooking is related positively to the water uptake and inversely to the
amylose content. Hence, after cooking for a definite time, samples may fail to give
a clear trend, for long-grain rice would often have a high amylose content but a
higher water uptake as compared to rounder and bigger grains, and vice versa.
When cooked for their individual optimum cooking times, however, low-amylose
rice should be expected to lose more solids. However, there may be more control-
ling solids loss than just water uptake and amylose content. This becomes clear
when one compares superior-processing U.S. long-grain cultivars (e.g., Bolivar and
L205) with other U.S. high-amylose types (e.g., Jodon and L202). These types both
have similar grain dimensions and amylose content, but the superior-processing
types have lower solids loss during processing. Thus, either chemical constituents
other than amylose cause this difference (e.g., differences in content of amylose-
lipid complex), or similar values determined by the amylose assay may result from
different chemical components or differing interactions among constituents.
Solids loss during cooking or after processing can be determined by collecting
the excess cooking water, adding the washings of cooked rice to it, making up the
volume, and then drying a portion of it in the oven and weighing (Batcher et al,
1956). Webb and Adair (1970) and Webb (1979) described a method for determin-
ing solids loss during the canning of rice. A method to calculate solids loss (s)
during cooking comes from the true (W) and apparent (W) water uptakes by the
formula (Indudhara Swamy et al, 1978):
V = : +-::
Gelatinization Temperature
During rice processing and cooking, starch granules absorb water and swell.
Over a critical temperature range, the starch granules undergo an irreversible proc-
ess known as gelatinization, which is marked by crystalline melting (loss of bire-
fringence) and starch solubilization. Birefringence is a phenomenon seen when
starch granules are viewed under a microscope with polarized light. The refraction
of polarized light by the crystalline areas in starch granules results in a characteris-
tic “Maltese cross” pattern. The disappearance of 90–95% of these crosses on heat-
ing a starch suspension is used to determine the endpoint GT (Schoch and Maywald,
1956). Rice has been reported to have GTs from 55 to 79°C (Bhattacharya, 1979).
Although, through the years, there has been much disagreement over the
biochemical basis of rice starch GT, amylopectin fine structure was recently shown
to be associated with this thermal property, as discussed below.
Determining starch GT with polarized light and a microscope is impractical for
large numbers of samples. Consequently, other methods have been developed that
either predict GT or measure other physical phenomena such as viscosity changes that
occur as starch takes up water and swells. GT methods previously reported were
reviewed by Bhattacharya (1979), and those currently in use are described below.
An amylograph method for determining rice GT was developed by Halick et al
(1960) and accepted by the AACC as an official method (AACC, 2000). This
method determines viscosity changes that occur during the heating and stirring of a
rice flour-water mixture. The temperature at which the viscosity of a 20% slurry
increases is correlated with a loss of starch birefringence. An instrument designed to
determine viscosity changes during the heating of relatively small samples (6 g) is
the Rapid ViscoAnalyser (RVA), made by FOSS NorthAmerica. The AACC Rice
Technical Committee is studying the association between the onset of GT deter-
mined using this instrument and the Brabender Amylograph (C. Earp, personal
communication).
Another instrument commonly used to determine GT is the differential scanning
calorimeter (DSC). The DSC consists of two calorimeters that are heated at a con-
stant rate and maintained at the same temperature. The reference calorimeter is
empty, while the other contains the sample under study. Water-rice samples that
have been sealed in pans are placed in a calorimeter, which is heated. The differ-
ence in power needed by the calorimeters to cause the identical temperature change
is measured. This difference in power is essentially the heat flow into the sample.
Data collected for rice kernels using the DSC are typically the onset temperature
(To), peak temperature (Tp), conclusion temperature (Tc), and enthalpy of gelatiniza-
tion (DH) (first inverted peak), as well as the same parameters for the amylose-lipid
complex melting transition (second inverted peak) (Fig. 3). Gelatinization within a
sample occurs across a range of temperatures and not at one exact temperature
(or time), even though typically one temperature (i.e., onset or peak) is reported
as a sample’s GT. Variations in this technique include the water-rice ratio
(1.5:1.0–2.0:1.0) used and the sample heating rate (i.e., 1°C/min to 10°C/min).
Also, some laboratories analyze whole kernels in pans able to hold up to 270 mg
of presoaked rice, while, more commonly, 5–10 mg of a rice flour or starch slurry
is studied (Normand and Marshall, 1989; Singh et al, 2000). After analysis,
sample pans can be stored and rescanned, and the enthalpy of the samples’
retrogradation (i.e., recrystallized amylopectin melting) can be studied (Karim et
al, 2000).
The alkali spreading (digestion) value assay is used as a prediction of GT. This
rapid method is primarily used by rice-breeding programs. Two variants of this test
exist. The first is based on work by Little et al (1958), who reported differences in
Fig. 3. Differential scanning calorimeter thermal curves of rice cultivars with low (Bengal, dotted line),
intermediate (Cypress, dashed line), and high (Jacinto, solid line) gelatinization temperatures.
disintegration of milled rice soaked in 1.7% KOH. Simpson et al (1965) found a

close association (r = –0.80) between milled rice GT and disintegration in KOH.
Genotypes are typically classified as high-GT (score of 2), intermediate-GT (score
of 3–5), or low-GT (score of 6–7) types, which correspond to >74, 70–74, and
<70C, respectively (IRRI, 1964). In some instances, the effect of growing
conditions on rice kernel GT results in all types having low alkali spreading
values. In such instances, the concentration of KOH can be reduced (e.g., to
1.4%) to improve the test’s power of discrimination. To reduce the sample prepa-
ration time required for making breeding selections, this method for milled rice
has been modified for use with dehulled (i.e., brown) rice. This modification
includes the use of 1.9% KOH and incubation overnight at 40C. Across 25
cultivars that varied from a score of 2.0 to 6.9, the correlation (r) between the
alkali spreading score determined using the method of Little et al (1958) versus
its modification reported above was 0.98 (C. Bergman, unpublished data). Figure
4 demonstrates brown rice samples that were scored as having alkali spreading
value scores from 2 to 7.
Another modification of the alkali spreading test was reported by Bhattacharya
and Sowbhagya (1972b). They observed that each rice cultivar was attacked by just
a certain KOH concentration and was fully digested by a KOH concentration 0.7–
0.8 percentage points higher. Also, each sample passed through a series of stages of
digestion over this KOH range. On this basis, they devised a nine-point score card
and proposed 1.4% KOH for the test, as it gave the widest range of scores among
the varieties. It reportedly provides a finer discrimination among cultivars, among
GT groups, and among grains in a given sample. The initial score card was slightly
modified and simplified for the purpose of an international cooperative test (Juliano
et al, 1982).
A simplified method to predict actual GT values was reported by Bhattacharya et
al (1982a). They developed equations with which to calculate the GT (y) from the
alkali score (x), using the formula:
Fig. 4. Samples representing each possible response (scores of 2–7) to the alkali spreading test
modified for brown rice.
\ = - [U = H[FOXGLQJZD[\YDULHWLHVQ =

RU \ = - [U = LQFOXGLQJZD[\YDULHWLHVQ =
The alkali spreading test is also useful for detecting segregation in a breeding
population or adulteration of a sample with a cultivar of a different GT type. Test-
ing 30 or more grains instead of the standard six grains is preferable in such cases.
Gel Consistency
The gel consistency test was developed to differentiate between high-amylose

types of rice with differing processing qualities, as measured by the Brabender
Amylograph (Cagampang et al, 1973). Ground rice is placed in a tube with bromo-
thymol blue and 0.2N KOH; the contents of the tube are vortexed; the tube is heated
in a boiling water bath; and the tubes are vortexed again and cooled in an ice bath.
The cooled tubes are laid flat on graph paper, and the length of the gel is deter-
mined. Gel consistency values are classified as soft (61–100 mm), medium (41–60),
or hard (26–40 mm). As with most rice quality measurements, degree of milling has
an effect on the measurement of gel consistency (Perez, 1979). Therefore, check
samples used to assess the method’s performance should be milled to the same
degree as the unknown samples. A crucial step in this test is to see that all material
is dispersed before the tubes are heated. Incomplete dispersion of samples results in
a softer gel. Juliano et al (1980) describe the results of an international cooperative
test. Reproducibility for the soft- and hard-gel types was good, but it was poor for
the medium-gel types. The method is affected by a large amount of operator error,
as well as by the age of the rice. Some breeding programs have replaced this test
with the RVA.
Pasting Characteristics
Rice exhibits a large degree of variation in starch quality, which affects the
texture of the cooked product and its use in various industrial processes.
Traditionally, the Brabender Viscoamylograph has been used to assess rice pasting
properties (Halick and Kelly, 1959; AACC, 2000). The amylograph measures and
records viscosity changes that occur during heating, holding, and cooling of a
starch-water slurry. This test requires 50 g of rice flour and run times of 1.5 hr.
More recently developed instruments that are able to simulate the Brabender
Viscoamylograph are the RVA and the Brabender Micro-Visco-Amylograph. These
instruments use a relatively small sample size (e.g., 3 g for the RVA) and a
shortened amount of time (e.g., 12.5 min for the RVA) (Blakeney et al, 1991,
Blakeney 1992).
In several countries, such as Australia, China, Japan, and the United States, the
RVA has become the standard method with which the rice processing industry and
breeding programs determine rice pasting properties. This method, like the
Brabender Viscoamylograph method, shows clear differences between various
styles of rice. Because of its small sample size requirement, it is suitable for
screening progeny in breeding programs. For example, Figure 5 displays RVA
pasting profiles for two cultivars with similar amylose content but of different proc-
essing qualities. Dixiebelle is a superior-processing type suitable for use in products
such as canned soup, while Jodon is not suitable for use in such products. A large
RVA setback is indicative of a superior-processing type. Also, the RVA is able to

determine the “degree of cook” after processing rice into precooked and extruded
products (Blakeney et al, 1991). Champagne et al (1999), studying a set (n = 87) of
rice samples of different quality types, found low associations between RVA data
and instrumental texture and trained sensory panel texture results. Thus, the RVA
was reported to be a poor predictor of cooked-rice texture. However, in this study,
the rice was cooked at different water-to-rice ratios depending on the type of rice,
while the RVA was used at a constant water-to-rice ratio across all samples studied.
Cooked-rice texture is known to be highly influenced by the amount of water in
which the rice is cooked (Juliano and Perez, 1983). In a later study (E. Champagne,
unpublished data), 17 diverse cultivars (with apparent amylose contents ranging
from 0 to 25%) were cooked at the same rice-water ratio (1:1.4). In this case, high
correlations were observed between some sensory texture attributes and RVA past-
ing properties. For example, correlations (r) between cold-paste viscosity and
stickiness to lips, stickiness between kernels, hardness, cohesiveness, uniformity of
bite, and cohesiveness of mass were –0.92, –0.90, 0.91, –0.85, –0.90, and –0.91,
respectively. Setback correlated well with roughness (r = 0.88) and hardness (r =
0.85). Pasting properties (i.e., final viscosity minus trough viscosity) have also been
reported to be a good predictor of the baking expansion of Japanese rice crackers
(Yamada et al, 1993).
In the initial stages of this procedure, the temperature is below the sample’s GT,
and the viscosity is low. When the temperature rises above the GT of the sample,
the starch granules begin to swell and the viscosity increases. The temperature at the
onset of this rise in viscosity is known as the “pasting temperature.” Granules swell
over a range of temperatures; the granules burst and the more-soluble starch leaches
out into solution. After the granules rupture, polymers align with each other due to
the mechanical shear, and the apparent viscosity of the paste is reduced. Peak
Fig. 5. Rapid Visco Analyser pasting profiles for two cultivars grown in the same location and with
similar amylose content (~25%). RVU = Rapid ViscoAnalyzer units. (Courtesy Newport Scientific
Pty. Ltd.)
viscosity is usually achieved soon after the heating cycle reaches 95°C. During the
holding period of the test, the sample is subjected for a time to constant temperature
(95°C) and mechanical shear stress. This further disrupts the granules, with accom-
panying starch leaching and subsequent polymer alignment. This period is accom-
panied by a reduction in viscosity; the minimum value reached is called the “hot
paste viscosity” or “trough.” The “breakdown” is the difference between the peak
and trough viscosities and is an indication of the stability of the starch to heat and
shear stress. As the mixture is subsequently cooled, reassociation between starch
molecules occurs to a greater or lesser degree. This results in the formation of a gel
and an increase in viscosity. The viscosity at the end of the test is called the “final”
or “cold paste” viscosity. This phase of the pasting curve, commonly referred to as
the “setback region,” involves retrogradation, or reordering, of the starch molecules.
Setback is often measured as the difference between final viscosity and peak
viscosity. Final viscosity is the most commonly used parameter to define a
particular sample’s quality, as it indicates the ability of the material to form a gel
after cooking and cooling.
The chemical components reported to affect starch pasting properties are amy-
lose, protein, and lipid contents, as well as amylopectin structure (Suzuki, 1979;
Juliano and Pascual, 1980; Han and Hamaker, 2001; Martin and Fitzgerald, 2002).
Pasting properties are known to be influenced by genotype and environment effects,
degree of milling, and postharvest handling (Suzuki, 1979; Zhou et al, 2003).
However, variance in pasting properties can also result from several steps in the
laboratory procedure. The use of different mills to prepare rice flour can result in
different particle sizes and damaged starch contents; both of these factors reportedly
influence pasting properties (Nishita and Bean, 1982). Some laboratories correct the
quantity of water they add for an amylograph procedure based on the sample’s
moisture content, and others do not. This can have an impact on rice flour pasting
properties. Storage age also affects the pasting properties of rice and, likewise, the
texture of cooked rice. Freshly harvested rice grains tend to become sticky and cling
together, whereas the aging process leads to a drier, firmer, and more flaky product.
These changes can be monitored by evaluating pasting properties, with age effects
evident in the peak and setback (as the difference between peak and final viscosity)
values (Blakeney, 1992; Sowbhagya and Bhattacharya, 2001). These effects were
similar to those obtained by Blakeney (1992) when the free fatty acid composition
of the flour was altered. Thus, when comparisons are made between the pasting
properties of various cultivars, samples should have been grown in the same location
using the same cultural management practices, be milled to the same degree, be
handled the same way, and be stored the same amount of time; in addition, the same
grinder should have been used to prepare all samples for analysis.
Some disagreement exists in the literature about the utility of the amyographic
procedures described above. The reasons are briefly explained below but have been
extensively discussed by Bhattacharya and Sowbhagya (1978, 1979), Sandhya Rani
and Bhattacharya (1995a), and Sowbhagya and Bhattacharya (2001). Mazurs et al
(1957) noticed that, when a flour or starch was viscographed in a series of very low
to very high slurry concentrations, the viscogram parameters went through a
sequence of changes. This can be seen from analyzing a similar set of curves, as
shown in Figure 6. First, the peak viscosity (P) increases exponentially with the
concentration. Second, there is no breakdown (BD = P – H, where H is the hot-
paste viscosity attained at the end of heating at 95°C) at very low concentrations,
Fig. 6. Brabender viscograms of one rice flour at different slurry concentrations (14% moisture basis,
total slurry = 500 g; amounts of rice flour are shown at the right of the curve). (Courtesy S. Z. Ali)
Fig. 7. Brabender viscograms of one rice cultivar from eight quality-type categories (groups I–VIII)
over a range of slurry concentrations. P = peak (solid line); H = hot-paste viscosity (x---x); C = cold-
paste viscosity (o– –o). (Reprinted, with permission, from Bhattacharya and Sowbhagya, 1979.
©Institute of Food Technologists)
but at higher concentrations, BD becomes evident. Third, the setback (SB = C – P,

where C is the final cold-paste viscosity after cooling to 50°C) is very high at low
concentrations (i.e., not in absolute value but as a ratio of P), then progressively
decreases, finally becoming negative at high concentrations. These changes are
brought out more dramatically if the P, H, and C values are plotted in log-log coor-
dinates against concentration, as shown in Figure 7. Clearly no cultivar as such is
intrinsically characterized by a high or a low breakdown, or a low or high or posi-
tive or negative setback, as generally believed. Any sample can be made to have
any of these characteristics depending on the slurry concentration chosen. Thus,
studying viscogram parameters at a fixed P value could provide useful information
about the functional differences between rice samples. This is what Mazurs et al
Fig. 8. “Relative breakdown” (breakdown/total setback) of rice of eight quality types (I–VIII) at
various peak viscosities. (Reprinted with permission from Bhattacharya and Sowbhagya, 1979.
©Institute of Food Technologists)
(1957) did using a graphing procedure, but the method was not practical for routine
work. Bhattacharya and Sowbhagya (1978, 1979) prepared viscograms at several
slurry concentrations and plotted the P, H, and C values in a log-log graph against
slurry concentration (Fig. 7). This figure brings out the gradation in properties from
quality type I to type VIII rice (referred to as Groups I through VIII in the figure).
Breakdown, setback, and total setback (SBt = C – H) at fixed P values were read off
from these plots. A new parameter called “relative breakdown” (BDr =
breakdown/total setback = BD/SBt) was proposed as a very sensitive parameter.
When the BDr values were plotted against the P values, the plots clearly
differentiated the different rice quality types from each other (Fig. 8).
Elongation Ratio
Relatively long cooked-kernel elongation is an important characteristic of the

basmati style of rice. The test commonly used to measure this trait is a modification
by Juliano and Perez (1986) of the original method by Azeez and Shafi (1966). The
principle is to presoak the uncooked rice in water, cook it for a set time, then
determine the ratio of the average cooked-kernel’s length to the average uncooked-
kernel’s length. Another modification of the original procedure was reported by
Hiranniah et al (2001). These authors used 0.05% tricalcium phosphate in the
cooking water to increase the ionic strength. The addition of this compound resulted
in less kernel breakage during cooking, thus facilitating accurate cooked-kernel
measurements.
Instrumental Measurement of Cooked-Rice Texture
Texture is the primary sensory property of cooked rice that controls its consumer
acceptance (Okabe, 1979). Textural properties can be measured via sensory analysis
or with instrumentation. Descriptive analysis of cooked rice texture, using a panel
trained in sensory profile terminology, provides the most reliable, complete data and
is considered a universal method (Stone and Sidel, 1993). Examples of the sensory
attributes of cooked rice measured using this technique include surface adhesiveness
to lips, cohesiveness of mass, and particle size (Champagne et al, 1998; Meullenet
et al, 1998). Other less-rigorous sensory evaluation methods have also been used to
study the texture of cooked rice (Perez and Juliano, 1979; Okabe, 1979; Chrastil,
1990; Rousset et al, 1995). However, the routine use of trained panels is hindered
by the large sample requirements and the cost of maintaining a trained sensory
panel. Consequently, researchers have evaluated instrumental methods designed to
predict the texture of cooked rice. Instrumental methods have also been developed
for use as quality assurance tools by the rice-processing industry and breeding
programs (Mohandoss and Pillaiyar, 1980; Lee and Peleg, 1988). These methods
have not been documented to correlate with trained sensory panel data, but rather
they have been designed to quantify, primarily, the hardness or firmness (and in
some instances stickiness, cohesiveness etc.) of cooked rice, as defined by the
instrumental method.
Kurasawa et al (1962, 1969) employed a table balance to measure the stickiness
of cooked rice. They measured the weight required to lift the pan of the balance
from its contact with the cooked rice to which it had been pressed. Manohar Kumar
et al (1976) also made a similar attempt but found the reproducibility poor. A par-
allel plate plastometer was used to study the viscoelasticity of cooked rice deforma-
tion (Chikubu et al, 1965; Chikubu, 1967; Endo et al, 1976). The cooked rice was
pressed between two parallel plates, and viscosity and elasticity were reported.
Nonglutinous rice had greater viscosity and elasticity than glutinous rice, and indica
rice more than japonica rice. Sugiyama et al (1990) and Yoshii et al (1993) further
refined this technique.
Arai et al (1981) proposed a six-element rheological model to predict textural
characteristics of cooked rice. Hampel (1961, 1965, 1966) in Germany used the
Haake consistometer to determine the consistency (i.e., the hardness) of cooked
rice. A plunger was pressed into the rice in a cylinder; when the rice back-extruded
through the plunger, the time needed to penetrate a given distance was measured.
The consistency was highly correlated with the amylose content. Manohar Kumar et
al (1976), Bhattacharya et al (1978), and Deshpande and Bhattacharya (1982) used
the same system and found the hardness of cooked rice to be very well correlated
with the total and water-insoluble amylose contents and the pasting characteristics.
Chikubu et al (1971) studied the General Foods Texturometer and suggested that
the index “hardness/adhesiveness” gave the best indication of rice texture. Okabe
(1979) evaluated the texture of Japanese rice samples with the above instrument and
demonstrated the measurement of hardness, cohesiveness, adhesiveness, gummi-
ness, etc., from the data. A texturogram was constructed from these readings that
showed a stickiness-hardness ratio of 0.15–0.20 to be the most acceptable to
Japanese consumers (Fig. 9). This instrument has been used in several other
research projects (Yoshikawa et al, 1974; Tanaka, 1975; Endo et al, 1976, 1980a,b;
Ebata and Hirasawa, 1982; Ebata et al, 1982; Suzuki et al, 1983a,b).
Fig. 9. Texturogram for cooked rice, showing zones of acceptability as a function of hardness and
stickiness. A = excellent; B = good; C = slightly poor, but acceptable; D = poor; E = unacceptable.
(Reprinted, with permission, from Okabe, 1979)
Scientists at the International Rice Research Institute, in the Philippines (Perez

and Juliano, 1979; Perez et al, 1979, 1993; Merca and Juliano, 1981; Juliano and
Perez, 1983) have used the Instron instrument for evaluating cooked-rice texture.
The maximum force generated during extrusion of cooked rice through the bottom
of a cell was used as a measure of hardness, while the force needed to detach a plate
after pressing a mass of cooked rice was measured as an index of stickiness.
Blakeney (1979) has reviewed the use of the Instron tester with various cells.
Mossman et al (1983) used the same method for measuring stickiness of rice, and
Fellers et al (1983) studied the effect of various heat treatments of rice on the
reduction of its stickiness.
The Chopin-INRA Viscoelastograph has been used to study firmness and elastic
recovery of cooked rice (Feillet et al, 1977; Laignelet and Alary, 1978; Laignelet
and Feillet, 1979). Cooked grains are pressed between two parallel plates, after
which the weight is withdrawn. The initial deformation and the final recovery of the
grains are noted. The values correlated well with the sensory attributes of rice as
well as with its amylose content. Sowbhagya et al (1987) and Sandhya Rani and
Bhattacharya (1995a) used the same instrument with a large number of samples and
found that both the parameters correlated with sensory, as well as several
physicochemical, parameters of rice.
Tsuji (1981, 1982, 1988) developed the Tensipresser instrument to measure
adhesiveness and hardness of cooked rice using a two-bite plunger. The ratio of the
adhesiveness/hardness value at 50% deformation to that at 90% deformation of the
rice grain provided the best index of the texture of rice. Ohtsubo et al (1998) and
Okadome et al (1998, 1999a) revised the method for this instrument using a
combination of a low (25%), a high (90%), and a continuous, progressive com-
pression system. This technique indicated that protein content played the most
important role in determining the surface hardness of cooked rice.
Manohar Kumar et al (1976), Bhattacharya et al (1978), and Deshpande and
Bhattacharya (1982) used a sieve test to measure cooked-rice stickiness. The
results correlated with sensory stickiness scores and were inversely related to
hardness scores. Lee and Peleg (1988) used a surface tensiometer to measure the
attractive force between cooked rice grains and observed differences between
sticky and firm rice samples. Chrastil (1990) also reported a method to measure
cooked rice stickiness. This author counted the number of various clusters (i.e.,
two, three, four, etc. grains lumped together) in a sample of cooked rice, and
plotted a frequency curve of the clusters. Stickiness was the mode (hmax) of this
frequency curve.
The results of an international cooperative test of several of the instrumental
methods described above were reported by Juliano et al (1981a). The authors
indicated that the instrumental methods were more sensitive than sensory results
and that they correlated well among themselves for hardness and stickiness
measurements.
The TA.XT2 texture analyzer (Stable Micro Systems, UK) is in use in many rice
laboratories of the world, particularly in the United States (Champagne et al, 1998,
1999; Meullenet et al, 1998, 2000; Meullenet and Gross, 1999; Lyon et al, 2000;
Sitakalin and Meullenet, 2000). Several of these laboratories are studying the ability
of the TA.XT2 to predict sensory data obtained using descriptive analysis method-
ology. Reports include the use of the instrument in compression mode as well as
with an extrusion cell. Weak to moderate correlations between these methods and
sensory texture profile data were found when using individual parameters such as
maximum load (Meullenet et al, 1998; Lyon et al, 2000). However, when Meullenet
et al (1999) used spectral stress-strain analysis, the predictive ability of data
obtained from the TA.XT2 in extrusion mode was superior to than that from previ-
ous studies. This approach treats force-distance curves as spectral data and creates
an instrumental “thumbprint” for each sample of rice tested. Models reported for the
two major texture characteristics, hardness and stickiness (i.e., adhesion to lips), had
relative predictive abilities of 0.85 and 0.76, respectively. However, several char-
acteristics such as toothpull and loose particles were not predicted using this tech-
nique. Sesmat and Meullenet (2001), using the TA.XT2 in compression mode,
along with partial least squares regression optimized with a stepwise method, were
able to predict (validation correlation coefficient >0.6) seven textural characteristics
evaluated by a trained sensory panel. The authors noted that, although compression
mode might be less reproducible and have a higher error of prediction than extru-
sion mode, the smaller sample size required makes it more likely to be used in a
breeding program. For any of these instrumental methods to be used by breeding
programs, however, textural targets to compare with breeding lines are needed (i.e.,
the desired range for each characteristic for each market class and specialty type of
rice must first be documented).
Cooking of Rice for Sensory or Instrumental Testing
The sensory and instrumental methods for evaluating rice texture require rice
that has been cooked in a reproducible and meaningful way. In fact, any method
designed to study cooked rice can only be as good as the cooking method it is
based on. Unfortunately there is no agreed-upon standard or official method for

preparing cooked rice. Juliano and Perez (1983) showed that the hardness of
cooked rice is dependent on its moisture content (shown as water-rice ratio in
Figure 10). So any nonuniform adjustment of the water ratio adds a significant
effect to a study. Yet one finds investigators using different water-rice ratios
dependent on the amylose class of each sample, and varied cooking methods are
also reported.
Juliano et al (1981a, 1982) and Juliano and Perez (1983) gave a lot of attention to
this issue, which is discussed in detail by Juliano (1985). This author mentions that
they adjusted the water-rice ratio for high-amylose rice to obtain an “acceptable soft
texture (less than 10 kg of Instron hardness).” It was suggested that the water-rice
ratio used for cooking should be 0.9–1.1 for waxy rice, 1.2–1.4 for low-amylose
rice, 1.5–1.6 for intermediate-amylose rice, and 1.7–2.0 for high-amylose rice. The
study of the effect of different water-rice ratios on the hardness values of cooked
rice (Fig. 10) reported that the above-mentioned ratios gave ranges of hardness val-
ues among the samples almost identical to those that would have resulted had an
identical ratio of 2.65 (final rice moisture, 75%) been chosen for all rices. None-
theless, these water-rice ratios are arbitrarily defined, and if the results are similar,
adjustment seems unnecessary. The ratios reported have not always been constant
either. For example, Juliano and Perez (1983) reported that in their program they
used ratios of 1.3, 1.7, 1.9, and 2.1 for the four types of rice mentioned above,
respectively. And in a survey, Juliano (1982) found that the water ratios being used
Fig. 10. Effect of water-rice ratio (uncorrected for steam loss) on cooked-rice Instron hardness of four
milled rices representing four amylose types. (Reprinted from Juliano, 1985)
by different investigators ranged from 0.8 to 1.3 (waxy), 1.1 to 1.7 (low-amylose),
and 1.5 to 2.5 (intermediate- and high-amylose rice). Since texture is dependent on
cooked-rice moisture content, the reported results are not comparable. An imaginary
example of the hazard of using arbitrary ratios can be shown by examining Figure
10. If one were to choose a water-rice ratio of approximately 1.6 for waxy rice, 2.1
for low-amylose rice, 2.3 for intermediate-amylose rice, and 2.7 for high-amylose
rice, all the samples would give an identical Instron hardness reading, leading the
investigator to proclaim that cooked-rice hardness is independent of amylose con-
stant, something known to be untrue.
The literature is not clear regarding the origin of the practice of adjusting the
water ratio to the amylose content. However, it may have originated from
researchers outside South Asia, those not accustomed to high-amylose rice. They
would have found this rice too firm, and therefore adding more water during
cooking would have resulted in a more “acceptable” texture. This practice then
may have gradually been transformed into a generalized rule of adjusting water
according to the amylose content of the rice. Del Mundo (1979) studied what was
described as the optimal amount of water to be used for cooking different types
of rice. Unfortunately the objective of the study was to cook rice to “similar
doneness,” which was defined by the subjective descriptions of “neither too dry
nor too soft.”
Based on the above discussion, the water-rice ratio used for studying cooked rice
should be decided on using the following considerations. First, as has been dis-
cussed before, all rices, regardless of type or composition, absorb approximately 2.5
times their weight of water when cooked in excess boiling water until the central
opaque core disappears. Based on this consideration, one logical approach would be
to cook each sample with precisely 2.5 times its weight of water. Second, Okadome
et al (1999a,b) have recently used a constant water-rice ratio of 1.6. While this ratio
looks arbitrary, it has the merit of being the same for all samples. So the
comparative position remains unchanged. Third, if consumer preference is under
evaluation, then cooking with adjusted water-rice ratios may be necessary if it has
been demonstrated that such is the actual practice in homes of different populations
or ethnic groups. An example that would justify varying the ratio of water to rice is
as follows. If one were studying the consumer preference for waxy-rice types in
northern Thailand and parts of Cambodia, the samples under study would need to be
cooked differently than in boiling water due to regional practices. In this region, rice
is soaked in ambient water overnight (the equilibrium water content would be
approximately 40%, including the adhering water), and then the soaked rice is
steamed on a perforated basket or cloth mesh. Thus, the final water uptake would lie
somewhere in the range of 0.5–1, and cooking waxy rice with such a ratio for its
testing could be justified.
The actual procedure used for cooking rice is another variable on which there
is little agreement. Cooking rice in a beaker, in an automatic cooker, and
indirectly by steam have all been reported. If a precise water-rice ratio is used,
then the cooking should preferably be indirectly by steam, so that the entire
amount of water is actually absorbed by the rice. Even in an automatic cooker,
some loss of steam occurs, thus altering the ratio. The uniformity of absorption of
water by all the rice grains in a sample can be ensured in one of two ways. One is
to cook the rice in a shallow layer (Bhattacharya et al, 1978; Deshpande and
Bhattacharya, 1982). The other way is to cook rice in excess water for the
cooking time of each sample when using a water-rice ratio of 2.5. The cooking
time would need to be predetermined using the method of Desikachar and
Subrahmanyan (1961) or Sowbhagya and Ali (1991). Some techniques require
that rice be cooked in a deep container such as a beaker or a cooker, but with a
given water-rice ratio. In such a case, a clear gradient of moisture content would
form from the top (least) to the bottom (most) rice layer. An attempt is made to
remove this error by discarding the top and the bottom layers and using only the
middle layer for the test or by collecting the rice in a plastic bag and allowing it
to equilibrate for some time.
BIOCHEMICAL PROPERTIES
Amylose Content
For several decades, rice end-use quality has been reported to be primarily con-
trolled by variation in grain amylose content, which is known to be associated with
mutations within the waxy gene. This starch fraction is also affected by environ-
mental factors, particularly temperature, during seed development. The same culti-
var grown in different environments may vary by up to 6% in amylose content
(Juliano and Pascual, 1980). For example, low-amylose types, which typically have
12–15% amylose content when grown at higher temperatures, have up to 18% when
grown at lower temperatures (Larkin and Park, 1999; Bao et al, 2000). Individual
kernels on a panicle reportedly can vary up to 3.3 percentage points (Matsue et al,
1994). Genetic analysis has shown that genotype × environmental effects on amy-
lose content are also significant (Shi et al, 1997). Consequently, breeders have
needed to evaluate progeny across multiple years and locations to obtain more accu-
rate estimates of a cultivar’s amylose type.
The initial colorimetric method reported for the determination of rice amylose
content was that of Williams et al (1958). That method is based on the fact that
iodine complexes with amylose, and this reaction can be measured spectropho-
tometrically. Simplification of this method has been reported by Juliano (1971b),
Sowbhagya and Bhattacharya (1971, 1979), Bolling and El Baya (1975), Perez and
Juliano (1978), and Juliano et al (1981b). These modifications centered primarily
around the following factors.
1. Method of solubizing the flour: either overnight soaking in an alkali solution,
at room temperature, or heating of the solution for a short period of time.
2. Whether to defat samples and how: either defatting rice flour by refluxing with
85% methanol (or 95% ethanol) (Juliano, 1971b) or defatting the dispersion in the
liquid phase (Sowbhagya and Bhattacharya, 1979). It has been suggested that defat-
ting can be omitted if the result is corrected by a conversion factor. Such factors
reported for nonwaxy rice have been 1.16 (Sowbhagya and Bhattacharya, 1971),
1.25 (Bolling and El Baya, 1975), and 1.29 (Rice Research and Development Cen-
ter, India, unpublished, based on analysis of a large number of samples widely dif-
fering in amylose). For waxy rice the factor was 1.07.
3. pH adjustment of the solution: neutralization of the sample solution with acetic
acid (Juliano, 1971b) or neutralization to internal phenolphthalein end-point
(Sowbhagya and Bhattacharya, 1971). The former method results in a slightly lower
amylose value than the second, but this difference is generally ignored and the ace-
tic acid system used for convenience.
4. Source of analytical standards: due to availability and “purity,” potato

amylose has historically been used to develop standard curves for quantifying rice
amylose content. However, a problem with using potato amylose has been that
sources differ in purity, and thus, results for rice amylose content can vary
dramatically. Amperometric or potentiometric titration methods can be used to
examine the purity of a potato amylose sample. After a sample of potato amylose is
determined to be of sufficient purity, it can be used to determine the amylose
content of several rice samples of varying amylose content. These rice samples of
predetermined amylose content can then be used for routine development of
standard curves. Waxy rice is generally included in such a set to act as the lowest
point in the curve. However, waxy rice has an amylose content of 2–5% against
potato amylose, rather than zero. Some users of this method force zero when using
waxy rice in a linear regression, while others use the actual amylose level. Juliano et
al (1981b) suggested that a mixture of amylose and amylopectin be used for
standard curve development to make it more closely aligned with these starch-
fraction proportions in rice. This method provides a “truer” amylose value.
Unfortunately, use of this procedure reduces rice amylose values by 3–5 percentage
points compared to the values that have been reported in the literature for the past
several decades, which may create confusion.
In an attempt to standardize the measurement of rice quality traits, the AACC
and the International Organization for Standardization have developed official
methods for milled-rice amylose content. Nonetheless, several factors in these
methods can create variation in the results independent of the amount of amylose in
a sample. Specifically, native lipids compete with iodine in forming a complex with
amylose. Thus whether the sample is defatted, and the method of defatting, can
change the amount of amylose found for a sample of rice. Long chains of
amylopectin also bind with iodine. Because rice samples vary in the amount of
amylopectin long chains they contain, the amount of iodine bound varies
accordingly, and the variation affects the resulting amylose value to differing
degrees. The amount of amylose found in a sample can also be affected by an
assumption that is made when calculating the amount of potato amylose and waxy
rice starch to use in the mixtures for standard curve development. This assumption
is that all rice samples contain 90% starch, which is not true. Because of the factors
described above, researchers have suggested that amylose content, determined
colorimetrically, should be reported as “apparent amylose” or “amylose equivalent”
(Takeda et al, 1987; Radhika Reddy et al, 1993). The former considers the errors
introduced by the method, and the latter includes the effect of iodine binding by
amylopectin. Throughout this chapter, when amylose content determined
spectrophotometrically is discussed, the reader should understand this to be
apparent amylose or amylose equivalent.
Juliano (1971b) and Webb (1972) described rapid procedures for amylose meas-
urement, using an auto-analyzer. Since these reports, this type of instrumentation
has been used for routine analysis in several rice-breeding programs around the
world. Another technique able to shorten the assay time for determination of rice
amylose content is a modification of the potato-amylose microtiter-plate method of
Cobb and Schoelles (1989) (C. Bergman, unpublished data). A similar method is in
use by Rice Tec, Inc. (C. Earp, personal communication).
Hot-water-insoluble amylose (“insoluble amylose” for short, i.e., the difference
between the total and the soluble amylose) is reported to be correlated with the
TABLE 3
Role of Insoluble Amylose in Texture of Cooked Ricea
Total Insoluble
No. of Amylose Amylose Stickiness Hardness
Amylose Class Varieties (% db) (% db) Score (105 cP)b
High 5 28.7 18.4 1.6 151
11 28.7 13.0 2.5 88
8 28.8 11.9 3.2 48
Intermediate 4 23.8 9.0 4.6 19
Low 4 20.6 8.2 6.4 12
a Data from Bhattacharya et al (1978). Used by permission.
b By Haake Consistometer.
pasting and textural properties of rice in general and of high-amylose types in par-
ticular (Table 3) (Manohar Kumar et al, 1976; Bhattacharya et al, 1978, 1982a;
Bhattacharya and Sowbhagya, 1978, 1979; Deshpande and Bhattacharya, 1982;
Sowbhagya et al, 1987). Kongseree and Juliano (1972), Juliano and Perdon (1975),
and Maniñgat and Juliano (1978) also observed that cultivars showing low solubility
of amylose generally showed very high setback and low breakdown, hard gel
consistency, and faster retrogradation. Kang et al (1994, 1995b,c) observed a
relationship between low solubility of amylose (i.e., high insoluble amylose) and
relatively hard and fluffy texture for low-amylose cooked rice.
The measurement of rice soluble amylose gives essentially the amylose content
of hot-water-solubles. Thus for defatting, pH adjustment, and reading against stan-
dard amylose, the same conditions apply as for amylose content determination,
except that the flour is not dispersed in alkali but is extracted in hot water. The
details are given in Shanthy et al (1980); it is suggested that defatting may be dis-
pensed with and the conversion factor for amylose content cited above be used.
Subtraction of the soluble amylose content value thus determined from the total
amylose content described above gives the hot-water-insoluble amylose content.
Amylopectin and Amylose Characteristics
After a study of 177 accessions reported to be a cross-section of the world’s rice

germ plasm, eight distinct quality types based on amylose content, gel consistency,
and Brabender relative breakdown (BDr) were reported (Bhattacharya et al, 1982a;
Bhattacharya and Sowbhagya, 1979) (Table 4). A few features of the above classifi-
cations are mentioned below. The first three types (I–III, all high-amylose) differ
sharply in their processing characteristics; this was correlated with differences in
their insoluble amylose. The next three types, the intermediate-amylose group (IV–
VI), although having similar total- and insoluble-amylose values, differed sharply in
their viscograms. The aromatic types (type IV) are especially interesting. Even
though collected from different parts of India, they all had virtually identical prop-
erties and a distinct BDr pattern; hence they were assigned a separate type despite
having total- and insoluble-amylose values similar to those of types V and VI.
Glaszmann (1987) classified rice based on isozyme polymorphisms and found that
aromatic rice accessions fell into a distinct class (isozyme group V) different from
indica types (isozyme group I). Similarly the bulu types (quality type VI) had a
pasting breakdown curve quite similar to that of low-amylose japonica types (qual-
ity type VII) (Fig. 7). And according to Glaszmann, bulu types belong to the same
isozyme group VI as japonica types.
Thus for several decades, insoluble amylose, amylose content, gel consistency,
and BDr were suggested to be the best predictors of cooked-rice texture. But there
was no underlying theory to explain why these indices were able to predict rice
texture. The last two were not entities and hence not determinants in any manner of
speaking, being themselves determined by something that determined rice texture.
Cooked-rice texture, gel consistency, and BDr, broadly speaking, reflected the same
property. Any one of them could not be claimed as a determinant of another. Both
amylose and insoluble amylose were undoubtedly entities and hence a determinant
of some sort; but what they determined no one knew for sure.
The analytical methods used thus far had reached their limit, and new approaches
were needed to permit understanding of the biochemical basis of cooked-rice tex-
ture and rice processing quality. Asaoka et al (1985) and Chinnaswamy (1985) initi-
ated a paradigm shift in the approach to the problem of the chemical basis of rice
quality. Rice starch was separated by gel-permeation chromatography (GPC)
through Sepharose 2B into two fractions: a higher molecular weight (MW) fraction
eluting at the void volume and another, smaller fraction that entered the gel (Fig.
11). The former was apparently amylopectin, since no residual carbohydrate
remained in the void volume once the mixture had been debranched. Yet, it was
noted that the fraction stained blue with iodine, and the color seemed to intensify
with increasing amylose content. The lmax of the iodine complex correlated well
with the insoluble amylose content of each sample (Fig. 12). These data indicated
that 1) what was being measured as amylose by iodine coloration was not com-
pletely amylose—a portion of the color was contributed by the amylopectin; 2)
something in the structure of amylopectin must have enabled it to bind iodine; and
3) the so-called “insoluble amylose” must have been related to that “something.”
Takeda et al (1987, 1989) and Hizukuri et al (1989) studied the chain-length dis-
tribution of chemically separated amylopectin from eight rice cultivars having low,
medium, and high iodine affinity. These samples were debranched and fractionated
by gel-permeation high-performance liquid chromatography (HPLC). Amylopectin
of indica types showed high iodine affinity and was found to have more long chains
TABLE 4
Quality Classification of Ricea
Rice Quality Amylose (% db) BDrb Cooked Rice
Type Designationc Total Insoluble (%) Hardness Stickiness
I HA: Hard >26 >15 0–5 Very high Very low
II HA: Intermediate >26 12.5–15 16–17
III HA: Soft >26 <12.5 31–55
IV IA: Aromatic 22–26 <10 56–81
V IA: Normal 22–26 <10 56–78
VI IA: Bulu 22–26 <10 134–157
VII LA 15–22 <10 111–153
VIII WX <5 – 252–333 Very low Very high
a Compiled from Bhattacharya et al (1982a). © Institution of Food Technologists. Used by permission.
b Relative breakdown, at a peak viscosity of 1,000 BU. Values are from Bhattacharya and Sowbhagya
(1979), n = 45. Used by permission.
c HA, high amylose; IA, intermediate amylose; LA, low amylose; WX, waxy.
and fewer short chains. On the other hand, amylopectin of japonica types, which
had low iodine affinity, had more short chains and fewer long chains. The different
amylopectin types, depending on their content of very long chains, had different
iodine affinity and intrinsic viscosity [h]. Some of these results are shown in Table
5.
Radhika Reddy et al (1993) took two to three representative cultivars of each of
the eight rice quality types described by Bhattacharya et al (1982a) and
Bhattacharya and Sowbhagya (1979). Amylopectin from each of these types was
isolated by preparatory GPC. These were debranched, and their chain profiles were
studied both before and after b-amylolysis. The authors reported trimodal chain
profiles for amylopectin, with some very long chains. Significantly, the amount of
Fig. 11. Elution pattern of rice starch on a Sepharose 2B column. The rice samples are identified by
their numbers, the first numeral of which stands for the quality type (1 = highest “total” and
“insoluble” amylose, 8 = waxy), and the second numeral for the sl. no. within the type. Vo = void
volume, Vt= total volume. (Reprinted with permission from Chinnaswamy and Bhattacharya, 1986)
Fig. 12. Relationship between “insoluble amylose” content as estimated in rice flour and the l
maximum of the iodine complex of fraction I of gel-permeation chromatography-fractionated rice
starch over Sepharose 2B. Fraction I is the high molecular weight fraction in Figure 11 that eluted at
the void volume. (Reprinted with permission from Chinnaswamy and Bhattacharya, 1986)
long chains decreased gradually from type I (highest insoluble-amylose content) to

type VIII (waxy) rice, showing a correlation between the rice types, the insoluble
amylose, the long chains of amylopectin, and cooked-rice texture. When the amy-
lopectin was subjected to b-amylolysis before debranching, the proportion of the
long chains decreased again in the same order, type I to VIII, showing that these
long chains were located mostly in the external part of the amylopectin molecule.
Additional reports of an association between rice texture and amylopectin chain
length have been reported by several others (Kang et al, 1995a; Ong and Blanshard,
1995; Yoshio et al, 1995, 1997; Lu et al, 1997; Jane et al, 1999). Interestingly,
differences in the amount of long chains have been reported not only among high-
and intermediate-amylose rice types, but also among low-amylose japonica types
(Kang et al, 1995a; Yoshio et al, 1995).
Ramesh et al (1999a) characterized the low-MW fraction II of the Sepharose
GPC separations, which had been previously speculated to be amylose. To their
surprise, they found that even this fraction was made up overwhelmingly of
branched molecules, i.e., amylopectin of low MW. These, too, had a chain profile
similar to that of the major amount of amylopectin fraction I, and their long B-
chains were also correlated with rice texture. Thus, it appeared that the “true” amy-
lose content in rice starch was lower than previously thought (i.e., not more than 7–
11%), which was confirmed later by Ramesh et al (1999b) by chromatographic
separations.
It appears, then, that amylopectin is not necessarily a very large molecule; and its
chain profile varies widely among species and genotypes. And amylose, if defined
as a straight, unbranched chain, if it exists at all, is only a minor component in
starch. The words of Young (1984) bear repeating:
Exclusion chromatography…produces fractions that raise the possibility that native
starch may be, not a mixture of branched and linear molecules of D-glucopyranosyl
units, but a mixture of covalently bound branched molecules having some extremely
long chains, a broad degree of branching, and broad molecular weight distributions.
The next question was: Why was rice texture related to the chain profile of its
amylopectin? Radhika Reddy et al (1994) studied the viscoelastic properties of rice
flour pastes in 15 cultivars with a Bohlin rheometer. The relaxation (G) and storage
(G) modulii of the pastes increased with increase in paste concentration in all culti-
vars as expected, but the increase was more in high-amylose than in low-amylose
rice (Fig. 13). Similarly, the relaxation time of the paste was maximum in high-
amylose paste and low in low-amylose paste. Thus, the starch granules in high-
amylose rice were reported to be more elastic than in low-amylose rice. In agree-
ment with this, when the pastes were heated at 95°C for another 60 min, the drop in
G and the relaxation time values at 60 min compared to those at 0 min were negli-
gible in high-amylose rice but high in low-amylose rice.
Sandhya Rani and Bhattacharya (1995a) studied 26 samples of rice of different
quality types by heating them up to 95°C and then heating the pastes at 95°C with
stirring for an additional 60 min. The fall in paste viscosity at 60 min compared to
that at 0 min of cooking was negligible in high-amylose rice of type I but very high
in low-amylose rice of type VII, with the remaining types more or less in the same
order. Waxy rice (type VIII) broke down even before reaching 95°C. This report
thus indicated that the strength and resilience of the high-amylose cultivars was
greater than that of low-amylose types. Microscopic examination of these pastes
confirmed these conclusions, for the starch granules in high-amylose rice remained
largely intact after 1 hr of heating, while the granules in low-amylose rice became
completely dispersed (Sandhya Rani and Bhattacharya, 1995b) (Fig. 14). These
findings suggest that the long chains of amylopectin, by intra- and intermolecular
interaction, help form a strong and resilient starch granule, while the deficiency of
long chains renders the granule weak and fragile. This difference in the strength of
the starch granules caused by the relative abundance of long chains in their starch
molecules thus was suggested to be at the root of differences in rice texture.
Evidence linking the chain length distribution of amylopectin to GT has recently
been reported. Rice amylopectin enriched with shorter chains reportedly has a lower
GT than that with longer chains (Nishi et al, 2001; Jahan et al, 2002). A report by
Umemoto et al (2002) indicated that amylopectin in japonica-type cultivars had a
greater proportion of short chains (degree of polymerization [DP] <11), while indica
types had a greater amount of longer chains (DP 11–24). Thus far, many of the
studies that have examined the association between amylopectin chain length and
rice end-use quality have been limited in terms of the rice germ plasm studied.
Thus, it is sometimes difficult to determine whether reported correlations are a phe-
notypic association due to the specific germ plasm studied or an actual causal event
(i.e., pleitrophy).
A current area of study that will help our understanding of the biochemistry
controlling rice end-use quality is focused on the determination of amylose structure
and the role that the amount and type of lipids and proteins play in controlling
starch solubility. These factors are hypothesized to influence the amount of amylose
that leaches from starch granules during cooking, which in turns affects starch
functionality (M. Fitzgerald, personal communication). Perhaps the next step in
understanding rice end-use quality will come from a study that examines the amy-
lopectin and amylose structure and various functional measurements of all rice end-
use quality types. Then relationships between lipids and proteins and rice function-
ality could also be examined using the same germ plasm.
The methods for studying amylopectin and amylose structure and molecular
weight are thus far research tools and have not reached the stage of being consid-
Fig. 13. Storage modules (G) (left column) and relaxation time (T0.75) (right column) of three rice
varieties pasted by heating up to 95°C and then further cooking at 95°C for 0 min (top) and 60 min
(bottom). Rice variety: Jaya, quality type I; Br 9, type IV; T 65, type VII. (Reprinted, with permission,
from Radhika Reddy et al, 1994)
Fig. 14. Photomicrographs of starch granules in 12% rice-flour pastes of four varieties (types I, III, V,
and VII). Flours were pasted by heating up to 95°C and were examined by light microscopy (A) and
scanning electron microscopy (B). For each, the left column = just pasted (0 min); the right column =
maintained at 95°C for another 60 min. After 60 min, the granules are only slightly affected in type I
but are completely degraded in type VII. (Reprinted, with permission, from Sandhya Rani and
Bhattacharya, 1995b.)
ered tools for routine rice-quality analysis. These methods begin with the prepara-
tion and solubilization of starch from milled-rice kernels. And with these steps
comes the potential for artifacts, that is, starch structural attributes that are created
from sample preparation techniques rather than actual characteristics in the native
starch. The various methods reported for starch preparation can themselves create
differences in starch structure and functionality, as reported by Chiou et al (2002).
In addition, the techniques reported to solubilize starch are varied and, as of yet,
have not been studied to determine any effects they may have on starch structure.
One thing is for sure and that is, rice starch is difficult to solubilize. Consequently,
reagents such as NaOH, DMSO, urea, LiBr, etc., in combination with increased
temperature or time, sonication, or elevated pressure, have been used in various
combinations to improve rice starch solubility. The literature is confusing as to
which combination of techniques results in superior starch solubility with the least
structural damage for all rice germ plasm, i.e., all levels of amylose content. In fact,
most reports do not mention whether samples were completely solubilized, what the
recoveries were after separation, or whether starch structure was affected by the
methodology.
Many of the studies of rice amylopectin structure have been performed using
size-exclusion HPLC (SE-HPLC) or GPC to separate debranched amylopectin into
partially separated groups of varying oligomer chain lengths. These separations are
based on differences in the hydrodynamic volume of oligomers. For rice, a trimodal
(or dimodal) distribution is generally reported, consisting of short (<25), medium
(25–65), and long (>65) oligomer chains (Lu et al, 1997; Han and Hamaker, 2001).
Although much work has been accomplished using these techniques, other
researchers have sought to achieve greater resolution of the polymeric distribution
of debranched amylopectin. Several promising separation techniques have been
reported to have improved resolution, namely, anion-exchange HPLC, capillary
electrophoresis, and matrix-assisted laser-desorption ionization–time of flight mass
spectrometry (MALDI-TOF). High-performance anion-exchange chromatography
with pulsed amperometric detection can resolve debranched amylopectin with DP
values approaching 60 glucose units (Koizumi et al, 1991; Beroft and Koch, 2000).
For capillary electrophoresis separation of amylopectin chains, up to 100 glucose
units have been achieved (O’Shea and Morell, 1996; Nakamura et al, 2002).
MALDI-TOF is reported to resolve rice oligmers up to 45 glucose units long, and
greater resolution is likely with the more-advanced instrumentation now available
(Fig. 15) (Grimm et al, 2003).
Separation techniques for native rice starch (i.e., not debranched) have been per-
formed using SE-HPLC and GPC. A difficulty reported with these techniques is
that molecular shearing can occur as the molecules pass through instrumentation
frits. Field-flow fractionation reportedly is able to minimize the shearing of very
large molecules but has not been used extensively in starch analysis (Hanselmann et
al, 1995). These separation techniques for native starch have been combined with
differential refractive index and multiangle laser light-scattering (MALLS)
detection to determine (or, some would say, estimate) amylose and amylopectin size
and mass distributions (Yokoyama et al, 1998). There is some concern that
associations between polymers can occur, and thus MALLS calculations can be
inaccurate. Other techniques with the future potential to assist in the
characterization of rice starch include zonal ultracentrifugation, proton nuclear
magnetic resonance, infrared spectroscopy, and capillary hydrodynamic fraction-
Fig. 15. Matrix-assisted laser-desorption ionization–time of flight mass spectrum of debranched rice
amylopectin. DP = degree of polymerization.
ation (Dos Ramos and Silebi, 1990; Aberle et al, 1997; Dunn et al, 1999; Majzoobi
et al, 2003; Sevenou et al, 2003).
Protein and Lipid Content
There is evidence that other factors besides starch contribute to the sensory and
functional characteristics of rice. Primo et al (1965, 1966) reported a close relation-
ship between sulfhydryl (-SH) and disulfide (-SS-) groups, especially in the outer
layers of the kernel, and the texture of cooked rice. The findings of Moritaka and
Yasumatsu (1972a,b) that the palatability of Japanese rice had a relationship to its -SH
content support these findings. Yanase et al (1984) suggested that the protein content
of rice was inversely related to the viscograph breakdown and cooked rice
adhesiveness. Hamaker and Griffin (1993) showed that rice pasting curves were
lowered when each slurry was treated with a reducing agent to break the -SS- bonds.
The hardness of cooked rice has been reported to be primarily determined by its starch
content, while surface hardness was found to be associated with protein content
(Okadome et al, 1999a,b). Kim et al (1997) have shown that protein content is
negatively correlated with Korean rice palatability. After the removal of protein from
rice flour via a protease treatment, Martin and Fitzgerald (2002) found dramatic
changes in RVA pasting curves. They concluded that proteins influence viscosity
curves through water binding and a network linked together by disulfide bonds.
Fitzgerald et al (2003) and Yang and Chang (1999) demonstrated that lipids have
an impact on rice-flour pasting properties. Also, an association exists between the
amount of lipid-amylose complex in rice and cooked-rice firmness. For example,
superior-processing types (e.g., Dixiebelle) have a significant amount of lipid
complexed with amylose, as determined by DSC, while soft-cooking low-amylose

types (e.g., Bengal) have little to no lipid-amylose complex. Since this complex
melts at temperatures higher than rice starch GT and higher than most processing or
cooking temperatures (in Figure 3, the second peak for the intermediate-GT type is
a lipid-amylose complex), it is possible that its association with cooked rice
firmness may be causal.
An issue yet to be resolved regarding some of the studies reported above is
whether the impact on cooked-rice properties reported is caused by variation in the
kernel’s protein content or is an indirect effect caused by a reduction in protein
content associated with an increase in amylose content. Also, assuming that protein
content does influence rice functionality, the work described above does not
indicate whether the effects are a consequence of variation in protein quality or,
perhaps, of interaction between starch structure and protein quality.
Aroma
More than 100 compounds that contribute to the aroma of rice have been identi-
fied (Tsugita, 1986; Widjaja et al, 1996). Some of these volatile compounds con-
tribute to consumer acceptance of certain samples of rice and some to the rejection
of others. The two issues currently of greatest importance to the industry are the
scent of aromatic rice and off-flavors due to rancidity. The aromatic compound 2-
acetyl-1-pyrroline (2-AP) has been reported to be the primary component of aro-
matic (scented) rice aroma (Buttery et al, 1983, 1988). Agronomic and postharvest
practices influence the levels of 2-AP in aromatic rice (Goodwin et al, 1994). Also,
there is evidence that genetic differences in the intensity of 2-AP exist, and thus
higher levels appear to be a feasible breeding target (Bergman et al, 2000). The
scent of aromatic rice is a highly heritable trait and reportedly is under the control
of from one to four genes, depending on the population studied (Dhulappanavar,
1976; Tripathi and Rao, 1979; Reddy and Sathyanarayanaiah, 1980; Pinson, 1994).
Tsuzuki and Shimokawa (1990) suggested that the lack of agreement regarding the
number of genes controlling aroma may, in some instances, be due to the empirical
methods used to assess rice aroma. Examples of those methods include biting ker-
nels, smelling vegetative tissue after warming or soaking in KOH, and eating
cooked rice (Dhulappanavar, 1976; Tripathi and Rao, 1979; Sood and Siddiq, 1979).
Clearly, sensory fatigue could be a source of error for such methods.
Quantification of 2-AP in rice has been described by Buttery et al (1986) and Lin
et al (1990) on sample sizes of 200 g using a simultaneous steam distillation-solvent
extraction apparatus. Tanchotikul and Hsieh (1991), using a micro-steam distilla-
tion-solvent extraction procedure, extracted 2-AP using 1-g samples. Bergman et al
(2000) and Grimm et al (2001) have reported rapid solvent and selective-phase
microextraction methods, respectively, for extracting 2-AP for analysis via gas
chromatography. Both methods require less than 1 g of sample; of the two, the for-
mer has superior recoveries and precision, while the latter requires shorter sample
preparation time and creates no toxic waste products.
Other volatiles associated with a specific aspect of rice aroma are the carbonyl
compounds. They cause stale or rancid odors in many foods (Bennion, 1980). Rice
lipids are primarily in the bran and germ; therefore, development of rancidity is of
greatest concern for industries focused on brown rice, rice bran, or its derivatives.
However, even with less than 1% lipid, milled rice is also susceptible to developing
rancid or stale smells. Consequently, efforts have been made to define storage con-
ditions and processing parameters that minimize lipid deterioration (Sowbhagya and
Bhattacharya, 1976; Ory et al, 1980; Sharp and Timme, 1986; Champagne and
Hron, 1993). Increased levels of carbonyl compounds are associated with lipid dete-
rioration. Of those compounds, hexanal increases the most during rice storage and
consequently has been used as an indicator of rancidity (Yasumatsu et al, 1966;
Shibuya et al, 1974; Tsugita et al, 1983; Gon Shin et al, 1986; Champagne and
Hron, 1993). Bergman et al (2000) reported a method for determining hexanal con-
tent that is rapid and thus suitable for use with large research projects, breeding pro-
grams, or quality assurance departments.
GENERAL TECHNIQUES
Near-Infrared Spectroscopy
All molecules vibrate when exposed to electromagnetic radiation. These vibra-

tions correspond to wavelengths in the NIR region. Thus when something is
exposed to radiation in the NIR region, each molecule of each given type absorbs
energy at the wavelength that causes it to vibrate. A mathematical relationship can
be established between the amount of energy absorbed at each wavelength of a
substance and the quantity of a constituent measured using an analytical procedure.
NIR methodology theoretically is a rapid predictive tool that requires little to
sometimes no sample preparation and is nondestructive and nonpolluting. However,
NIR is a secondary method that relies on primary laboratory data. Consequently,
NIR prediction error rates can be only as low as the error level of the laboratory
procedure that the prediction is based on and no lower.
NIR analysis should be regarded as empirical (Murray and Williams, 1987). That
is, results are based on correlations (not cause and effect) between reflectance at
certain wavelenth(s) and laboratory data. Experimentation is required for calibration
equation development and for calibration updating. The physical and biochemical
traits that each sample possesses are consequences of a unique combination of
genetic and environmental effects. The properties of a sample are also affected by
how they are harvested, handled, and stored. Each sample thus has a unique NIR
spectra, even though several samples will have similar values for the trait of inter-
est. Consequently, a researcher must use various strategies to “teach” the software
how to see through the spectral variation and find aspects of the data that are corre-
lated with the trait being predicted. Some of the strategies that have been used in
development of NIR calibration equations (NIR-CEs) include using various mathe-
matical treatments, obtaining samples with much spectral variation, and the use of
repeatability files (Shenk and Westerhaus, 1991). Assessing the strategy best suited
for the development of an NIR-CE is currently an effort of trial and error (Sinnaeve
et al, 1994).
Predictions of nitrogen (or total protein) and moisture contents are the most
common use for NIR in the grain industry (Williams, 1975; Kawano and Chie,
1999). This technology has also been evaluated for use in predicting rice lipid acid-
ity and surface lipid content (Chen et al, 1997; Li and Shaw, 1997), chemical com-
position (Hwang et al, 1994), and grain quality in general (Shu et al, 1999a;
Natsuga, 1999). Several attempts to evaluate the ability of NIR to predict milled-
rice amylose content have been reported (Kobayashi et al, 1994; Delwiche et al,
1995; Shimizu et al, 1998, 1999, 2000; Shu et al, 1999b). The strength of these pre-
dictions were from weak to strong. NIR has also been studied to evaluate whether
this technology can be used to differentiate between different cultivars of rice
(Krzanowski, 1995; Kim et al, 2003), although genetic markers are likely the only
analytical method able to accurately perform such tasks.
Perhaps the most ambitious NIR projects have been those studying the use of this
technology to predict the texture of cooked rice as determined by descriptive ana-
lytical methods (Windham et al, 1997; Barton et al, 1998, 1999; Meullenet et al,
1999, 2000, 2002; Champagne et al, 2001). These projects were very challenging, in
part, because the primary rice fraction that reportedly creates rice textural variation
(amylose content) cannot easily be predicted using NIR technology (see discussion
below). Thus, any traits strongly influenced by amylose content will also likely be
difficult to model. Also, updating these calibration equations to function with each
new set of samples not represented in the calibration set requires additional analysis
by trained sensory panelists. Such analysis is extremely expensive and time-
consuming; thus, the original equation must be as spectrally broad as possible.
For the NIR methods described above to be fully validated, practical use of the
technology to predict these characteristics in a real-world setting is needed. To date,
the only commercial use of this technology to predict rice properties has been in
Japanese mills and marketing agencies (i.e., testing consumer acceptance/sensory
properties) and rice breeding programs (i.e., testing apparent amylose, moisture, and
protein contents), as described below.
In the Japanese rice industry, some workers such as millers use NIR technology
to evaluate milled rice for characteristics related to palatability (Hosaka, 1987;
Tanaka et al, 1999). The instruments used are called rice-taste analyzers or meters,
and, although the technology is proprietary, they are thought to use NIR to predict
amylose, protein, moisture, and lipid acidity. These data are then converted into
preference testing scores, or such scores are predicted directly (Iwamoto, 1980;
Chikubu et al, 1983, 1985). Kawamura et al (1996), studying Japanese rice samples,
and Champagne et al (1996), evaluating U.S. conventional medium-grain rice sam-
ples, reported that these instruments are not able to replace sensory panel evalua-
tions. This is not surprising since the NIR-CE these instruments rely on cannot be
updated, and the instruments cannot be used to predict the sensory scores of rice
samples not adequately represented in the calibration equation sample set. Addi-
tionally, rice-taste analyzers, as Champagne et al (1996) point out, are designed to
“measure” rice preference scores of Japanese consumers. Thus, they are completely
unsuited to determine the preferences of consumers other than those Japanese peo-
ple used to develop their calibration equations.
A calibration database for a grain or feed constituent should contain a minimum
of 500 samples (Burns and Ciurczak, 2001). No reports could be found of this many
samples being used to develop NIR-CE to predict rice physical or chemical
properties. Very few NIR-based research publications utilize this many samples
when developing and validating calibration equations. In most reports, the number
of samples used for calibration-equation development has been less than 100.
Usually, the greater the number of samples and the more spectrally different they
are, the less equation updating is required when new samples are presented to the
equation. However, even if a calibration equation has been developed using a large
number of diverse samples, it will require future updating for most constituents, to
include samples representative of each new combination of cultivars, growing con-
ditions, postharvest handling and processing conditions, and their interactions.

Marks et al (1997) reported that cultivar had such a large effect on the robustness of
their calibrations for pasting viscosity, water absorption, and expansion that
cultivar-specific equations would be required for NIR predictions of these
characteristics. This work used only three cultivars subjected to different
postharvest handling conditions.
NIR has been used to predict various chemical constituents (e.g., apparent amy-
lose, protein, and moisture) of breeding material across the world. The spectral
database developed by the USDA-ARS Rice Research Unit for predicting the
apparent amylose content of milled rice of U.S. rice genotypes contains over 1,000
sample scans, along with their laboratory values. Achieving reliable predictions has
required that this calibration set fully represent all samples that will be predicted in
terms of their genetics, growing location, and postharvest handling practices. Two
methods have been evaluated for determining which samples from a new set to use
to update this calibration equation. A random set of 10% of the samples was chosen,
or the Select algorithm was used (Shenk and Westerhaus, 1991) to choose samples
that spectrally represent the sample set. For moisture and protein content, the latter
method generally resulted in fewer samples needing to be analyzed via wet-
chemistry techniques to create NIR-CEs with acceptable accuracy. However, nei-
ther method has been able to consistently produce accurate NIR-CEs for apparent
amylose content of breeding samples of diverse genetic backgrounds and growing
conditions. Consequently, this program, as well as others, such as the one in
Australia, have stopped using NIR technology to predict apparent amylose content
and have gone back to using the wet-chemistry method with the addition of a Waxy-
gene molecular-marker assay (discussed below) for use in early-generation screen-
ing (M. Fitzgerald, personal communication). Others continue to use this
technology and are satisfied with its performance in predicting apparent amylose
content when specific method modifications are incorporated. For example,
RiceTec, Inc. (C. Earp, oral communication) conditions samples to a narrow mois-
ture range (i.e., 11.5–12.5%) and reports improved prediction accuracy compared to
using samples on an as-is-basis moisture basis. The USDA-ARS Rice Research
Unit tested their NIR-CE for apparent amylose with samples conditioned to have a
range in moisture content and found the accuracy of the predictions to be signifi-
cantly affected only when the levels were outside the range of 10–14%. Others,
such as the California Cooperative Rice Research Foundation, find NIR to satisfac-
torily predict apparent amylose in their breeding program when calibration equa-
tions are built to function for a relatively narrow germ-plasm base (K. McKenzie,
oral communication). However, in the latter case, calibration equations are updated
with wet-chemistry values from approximately 100 samples from each new breed-
ing nursery.
Genetic Markers
QUALITY TRAIT IDENTIFICATION

The methods described above all have one thing in common, besides being ana-
lytical methods for assessing rice end-use quality. They are phenotypic assays, that
is, they measure grain characteristics that result from the effects of genetics and the
environment, which includes the effects of differences in growing temperatures, soil
nutrition, and insect and disease pressure. Storage and processing can also affect
phenotype. These tests are thus ideal for measuring or predicting the sensory and
processing quality of rice. However, they are not optimum breeding tools.
The effects of a rice cultivar or breeding line’s growing conditions make identi-
fication of its processing or sensory quality challenging. For example, the amylose
content of a cultivar or breeding line varies depending on its growing conditions.
Figure 16 shows the amylose content of U.S. cultivars and breeding lines grown in
the Uniform Regional Rice Nursery trials in Arkansas, Louisiana, Texas, and Mis-
sissippi during four growing seasons (Bergman et al, 2001). In many instances, the
amylose content of these cultivars fell outside of the industry-defined range in
amylose content for the various rice market classes, as displayed in Figure 1. Con-
sequently, when using traditional analytical assays, breeding lines must be evaluated
under multiple growing conditions to confidently determine their market class.
Using genetic markers as breeding tools, however, allows selections to be made
based on one measurement only, because the results are not influenced by
environmental conditions. DNA marker analysis can be performed on as little as
one milled- or brown-rice kernel or a few pieces of leaf tissue, and most results can
be completed in two days. Currently, genetic markers have been developed and
tested in breeding populations for the following rice end-use quality traits: apparent
amylose content, 2-acetyl-1-pyrroline, cooked-kernel elongation, and gelatinization
temperature.
A microsatellite marker (a short sequence of nucleotides that are repeated multi-
ple times in tandem) associated with amylose class has been developed and is being
used by various breeding programs around the world, such as those in the United
States and Australia (Bligh et al, 1995; Ayers et al, 1997; Bergman et al, 2001).
This marker identifies the particular number of cytosine and thymine sequence
Fig. 16. Apparent amylose content and cytosine-and-thymine-sequence-repeat (CT)n class of 142
progeny developed from a cross between Dellmont and B8462T3-710. Mean apparent amylose values
in each of the two (CT)n classes of the progeny are listed in the legend. LSD = least significant
difference (a = 0.05).
repeats (CT)n. In the Bergman et al (2001) study, the amylose content of the
breeding progeny from a cross between a conventional U.S. long-grain (intermedi-
ate amylose) variety and a long-grain variety with superior processing quality (high
amylose) was assayed (Fig. 16). The intermediate-amylose progeny had a mean of
22.3%, while the superior-processing types had a higher mean of 23.5%. Although
the mean amylose values for the two quality types were significantly different,
many of the lines grown in the particular environment had a value that would cause
them to be misclassified. Selecting for the superior-processing types in this
population would require that the cross be grown in several additional environments
and that the amylose content and pasting properties of the progeny be evaluated.
However, using the genetic marker, known as the waxy microsatellite, the
population can be separated into two distinct groups, those with a (CT)20 allele, like
the conventional long-grain parent, and those with a (CT)11 allele, like the parent
with superior processing quality. Only one assay is needed in order to confidently
assign each progeny to an amylose class. Several methods are now available to
analyze genetic markers (i.e., polymerase chain reaction [PCR] products). An
example of the results from a relatively inexpensive method, gel electrophoresis, is
displayed in Figure 17.
GT also varies greatly as a result of environmental effects. Consequently, rice
breeders must analyze the GT of breeding material via the alkali spreading test
multiple times during cultivar development. One or more genetic markers for this
trait would be a great advantage to rice breeders across the world. The structure of
amylopectin is reportedly involved in the control of starch GT (Jane et al, 1999;
Nakamura et al, 2002). Amylopectin structure is controlled by several enzymes, one
of which is soluble starch synthase IIa (SSSIIa). In a comprehensive genotype
analysis of 196 international rice accessions, Chen et al (2003) identified two sepa-
Fig. 17. Polymerase chain reaction products from a microsatellite showing number of repeats of
cytosine and thymine (CT)n associated with the waxy gene separated using polyacrylamide plus
Spreadex gel electrophoresis. Numbers above bands indicate the lane number, while numbers
underneath designate the band’s microsatellite class. Lanes 1 and 18 (microsatellite classes 10, 17, and
20) = controls used for scoring the gel, lanes 2–10 = one band for each known allele for the waxy
microsatellite, lanes 11–13 = cv. Bengal (leaf material, hulled kernels, and milled kernels), lane14 =
cv. Dixiebelle (milled parboiled kernels), lane 15 = cv. L205 (stabilized bran), lane 16 = cv. Cypress
(instant rice kernels), and lane 17 = cv. Bengal (crisp rice).
rate mutations in the SSSIIa that were associated with low GT. One mutation was
common in Japanese accessions, while the other was common in low-GT accessions
from the United States. Both presumably encode an SSSIIa enzyme with lower
activity, which results in the production of shorter chains of branched amylopectin
in these accessions, causing them to have low GT. These mutations in the SSSIIa
gene are likely to become analyzed routinely to permit rapid differentiation between
breeding lines with low, versus intermediate or high, GT.
Specialty rices are highly valued and are growing in popularity throughout the
world. Cooked kernel elongation and aroma are some of the end-use quality traits that
differentiate these types of rice from those in conventional market classes.
Development of agronomically superior specialty cultivars has proven challenging, in
part due to the difficulty breeders face in recovering the quality attributes of aroma and
cooked-kernel elongation. PCR-based genetic markers associated with 2-AP content
and cooked kernel elongation have been identified and their utility as breeding tools
verified using U.S. and Australian germ plasm (Bergman et al, 2002; Garland et al,
2000). Surveys of these markers in U.S. and Australian rice germ plasm showed that
aromatic and nonaromatic germ plasm have many genotypes with similar marker
alleles. Since no unique marker alleles are found in aromatic germ plasm, markers for
selection must be chosen in respect to parental polymorphism and genetic distance
from the fgr locus, which is associated with 2AP levels.
CULTIVAR IDENTIFICATION
Certain cultivars are traded at a relatively high price, as “premium rice,” because
of their superior organoleptic properties or processing quality, etc. Because these
types sell at a higher price, some dishonest rice wholesalers or retailers blend less-
expensive rice of poorer quality with more-expensive premium-quality rice and
mislabel it as “high-quality premium rice.” Consumers in some countries (e.g.,
Japan) purchase rice by cultivar name and want to be assured that they are buying
what the label states. Also, some cultivars may possess proprietary traits, and thus,
buyers want to be guaranteed that their purchase contains the desired cultivar.
Therefore, methodology able to identify rice cultivars is important for the entire rice
industry, i.e., from breeders to consumers.
Rice cultivars have previously been identified on the basis of phenotype, such as
the morphological characteristics of rice plant or grains, days to flowering after
emergence, etc. These methods can be used to differentiate between cultivars with
large differences in phenotype, such as low-GT types vs. high-GT types. But, many
breeding programs make crosses using a narrow germ-plasm base, which makes it
difficult to phenotypically distinguish among closely related cultivars. Thus, having
methods that reliably differentiate between cultivars is important to the rice industry
so it can maintain high quality-control standards.
The rice genome has approximately 430,000,000 base pairs in total; thus, a tre-
mendous amount of information in the DNA can be used for cultivar identification.
In randomly amplified polymorphic DNA (RAPD) analysis, short oligonucleotides
of arbitrary sequence are used to amplify regions of the test-plant genome, and am-
plification products are separated by gel electrophoresis. Mackill (1995) studied the
genetic diversity of 134 predominantly japonica cultivars with RAPD markers. The
indica and japonica cultivars were classified into separate groups by cluster analy-
sis. The author concluded that RAPD markers are useful for classification of
japonica vs. indica cultivars, but many primers would be needed to resolve closely
related japonica cultivars. RAPD data were obtained from 26 U.S. cultivars and
breeding lines by Cao and Oard (1997). Seven primers among the 69 studied could
differentiate all cultivars and lines. Stobdan et al (1999) have also reported the
development of RAPD markers able to “fingerprint” basmati types of rice. RAPD
markers, however, often display poor reproducibility and are difficult to score.
In Japan, the leading cultivar is Koshihikari, which, for several years, has been
planted on more than one third of the total land under rice cultivation. This cultivar
is extremely popular because of its desirable visual appeal, flavor, and texture. Con-
sequently, it trades at a relatively high price. Japanese packaged rice is required to
display the name of the cultivar, the location of its cultivation, and the year of its
production, as specified by the Agricultural Standard Act. Therefore, it has become
necessary to develop methodology for identification of various cultivars.
A sequence-tagged site (STS) is a DNA sequence with a chromosomal location
that can be mapped to a specific contig. Several STS primers have been identified
and commercialized by Ohtsubo et al (2002) as able to reliably identify Koshihikari.
Also, to detect other cultivars blended with Koshihikari, a “detection kit” has been
developed. The National Food Agency in Japan uses these two primer sets along
with a DNA extraction method, suitable for use with single kernels of milled or
cooked rice, for the inspection of rice (Fig. 18; Ohtsubo et al, 1999).
Fig. 18. Electrophoretic results from using primers designed to identify cv. Koshihikari or detect other
cultivars. A, primer set for identification of Koshihikari; lane 1 = Koshihikari, lanes 2–50 = other
cultivars. B, primer set for detection of other cultivars blended with Koshihikari; lane 1 = Koshihikari
(no band), lanes 2–50 = other cultivars. Lane M = control. The arrows designate the distance traveled
by each of the four possible electrophoretic bands.
Jellstrom and McClung (2002) screened approximately 200 microsatellite

markers for suitable levels of polymorphism and reproducibility in current U.S.
germ plasm. Thirty-seven cultivars currently being grown in the United States were
selected for study, including cultivars from the southern United States and from
California. Markers were chosen to represent each of the 12 rice chromosomes. A
set of 14 markers able to distinguish among these cultivars was identified. The use
of three to four markers was able to separate one cultivar from another. It is likely
that, as rice end-use quality becomes of increasing importance to the food industry,
technology to identify cultivars will increase in importance and use.
FUTURE RESEARCH NEEDS
Many of the end-use quality assays developed decades ago are still in use. This
is true because methods such as the alkali spreading test are rapid and require a
small amount of sample. However, it is also true that many of the traditional end-
use quality procedures have not been replaced or modified because little research
has focused on developing improved assays. This appears to be changing as a
renewed focus on rice end-use quality develops in various countries. Improved
end-use quality is now considered an arena that can give one rice-growing area an
advantage over another. Consequently, new public research programs on rice
quality have been created recently as competition for world rice markets
increased.
The food-processing industry has a future role in improving the methods
available for describing end-use quality. Those of us in the rice-quality field often
hear comments from the processing industry such as “that cultivar has poor quality”
or “the rice from last year’s crop was of poor quality.” When asked what is meant
by the term “poor quality,” these colleagues often have difficulty describing their
observations. This may be because of corporate rules about proprietary information,
which keeps workers from divulging the details of the processing methods, or
because they do not have methods to quantify what they mean by “poor quality.” In
the latter case, partnerships between industry workers and public researchers could
lead to the development of improved methods able to describe aspects of rice
functionality.
Several areas need further research before new end-use quality methods can be
developed.
One is identification of volatile compounds that confer consumer acceptabil-
ity to both milled and brown rice. Probably, other volatile compounds besides
2AP and lipid breakdown products influence consumer preferences for cooked
and processed rice. And it may not be individual compounds, but rather an aroma
profile, that triggers consumer preference.
Mills designed for small samples are available, but the rice they produce is
not well suited for tests such as instrumental texture measurements. Either the
rice produced is damaged (e.g., blunt ends produced) or a milling aid such as
calcium carbonate is left on the grains. Therefore, an instrument is needed that
can produce well-milled rice from small samples and also allow accurate milling
yields to be determined. Development of an automated system that could take a
small sample all the way from dehulling to milling and provide data on the ker-
nel dimensions, percent chalk, and milling yield would also be useful, especially
for breeding programs.
Processing quality for groups of product types needs to be defined and then
rapid methods developed to quantify variation in these properties.
We need to know what physical and chemical attributes control stable milling
yield, or learn what specific genetic mutations are associated with this trait. This
information then must be translated into a method that breeders can use to rap-
idly determine the milling stability of their breeding lines.
The sensory arena has produced information regarding the degree to which
cultivar differences and postharvest handling practices affect rice flavor and
texture. But what is still lacking are links between consumer preference and dif-
ferent styles of rice. For example, how firm should rice from a conventional
long-grain cultivar be to meet the demands of the Mexican-American consumer?
What firmness-to-stickiness ratio for jasmine-style rice is desired by people of
Thai descent vs. non-Thai patrons who frequent Thai restaurants? Is greater
firmness after processing desired by consumers who eat rice in canned soups?
How firm is too firm? When this sensory information is known, it then needs to
be translated into instrumental texture language. Such data would facilitate the
breeding of rice for specific markets. As things are now, the rice industry has
reported to rice-breeding programs only the ranges in amylose content and GT to
use as targets during cultivar development.
The determinants of cooked rice texture are still controversial. Clearly, an
area not fully explored is how the structure of “amylose” and “amylopectin”
(however they are defined) differ in soluble versus insoluble starch and how
these fractions associate with each other and with various types of lipids and
proteins to lead to differences in cooked-rice texture and product quality.
This list could certainly go on. Suffice it to say: with the current explosion in rice
end-use quality research, the next time this subject is reviewed, we will likely know
much more about the physical and chemical properties controlling rice cooking and
processing characteristics and have many more superior techniques for measuring
these attributes.
REFERENCES
AACC. 2000. Approved Methods of the Ameri- Arai, T., Sawayama, S., Kawabata, A., Tanimura,
can Association of Cereal Chemistry, 10th ed. W., and Obara, T. 1981. Studies on gelatinized
Methods 61-01 and 61-03. The Association, rice. III. Physical properties and sensory
St. Paul, MN. attributes of gelatinized rice. J. Jpn. Soc. Food
Aberle, T., Burchard, W., Galinsky, G., Sci. Technol. 28:444-450.
Hanelmann, R., Klingler, R. W., and Michael, Asaoka, M., Okuno, K., Sugimoto, Y., and Fuwa,
E. 1997. Particularities in the structure of H. 1985. Developmental changes in the struc-
amylopectin, amylose, and some of their ture of endosperm starch of rice (Oryza sativa
derivatives in solution. Macromol. Symp. L.). Agric. Biol. Chem. 49:1973-1978.
120:47-63. Ayres, N. M., McClung, A. M., Larkin, P. D.,
Adair, C. R., Bollich, C. N., Bowman, D. H., Bligh, H. F. J., Jones, C. A., and Park, W. D.
Jodon, N. E., Johnston, T. H., Webb, B. D., 1997. Microsatellites and a single-nucleotide
and Atkins, J. G. 1973. Rice breeding and polymorphism differentiate apparent amy-
testing methods in the United States. Pages 22- lose classes in an extended pedigree of US
75 in: Rice in the United States: Varieties and rice germ plasm. Theor. Appl. Genet.
Production. Handbook 289 (revised). U.S. 94:773-781.
Dept. Agric., Washington, DC. Azeez, M. A., and Shafi, M. 1966. Quality in
Andrews, S. B., Siebenmorgen, T. J., and Rice. Tech. Bull. 13. Dep. Agric. Govt. West
Mauromoustakis, M. 1992. Evaluation of the Pakistan, Kala Shah Kaku.
McGill No. 2 Rice Miller. Cereal Chem. Banaszek, M. M., and Siebenmorgen, T. J. 1990.
69:35-43. Head rice yield reduction rates caused by
moisture adsorption. Trans. ASAE 33:1263- Work. Group Meet., 29th. La. Agric. Exp. Stn.,
1269. Rice Res. Stn., Crowley, LA.
Bao, J. S., Xu, J. J., Wu, W. Q., Shu, Q. Y., and Beroft, E., and Koch, K. 2000. Composition of
Xia, Y. W. 2000. The changes of eating and chains in waxy-rice starch and its structural
cooking quality of indica early rice in different units. Carbohydr. Polym. 41:121-132.
cropping seasons. J. Zhejiang Univ. (Agric. Bhattacharya, K. R. 1979. Gelatinization tem-
Life Sci.) 26:103-106. perature of rice starch and its determination.
Barber, S. and Benedito de Barber, C. 1979. Pages 231-249 in: Chemical Aspects of Rice
Outlook for rice milling quality evaluation Grain Quality. Int. Rice Res. Inst., Los Baños,
systems. Pages 209-221 in: Proc. Workshop Laguna, Philippines.
on Chemical Aspects of Rice Grain Quality. Bhattacharya, K. R. 1980. Breakage of rice dur-
Int. Rice Res. Inst., Los Baños, Laguna, Phil- ing milling: A review. Trop. Sci. 22:255-276.
ippines. Bhattacharya, K. R., and Sowbhagya, C. M.
Barton, F. E., II, Windham, W. R., Champagne, 1971. Water uptake by rice during cooking.
E. T., and Lyon, B. G. 1998. Optimal geome- Cereal Sci. Today 16:420-424.
tries for the development of rice quality spec- Bhattacharya, K. R., and Sowbhagya, C. M.
troscopic chemometric models. Cereal Chem. 1972a. A colorimetric bran pigment method
75:315-319. for determining the degree of milling of rice. J.
Barton, F. E., II, Himmelsbach, D. S., McClung, Sci. Food Agric. 23:161-169.
A. M., and Champagne, E. T. 1999. Two Bhattacharya, K. R., and Sowbhagya, C. M.
Dimensional Spectroscopy of Rice. Proc. U.S.- 1972b. An improved alkali reaction test for
Jpn. Coop. Program Nat. Resour. (UJNR). rice quality. J. Food Technol. 7:323-331.
National Food Resesearch Institute, Tsukuba, Bhattacharya, K. R., and Sowbhagya, C. M.
Japan. 1976. An alkali degradation test and an alco-
Batcher, O. M., Deary, P. A., Helmintoller, K. F., holic alkali bran-staining test for determining
and Dawson, E. H. 1956. Development and the approximate degree of milling of rice. J.
application of methods for evaluating cooking Food Technol. 11:309-312.
and eating quality of rice. Rice J. 59:4-8, 32. Bhattacharya, K. R., and Sowbhagya, C. M.
Batcher, O. M., Deary, P. A., and Dawson, E. H. 1978. On viscograms and viscography, with
1957. Cooking quality of 26 varieties of special reference to rice flour. J. Texture Stud.
milled white rice. Cereal Chem. 34:277-285. 9:341-351.
Beachell, H. M., and Halick, J. V. 1956. Bhattacharya, K. R., and Sowbhagya, C. M.
Research being done to improve milling, proc- 1979. Pasting behavior of rice: A new method
essing and cooking qualities of rice. Rice J. of viscography. J. Food Sci. 44:797-800,
59:20, 22, 30-31. 804.
Beachell, H. M., and Halick, J. V. 1957. Breed- Bhattacharya, K. R., Desikachar, H. S. R., and
ing for improved milling, processing and Subrahmanyan, V. 1964. Curing of freshly
cooking characteristics of rice. Int. Rice harvested rice by heat treatment. Indian J.
Comm. Newsl. 6:1-7. Technol. 2:378-380.
Beachell, H. M., and Stansel, J. W. 1963. Bhattacharya, K. R., Sowbhagya, C. M., and
Selecting rice for specific cooking characteris- Indudhara Swamy, Y. M. 1972. Interrelation-
tics in a breeding program. Int. Rice Comm. ship between certain physicochemical proper-
Newsl. (Spec. Issue):25-40. ties of rice. J. Food Sci. 37:733-735.
Bennion, M. 1980. The Science of Food. Harper Bhattacharya, K. R., Sowbhagya, C. M., and
and Row, San Francisco, CA. Indudhara Swamy, Y. M. 1978. Importance
Bergman, C. J., Delgado, J. T., Bryant, R., of insoluble amylose as a determinant of rice
Grimm, C., Cadwallader, K. R., and Webb, B. quality. J. Sci. Food Agric. 29:359-364.
D. 2000. Rapid gas chromatographic tech- Bhattacharya, K. R., Indudhara Swamy, Y. M.,
nique for quantifying 2-acetyl-1-pyrroline and and Sowbhagya, C. M. 1979. Varietal differ-
hexanal in rice (Oryza sativa L.). Cereal ence in equilibrium moisture content of rice
Chem. 77:454-458. and effect of kernel chalkiness. J. Food Sci.
Bergman, C. J., Delgado, J. T., McClung, A. M., Technol. 16:214-215.
and Fjellstrom, R. G. 2001. An improved Bhattacharya, K. R., Sowbhagya, C. M., and
method for using a microsatellite in the rice Indudhara Swamy, Y. M. 1982a. Quality pro-
waxy gene to determine amylose class. Cereal files of rice: A tentative scheme for classifica-
Chem. 78:257-260. tion. J. Food Sci. 47:564-569.
Bergman, C. J., McClung, A. M., Pinson, S. R., Bhattacharya, K. R., Ramesh, B. S., and
and Fjellstrom, R. G. 2002. Development of Sowbhagya, C. M. 1982b. Dimensional classi-
PCR-based markers associated with cooked fication of rice for marketing. J. Agric. Eng.
kernel elongation and aroma. Proc. Rice Tech. 19:69-76.
Blakeney, A. B. 1979. Instron measurement of U.S. medium-grain rice using a Japanese taste
cooked-rice texture. Pages 343-353 in: Chemi- analyzer. Cereal Chem. 73:290-294.
cal Aspects of Rice Grain Quality. Int. Rice Champagne, E. T., Bett, K. L., Vinyard, B. T.,
Res. Inst., Los Baños, Laguna, Philippines. Webb, B. D., McClung, A., Barton, F. E.,
Blakeney, A. B. 1992. Developing rice varieties Lyon, B. G., Moldenhauer, K., Linscombe, S.,
with different textures and tastes. Chem. Aust. and Kohlwey, D. 1997. Effect of drying con-
59:475-476. ditions, final moisture content and degree of
Blakeney, A. B., Welsh, L. A., and Bannon, D. milling on rice flavor. Cereal Chem. 74:566-
R. 1991. Rice quality analysis using a com- 570.
puter controlled RVA. Pages 180-182 in Cere- Champagne, E. T., Lyon, B. G., Min, B. K.,
als International. D. J. Martin and C. W. Vinyard, B. T., Bett, K. L., Barton, F. E.,
Wrigley, Eds. R. Aust. Chem. Inst., II, Webb, B. D., McClung, A. M.,
Melbourne. Moldenhauer, K. A., Linscombe, S.,
Bligh, H. F. J., Till, R. I., and Jones, C. A. 1995. McKenzie, K. S., and Kohlwey, D. E. 1998.
A microsatellite sequence closely linked to the Effects of postharvest processing on texture
waxy gene of Oryza sativa. Euphytica 86:83- profile analysis of cooked rice. Cereal
85. Chem. 75:181-186.
Bolling, H., and El Baya, A. W. 1975. Effect of Champagne, E. T., Meullenet, I. F., Bett, K. L.,
lipids on the determination of amylose content McClung, A. M., and Kauffmann, D. 1999. A
in rice and wheat. (In German, with English novel instrumental method for rapid assess-
summary) Chem. Microbiol. Technol. ment of cooked rice texture characteristics. In:
Lebensm. 3:161-163. Two Dimensional Spectroscopy of Rice. Proc.
Borasio, L., De Rege, F., Ranghino, F., and U.S.-Jpn. Coop. Program Nat. Resour.
Leonzio, L. 1964. Cooking qualities of rice as (UJNR). National Food Resesearch Institute,
revealed by different tests: Iodine test, Refai Tsukuba, Japan.
test, alkali test, etc. (In Italian) Riso 13:58-78. Champagne, E. T., Bett, K. L., Grimm, C. C.,
Burns, D. A., and Ciurczak, E. W. 2001. Page McClung, A., Moldenhauer, K., Linscombe,
820 in: Handbook of Near-Infrared Analysis, S., McKenzie, K. S., and Barton, F. E. 2001.
2nd ed. Practical Spectroscopy, Vol. 27. Near-infrared reflectance analysis for predic-
Marcel Dekker, New York. tion of cooked rice texture. Cereal Chem.
Buttery, R. G., Ling, L. C., Juliano, B. O., and 78:358-362.
Turnbaugh, J. G. 1983. Cooked rice aroma and Chen, H., Marks, B. P., and Siebenmorgen, T. J.
2-acetyl-1-pyrroline. J. Agric. Food Chem. 1997. Quantifying surface lipid content of
31:823-826. milled rice via visible/near-infrared spectros-
Buttery, R. G., Ling, L. C., and Mon, T. R. 1986. copy. Cereal Chem. 74:826-831.
Quantitative analysis of 2-acetyl-1-pyrroline in Chen, M.-H., Bergman, C. J., and Fjellstrom, R.
rice. J. Agric. Food Chem. 34:112-114. G. 2003. SSSIIA locus genetic variation asso-
Buttery, R. G., Turnbaugh, J. G., and Ling, L. C. ciated with alkali spreading value in interna-
1988. Contribution of volatiles to rice aroma. tional rice germplasm. Page 314 in: Proc. Plant
J. Agric. Food Chem. 36:1006-1009. and Animal Genomes Conf., XI.
Cagampang, G. B., Perez, C. M., and Juliano, B. Chikubu, S. 1967. Chemical composition and
O. 1973. A gel consistency test for eating quality of rice in Japan. Int. Rice Comm.
quality of rice. J. Sci. Food Agric. 24:1589- Newsl. (Spec. Issue):146-160.
1594. Chikubu, S., Iwasaki, T., and Tani, T. 1960.
Cao, D., and Oard, J. H. 1997. Pedigree and Studies on cooking and eating qualities of
RAPD-based DNA analysis of commercial white rice. Part I. Comparison between Japa-
U.S. rice cultivars. Crop Sci. 37:1630-1635. nese rice and foreign rice. J. Jpn. Soc. Food
Champagne, E. T., and Hron, R. J., Sr. 1993. Nutr. 13:137-140.
Utilizing ethanol containing an antioxidant or Chikubu, S., Endo, I., and Tani, T. 1965. Studies
chelator to produce stable brown rice products. on cooking and eating qualities of white rice.
Cereal Chem. 70:562-567. (Part 2) Measurement of visco-elasticity
Champagne, E. T., Marshall, W. E. and Goynes, behavior of cooked rice by the parallel plate
W. R. 1990. Effects of degree of milling and plastometer. (In Japanese) Food Res. Inst.
lipid removal on starch gelatinization in the Bull. Jpn. 19:251-254.
brown rice kernel. Cereal Chem. 67:570-574. Chikubu, S., Endo, I., Shratori, T., Yoshikawa, S.,
Champagne, E. T., Richard, O. A., Bett, K. L., and Tani, T. 1971. Research on cooking quality
Grimm, C. C., Vinyard, B. T., Webb, B. D., of rice. Part 6. The application of texturometer.
McClung, A. M., Barton, F. E., II, Lyon, B. G., Proc. Agric. Chem. Soc. Jpn. 4F-09.
Moldenhauer, K., Linscombe, S., Mohindra, R., Chikubu, S., Watanabe, S., Sugimoto, T., Sakai,
and Kohlwey, D. 1996. Quality evaluation of F., and Taniguchi, Y. 1983. Relation between
palatability evaluations of cooked rice and types of amylopectin. Macromol. Symp.

physicochemical properties of rice. Part I. J. 140:179-186.
Jpn. Soc. Starch Sci. 30:333-341. Ebata, M., and Hirasawa, K. 1982. I. Textural
Chikubu, S., Watanabe, S., Sugimoto, T., parameters in relation to palatability. (In Japa-
Manabe, N., Sakai, F., and Taniguchi, Y. nese) Jpn. J. Crop Sci. 51:235-241.
1985. The relationship between palatability Ebata, M., Hirasawa, K., and Shibata, S. 1982.
evaluations of cooked rice and physicochemi- Studies on the texture of cooked rice. II.
cal properties of rice. II. Multiple regression Effects of kernel size, apparent quality of
analysis. J. Jpn. Soc. Starch Sci. 32:51-60. brown rice and degree of kernel maturation.
Chinnaswamy, R. 1985. Studies on expanded Jpn. J. Crop Sci. 51:242-247.
rice. Ph.D. thesis, University of Mysore, India. Endo, I., Chikubu, S., Suzuki, M., Kobayashi, K.,
Chinnaswamy, R., and Bhattacharya, K. R. 1986. and Naka, M. 1976. Palatability evaluation of
Characteristics of gel-chromatographic frac- cooked milled rice by physicochemical mea-
tions of starch in relation to rice and expanded surement. Rep. Natl. Food Res. Inst. Tokyo
rice-product qualities. Starch/Staerke 38:51- 31:1-11.
57. Endo, I., Yanase, H., Ishima, T., and Chikubu, S.
Chiou, H., Martin, M., and Fitzgerald, M. 2002. 1980a. Texture measurement of a very small
Effect of purification methods on rice starch amount of cooked rice by texturometer. I.
structure. Starch/Staerke 54:415-420. Investigation of measurement conditions and
Chrastil, J. 1990. Chemical and physicochemical application to main rice varieties. Rep. Natl.
changes of rice during storage at different Food Res. Inst. Tokyo 37:1-8.
temperatures. J. Cereal Sci. 11:71-74. Endo, I., Yanase, H., and Chikubu, S. 1980b. Use
Cobb, B. G., and Schoelles, D. B. 1989. Rapid of a new arm in a texturometer for simplifica-
method for determining amylose using a tion of the evaluation of cooked rice palatabil-
microplate reader. Hort. Sci. 24:487-488. ity and comparison with an ordinary arm. J.
Del Mundo, A. M. 1979. Sensory assessment of Jpn. Soc. Food Sci. Technol. 27:92-96.
cooked milled rice. Pages 313-325 in: Chemi- Feillet, P., Abecassis, J., and Alary, R. 1977.
cal Aspects of Rice Grain Quality. Int. Rice Description of a new method for measuring
Res. Inst., Los Baños, Laguna, Philippines. the viscoelastic properties of cereal products.
Delwiche, S. R., Bean, M. M., Muller, R. E., Bull. Anc. Eleves Ec. Fr. Meun. 278:97-101.
Webb, B. D., and Williams, P. C. 1995. Fellers, D. A., Mossman, A. P., and Suzuki, H.
Apparent amylose content of milled rice by 1983. Rice stickiness. II. Application of an
near-infrared reflectance spectrophotometry. Instron method to make varietal comparisons
Cereal Chem. 72:182-187. and to study modification of milled rice by
Deshpande, S. S., and Bhattacharya, K. R. 1982. hot-air treatment. Cereal Chem. 60:292-295.
The texture of cooked rice. J. Texture Stud. Fitzgerald, M. A., Martin, M., Ward, R. M., Park,
13:31-42. W. D., and Shead, H. J. 2003. Viscosity of rice
Desikachar, H. S. R. 1956. Changes leading to flour: A rheological and biological study. J.
improved culinary properties of rice on stor- Agric. Food Chem. 51:2295-2299.
age. Cereal Chem. 33:324-328. Fjellstrom, R. G., and McClung, A. M. 2002.
Desikachar, H. S. R., and Subrahmanyan, V. Development of DNA markers to distinguish
1961. The formation of cracks in rice during commercial U.S. rice cultivars. Page 60 in: Proc.
wetting and its effect on the cooking charac- Rice Tech. Work. Group Meet., 29th. La. Agric.
teristics of the cereal. Cereal Chem. 38:356- Exp. Stn., Rice Res. Stn., Crowley, LA.
364. Gangidi, R. R., Proctor, A. and Meullenet, J. F.
Desikachar, H. S. R., Raghavendra Rao, S. N., 2002. Milled rice surface lipid measurements
and Ananthachar, T. K. 1965. Effect of degree by diffuse Fourier transformed infrared spec-
of milling on water absorption of rice during troscopy. J. Am. Oil Chem. Soc. 79:7-12.
cooking. J. Food Sci. Technol. 2:110-112. Garland, S., Lewin, L., and Blakeney, A. 2000.
Dhulappanavar, C. V. 1976. Inheritance of scent PCR-based markers for the fragrance gene in
in rice. Euphytica 25:659-662. rice (Oryza sativa L.). Theor. Appl. Genet.
Dos Ramos, J. G., and Silebi, C. A. 1990. The 101:364-371.
determination of particle size distribution of Glaszmann, J. C. 1987. Isozymes and classifica-
submicrometer particles by capillary hydrody- tion of Asian rice varieties. Theor. Appl.
namic fractionation. J. Colloid Interface Sci. Genet. 74:21-30.
135:165-177. Goodwin, H. L., Rister, M. E., Koop, L. L.,
Dunn, L. B., and Krueger, W. J. 1999. Branching McClung, A. M., Miller, R. K., Bett, K. D.,
ratios of starch via proton nuclear magnetic Webb, B. D., Stansel, J. W., Gates, C. E.,
resonance and their use in determining amy- Dahm, P. F., Cadwallader, K. R., Kohlwey, D.,
lose/amylopectin content: Evidence for three and Dornak, J. 1994. Impacts of various cul-
tural, harvest and postharvest handling prac- Hizukuri, S. Takeda, Y., Maruta, N., and Juliano,
tices on rice quality attributes of Jasmine 85. B. O. 1989. Molecular structures of rice starch.
In: Proc. Rice Tech. Work. Group, 26th Texas Carbohydr. Res. 189:227-235.
Agric. Exp. Stn., Texas A&M University, Hogan, J. T., and Deobald, H. J. 1961. Note on a
College Station, TX. method of determining the degree of milling
Grimm, C. C., Bergman, C. J., Delgado, J. T., of whole milled rice. Cereal Chem. 38:291.
and Bryant, R. 2001. Screening for 2-acetyl-1- Hogan, J. T., and Planck, R. W. 1958. Hydration
pyrroline in the headspace of rice using characteristics of rice as influenced by cultivar
SPME/GC-MS. J. Agric. Food Chem. 49:245- and drying method. Cereal Chem. 35:469-482.
249. Hosaka, Y. 1987. Evaluation of taste of rice by
Grimm, C. C., Bergman, C. J., and Grimm, D. A. means of near infrared. Proc. Jpn. Soc. Agric.
2003. The analysis of oligosaccharides by Mach. pp. 357-360.
mass spectrometry. Pages 32-42 in Oligosac- Husain, A., Agrawal, K. K., and Pandya, A. C.
charides in Food and Agriculture. Symp. Ser. 1968. Physical properties of wheat and paddy.
849. G. Eggleston and G. L. Côté, Eds. Harvester (India), Anniversary No. 66.
American Chemical Society, Washington, DC. Hwang, H. G., Cho, R. K., Sohn, J. K., and Lee,
Halick, J. V., and Kelly, V. J. 1959. Gelatiniza- S. K. 1994. Rapid evaluation of chemical
tion and pasting characteristics of rice varieties components of rice grain near infrared spec-
as related to cooking behavior. Cereal Chem. troscopy. Korean J. Crop Sci. 39:7-14.
36:91-98. Ikehashi, H., and Khush, G. S. 1979. Methodol-
Halick, J. V., and Keneaster, K. K. 1956. The use ogy of assessing appearance of the rice grain,
of starch-iodine-blue test as a quality indicator including chalkiness and whiteness. Pages
of white milled rice. Cereal Chem. 33:315- 223-229 in: Proc. Workshop on Chemical
319. Aspects of Rice Grain Quality. Int. Rice Res.
Halick, J. V., Beachell, H. M., Stansel, J. W., and Inst., Los Baños, Laguna, Philippines.
Kramer, H. H. 1960. A note on the determina- Indudhara Swamy, Y. M., Sowbhagya, C. M.,
tion of gelatinization temperature of rice and Bhattacharya, K. R. 1978. Changes in the
varieties. Cereal Chem. 37:670. physicochemical properties of rice with aging.
Hamaker, B. R., and Griffin, V. K. 1993. Effect J. Sci. Food Agric. 29:627-639.
of disulfide bond-containing protein on rice IRRI. 1964. 1963 Annual Report. Int. Rice Res.
starch gelatinization and pasting. Cereal Inst., Los Baños, Laguna, Philippines.
Chem. 70:377-380. Iwamoto, M. 1980. Nondestructive quality meas-
Hampel, G. 1961. Determination of the viscosity urement of food by near infrared spectroscopic
of swollen rice with the help of Haake’s con- technique. J. Soc. Food Sci. Technol. 27:464-
sistometer. Getreide Mehl 11:133-135. 472.
Hampel, G. 1965. Investigations on the quality of Jahan, M. S., Nishi, A., Hamid, A., and Satoh, H.
white rice. I. Cooking characteristics, amylose 2002. Variation in amylopectin fine structure
content and viscosity. Getreide Mehl 15:132- of Bangladesh rice cultivars. Rice Genet.
136. Newsl. 19:30-32.
Hampel, G. 1966. Cooking characteristics and Jane, J., Chen, Y. Y., Lee, L. F., McPherson, A.
organoleptic evaluation of white rice varieties E., Wong, K. S., Radosavljevic, M., and
imported during the year 1964. Getreide Mehl Kasemsuwan, T. 1999. Effects of amylopectin
(Annu. Rep.) 16:56. branch chain length and amylose content on
Hampel, G. 1967. Investigations on the quality of the gelatinization and pasting properties of
white rice, alkali test, iodine blue value, rice starch. Cereal Chem. 76:629-637.
swelling number. Getreide Mehl 17:70-72. Jia, C. C., Yang, W., Siebenmorgen, T. J.,
Han, X. Z., and Hamaker, B. R. 2001. Amy- Bautista, R. C., and Cnossen, A. G. 2002. A
lopectin fine structure and rice starch paste study of rice fissuring by finite-element simu-
breakdown. Cereal Sci. 34:279-284. lation of internal stresses combined with high-
Hanselmann, R., Ehrat, M., and Widmer, H. M. speed microscopy imaging of fissure appear-
1995. Sedimentation field flow fractionation ance. Trans. ASAE 45:741-749.
combined with multiangle laser light scatter- Jodari, F., and Linscombe, S. D. 1996. Grain
ing applied for characterization of starch fissuring and milling yields of rice cultivars as
polymers. Starch/Staerke 46:345-349. influenced by environmental conditions. Crop
Hirannaiah, B. V., Bhashyam, M. K., and Ali, S. Z. Sci. 36:1496-1502.
2001. An improved cooking quality test for Juliano, B. O. 1971a. Relation of physico-
basmati rice. J. Food Sci. Technol. 38:116-119. chemical properties to processing character-
Hizukuri, S. 1986. Polymodal distribution of the istics of rice. Pages 21-27 in: Proc. World
chain lengths of amylopectins, and its signifi- Cereal and Bread Congress, 5th, Vol. 4. VEB
cance. Carbohydr. Res. 147:342-347. Fachbuchverlag, Leipzig.
Juliano, B. O. 1971b. A simplified assay for and Webb, B. D. 1981a. International coop-
milled-rice amylose. Cereal Sci. Today erative comparison of instrument methods for
16:334-336, 338, 360. cooked rice texture. J. Texture Stud. 12:17-38.
Juliano, B. O. 1972. Recent developments in rice Juliano, B. O., Perez, C. M., Blakeney, A. B.,
grain research. Pages 57-63 in: Reports of Castillot, D., Kongseree, N., Laignelet, B.,
Working and Discussion Meetings, 7th. Int. Lapis, E. T., Murty, V. V. S., Paule, C. M.,
Assoc. Cereal Chem., Vienna. and Webb, B. D. 1981b. International coop-
Juliano, B. O. 1979. The chemical basis of rice erative testing on the amylose content of
grain quality. Page 69-90 in: Chemical milled rice. Starch/Staerke 33:157-162.
Aspects of Rice Grain Quality. Int. Rice Res. Juliano, B. O., Blakeney, A. B., Butta, I.,
Inst., Los Baños, Laguna, Philippines. Castillo, D., Choudhury, N. H., Iwasaki, T.,
Juliano, B. O. 1982. An international survey of Shibuya, N., Kongseree, N., Lapis, E. T.,
methods used for evaluation of the cooking Murty, V. V. S., Paule, C. M., Perez, C. M.,
and eating qualities of milled rice. IRRI Res. and Webb, B. D. 1982. International
Paper Ser. 77. Int. Rice Res. Inst., Los Baños, cooperative testing of the alkali digestibility
Laguna, Philippines. values for milled rice. Starch/Staerke 34:21-
Juliano, B. O. 1985. Criteria and tests for rice 26.
grain qualities. Pages 443-524 in: Rice: Juliano, B. O., Perez, C. M., and Cuevas-Perez,
Chemistry and Technology, 2nd ed. B. O. F. 1993. Screening for stable high head rice
Juliano, Ed. Am. Assoc. Cereal Chem., St. yields in rough rice. Cereal Chem. 70:650-
Paul, MN. 655.
Juliano, B. O. 1998. Varietal impact on rice Kang, K. J., Kim, K., Kim, S. K., and Murata, A.
quality. Cereal Foods World 43:207-211, 214- 1994. Relationship between molecular struc-
216, 218-222. ture of amylose and texture of cooked rice of
Juliano, B. O., and Pascual, C. G. 1980. Quality Korean rices. J. Appl. Glycosci. 41:35-40.
characteristics of milled rice grown in different Kang, K. J., Kim, K., and Kim, S. K. 1995a.
countries. IRRI Res. Paper Ser. 48:1-25. Int. Relationship between molecular structure of
Rice Res. Inst., Los Baños, Laguna, Philip- rice amylopectin and texture of cooked rice.
pines. Korean J. Food Sci. Technol. 27:105-111.
Juliano, B. O., and Perdon, A. A. 1975. Gel and Kang, K. J., Kim, K., and Kim, S. K. 1995b.
molecular properties of nonwaxy rice starch. Structure and properties of hot water-insoluble
Staerke 27:115-120. rice starch. Korean J. Food Sci. Technol.
Juliano, B. O., and Perez, C. M. 1983. Major 27:631-634.
factors affecting cooked milled rice hardness Kang, K. J., Kim, K., and Kim, S. K. 1995c.
and cooking time. J. Texture Stud. 14:235- Structure of hot-water soluble rice starch in
243. relation to the structure of rice starch and
Juliano, B. O., and Perez, C. M. 1986. Kinetic texture of cooked rice. Korean J. Food Sci.
studies on cooking of tropical milled rice. Technol. 27:757-761.
Food Chem. 20:97-105. Karim, A. A., Norziah, M. H., and Seow, C. C.
Juliano, B. O., Oate, L. U., and Del Mundo, A. 2000. Methods for the study of starch retro-
M. 1965. Relation of starch composition, pro- gradation. Rev. Food Chem. 71:9-36.
tein content and gelatinization temperature to Kawamura, S., Natsuga, M., Kouno, S. and Itoh,
cooking and eating qualities of milled rice. K. 1996. Instrument analysis and sensory test
Food Technol. 19:1006-1011. for rice taste evaluation. Jpn. Agric. Eng.
Juliano, B. O., Nazareno, M. B., and Ramos, N. 58:95-100.
B. 1969. Properties of waxy and isogenic Kawano, S., and Chie, I. 1999. Determination of
nonwaxy rices differing in starch gelatiniza- moisture and protein contents of single kernels
tion temperature. J. Agric. Food Chem. of brown rice by NIR transmittance spectros-
17:1364-1369. copy. In: Two Dimensional Spectroscopy of
Juliano, B. O., Perez, C. M., Blakeney, A. B., Rice. Proc. U.S.-Jpn. Coop. Program Nat.
Breckenridge, C., Toro, D. C., Choudhury, N. Resour. (UJNR). National Food Resesearch
H., Kongseree, N., Laignelet, B., Merca, F. E., Institute, Tsukuba, Japan.
Paule, C. M., and Webb, B. D. 1980. Report of Kim, K. A., and Jeon, B. R. 1996. Physicochemi-
the international cooperative testing on gel cal properties and hydration of rice on various
consistency of milled rice. Riso 29:233-237. polishing degrees. Korean J. Food Sci. Tech-
Juliano, B. O., Perez, C. M., Barber, S., nol. 28:959-964.
Blakeney, A. B., Iwasaki, T., Shibuya, N., Kim, H. J., Hong, S. P., Park, K. H., Kim, I. H.,
Keneaster, K. K., Chung, S., Laignelet, B., and Lee, S. H. 1997. Comparative study of
Launay, B., Del Mundo, A. M., Suzuki, H., japonica rice protein with different eating
Shiki, J., Tsuji, S., Tokoyama, J., Tatsumi, K., quality. Foods Biotechnol. 6:149-153.
Kim, S. S., Rhyu, M.-R., Kim, J. M., and Lee, S.- Lan, Y., Fang, Q., Kocher, M. F., and Hanna, M.
H. 2003. Authentication of rice using near- A. 2002. Detection of fissures in rice grains
infrared reflectance spectroscopy. Cereal using imaging enhancement. Int. Food Prop.
Chem. 80:346-349. 5:205-215.
Kobayashi, A., Shimizu, T., and Ohtsubo, K. Larkin, P. D., and Park, W. D. 1999. Transcript
1994. Rapid determination of amylose con- accumulation and utilization of alternate and
tents of newly bred characteristic rices by near non-consensus splice sites in rice granule-
infrared spectroscopy. Rep. Natl. Food Res. bound starch synthase are temperature-sensi-
Inst. 58:21-26. tive and controlled by a single-nucleotide
Kohlwey, D. E. 1994. New methods for the polymorphism. Plant Mol. Biol. 40:719-727.
evaluation of rice quality and related termi- Lee, S. J., and Peleg, M. 1988. Direct measure-
nology. Pages 113-138 in: Rice Science ment of the attractive force between individual
and Technology. W. E. Marshall and J. I. cooked rice grains of sticky and flaky culti-
Wadsworth, Eds. Marcel Dekker, New vars. J. Food Sci. 53:1113-1115.
York. Li, W. S., and Shaw, J. T. 1997. Determining the
Koizumi, K., Fukuda, M., and Hizukuri, S. 1991. fat acidity of rough rice by near-infrared
Estimation of the distributions of chain length reflectance spectroscopy. Cereal Chem.
of amylopectin by high-performance liquid 74:556-560.
chromatography with pulsed amperometric Lin, C. F., Hsieh, T. C.-Y., and Huff, B. J. 1990.
detection. J. Chromatogr. 585:233-268. Identification and quantification of the “pop-
Kongseree, N., and Juliano, B. O. 1972. Physico- corn”-like aroma in Louisiana aromatic Della
chemical properties of rice grain and starch rice (Oryza sativa L.). J. Food Sci. 55:1466-
from lines differing in amylose content and 1467, 1469.
gelatinization temperature. J. Agric. Food Lisle, A. J., Martin, M., and Fitzgerald, M. A.
Chem. 20:714-718. 2000. Chalky and translucent rice grains differ
Krzanowski, W. J. 1995. The authentication of in starch composition and structure and cook-
basmati rice using near infrared spectroscopy: ing properties. Cereal Chem. 77:627-632.
Some further analysis. J. Near Infrar. Spec. Little, R. R., Hilder, G. B., and Dawson, E. H.
3:111-117. 1958. Differential effect of dilute alkali on 25
Kurasawa, H., Igaue, I., and Hayakawa, T. 1962. varieties of milled white rice. Cereal Chem.
Study on the eating-quality (especially sticki- 35:111-126.
ness) of white-milled non-waxy rice. (Part III). Liu, W., Tao, Y., Siebenmorgen, T. J., and Chen,
A starch-iodine blue test of rice powder as a H. 1998. Digital image analysis method for
quality indication of the rice. J. Jpn. Soc. Food rapid measurement of rice degree of milling.
Nutr. 15:84-87. Cereal Chem. 75:380-385.
Kurasawa, H., Kanauti, Y., Yamamoto, I., Lloyd, B. J., Cnossen, A. G., and Siebenmorgen,
Hayakawa, T., and Igaue, I. 1969. Some T. J. 2001. Evaluation of two methods for
physicochemical properties of nonwaxy paddy separating head rice from brokens for head
rice starch in Niigata prefecture. Part I. Rela- rice yield determination. Appl. Eng. Agric.
tion of properties of starch to eating and 17:643-648.
cooking qualities of milled rice. Agric. Biol. Lu, R., and Siebenmorgen, R. 1992. Moisture
Chem. 33:798-806. diffusivity of long-grain rice components.
Laignelet, B., and Alary, R. 1978. Determination Trans. ASAE 35:1955-1961.
of viscoelastic characteristics of cooked rice Lu, S., Chen, L. N., and Lii, C. Y. 1997. Correla-
with the aid of a viscoelastograph. Ann. Tech- tions between the fine structure, physico-
nol. Agric. 27:783-797. chemical properties, and retrogradation of
Laignelet, B., and Feillet, P. 1979. Use of the amylopectins from Taiwan rice varieties.
viscoelastograph for measuring the texture of Cereal Chem. 74:34-39.
cooked rice. Pages 355-361 in: Chemical Lyon, B. G., Champagne, E. T., Vinyard, B. T.
Aspects of Rice Grain Quality. Int. Rice Res. and Windham, W. R. 2000. Sensory and
Inst., Los Baños, Laguna, Philippines. instrumental relationships of texture of
Lam, H. S., and Proctor, A. 2001. Rapid methods cooked rice from selected cultivars and
for milled rice surface lipid and free fatty acid postharvest handling practices. Cereal Chem.
determination. Cereal Chem. 78:498-499. 77:64-69.
Lan, Y., and Kunze, O. R. 1996a. Fissure resis- Mackill, D. J. 1995. Classifying Japonica rice
tance of rice varieties. Appl. Eng. Agric. cultivars with RAPD markers. Crop Sci.
12:365-368. 35:889-894.
Lan, Y., and Kunze, O. R. 1996b. Fissure char- Majzoobi, M., Rowe, A. J., Connock, M., Hill, S.
acteristics related to moisture adsorption E., and Hardings, S. E. 2003. Partial fractiona-
stresses in rice. Trans. ASAE 39:2169-2174. tion of wheat starch amylose and amylopectin
using zonal ultracentrifugation. Carbohydr. Meullenet, J. F., Horner, T. B., and Marks, B. P.
Polym. 52:269-274. 2002. Prediction of the texture of cooked
Maniñgat, C. C., and Juliano, B. O. 1978. Alkali white rice by near-infrared reflectance analysis
digestibility pattern, apparent solubility and of whole-grain milled samples. Cereal Chem.
gel consistency of milled rice. Starch/Staerke 79:52-57.
30:125-127. Mohandoss, R., and Pillaiyar, P. 1980. An extru-
Manohar Kumar, B., Upadhyay, J. K., and sion test for determining the palatability of
Bhattacharya, K. R. 1976. Objective tests for parboiled rices. J. Food Sci. Technol. 17:244-
the stickiness of cooked rice. J. Texture Stud. 246.
7:271-278. Moritaka, S., and Yasumatsu, K. 1972a. Studies
Marks, B. P., Chen, H., Bellman, T. L., Daniels, on cereals. 9. Sulphydryl and disulphide con-
M. J., Perdon, A. A., and Siebenmorgen, T. J. tents of milled rice. (In Japanese) J. Jpn. Soc.
1997. Cooking and amylograph properties of Food Nutr. 25:42-45.
rice via near-infrared spectroscopy. (Abstr.) Moritaka, S., and Yasumatsu, K. 1972b. Studies
Cereal Foods World 42:636. on cereals. 10. The effect of sulphydryl groups
Marshall, W. E. 1992. Effect of degree of milling on storage deterioration of milled rice. (In
of brown rice and particle size of milled rice Japanese) J. Jpn. Soc. Food Nutr. 25:59-62.
on starch gelatinization. Cereal Chem. 69:632- Mossman, A. P., Fellers, D. A., and Suzuki, H.
636. 1983. Rice stickiness. I. Determination of rice
Martin, M., and Fitzgerald, M. A. 2002. Proteins stickiness with an Instron tester. Cereal Chem.
in rice grains influence cooking properties. 60:286-292.
Cereal Sci. 36:285-294. Murray, I., and Williams, P. C. 1987. Chemical
Matsue, Y., Odahara, K. and Hiramatsu, M. principles of near-infrared technology. Pages
1994. Differences in protein content, amylose 17–34. in Near-Infrared Technology in the
content, and palatability in relation to location Agricultural and Food Industries, 1st ed. P. C.
of grains within a rice panicle. Jpn. J. Crop Williams and K. H. Norris, Eds. Am. Assoc.
Sci. 63:271-277. Cereal Chem., St. Paul, MN.
Mazurs, E. G., Schoch, T. J., and Kite, F. E. Nakamura, Y., Sakurai, A., Inaba, Y., Kimura,
1957. Graphical analysis of the Brabender vis- K., Iwasawa, N., and Nagamine, T. 2002. The
cosity curves of various starches. Cereal fine structure of amylopectin endosperm from
Chem. 34:141-152. Asian cultivated rice can be largely classified
McClung, A. M., and Castro, E. M. 1994. into two classes. Starch/Staerke 54:117-131.
Screening for milling yield using rough rice Natsuga, M. 1999. Grain quality determination
and reduced sample size. Page 49 in: Proc. by near-infrared spectroscopy. 4. Grain quality
Rice Technical Working Group Meet., 25th. determination by near-infrared spectroscopy.
Tex. Agric. Exp Stn., Texas A&M University, Mem. Fac. Agric. Hokkaido Univ. 22:127-
College Station. 168.
Merca, F. E., and Juliano, B. O. 1981. Physico- Nishi, A., Nakamura, Y., Tanaka, N., and Satoh,
chemical properties of starch of intermediate- H. 2001. Biochemical and genetic analysis of
amylose and waxy rices differing in grain the effects of amyloseextender mutation in rice
quality. Starch/Staerke 33:253-260. endosperm. Plant Physiol. 127:459-472.
Meullenet, J. F., and Gross, J. 1999. Instrumental Nishita, K. D., and Bean, M. M. 1982. Grinding
single and double compression tests to predict methods: Their impact on rice flour properties.
sensory texture characteristics of foods. J. Cereal Chem. 61:63-68.
Texture Stud. 30:167-180. Normand, F. L., and Marshall, W. E. 1989. Dif-
Meullenet, J. F., Gross, J., Marks, B. P., and ferential scanning calorimetry of whole grain
Daniels, M. 1998. Sensory descriptive texture milled rice and milled rice flour. Cereal Chem.
analyses of cooked rice and its correlation to 66:317-321.
instrumental parameters using an extrusion Ohtsubo, K., Toyoshima, H., and Okadome, H.
cell. Cereal Chem. 75:714-720. 1998. Quality assay of rice using traditional
Meullenet, J. F., Marks, B. P., Griffin, K., and and novel tools. Cereal Foods World 43:203-
Daniels, M. J. 1999. Effect of rough rice dry- 206.
ing and storage conditions on sensory pro- Ohtsubo, K., Nakamura, S., Morooka, H., Fujii,
files of cooked rice. Cereal Chem. 76:483- T., Fuse, T., and Kawasaki, S. 1999. Cultivar
486. identification of rice using a single kernel of
Meullenet, J. F., Champagne, E. T., Bett, K. L., boiled rice. (In Japanese with English sum-
McClung, A. M., and Kauffmann, D. 2000. mary) Nippon Shokuhin Kagaku Kogaku Kai-
Instrumental assessment of cooked rice texture shi 46:262-267.
characteristics: A method for breeders. Cereal Ohtsubo, K., Nakamura, S., and Imamura, T.
Chem. 77:512-517. 2002. Development of the primer sets for
identification of a rice cultivar, Koshihikari, on rice pasting properties. Cereal Chem.

by PCR. (In Japanese with English summary) 78:205-209.
Nippon Nogeikagaku Kaishi 76:388-397. Perez, C. M. 1979. Gel consistency and viscosity
Okabe, M. 1979. Texture measurement of cooked of rice. Pages 293-302 in: Chemical Aspects
rice and its relationship to the eating quality. J. of Rice Grain Quality. Int. Rice Res. Inst., Los
Texture Stud. 10:131-152. Baños, Laguna, Philippines.
Okadome, H., Toyoshima, H., Suto, M., Ando, I., Perez, C. M., and Juliano, B. O. 1978. Modifica-
Numaguchi, K., Horisue, N., and Ohtsubo, K. tion of the simplified amylose test for milled
1998. Development of advanced physical rice. Starch/Staerke 30:424-426.
measurement method for individual cooked Perez, C. M., and Juliano, B. O. 1979. Indicators
rice. II. Palatability evaluation for japonica of eating quality for non-waxy rices. Food
rice grains based on multiple physical meas- Chem. 4:185-195.
urements of individual cooked rice grain. Jpn. Perez, C. M., Pascual, C. G., and Juliano, B. O.
J. Soc. Food Sci. Technol. 45:398-407. 1979. Eating quality indicators for waxy rices.
Okadome, H., Toyoshima, H., and Ohtsubo, K. Food Chem. 4:179-184.
1999a. Multiple measurements of physical Perez, C. M., Juliano, B. O., Bourne, M. C., and
properties of individual cooked rice grains Morales, A. A. 1993. Hardness of cooked
with a single apparatus. Cereal Chem. 76:855- milled rice by instrumental and sensory meth-
860. ods. J. Texture Stud. 24:81-94.
Okadome, H., Kurihara, M., Kusuda, O., Pinson, S. R. M. 1994. Inheritance of aroma in
Toyoshima, H., Shimotsubo, K., Matsuda, T., six rice cultivars. Crop Sci. 34:1151-1157.
and Ohtsubo, K. 1999b. Changes in physical Primo, E., Barber, S., and Benedito de Barber, C.
properties of cooked rice grains cultivated 1965. Quality factors of rice. XVI. Influence
with different nitrogenous fertilizers. Pages of sulphydryl and disulphide groups. Rev.
43-48 in: Two Dimensional Spectroscopy of Agroquim. Technol. Aliment. 5:232-238.
Rice. Proc. U.S.-Jpn. Coop. Program Nat. Primo, E., Barber, S., Benedito de Barber, C., and
Resour. (UJNR). National Food Resesearch Hernandez, D. 1966. Quality factors of rice.
Institute, Tsukuba, Japan. XVII. Influence of the sulphydryl and disul-
Ong, M. H., and Blanshard, J. M. V. 1995. Tex- phide groups of the outer layer of the grain.
ture determinants in cooked, parboiled rice. I: Rev. Agroquim. Technol. Aliment. 6:476-483.
Rice starch amylose and the fine structure of Radhika Reddy, K., Ali, S. Z., and Bhattacharya,
amylopectin. J. Cereal Sci. 21:251-260. K. R. 1993. The fine structure of rice-starch
Ory, R. L., Delucca, A. J., Angelo, A. J., and amylopectin and its relation to the texture of
Dupuy, H. P. 1980. Storage quality of brown cooked rice. Carbohydr. Polym. 22:267-275.
rice as affected by packaging with and without Radhika Reddy, K., Subramanian, R., Ali, S. Z.,
carbon dioxide. J. Food Prot. 43:929-932. and Bhattacharya, K. R. 1994. Viscoelastic
O’Shea, M. G., and Morell, M. K. 1996. High properties of rice-flour and their relationship
resolution gel electrophoresis of 8-amino- to amylose content and rice quality. Cereal
1,3,6-pyrentrisulfonic acid (APTS) tagged Chem. 71:548-552.
oligiosaccharides using a DNA sequencer. Raju, G. N., Manjunath, N., and Srinivas, T.
Electrophoresis 17:681-688. 1991. Grain chalkiness in cereals. Trop. Sci.
Paige, D., Geng, S., and Jongkaewwattana, S. 31:407-415.
1991. Apparatus for automatic measurement of Ramesh, M., Ali, S. Z., and Bhattacharya, K. R.
kernel length, width and thickness. Crop Sci. 1999a. Structure of rice starch and its relation
31:1314-1318. to cooked-rice texture. Carb. Polym. 38:337-
Park, J. K., Kim, S. S., and Kim, K. O. 2001. 347.
Effect of milling ratio on sensory properties of Ramesh, M., Mitchell, J. R., Jumel, K., and
cooked rice and on physicochemical properties Harding, S. E. 1999b. Amylose content of rice
of milled and cooked rice. Cereal Chem. starch. Starch/Staerke 51:311-313.
78:151-156. Ranghino, F. 1966. Evaluation of the resistance
Parthasarathi, N. V. V., and Nath, N. 1953. of rice to cooking as a function of the time of
Cooking and energy requirements. J. Sci. gelatinization of the grains. Riso 15:117-127.
Indus. Res. 12B:224-226. Reddy, P. R., and Sathyanarayanaiah, K. 1980.
Patil, N. D.1988. Evaluation of diffusion equa- Inheritance of aroma in rice. Indian J. Genet.
tion for simulating moisture movement within Plant Breed. 40:327-329.
an individual grain kernel. Drying Technol. Reece, J., E., and Blakeney, A. B. 1993. Image
6:21-42. analysis for the assessment of chalk in milled
Perdon, A. A., Siebenmorgen, T. J., rice. Pages 21-25 in: Proc. Aust. Cereal Chem.
Mauromoustakos, A., Griffin, V. K., and Conf., 43rd. C. W. Wrigley, Ed. Royal Aus-
Johnson, E. R. 2001. Degree of milling effects tralian Chemical Institute, Melbourne.
Rousset, S., Pons, B., and Pilandon, C. 1995. cooking quality traits of indica rice in different
Sensory texture profile, grain physico-chemical environments. Theor. Appl. Genet. 95:294-
characteristics and instrumental measurements 300.
of cooked rice. J. Texture Stud. 26:119-135. Shibuya, N., Iwasaki, T., Yanase, H., and
Sandhya Rani, M. R., and Bhattacharya, K. R. Chikubu, S. 1974. Studies on deterioration of
1995a. Rheology of rice-flour pastes: Rela- rice during storage. I. Changes of brown rice
tionship of paste breakdown to rice quality, and milled rice during storage. J. Jpn. Soc.
and a simplified Brabender Viscograph test. J. Food Sci. Technol. 21:597-603.
Texture. Stud. 26:587-598. Shimizu, N., Katsura, J., Yanagisawa, T., Tezuka,
Sandhya Rani, M. R., and Bhattacharya, K. R. B., Maruyama, Y., Inoue, S., Eddison, C. G.,
1995b. Microscopy of rice starch granules Cowe, I. A., Withey, R. P., Blakeney, A. B.,
during cooking. Starch/Staerke 47:334-337. Kimura, T., Yoshizaki, S., Okadome, H.,
Sanjiva Rao, B. 1938. Investigation on rice. Curr. Toyoshima, H., and Ohtsubo, K. 1998. Evalu-
Sci. India 6:446-447. ating techniques for rice grain quality using
Sanjiva Rao, B. 1948. Rice: Physico-chemical near infrared transmission spectroscopy. J.
aspects of its curing to secure vitamin conser- Near Infrared Spectrosc. 6:A111-A116.
vation and improvement in storage and cooking Shimizu, N., Katsura, J., Yanagisawa, T., Inoue,
quality. Pages 259-270 in: Proc. Indian Sci. S., Withey, R. P., Cowe, I. A., Eddison, C. G.,
Congr., 35th. Part II: Presidential addresses. Blakeney, A. B., Kimura, T., Yoshizaki, S.,
Indian Science Congress Assoc., Calcutta. Okadome, H., Toyoshima, H., and Ohtsubo, K.
Sanjiva Rao, B., Vasudeva Murthy, A. R., and 1999. Determination of apparent amylose
Subrahmanya, R. S. 1952. The amylose and content in Japanese milled rice using near-
the amylopectin contents of rice and their infrared transmittance spectroscopy. Food Sci.
influence on the cooking quality of the cereal. Technol. Res. 5:337-342.
Proc. Indian Acad. Sci. 36B:70-80. Shimizu, N., Kimura, T., Yanagisawa, T.,
Sarker, N. N., Kunze, O. R., and Strouboulis, T. Andren, H., Okadome, H., Toyoshima, H., and
1996. Transient moisture gradients in rough Ohtsubo, K. 2000. Analysis of practical cali-
rice mapped with finite element model and bration models characteristics for rice quality
related to fissures after heated air drying. evaluation using near-infrared spectroscopy.
Trans. ASAE 39:625-631. Paper 006121. ASAE Annual International
Schoch, T. J., and Maywald, E. C. 1956. Micro- Meeting. American Society of Agricultural
scopic examination of modified starches. Anal. Engineers, St. Joseph, MI.
Chem. 28:382-387. Shin, M.-G., Yoon, S.-H., Rhee, J.-S., and Kwon,
Sesmat, A., and Meullenet, J.-F. 2001. Prediction T.-W. 1986. Correlation between oxidative
of rice sensory texture attributes from a single deterioration of unsaturated lipid and hexanal
compression test, multivariate regression, and during storage of brown rice. J. Food Sci.
a stepwise model optimization method. J. 51:460-463.
Food Sci. 66:124-131. Shu, Q. Y., Wu, D. X., Xia, Y. W., Gao, M. W.,
Sevenou, O., Hill, S. E., Farhat, J. A., and and McClung, A. 1999a. Analysis of grain
Mitchell, J. R. 2003. Organization of the quality characters in small ground brown rice
external region of the starch granule as deter- samples by near infrared reflectance spectros-
mined by infrared spectroscopy. Int. J. Biol. copy. Sci. Agric. Sin. 32:92-97.
Macromol. 31:79-85. Shu, Q. Y., Wu, D. X., Xia, Y. W., Gao, M. W.,
Shams-Ud-Din, M. D., and Bhattacharya, K. R. and McClung, A. 1999b. Calibration optimi-
1978. On the meaning of the degree of milling zation for rice apparent amylose content by
of rice. J. Food Technol. 13:99-105. near infrared reflectance spectroscopy (NIRS).
Shenk, J. S., and Westerhaus, M. O. 1991. Popu- J. Zhejiang Agric. Univ. 25:343-346.
lation definition, sample selection and calibra- Simpson, J. E., Adair, C. R., Kohler, G. O.,
tion procedures for near infrared reflectance Dawson, E. H., Deobald, G. J., Kester, E. B.,
spectroscopy. Crop Sci. 31:469-474. Hogan, J. T., Batcher, O. M., and Halick, J.
Shanthy, A. P., Sowbhagya, C. M., and V. 1965. Quality evaluation studies of
Bhattacharya, K. R. 1980. Simplified determi- foreign and domestic rices. Tech. Bull. 1331.
nation of water-insoluble amylose content of Agric. Res. Serv. U.S.-Dep. Agric.,
rice. Starch/Staerke 32:409-411. Washington, DC.
Sharp, R. N., and Timme, L. K. 1986. Effects of Singh, V., Okadome, H., Toyoshima, H., and
storage time, storage temperature and pack- Ohtsubo, K. 2000. Thermal and physico-
aging methods on the shelf-life of brown rice. chemical properties of rice grain, flour and
Cereal Chem. 63:247-251. starch. J. Agric. Food Chem. 48:2639-2647.
Shi, C. H., Zhu, J., Zang, R. C., and Chen, G. L. Sinnaeve, G., Dardenne, P., and Agneessens, R.
1997. Genetic and heterosis analysis for 1994. Global or local? A choice for NIR cali-
brations in analysis of forage quality. J. Near Suzuki, H. 1979. Amylography and alkali vis-
Infrared Spectrosc. 2:163-175. cography of rice. Pages 261-273 in: Chemical
Sitakalin, C., and Meullenet, J. F. 2000. Predic- Aspects of Rice Grain Quality. Int. Rice Res.
tion of cooked rice texture using extrusion and Inst., Los Baños, Laguna, Philippines.
compression tests in conjunction with spectral Suzuki, H., Shiki, J., and Inoue, T. 1983a. Hard-
stress strain analysis. Cereal Chem. 77:501- ness and stickiness of Indica waxy and non-
506. waxy rices and Japonica nonwaxy rices
Sood, B. G., and Siddiq, E. A. 1979. A rapid cooked by excess water method and rice
technique for scent determination in rice. cooker method. J. Jpn. Soc. Nutr. Food Sci.
Indian J. Genet. Plant Breed. 38:268-271. 36:389-392.
Sowbhagya, C. M., and Ali, S. Z. 1991. Effect of Suzuki, H., Shiki, J., and Osugi, M. 1983b. Con-
pre-soaking on cooking time and texture of ditions for measuring hardness and stickiness
raw and parboiled rice. J. Food Sci. Technol. of cooked rice grains by texturometer. J. Jpn.
28:76-80. Soc. Nutr. Food Sci. 36:199-202.
Sowbhagya, C. M., and Bhattacharya, K. R. Takeda, Y., Hizukuri, S., and Juliano, B. O.
1971. A simplified colorimetric method for 1987. Structures of rice amylopectins with low
determination of amylose content in rice. and high affinities for iodine. Carbohydr. Res.
Staerke 23:53-56. 168:79-88.
Sowbhagya, C. M., and Bhattacharya, K. R. Takeda, Y., Maruta, N., Hizukuri, S., and Juliano,
1976. Lipid auto-oxidation in rice. J. Food Sci. B. O. 1989. Structures of indica rice starches
41:1018-1023. (IR 48 and IR 64) having intermediate affini-
Sowbhagya, C. M., and Bhattacharya, K. R. ties for iodine. Carbohydr. Res. 187:287-294.
1979. Simplified determination of amylose in Tanaka, M. 1975. General Foods Texturometer
milled rice. Starch/Staerke 31:159-163. application to food texture research in Japan.
Sowbhagya, C. M., and Bhattacharya, K. R. J. Texture Stud. 6:101-116.
2001. Changes in pasting behavior of rice Tanaka, K., Jinnouchi, N., and Harada, K. 1999.
during aging. J. Cereal Sci. 34:115-124. How to use NIR spectrophotometer for evalu-
Sowbhagya, C. M., Ramesh, B. S., and ating rice palatability. Kyushu Agric. Res.
Bhattacharya, K. R. 1984. Improved indices 61:9.
for dimensional classification of rice. J. Food Tanchotikul, U., and Hsieh, T. C.-Y. 1991. An
Sci. Technol. 21:15-19. improved method for quantification of 2-acetyl-
Sowbhagya, C. M., Ramesh, B. S., and 1-pyrroline, a “popcorn”-like aroma, in aromatic
Bhattacharya, K. R. 1987. The relationship rice by high-resolution gas chromatogra-
between cooked rice texture and the physico- phy/mass spectrometry/selected ion monitoring.
chemical characteristics of rice. J. Cereal Sci. J. Agric. Food Chem. 39:944-947.
5:287-297. Tani, T., Chikubu, S. and Shikano, T. 1952. A
Sreenivasan, A. 1938. Investigations on rice. new M.G. method of staining for grain. I.
Current Sci. India 6:615-616. Application to pressed barley and milled rice.
Sreenivasan, A. 1939. Studies on quality in rice. Shokuryo Kenkyusho Kenkyu Hokoku 6:75-
IV. Storage changes in rice after harvest. 78.
Indian J. Agric. Sci. 9:208-222. Tashiro, T., and Ebata, M. 1975. The effect of
Srinivas, T., Bhashyam, M. K., Mahadevappa, ripening conditions on occurrence of white
M., and Desikachar, H. S. R. 1977. Varietal belly kernel. Proc. Crop Sci. Soc. Jpn. 44:86-
differences in crack formation due to weath- 92.
ering and wetting stress in rice. Indian J. Tashiro, T., and Wardlaw, I. F. 1991. The effect
Agric. Sci. 47:27-31. of high temperature on kernel dimensions and
Steffe, J. F., and Singh, R. P. 1980. Liquid diffu- the type and occurrence of kernel damage in
sivity of rough rice components. Trans. ASAE rice. Aust. J. Agric. Res. 42:485-496.
23:767-774, 782. Tripathi, R. S., and Rao, M. J. 1979. Inheritance
Stobdan, T., Khanna, V. K., Singh, N. K. and and linkage relationship of scent in rice.
Mani, S. C. 1999. DNA fingerprinting of Euphytica 28:319-323.
Indian aromatic rice genotypes by RAPD Tsugita, T. 1986. Aroma of cooked rice. Food
markers. Rice Biotech. Q. 40:18-19. Rev. Int. 1:497-520.
Stone, H., and Sidel, J. L. 1993. Sensory Tsugita, T., Ohta, T., and Kato, H. 1983. Cook-
Evaluation Practices. 2nd ed. Academic Press, ing flavor and texture of rice stored under dif-
San Diego CA. ferent conditions. Agric. Biol. Chem. 47:543-
Sugiyama, J., Kurokouchi, K., and Horiuchi, H. 549.
1990. A system for measuring dynamic vis- Tsuji, S. 1981. Texture measurement of cooked
coelasticity of a small irregularly shaped sam- rice kernels using the multiple-point mensura-
ple. J. Jpn. Soc. Food Sci. Technol. 37:61-67. tion method. J. Texture Stud. 12:93-105.
Tsuji, S. 1982. Texture profile analysis of proc- McKenzie, K. S. 1997. Prediction of cooked
essed foods using the Tensipresser and the rice texture quality using near-infrared reflec-
multiple-point mensuration method. J. Texture tance analysis of whole-grain milled samples.
Stud. 13:135-186. Cereal Chem. 74:626-632.
Tsuji, S. 1988. Changes in physical properties of Yadav, B. K., and Jindal, V. K. 2001. Monitoring
pastes of glutinous and non-glutinous rice milling quality of rice by image analysis.
flours by additives. Measurements of physical Comput. Electron. Agric. 33:19-33.
properties of pastes of rice flours using the Yamada, H., Sasagawa, N., and Kitazawa, K.
multi-biting test-II. J. Jpn. Soc. Food Sci. 1993. Suitabilities of improved rice varieties
Technol. 35:742-747. and lines for rice crackers. J. Jpn. Soc. Food
Tsuzuki, E., and Shimokawa, E. 1990. Inheri- Sci. Technol. 40:278-286.
tance of aroma in rice. Euphytica 46:157-159. Yanase, H., Ohtsubo, K., Hashimoto, K., Sato,
Umemoto, T., Yano, M., Satoh, H., Shomura, A., H., and Teranishi, T. 1984. Correlation
and Nakamura, Y. 2002. Mapping of a gene between protein contents of brown rice and
responsible for the difference in amylopectin textural parameters of cooked rice and cooking
structure between japonica-type and indica- quality of rice. Rep. Natl. Food Res. Inst.
type rice varieties. Theor. Appl. Genet. 104:1- 45:118-122.
8. Yang, C. H., and Chang, W. H. 1999. Effects of
USDA. 1982. Rice Inspection Handbook. Fed. protein and lipid binding to starch on the
Grain Inspect. Serv., U.S. Dep. Agric., physicochemical and pasting properties of rice
Washington, DC. flour. Food Sci. Agric. Chem. 1:277-285.
USDA. 1994. Rice Inspection Handbook. Fed. Yasumatsu, K., Moritaka, S., and Wada, S. 1966.
Grain Inspect. Serv., U.S. Dep. Agric., Studies on cereals. V. Stale flavor of stored
Washington, DC. rice. Agric. Biol. Chem. 30:483-486.
USDA. 1995. United States Standards for Rice. Yokoyama, W., Renner-Nantz, J. J., and
U.S. Dep. Agric., Washington, DC. Shoemaker, C. F. 1998. Starch molecular mass
Warth, F. J., and Darabsett, D. B. 1914. Disinte- and size by size-exclusion chromatography in
gration of rice grains by means of alkali. Bull. DMSO-LiBr coupled with multiple angle laser
no. 38. Agric. Res. Inst., Pusa, India. light scattering. Cereal Chem. 75:530-535.
Webb, B. D. 1972. A totally automated proce- Yoshii, Y., Otobe, K., Sugiyama, J., Arisaka, M.,
dure for amylose analysis in whole kernel and Kikuchi, Y. 1993. Evaluation of varietal
milled rice. (Abstr.) Pages 62-63 in: Proc. Rice properties of cooked rice grains by rheological
Technical Working Group, 14th. Agric. Exp. measurements. J. Jpn. Soc. Food Sci. Technol.
Stn., Davis, CA. 40:236-243.
Webb, B. D. 1979. Assessing parboil-canning Yoshikawa, S., Endo, I., and Okabe, M. 1974.
stability of rice in a varietal improvement pro- New approaches in texturometer evaluation of
gram. Pages 283-292 in: Chemical Aspects of sensory properties. Food Sci. Technol., Proc.
Rice Grain Quality. Int. Rice Res. Inst., Los Int. Congr., 4th. 2:144-149.
Baños, Laguna, Philippines. Yoshio, N., Maeda, I., Teranishi, K., Hisamatsu,
Webb, B. D., and Adair, C. R. 1970. Laboratory M., and Yamada, T. 1995. Molecular
parboiling apparatus and methods of evaluat- structures and properties of starches from
ing parboil-canning stability of rice. Cereal different cultivars of rice (Oryza sativa L.
Chem. 47:708-714. japonica) produced in Japan. J. Appl.
Widjaja, R., Craske, J. D., and Wootton, M. Glycosci. 42:365-374.
1996. Comparative studies on volatile compo- Yoshio, N., Lin, Q., Hisamatsu, M., Teranishi,
nents of non-fragrant and fragrant rices. J. Sci. K., Yamada, T., and Maeda, I. 1997. Com-
Food Agric. 70:151-161. parative study on amylopectin molecules of
Williams, P. C. 1975. Application of near infra- Koshihikari and Reiho (Oryza sativa L. japonica).
red reflectance spectroscopy to the analysis of J. Appl. Glycosci. 44:169-173.
cereal grains and oilseeds. Cereal Chem. Young, A. H. 1984. Fractionation of starch.
52:561-576. Pages 249-283 in: Starch Chemistry and
Williams, V. R., Wu, W.-T., Tsai, H. Y., and Technology, 2nd ed. R. L. Whistler, J. N.
Bates, H. G. 1958. Varietal differences in BeMiller, and E. F. Paschall, Eds. Academic
amylose content of rice starch. J. Agric. Food Press, London.
Chem. 6:47-48. Zhou, Z., Robards, K., Helliwell, S., and
Windham, W. R., Lyon, B. G., Champagne, E. Blanchard, C. 2003. Effect of rice storage on
T., Barton, F. E., Webb, B. D., McClung, A. pasting properties of rice flour. Food Res. Int.
M., Moldenhauer, K. A., Linscombe, S., and 36:625-634.
CHAPTER 16
VALUE-ADDED RICE PRODUCTS
Heidi C. Wilkinson
USA Rice Federation
Houston, Texas
Elaine T. Champagne
Overall rice consumption in the United States has been rising steadily for the last
few decades. Per capita consumption has grown from around 15.75 lb per person in
marketing year 1984-1985 (USA Rice Federation, 1996) to 27.09 lb in 1999-2000
(USA Rice Federation, 2000) Use in all areas measured in the Federation’s report,
including for direct food use (consumer and foodservice use as table rice), in
processed foods, and in beer, has generally risen (Table 1). The use of rice in
various value-added rice products is in large part responsible for the increase in
U.S. rice consumption since the mid-1990s.
GRAIN TYPES
One of the oldest food crops in the world, rice exists in a variety of types and
forms, many of which have unique cooking, texture, and aroma characteristics.
Table 2 lists the amounts of rice types shipped to processors for use in processed
foods. The amounts of each type used in various food product categories are listed
in Table 3. The following sections describe some basic categories for the myriad
types of rice.
Brown Rice
Rice from which only the outer hull has been removed is brown rice, the most
nutritious version of rice. Its light brown color is caused by the presence of bran
layers that are rich in minerals and vitamins, especially the B-complex group (see
Chapter 4). Brown rice may be eaten as is or milled into white rice. Cooked brown
rice has a slightly chewy texture and a nutlike flavor (Kuntz, 1997). Because wax in
the outer bran layer reduces the rate of water absorption, brown rice takes longer to
cook than milled white rice (Desikachar et al, 1965). When cooked, the bran layers
often detach and split irregularly (Little and Dawson, 1960).
473
Regular Milled White Rice
Regular milled white rice has the hull and bran layers removed. It is sometimes
called milled rice, raw milled rice, milled white rice, polished rice, or polished
white rice (Knehr, 1998). In the United States, rice is classified into three grain
types: long, medium, and short, depending on length-width ratios (USDA, 1989).
LONG-GRAIN
Long-grain rice has a slender kernel three to five times longer than its width.
Long-grain varieties in the United States are characterized by intermediate appar-
ent-amylose content (21–23%) and intermediate gelatinization temperatures (69–
73°C). Cooked grains are separate, light, and fluffy. Long-grain varieties are mar-
keted primarily as milled rice, parboiled rice, quick-cooking rice, and in packaged
rice mixes. Specific varieties (e.g., Newrex, Rexmont, Dixiebelle, and L-205) are
extra-high amylose cultivars (>24%) with superior resistance to overcooking and
abuse. They are recommended for canning applications, quick-cooking, and frozen
rice products (Webb, 1991).
MEDIUM- AND SHORT-GRAIN

Medium- and short-grain rice types have shorter and wider kernels than long-
grain rice, with length-width ratios of 2.3:1–3.3:1 for medium-grain and 2.2:1 and
less for short-grain. Typical U.S. medium- and short-grain varieties have low
apparent-amylose content (15–20%) and relatively low gelatinization temperatures
(63–68C). Cooked grains of these types are more moist and tender and have a
greater tendency to cling together than long grains do. In addition to use in dishes
intended for chopsticks, medium- and short-grain varieties are preferred for ready-
to-eat breakfast cereals, baby foods, and brewing. They are also used for products
that are served cold, because the starch tends not to retrograde as readily as it cools
compared to that of typical long-grain varieties (Kelly, 1985).
Sweet or Waxy Rice
Waxy (also referred to as sweet or glutinous) rice types are the staple food in
China, Laos, and Thailand (Luh, 1991a). Waxy varieties are characterized by
opaque endosperms, with the starch component being all amylopectin. They
TABLE 1
Distribution of Milled Ricea to Principal Domestic Outletsb
Marketing Year Direct Food Use Processed Foods Beer Total Domestic
1988-1989 25.050 8.621 11.150 44.821
1990-1991 27.970 12.180 11.000 51.150
1994-1995 31.506 16.134 10.707 58.347
1995-1996 36.282 14.900 11.177 62.359
1996-1997 35.780 14.133 10.820 60.733
1997-1998 37.556 15.565 11.088 64.209
1998-1999 38.104 16.146 10.699 64.949
1999-2000 39.217 16.859 11.181 67.257
a In million hundredweights (mm cwt).
b Source: USA Rice Federation (2000).
Value-Added Products / 475
generally have low gelatinization temperatures (61–63C) and relatively low

amylographic-peak, hot-paste and cool-paste viscosities (Webb, 1991). When
cooked, they tend to lose their shape and are very sticky.
Aromatic Rice
Aromatic rice has a flavor and aroma similar to that of roasted nuts or popcorn
(Juttelstad, 1999). The major compound that gives aromatic varieties their charac-
teristic aromas and flavors, 2-acetyl-1-pyrroline, is present in all rice, but it is at
higher concentrations in aromatic varieties (Buttery et al, 1988). The most popular
U.S.-grown aromatic rices include Della, a basmati cross developed in the late
1970s to early 1980s, and its successors, Dellmont and Dellrose, which all cook dry,
separate, and fluffy; Jasmine85, a version of a Southeast Asian variety bred to grow
in southern U.S. climates, having grains that cook moist and tend to cling together;
and Dellmati, Calmati, and Sierra, U.S.-bred versions of basmati rices from India
and Pakistan, which cook into elongated, slender grains that are dry, separate, and
fluffy. These and other aromatic varieties are further described in Chapter 3.
TABLE 2
Type of Milled Rice Used in Processed Fooda,b
Marketing Medium Long Broken Specialty Rice Short
Year Grain Grain Rice Rice Flour Grain Other Total
1995-1996 8.192 5.328 6.410 3.303 1.174 0.091 0.848 25.346
1996-1997 8.881 3.449 6.072 2.112 1.112 0.461 0.810 22.897
1997-1998 9.932 4.170 6.172 1.838 1.351 0.560 1.034 25.057
1998-1999 8.939 6.738 7.060 2.174 1.422 0.188 0.323 26.844
1999-2000 8.562 8.461 5.532 3.381 1.179 0.605 0.320 28.040
a Rice types shipped to processors for use in processed foods are listed left to right in order of tonnage
importance for 1999-2000. Shipments are in million hundredweights.
TABLE 3
Amounts of Various Rice Types Used in Processed Foodsa,b
Food Long Medium Short Broken Rice Specialty
Product Grain Grain Grain Rice Flour Other Rice Total
Cereal 0.218 2.995 0.179 0.297 0.360 0.004 0.000 4.053
Soup 0.000 0.013 0.000 0.000 0.005 0.006 0.001 0.025
Baby food 0.000 0.003 0.000 0.000 0.389 0.000 0.002 0.394
Beer 5.638 3.275 0.000 0.441 0.000 0.000 0.230 9.584
Rice cakes 0.168 0.020 0.060 0.000 0.000 0.000 0.005 0.253
Crackers/snacks/crispies 0.000 0.601 0.000 0.011 0.218 0.000 0.053 0.883
Package mixes 0.818 0.363 0.000 0.001 0.002 0.000 1.334 2.518
Frozen dinners/sides 0.001 0.051 0.000 0.000 0.001 0.000 0.300 0.353
Candy 0.000 0.071 0.000 0.105 0.000 0.000 0.006 0.182
Pet food 1.010 0.276 0.094 4.069 0.012 0.278 1.163 6.902
Pudding/dessert 0.002 0.038 0.000 0.000 0.000 0.000 0.001 0.041
Other 0.128 0.372 0.238 0.295 0.126 0.001 0.095 1.255
Total 7.983 8.078 0.571 5.219 1.113 0.289 3.190 26.443
a For marketing year 1999-2000 on a milled basis, in million hundredweights.
Arborio Rice
Arborio rice has a large, bold kernel with a characteristic chalky center. Based
on length-width ratio and starch characteristics, it is classified as a medium-grain
rice. Typically used in risotto, this rice develops a creamy texture around a chewy
center and has exceptional ability to absorb flavors.
FORMS OF RICE
Unprocessed Forms
HEAD RICE
Head rice is defined by U.S. millers as whole kernels of milled rice or broken
kernels that meet the standard of measuring at least three-fourths the length of a
whole kernel. After milling, 58–64% of kernels emerge in this form (USDA, 2000).
SECOND HEADS
Broken kernels from the milling process—at least one-half as long as the whole
kernel, but less than three-fourths of the original length—are second heads. These
partial kernels can be used for blending with other rice grains for a variety of prod-
ucts or may be ground for rice flour (Deis, 1997).
BREWERS’ RICE
The smallest broken fragments from the milling process—usually about one-
fourth of a full kernel—are brewers’ rice. These grains are commonly used in
brewing beer and other fermented products, and also in pet foods (Deis, 1997).
Processed Forms
Rice is produced in a number of processed forms for different value-added appli-

cations; some cook quickly, others sustain harsh processing techniques. Some pro-
vide texture or flavor for finished products, like crisped rice in breakfast cereals
(Deis, 1997; Burrington, 2001).
PARBOILED RICE
Rough rice (the kernel within the hull) that has gone through a steam-pressure
process before milling is called parboiled rice (see Chapter 13). This procedure
gelatinizes the starch in the grain and ensures a firmer, more separate grain when it
is milled to white rice and cooked. Parboiled rice satisfies consumers and chefs who
desire extra-fluffy and separate grains (Kreuzer, 2000). This form takes longer to
cook because the gelatinized starch is more resistant to absorbing water. The cooked
kernels are firmer in texture than cooked raw-milled rice and appear wider (see
Chapter 13). Parboiled rice holds up well in retort, canning, and steam-table
applications, maintaining the grain shape (Kester, 1959; Burns, 1972).
QUICK-COOKING RICE
Milled white and brown rice take 15–25 min and 45–60 min, respectively, to
cook (Keneaster, 1974). Quick-cooking processes reduce the cooking time to within
5 min for white rice and 10–15 min for brown rice. Many quick-cooking processes
have been developed and patented; these are discussed in reviews by Roberts
(1972), Roberts et al (1980), Juliano and Sakurai (1985), and Luh (1991b). Usually
the grain is precooked in water, steam, or both, followed by drying so that the grains
retain a porous and open structure without clumping (Juliano and Sakurai, 1985).
Other methods use dry heat treatments and abrading of the kernel, in which the
starch is not gelatinized. The following paragraphs summarize several types of
processes for preparing quick-cooking white, brown, and parboiled rice as described
by Roberts (1972) and Juliano and Sakurai (1985). The details of these processes
are presented by Luh (1991b).
Soak-Boil-Steam-Dry Methods. This type of method is used in producing Minute
Rice. Raw milled rice is soaked to about 30% moisture, cooked in hot water to
about 50–60% moisture with or without steaming, further boiled or steamed to 60–
70% moisture, and dried to 8–14% moisture in a manner that maintains a porous
structure. In a modification, grains can be fissured (cracked) using dry heat before
the cooking and drying.
Gelatinize-Dry-Puff Methods. Rice is soaked, boiled, steamed, or pressure-
cooked, and then dried at low temperatures. The dense, glassy grains produced are
then puffed at high temperature to yield a product with a porous structure for rapid
water uptake.
Gelatinize-Roll or Bump-Dry Methods. Rolling or bumping flattens the gelati-
nized grains, yielding a relatively hard, glassy product after drying.
Dry-Heat Methods. Milled white or brown rice is heated with 57–82°C air for
10–30 min or with 272°C air for 17.5 sec. The grains fissure, absorb water more
readily, and thus cook in less time than untreated rice.
Freeze-Thaw Methods. Rice is precooked, frozen at 0°C, thawed, and dried.
Gun Puffing. Milled rice conditioned to 20–22% moisture is steamed under a
pressure of 3.5–5.6 kg/cm2 for 5–10 min, following which the pressure is released
and the product puffs.
Freeze-Drying. Cooked rice is frozen in a blast freezer and dried by sublimation
under vacuum.
Chemical Treatments. NaCl, NaOH, and disodium phosphate are used in the
cooking process.
Processes have also been developed that subject the rice grain to mechanical
action to modify their physical structure. These technologies do not require the
cooking and drying of traditional quick-cooking processes. One method passes
brown rice through a pair of perforation rollers that create multiple small holes in
the bran, thereby facilitating the absorption of water (Shinzo, 1987). Another
method improves water absorption by making multiple, random cuts with blades
into the outer pericarp of the brown rice (Kinya, 1988). A new, economical tech-
nique to make quick-cooking brown rice changes the cooking time from around 45–
60 min to 15–20 min (Guraya, 2003). The process, developed at the USDA’s
Southern Regional Research Service in New Orleans, LA, “sandblasts” brown rice
with rice flour. The bombardment creates microperforations in the waxy layer of the
bran, allowing the kernel to absorb water more readily and cook more rapidly.
INDIVIDUALLY QUICK-FROZEN RICE

Another recent addition to quick-cooking products is individually quick-frozen
(IQF) rice. IQF rice is fully cooked to about 64% moisture and then quickly frozen.
The process provides a free-flowing ingredient for use in frozen and prepared food
products, as well as in foodservice operations to decrease preparation time and

equipment needs (Pehanich, 2002). The rice can be heated in a microwave oven in
2–4 min for consumption. Unlike some products for quick cooking, IQF kernels do
not crack or fissure.
PUFFED AND CRISPED RICE

Rice kernels can be processed in several ways to create desired textures for cere-
als, candy bars, and energy/sports bars. These include processes for making puffed
and crisped rice.
Oven-Puffed. Oven-puffed rice is prepared from whole kernels of short-grain,
fully milled rice (frequently parboiled). Plump, translucent grains of uniform vari-
ety, size, and moisture are preferred. In one process, the rice is cooked in a rotary
cooker with sugar, malt syrup, and salt (Brockington and Kelly, 1972). The hot
cooked rice is then compressed (bumped) by passage through smooth rolls. The
bumped rice is surface-dried to about 15% moisture and tempered for 12–15 hr at
ambient temperature. The rice may be sprayed with malt flavor and is then passed
through a hot toasting oven, where the kernels expand five to six times in size.
Another described process cooks the rice in a retort for 5 hr under 100–150 kPa
steam pressure, followed by drying to 25–30% moisture content in a rotating louver
drier (Hsieh and Luh, 1991). This is followed by 15 hr of storage to equilibrate the
moisture, a step that reduces stickiness and toughens the kernel for bumping. Ker-
nels are then dried to 18–20% moisture and passed under a radiant heater to plasti-
cize the outer layers to prevent splitting when they are run through rolls. The
bumped grains are tempered 24 hr and then passed through rotating roasting ovens
at 232.2–301.7°C for 30–45 sec. The rice emerges puffed.
Gun-Puffed. Gun-puffed rice is prepared by feeding raw, short-grain, well-milled
rice (approximately 13% moisture) that has been preheated with air into a sealed
chamber or gun (Juliano and Sakurai, 1985) or by preheating said rice in the gun
(Hsieh and Luh, 1991). The rice is then cooked with superheated (dry) steam at
pressures up to 15.1 kg/cm2 and temperatures up to 242°C (Brockington and Kelly,
1972). Following a short cooking time to allow the steam to cook the rice to a
semiplastic state, the gun is suddenly opened, releasing the pressure. The puffed
grain is separated from unpuffed and clumped grains and dried to a maximum of 3%
moisture. Successful puffing depends on the moisture content of the raw rice, the
moisture involved in steaming within the gun, the time and temperature at which the
rice is preheated, and attaining the temperature at which starch exhibits plastic flow
under pressure (Brockington, 1967; Brockington and Kelly, 1972). Steam pressure is
critical. If the pressure is too high, the grains shatter when the gun is opened. Too low
a pressure results in tough texture and lack of crispness (Juliano and Sakurai, 1985).
Extruder-Puffed. Extruder-puffed rice is made by extruding superheated and
pressurized dough through an orifice of a single- or twin-screw extruder into the
atmosphere (Hseih and Luh, 1991). The sudden expansion of water vapor as the
pressure is released results in puffing. This process has several advantages over
other puffing methods, such as higher production throughput, more versatility in
product shape, and easier control of product density (Hseih and Luh, 1991). Details
of the extrusion process and extruder configurations are discussed in Chapter 17 and
by Hseih and Luh (1991) and Mulvaney and Hseih (1988).
Crisped Rice. Crisped rice can be manufactured by oven puffing or high-pressure
extrusion puffing (Hseih and Luh, 1991).
FLOUR
Whole or broken kernels of rice can be ground into flour. Since each variety has
unique viscosity, gelatinization temperatures and other characteristics, applications
vary depending on the type of rice used (Hegenbart, 1995). The milling method
employed and the type of mill used also influence functional properties. Chapter 17
discusses the methods for preparing rice flours, their functional properties, and their
applications, with an emphasis on Asian food products. Rice flour is often the pre-
ferred choice of manufacturers creating products targeted to consumers who require
a hypoallergenic alternative to gluten and wheat-flour products (Deis, 1997).
Typically, long-grain rice flours impart a firm, dry, or crisp texture, leading to
good performance in ready-to-eat breakfast cereals, coatings, and extruded crisp
rice and snacks (Kreuzer, 2000). They also perform well in applications such as hot
breakfast cereals, baby foods, baking mixes, pasta, and malt beverages and as a
dusting agent for frozen food or bakery dough (Parlin, 1997). Medium-grain flours
are used as thickeners in soups, gravies, and baby foods and as binders in sausage
and chopped meats. Medium-grain flours are favored in gluten-free breads and in
traditional Asian-style rice crackers (Deis, 1997).
Waxy rice flour is frequently used in commercial product formulations for its
functional attributes, which include a freeze-thaw stability superior to that of
nonwaxy and waxy starches from other crops (Hanson et al, 1951, 1953; Hanson
and Fletcher, 1961) and resistance against high-temperature processing (e.g.,
sterilization, ultrahigh temperature, or microwave). Waxy flour has lower peak
viscosity compared to some short-grain rice and practically no setback viscosity
(Luh and Liu, 1991). These characteristics make it a good fit for thickening soups,
sauces, gravies, baby foods, puddings, and the like, especially if they will be frozen
and later thawed (Ennen, 2001). Waxy rice flours are sought after in the natural-
foods market for freeze-thaw applications because they contribute to a “clean”
ingredients label (Brandt, 2002). The freeze-thaw stability provided by waxy flour
also applies to rice-wheat flour mixtures at 40–60% rice levels (Luh and Liu, 1991).
Asian foods prepared using waxy rice flour are described in Chapter 17.
Brown rice flours are favored in health and organic foods (Parlin, 1997). Brown
rice flour has a nutty flavor and the fiber, vitamins, and minerals associated with the
bran layers. Flours prepared from raw brown rice have a limited shelf life because
of the lipolytic hydrolysis resulting from lipase comingling with the oil substrate
following grinding. Flour prepared from parboiled rice does not have this instability
problem, but the starch is pregelatinized and the functional properties of the flour
differ from those of flour prepared from raw brown rice. A stable, raw brown rice
flour can be prepared by mixing stabilized rice bran with white rice flour at appro-
priate levels.
Pregelatinized white rice flour can be prepared from parboiled rice by passing
white rice flour through an extruder, or from puffed or roasted rice. Pregelatinized
rice flours have applications as crispness-producing agent in crackers, thickeners in
soups and gravies, carriers for flavors and seasonings, binding agents in meat prod-
ucts, and in rice pasta formulations.
Rice meal differs from flour by having a larger particle-size distribution. Flour
granulation is typically 95–100% through U.S. 50 mesh, 45–65% through U.S. 100
mesh, and 25–40% through U.S. 140 mesh. Meal is 95–100% through 20 mesh, 15–
45% through 40 mesh, and 30% maximum through 60 mesh. Rice meal finds appli-
cations in breadings (where the coarser texture is preferred), extruded rice-based
snacks, and pet foods. Rice meal may also be preferred for applications in which
minimal starch damage is desired.
SYRUP
Rice syrup can be made from white or brown rice by enzymatic or chemical
hydrolysis, using processes similar to those used in making corn syrup. Rice syrup
on the market in the United States is mainly a natural- or health-food item with lim-
ited distribution (Deis, 2001). The demand for it has expanded as the demand for
soy milk products has grown; rice syrup is often used as a sweetener in soy milk to
mask beany flavors (Klahorst, 2000). Syrups range from 26 dextrose equivalent
(DE) to 70 DE and a minimum 78° Brix (Knehr, 1998).
COMPONENTS AND CO-PRODUCTS
Kernel components and co-products of rice milling incorporated into value-

added food products include the bran, bran oil, starch, and protein. Table 4 lists
amounts of bran and oil shipped by millers in marketing year 1999-2000.
BRAN
The germ and bran layers (about 10% of a rice kernel) removed during milling
are rich in protein, lipids, fiber, minerals, and B-complex vitamins (see Chapters 4
and 19). Additionally, rice bran has oryzanol, a phytosterol esterified to ferulic acid,
which has been demonstrated to be responsible for the lowering of serum choles-
terol in humans (Hegsted et al, 1993; Lichtenstein et al, 1994).
Rice bran is subject to rapid deterioration resulting from hydrolytic and oxidative
rancidity and thus must be stabilized for food use. Extrusion processing is generally
employed for stabilization. Stabilized rice bran is used in health foods, cereals,
baked goods, and dietary supplements in which the addition of dietary fiber is
desired. Rice flour mixed with bran has been used as a meat tenderizer; it also helps
maintain good color (Klahorst, 2000). An emulsifier made from rice bran helps
machineability for pasta products (Kobs, 2000). Rice bran can also be used as a fat
replacer (Busetti, 1995). Stabilized rice bran is used in pet foods and horse feeds;
raw rice bran is mixed with other feed components for livestock feeds. The high fat
and vitamin E levels in bran make it appealing for use in feeds for performance
horses.
TABLE 4
Shipments of Rice Bran and Oil by Origina,b
Origin Rice Bran Rice Bran Oil
Arkansas-Missouri 5.726 0.323
Louisiana-Florida 1.044 0.000
Texas-Mississippi 2.478 0.200
Total South 9.248 0.523
California 2.353 0.000
Total United States 11.601 0.523
a For marketing year 1999-2000, in million hundredweights.
RICE BRAN OIL

Extracted from rice bran, rice bran oil is a high-quality cooking oil with an
excellent balance of fatty acids (McCaskill and Zhang, 1999). Rice bran oil has
higher oxidative stability than soybean or cottonseed oil, due in part to its low lino-
lenic acid content (Kao and Luh, 1991; McCaskill and Zhang, 1999). It also has
high thermostability (Yoon et al, 1987). Frying test results showed good perform-
ance with low formation of peroxides, foam, free fatty acids, and polymers (Lynn et
al, 1968; Kim and Yum, 1983). Its high flash point has made it a popular cooking
oil in Asia for centuries (Kreuzer, 2000).
Rice bran oil has several properties (e.g., tendency to form a stable bcrystal
lattice, high palmitic acid content) that contribute to its functioning well in
margarine (McCaskill and Zhang, 1999). It allows margarine to be produced with a
low degree of hydrogenation and thus with low trans-fatty acid levels.
Rice bran oil has potential for being marketed as a nutritional oil. The unsaponi-
fiable fraction of rice oil contains a wide range of antioxidants, tocopherols,
tocotrienol, oryzanol, ferulic acid and other functional food ingredients that have
health benefits (see Chapters 4, 7, 19, and 20). Both animal and human studies have
shown that feeding rice bran oil lowered low-density lipoprotein (LDL, or “bad”)
cholesterol levels as well as corn oil did and better than olive oil (Nicolosi et al,
1991; Lichtenstein et al, 1994). Oryzanol is believed to be a key ingredient in this
cholesterol-lowering ability. Rice bran oil has also been demonstrated to reduce
aortic fatty streaks (Rong et al, 1997).
STARCH
Rice starch is the major component (90–93% of the dry weight) of milled rice.
Rice has one of the smallest starch granules among the cereal grains, ranging in size
from 3 to 5 mm (see Chapter 5). In comparison, starch granules are about 15 mm in
corn, 20 mm in tapioca, and 15–100 mm in potato. Commercial native rice starch is a
very fine, very white powder that provides applications as a cosmetic dusting pow-
der, laundry stiffening agent, paper and photographic paper powder, sugar coating
in confectionery, and excipient for pharmaceutical tablets (Champagne, 1996). In
gelatinized form, rice starch has a bland flavor, is an excellent thickening agent for
sauces and desserts, and provides a superior creamy texture for custard. Because
rice starch granules are about the same size as fat globules, rice starch provides a
texture that is perceived as similar to that of fat. This leads to applications as a fat
replacer. The addition of 4–5% rice starch to a no-oil mayonnaise increases the
product’s opacity to a level more typical of a full-fat product (Bakal, 1994). Waxy
rice starches, with their superior freeze-thaw stability, serve well as fat replacers in
frozen desserts such as ice cream.
Rice starch is considered to be nonallergenic because of the hypoallergenicity of
its associated residual protein. The starch can be used to replace gelatin for religious
or vegetarian applications.
PROTEIN
Uncooked rice, brown or white, contains about 7–8% protein. Rice is unique
among cereals in that its storage protein is primarily glutelin, while in most other
cereals it is prolamin (Hamaker, 1994). This causes lysine, the first limiting essen-
tial amino acid in cereals, to be higher in rice than in other cereals. The protein effi-
ciency ratio (PER) for milled rice is about 2.0 (de Lumen and Chow, 1991); pro-
teins with PER of 2.0 or higher are considered to be of good nutritional quality.
Rice proteins are considered to be hypoallergenic.
The processing of protein concentrates and isolates from the bran and from
milled white rice is discussed in Chapter 6; enzymatic, chemical (alkali), and physi-
cal methods are used. Production of rice syrups and maltodextrins also yields pro-
tein concentrates. The functional properties of isolated protein concentrates or iso-
lates are dependent upon the processing method. Rice proteins can act as emulsifiers
or viscosifiers, especially in fat-reduction systems (Kuntz, 1997). They can provide
opacity and can enhance structural and water-binding effects of rice flours (Kuntz,
1997). Rice protein can be used in meat analogs to help hold the product together
(Decker, 2002).
RICE IN PROCESSED FOODS
The food industry in the United States was awash in new product introductions
between 1997 and 2002 (Fig. 1). Data show that rice has played a growing role in a
variety of food categories (Table 5). The major categories of foods that use rice
include baby food, bakery goods, breakfast cereals, confections, beverages and
dairy items, desserts, side dishes, meal centers (frozen, shelf-stable, chilled), meats
and meat substitutes, sauces and spreads, snacks, soups, and pet foods (Kuntz, 1997;
Pehanich, 2002). Many of these categories grew rapidly since the 1980s, leading to
increased rice usage (Tables 6–8). For example, rice usage in meal centers and
snacks increased 10-fold and tripled, respectively (USA Rice Federation, 2000 and
other years). Energy bars, a new entry in the snack category in the 1990s, account
for about 80–100 new product launches a year (Mintel International Group Ltd.,
2003). Overall, rice use by food processors in the United States tripled during the
1990s (USA Rice Federation, 2000 and other years). More growth is predicted for
the remainder of this decade (Backas, 2002).
Increased usage of rice by food processors has been in response to various con-
sumer demands, including more healthful foods, more convenient products, and
growing interest in a wide variety of ethnic foods (Frank, 2002; Pehanich, 2002).
The popularity of Mexican, Spanish, Cajun, and different Asian cuisines among
Americans has helped boost rice consumption (Anonymous, 1998). Food trends
Fig. 1. Total new products containing rice, by year. (Data from Mintel International Group Ltd, 2003)
TABLE 7
High and Low Usea of Rice by Categoryb
Highest Level Lowest Level
Category Use Year Use Year
Cereal 5.934 1994-1995 2.503 1982-1983
Soup 0.241 1984-1985 0.025 1999-2000
Baby food 1.106 1997-1998 0.133 1980-1981
Packaged mixes 3.274 1994-1995 0.221 1982-1983
Frozen entrees 0.685 2000-2001 0.061 1986-1987
Desserts 0.068 2000-2001 0.034 1998-1999
Snacks 1.578 2000-2001 0.323 1995-1996
Rice cakes 0.707 1988-1989 0.197 2000-2001
Candy 0.298 1995-1996 0.105 1990-1991
Pet food 6.902 1999-2000 0.431 1986-1987
a In million hundredweights.
b Source: data from USA Rice Federation (2000 and other years).
TABLE 8
Number of New Products with Rice, Activity by Categorya
Highest Level Lowest Level
Category No. of New Products Year No. of New Products Year
Cereal 76 2001 7 1997
Soup 33 2002 5 1997
Baby food 8 2001 2 1997 and 1998
Sides 81 2002 25 1997
Meal centers 103 2002 20 1997
Desserts 24 2002 2 1997
Snacks 130 2002 12 1997
Candy 30 2002 3 1997
Pet food 62 2002 12 1998
a Source: data from Mintel International Group Ltd. (2003).
(i.e., home meal replacement and less preparation done on-site in restaurants) have
created opportunities for food processors to provide both home and foodservice
customers with finished meals instead of the means to make them, with rice playing
an important role (Pehanich, 2002).
Breakfast Cereals
Rice offers breakfast-cereal makers a neutral flavor, hypoallergenicity, and a

good nutritional profile. Ready-to-eat breakfast cereals include puffed rice (by oven,
gun, or extruder), rice flakes, shredded-rice cereal, and multigrain cereal. The proc-
esses for puffed rice are described above under “Processed Forms.”
Rice Flakes. Rice flakes are processed in a manner similar to that of oven-
puffed rice. The rice is cooked, coated with nutritional adjuncts (e.g., wheat
gluten, wheat-germ meal, dried skim milk, or debittered brewer’s yeast), partially
dried, tempered, run through flaking rolls, and then oven toasted (Hsieh and Luh,
1991). The flakes emerge from the oven toasted, blistered, and dehydrated. In
comparison with oven-puffed rice, rice flakes tend to soften more quickly in milk
(Hsieh and Luh, 1991).
Shredded Rice Cereal. Shredded rice is prepared by cooking washed, whole-

kernel or broken rice in a rotary cooker with sugar, salt, and malt syrup under 100–
150 kPa steam pressure until the rice is uniformly cooked (1–2 hr) (Hseih and Luh,
1991). The cooked rice, with moisture content of about 40%, is partially dried to
about 25–30% moisture and then allowed to temper. The tempered rice is passed
through a pair of shredding rolls, one of which has a series of shallow square or
rectangular corrugations running around the periphery; the other roll is smooth.
Individual rice particles are drawn and compressed between the rolls, resulting in a
latticework ribbon. The resultant dough pieces are passed through an oven and
toasted.
Hsieh and Luh (1991) describe the process for making Rice Chex and Crispix,
which were manufactured at that time by Ralston Purina. Two pairs of shredding
rolls are used for these products. The dough sheet from the second pair of rolls is
laid on top of the first sheet on the same moving belt. The layered sheets can be cut
by cutting rolls to fuse the sheets at regular intervals, forming a continuous biscuit
matrix. The wet biscuits are placed on another belt and passed through a gas-fired
oven for toasting.
Rice Flour. Rice flour is often used as an ingredient in cereals made from other
grains. In addition to providing a more tender bite, rice flour in flaked cereal makes
the flakes expand better. Waxy rice flour improves bowl life, because the higher
amylopectin level produces a more hydrophobic gel (Pszczola, 2003). Extruded
cereals formulated with other grains often use rice flour to improve crispness and
expansion as well as to reduce breakage (Knehr, 1998). Pregelatinized rice flour can
also be a functional aid, holding extra water in the product for more expansion and
absorbing water faster in the extrusion process, thus decreasing dwell time (Knehr,
1998).
Breakfast cereals is a category that has been hit hard by convenience foods. This
is ironic, considering that cold cereal and milk were originally touted as quick and
easy and, in many ways, created “convenience foods” (Roberts, 2002). The compa-
nies involved have addressed this with various breakfast bars designed to be eaten
on the move, many of which use puffed or crisped rice (Mintel International Group
Ltd., 2003). One example is a breakfast bar introduced in 2003 that contains milk
infused into the grains composing the bar (Pszczola, 2003). These breakfast bars
distort the breakfast cereal category, since they are usually counted as snacks by
most tracking systems (Roberts, 2002).
Breakfast cereals is an area in which organic products and those with health and
functional claims have been making big strides in new-product introductions in
recent years (Mintel International Group Ltd., 2003).
Snacks
Snack Bars. Snack bars include a variety of granola, breakfast, and energy bars,
and many of these use rice. Energy bars have been a booming business in the
United States since the early 1990s, with about $500–700 million in sales in 2000
(Kreuzer, 2001). They target active consumers and often use health claims with a
focus on enrichment (Mintel International Group Ltd., 2003). Some are designed
with ingredients specifically chosen for health purposes, such as to reduce choles-
terol (Brandt, 2000). Organic and natural offerings are common in all snack bars
(Mintel International Group Ltd., 2003).
Eating occasions for these snack bars have expanded. They are consumed as
portable alternatives to breakfast cereal, as an alternative to candy bars, and even as
an opportunity to get vegetable servings, fiber, or other diet essentials. Special
products designed for children, women, vegetarians, and other groups have been
launched in recent years (Mintel International Group Ltd., 2003). Some even target
particular sports, with different types of bars for golfers, bicyclists, and runners.
Market growth for snack bars is expected to continue because of their convenience;
developments to further the “on-the-go” aspects are likely (Brandt, 2000).
Crisped rice is used in granola bars, nutrition or energy bars, and crisped-rice
treats (Kuntz, 1997). Crisped brown rice is often seen in varieties that promote
health benefits (LaBell, 1996). High-pressure extrusion processing is generally used
for the manufacture of crisped rice for snack bars.
Rice Cake. A U.S.-manufactured rice cake is a disk-shaped, puffed product
weighing 9–10 g that is low in calories (35–40 kcal per cake) (Hseih and Luh,
1991). The procedure for making rice cakes, detailed by Hseih and Luh (1991),
consists of adjusting the moisture content of long- or medium-grain brown rice to
14–18%, tumbling the grains for several hours to temper, and then putting them
into the ring mold of a rice cake machine preheated to 200°C or higher. The rice
is then pressed between upper and lower movable platens. At the end of a
specified heating time, the upper platen is lifted and stopped at the upper edge of
the ring. As a result of the sudden release of water vapor, the kernels puff and
fuse together to form the rice cake. An extrusion process can also be used to
make a similar product.
Although rice cakes have declined in popularity in the United States in recent
years, their manufacture is still a sizeable market for rice, especially brown rice, as
these products tend to appeal to health-conscious consumers (Hegenbart, 1995).
Their popularity, however, rose again with the introduction of several new products
in 2000, and the market for rice cakes, in particular mini-rice cakes, is expected to
grow (Mintel International Group Ltd., 2003). Flavored rice cakes take advantage of
rice’s flavor-carrying abilities (LaBell, 1996).
Crackers and Chips. Positioning of crackers as a light meal, particularly as part
of meal kits, contributed to increased production in the late 1990s. New versions of
meal kits based on crackers now seek the adult consumer, not just the child’s
lunchbox. Ethnic offerings are also helping expand this snack category (Mintel
International Group Ltd., 2003).
Rice flour is often used as an ingredient in crackers and chips for many of the
same reasons that rice is used in multigrain breakfast cereals: to control moisture,
bite, texture, and flow in extruders (Huber, 2001). Long-grain rice flour increases
crispness; pregelatinized or partially gelatinized flour improves water absorption
and improves dough sheeting (Kreuzer, 2000). Chips made with rice flour absorb
less fat during frying than those made from potato (Pszczola, 2001). Baked or fried
chips with 5–10% waxy rice flour experience less breakage (Knehr, 1998).
Side Dishes and Packaged Mixes
Flavored rice mixes usually combine rice with flavored broths, seasonings,
dehydrated vegetables, dehydrated cheese sauce, or other ingredients (Knehr, 1998).
Traditionally, these are dried, dehydrated mixes, but IQF versions have been grow-
ing, both for food processing and foodservice use (Pehanich, 2002).
Consumption of rice for packaged mixes roughly doubled during the 1990s,
according to the U.S. Rice Distribution Reports from the periods 1995-1996 and 1999-
2000 (USA Rice Federation, 1996, 2000). This can be attributed, in part, to an increase
in consumer interest in ethnic foods (Mintel International Group Ltd., 2003). Mintel
International Group Ltd. (2003) notes that about one in five rice side-dish
launches are ethnic offerings. Other key reasons for this growth in the 1990s
include the rising popularity of organic foods, an increase in quick-cook varieties,
and an ability to recreate in the processing plant food products generally
considered difficult to make, such as risotto, giving consumers the opportunity to
make these at home through a mix that simplifies preparation (Juttelstad, 1999;
Decker, 2002). Mintel International Group Ltd. (2003) forecasts that this category
will continue to grow.
Rice for a dry mix requires a low moisture level, around 6.5%, down from nor-
mal levels of around 11.0–12.5% (Knehr, 1998). For dry products, either regular or
parboiled rice is used, depending on the desired texture of the final product. For a
product in which distinct grains are important, such as a pilaf, or when a sauce will
add moisture, parboiled often performs better (Knehr, 1998).
Entrées
Rice usage in frozen entrees for U.S. consumers multiplied 10-fold from 1980 to
2000 (USA Rice Federation, 1996, 2000). Interest in ethnic foods boosted this
growth. Frozen ethnic entrées were worth about $2.2 billion in 2000 (Brandt, 2001).
Bowl meals—a frozen, entire entrée in a microwaveable bowl, usually protein
and vegetables over rice or pasta—have helped contribute to this growth. An overall
trend for one-dish meals, plus the ease of preparation for these bowl meals, helped
drive this category, which started at Asian-style “bowl” restaurants. Prepared frozen
versions quickly followed (Decker, 2002).
Rice used in frozen entrées is usually cooked in excess water in a continuous
process (Hegenbart, 1995). Typical processing steps include soaking, cooking,
cooling, draining, blending with flavorings and other ingredients, filling, sealing,
and freezing. The actual cook time is often shorter than that normally stated for the
type of rice used, to compensate for continued softening of the rice during subse-
quent processing steps and reconstitution by the consumer.
A second method, as described by Tressler et al (1968) for commercial processing
of frozen cooked rice, involves cooking rice in a closed vessel. Rice that has been
soaked in citric acid to reach a pH of 4.0–5.5 is placed on a screen above boiling water
in a pressure cooker and steamed for 12–15 min under a pressure of 2.09 kg/cm2. The
steamed rice is then placed in excess hot (93.3–98.9°C) water flowing through a
perforated vessel to allow the kernels to imbibe water. The hot water is drained, and
the rice is rinsed with cold water that has been pH-adjusted to 4.0–5.5. The cooked
rice is then conveyed by belt through an air-blast cooler, packaged with other
ingredients, and frozen in air-blast freezers. Alternatively, the rice may be frozen as
IQF products before being packaged in a fluidized bed freezer (Luh, 1991c).
Soups
The U.S. rice industry has seen the soup category decline, from the highest use
of rice in the mid-1980s to an all-time low around 2000 (USA Rice Federation,
1996, 2000). This is due to an overall decline in the category (Mintel International
Group Ltd., 2003).
Opportunities for rice usage in soups are in niche markets, such as organic, and
in a variety of ethnic offerings. Soup packaging designed for on-the-go usage has
also come into vogue as a way to increase sales. Rice has an advantage in that it is
already the right size to fit through a hole designed to drink from—unlike many
noodles. The bowl trend that hit entrées in the late 1990s has also extended to soups,
presenting another opportunity for rice.
Varieties that demonstrate canning stability have low solids loss and do not split
and fray upon processing. Quality characteristics leading to canning stability
include high apparent amylose content, low amylographic breakdown, high setback
viscosity, and strong gel consistency. Superior-processing varieties for canning
include Newrex, Rexmont, Dixiebelle, L-205, and Bolivar (Chapter 3).
Raw or parboiled rice may be processed for soups. The rice is precooked or
blanched in boiling water to near-completion, drained, rinsed with cold water, blended
with other ingredients, and then transferred to cans that are sealed and retorted (Jones
et al, 1946). During the retort process, the rice takes up additional water.
Desserts
Desserts with rice grew in the late 1990s, with a variety of instant and ready-to-eat
products introduced. Desserts, which were not measured in the U.S. Rice Distribution
Reports as recently as 1997, grew from 0.002 million hundredweights (mm cwt) in
1997-1998 to 0.041 mm cwt in 1999-2000 (USA Rice Federation, 2000).
Chilled desserts are a recent area of growth for rice. As discussed earlier, waxy
rice starch serves well as a fat replacer in ice cream. The addition of 1.5% waxy
rice starch improves the creamy mouthfeel and overall texture, allowing the product
to mimic premium (higher-fat) ice cream. The excellent freeze-thaw stability of
waxy rice starch helps reduce iciness during storage (Bakal, 1994).
Crisped rice provides texture to many chocolate dessert products. One trend here,
in contradiction to the overall trend of indulgence traditionally associated with
chocolate, is the growth of more-healthful, lower-fat products; adding crisped rice is
a common way to accomplish this (Mintel International Group Ltd., 2003).
Baby Food
The baby food category does not traditionally see many new product introduc-
tions. However, trends like organic foods, extension of the time a baby uses baby
foods into the toddler years, and expansion of the convenience of baby foods helped
increase product launches in the late 1990s (Mintel International Group Ltd., 2003).
In addition to products that target older babies, products that target the taste buds of
parents are also a growing trend (Mintel International Group Ltd., 2003).
Babies have unique nutritional requirements, and rice cereal is often the first
solid food introduced into the baby’s diet. This is because it is easily digested,
hypoallergenic, low in fat, and has a bland flavor. Rice flour is often used to thicken
baby food products. It blends well with other ingredients such as vegetables and
proteins and does not add unfamiliar ingredient names to the label. As more organic
foods were introduced into the marketplace in 1997–2002, brown rice started
appearing in baby foods.
Rice baby cereals are marketed as dry, precooked flakes for rehydration and as
ready-to-eat cereals. Precooked flakes can be manufactured by drying a cereal
slurry on a double-drum atmospheric dryer followed by flaking (Johnson et al,
1935; Abeling, 1959; Hogan 1967, 1977) or by extrusion cooking (Mercier et al,
1989). Ingredients used in the formulation of baby cereals include rice flour pre-
pared from medium and short head and second-head grains, sugar, dibasic calcium
phosphate, iodized salt, sodium iron pyrophosphate, glycerol monostearate (emulsi-
fier), rice oil, thiamin hydrochloride, riboflavin, and niacin or niacinamide (Kelly,
1985; Hsieh and Luh, 1991). Hsieh and Luh (1991) and Juliano and Sakurai (1985)
discuss processing factors influencing the textural properties of reconstituted rice
cereals prepared from precooked flakes manufactured by the drum-drying and
extrusion-cooking processes, respectively.
In ready-to-eat baby cereals, the choice of rice variety is important. Medium-
and short-grain varieties with lower amylose contents are used to avoid retrograda-
tion during storage, which results in formation of a very rigid gel and separation of
water (Williams et al, 1958).
Beer
Beer use was relatively flat in the 1990s but increased from just under 8 mm cwt
in 1980-1981 to around 11 mm cwt in 1999-2000 (USA Rice Federation, 1996,
2000) (Table 1). Chapter 18 discusses the use of rice in brewing in detail.
Pet Foods
The largest category for rice use is pet food, which used 6.902 mm cwt in 1999–
2000, up 64% from the 4.425 mm cwt in 1994-1995 (USA Rice Federation, 1996,
2000). The pet food industry in the United States is booming, with 2001 sales at
$11.8 billion, up almost 6% from the year before and projected to keep growing
(Gurkin, 2002). Rice usage in pet foods is expected to follow suit.
Rice has found a place in pet foods because it is highly digestible and hypoallergenic.
It has higher digestibility than maize. Substitution of brewer’s rice for maize in dry adult
dog food showed a linear decrease of mean daily dry matter (DM) and wet stool output
with increasing brewer’s rice levels (Belay et al, 1997). As a result, apparent DM, fat,
carbohydrate, and energy digestibility also increased linearly. Intestinal gas and volatile
fatty acid production are lower in rice-containing diets (Grubler et al, 1989). Rice-
containing pet foods with selected protein sources help in long-term management of
pruritus caused by adverse food reactions (Leistra et al, 2001). Some dogs are susceptible
to gluten-induced enteropathy; rice diets can be fed successfully (Batt and Hall, 1992).
Broken or brewers’ rice is most commonly used in pet food because of price
considerations, but flour, bran, and whole long- or medium-grain rices are also
common ingredients (Kuntz, 1997). Pet foods are formulated with anywhere from
10 to 50% grain products by weight. Extrusion processing is commonly used.
UNUSUAL APPLICATIONS
Meat and Analogs

Changing the profile of meat products to be more healthful is a growing trend
among processors (Mintel International Group Ltd., 2003). Rice incorporated into a
product can help reduce fat content, for instance in the making of “light” sausages.
Rice batters have been developed that, when applied to chicken, absorb 60% less oil
than a traditional wheat-based batter (Shih and Daigle, 2002). Lipid oxidation in
ground beef can be inhibited by incorporation of defatted brown rice flour or defat-
ted rice bran (F. Shih and K. Daigle, unpublished results).
Rice starch can be used to bind water in poultry products during vacuum tum-
bling or injection (Pszczola, 2003). Meat analogs have risen in popularity in recent
years, with nearly equal consumption by both vegetarians and nonvegetarians. Rice
and rice protein are often used in meat analogs; short-grain rice is usually used to
help hold the other ingredients together.
Milk and Dairy Alternatives
Dairy alternatives for people with allergies frequently use rice for its hypoaller-
genic properties. Rice is used in a variety of lactose-free “dairy” products. Cheese-
alternative products made with rice include forms of cheddar, sour cream, cream
cheese, mozzarella, and other typical cheese flavors. Rice beverages offer an alter-
native to cow and soy products and are more frequently seen in mainstream U.S.
supermarkets than they were in the early 1990s.
Rice and rice syrup often help mask “beany” flavors in soy milk (Burrington,
2002). Hydrolyzed rice flour helps stabilize soy beverages and keeps protein insolu-
bles suspended in solution (Klahorst, 2000).
Electrolyte-replacement drinks use rice. The long-chain carbohydrates and pro-
tein in rice speed the absorption of fluid and electrolyte salts into the body’s cells
faster and more efficiently than traditional glucose solutions (Greenough, 1998).
CONCLUSION
Thanks to a wide variety of innovative product ideas, both on the ingredient side
and on the food-product side, rice saw tremendous growth in the 1980s and 1990s
and is poised to continue to rise in the years to come. The use of rice as a food
ingredient, in addition to its traditional use as a table food, has helped to increase
rice consumption in the United States to an all-time high. Continued demand for
rice’s functional benefits, health attributes, and versatility seem ensured by demand
for food products that are easy to use, satisfy the desire for ethnic flavors, are port-
able, and are unlikely to produce allergic reactions. Its ability to improve product
texture and to offer no surprises in ingredient labeling are also reasons that rice will
continue to be in high demand in the food-processing industry.
REFERENCES
Abeling, E. J. 1959. Manufacture of special dietary Batt, R. M., and Hall, E. J. 1992. Gluten-induced
foods. Pages 462-474 in: The Chemistry and enteropathy in the dog. Wien. Tierarztl.
Technology of Cereals as Food and Feed. S. A. Monat. 79:242-247.
Matz, Ed. AVI Pub. Co., Inc. Westport, CT. Belay, T., Shields, R. G., Jr., Wiernusz, C. J.,
Anonymous. 1998. Rice is right. Prep. Foods Kigin, P. D., and Brayman, C. A. 1997. Vet.
167:81-84. Clin. Nutr. 4:122-129.
Backas, N. 2002. The grand finale—Finishing it Brandt, L. A. 2001. Entrees take the orient
off. Food Prod. Design 12:99-108. express. Prep. Foods 170:65-68.
Bakal, A. 1994. The lowdown on formulating Brandt, L. A. 2002. Specifying starches. Prep.
lowfat. Prep. Foods 163:75-78. Foods 171:43-46.
Brandt, L. A. 2000. Wonder bars. Prep. Foods Science and Technology. W. E. Marshall and
169:45. J. I. Wadsworth, Eds. Marcel Dekker, Inc.,
Brockington, S. F. 1967. Puffed rice products. New York.
Pages 56-58 in: Proc. Natl. Rice Util. Conf., Hanson, H. L. and Fletcher, L. R. 1961. Salad
1966. ARS 72-53. U.S. Dep. Agric., Agric. dressings stable to frozen storage. Food Tech-
Res. Serv., New Orleans, LA. nol. 15:256.
Brockington, S. F., and Kelly, V. J. 1972. Rice Hanson, H. L., Campbell, A., and Lineweaver, H.
breakfast cereals and infant foods. Pages 400- 1951. Preparation of stable frozen sauces and
418 in: Rice: Chemistry and Technology, 1st gravies. Food. Technol. 5:432.
ed. D. F. Houston, Ed. Am. Assoc. Cereal Hanson, H. L., Nishita, K. D., and Lineweaver,
Chem., St. Paul, MN. H. 1953. Preparation of stable frozen pud-
Burns, E. E. 1972. Canned rice foods. Pages 419- dings. Food Technol. 7:462.
427 in: Rice: Chemistry and Technology, 1st Hegenbart S. 1995. Truth of grain: Formulating
ed. D. F. Houston, Ed. Am. Assoc. Cereal with rice. Food Prod. Design 5(Suppl. 1):1-12.
Chem., St. Paul, MN. Hegsted, M., Windhauser, M. M., Morris, S. K.,
Burrington, K. J. 2001. Keeping the crunch in and Lester, S. B. 1993. Stabilized rice bran
breakfast cereals. Food Prod. Design 11:63-83. and oat bran lower cholesterol in humans.
Burrington, K. J. 2002. More than just milk. Nutr. Res. 13:387-398.
Food Prod. Design 12:37-64. Hogan, J. T. 1967. Processed rice products. Rice
Busetti, M. 1995. Building better reduced-fat J. 70:25, 28-31.
baked goods and snacks. Prep. Foods 164:77- Hogan, J. T. 1977. Rice and rice products. In:
81. Elements of Food Technology. N. W.
Buttery, R. G., Turnbaugh, J. G., and Ling., L. C. Desrosier, Ed. AVI Pub. Co., Inc. Westport,
1988. Contribution of volatiles to rice aroma. CT.
J. Agric. Food Chem. 36:1006-1009. Hsieh, F. and Luh, B. S. 1991. Rice snack foods.
Champagne, E. T. 1996. Rice starch composition Pages 225-250 in: Rice Utilization, Vol. II. B.
and characteristics. Cereal Foods World S. Luh, Ed. Van Nostrand Reinhold, New
41:833-838. York.
Decker, K. J. 2002. Entrees for a small planet. Huber, G. R. 2001. Developments and trends in
Food Prod. Design 12:36-66. extruded snacks. Food Prod. Design 11:123-
Deis, R. C. 2001. Sweeteners for health foods. 143.
Food Prod. Design 10:48-68. Jones, J. W., Zeleny, L., and Taylor, J. W. 1946.
Deis, R. C. 1997. Functional ingredients from Effect of parboiling and related treatments on
rice. Food Prod. Design 6:45-56. the milling nutritional and cooking quality of
de Lumen, B. O., and Chow, H. 1991. Nutritional rice. Circ. 752. U.S. Dep. Agric., Washington,
quality of rice endosperm. Pages 363-395 in: DC.
Rice Utilization, Vol. II. B. S. Luh, Ed. Van Johnson, L. D., True, N. F., and Engel, H. H.
Nostrand Reinhold, New York. 1935. Food product and method. U.S. patent
Desikachar, H. S. R., Raghavendra Rao, S. N., 1,990,329.
and Ananthachar, T. K. 1965. Effect of degree Juliano, B. O., and Sakurai, J. 1985. Miscellane-
of milling on water absorption of rice during ous rice products. Pages 569-618 in: Rice:
cooking. J. Food Sci. Technol. 2:110-112. Chemistry and Technology, 2nd ed. B. O.
Ennen, S. 2001. New organic tools in starches Juliano, Ed. Am. Assoc. Cereal Chem., St.
and gums. Food Process. 62:33. Paul, MN.
Frank, P. 2002. World on a plate: Refrigerated Juttelstad A. 1999. Pilaf and risotto—Grain-
and frozen foods. 13:38-41. based delights. Food Prod. Design 8:77-87.
Greenough, W. B., III. 1998. Oral rehydration Kao, C., and Luh, B. S. 1991. Rice oil. Pages
therapy: Something new; something old. Infec. 295-312 in: Rice Utilization, Vol. II. B. S.
Dis. Clin. Pract. 7:97-100. Luh, Ed. Van Nostrand Reinhold, New
Grubler, B., Ingwersen, M., and Meyer, H. 1989. York.
In vitro gas generation in colonic and ileal Kelly, V. J. 1985. Rice in infant foods. Pages
digesta depending on feed. Adv. Anim. 525-538 in: Rice: Chemistry and Technology,
Physiol. Anim. Nutr. 19:101-108. 2nd ed. B. O. Juliano, Ed. Am. Assoc. Cereal
Guraya, H. S. 2003. Process for increasing the Chem., St. Paul, MN.
rate of hydration of food crop seeds. U.S. pat- Keneaster, K. K. 1974. Quick-cooking rice proc-
ent 6,586,036. esses. Activities Rep. Res. Dev. Associates
Gurkin, A. 2002. The petfood report. Petfood Military Food Packaging Systems, Inc. 26:38-
Ind. 44:24-26. 50.
Hamaker, B. R. 1994. The influence of rice pro- Kester, E. B. 1959. Rice processing. Pages 427-
tein on rice quality. Pages 177-194 in: Rice 461 in: The Chemistry and Technology of
Cereals as Food and Feed. S. A. Matz, Ed. Mintel International Group Ltd. 2003. Global
AVI Pub. Co., Inc., Westport, CT. new products database. Category review.
Kim, G. S, and Yum, C. A. 1983. Studies on the Online: www.gnpd.com. (See categories
properties and frying performance of domestic Snack bars, snack mixes and energy bars; Cold
rice bran oil. Korean J. Food Sci. Technol. cereal; Rice; Soup.)
15:77-89. Mulvaney, S. J., and Hseih, F.-H. 1988. Process
Kinya, N. 1988. Processed unpolished rice. Japan control for extrusion processing. Cereal Foods
patent 63-173548. World 33:971-977.
Klahorst, S. J. 2000. Specialty ingredients from Nicolosi, R. J., Ausman, L. M., and Hegsted, D.
grains. Food Prod. Design 10:85-102. M. 1991. Rice bran oil lowers serum total and
Knehr E. 1998. Reaching for rice. Food Prod. low density lipoprotein cholesterol and apo B
Design 8(Suppl. 1):1-8. levels in non-human primates. Atherosclerosis
Kobs, L. 2000. Frozen pasta and rice dishes. 88:133-142.
Food Prod. Design 10:124-143. Parlin S. 1997. Rice flour makes its mark. Food
Kreuzer, H. 2000. Dividends from rice. Food Process. 58(10):60.
Prod. Design 10:1-8. Pehanich M. 2002. Ready, set… on-the-go rice.
Kreuzer H. 2001. High-octane bars and bever- Food Prod. Design 12(Suppl. 1):1-8.
ages. Food Prod. Design 11:33-58. Pszczola, D. E. 2003. New ingredient
Kuntz, L. A. 1997. Rhapsody on rice. Food Prod. developments are going with the grain. Food
Design 7(Suppl. 1):1-8. Technol. 57(2):46-61.
LaBell, F. 1996. New horizons for rice. Prep. Pszczola, D. E. 2001. Rice: Not just for throwing.
Foods 165:49. Food Technol. 55(2):53-54, 56-59.
Leistra, M. H. G., Markwell, P. J., and Willemse, Roberts, R. L. 1972. Quick-cooking rice. Pages
T. 2001. Evaluation of selected protein source 381-399 in: Rice: Chemistry and Technology,
diets for management of dogs with adverse 1st ed. D. F. Houston, Ed. Am. Assoc. Cereal
reactions to foods. J. Am. Vet. Med. Assoc. Chem., St. Paul, MN.
219:1411-1414. Roberts, R. L., Carlson, R. A., and Farkas, D. F.
Lichtenstein, A. H., Ausman, L. M., Carrasco, 1980. Preparation of a quick-cooking brown
W., Gaultieri, L. J., Jenner, J. L., Ordovas, J. rice product using a centrifugal fluidized bed
M., Nicolosi, R. J., Goldin, B. R., and drier. J. Food Sci. 45:1080-1081.
Schaefer, E. J. 1994. Rice bran oil consump- Roberts, W. A., Jr. 2002. A vulnerable position.
tion and plasma lipid levels in moderately Prep. Foods 171:13.
hypercholesterolemic humans. Arterioscler. Rong, N., Ausman, L. M., and Nicolosi, R. J.
Throm. 14:549-556. 1997. Oryzanol decreases cholesterol absorp-
Little, R. R., and Dawson, E. H. 1960. Histology tion and aortic fatty streaks in hamsters. Lipids
and histochemistry of raw and cooked rice 32:303-309.
kernels. Food Res. 25:611-622. Shih, F. and Daigle, K. 2002. Preparation and
Luh, B. S., and Liu, Y. K. 1991. Rice flours in characterization of low-oil uptake rice cake
baking. Pages 9-33 in: Rice Utilization, Vol. donuts. Cereal Chem. 70:745-748.
II. B. S. Luh, Ed. Van Nostrand Reinhold, Shinzo, F. 1987. Method of multiply perforating
New York. bran layer of brown rice. Japan patent-62-
Luh, B. S. 1991a. Introduction. Pages 1-7 in: 111652.
Rice Utilization, Vol. II. B. S. Luh, Ed. Van Tressler, D. K., Van Arsdel, W. B., and Copley,
Nostrand Reinhold, New York. M. J. 1968. The Freezing Preservations of
Luh, B. S. 1991b. Quick-cooking rice. Pages Foods, Vol. 4, 4th ed. AVI, Westport, CT.
121-146 in: Rice Utilization, Vol. II. B. S. USA Rice Federation. 2000. U.S. rice dis-
Luh, Ed. Van Nostrand Reinhold, New tribution patterns, 1999-2000 report. The
York. Federation, Houston, TX. (Also 1996 and
Luh, B. S. 1991c. Canning, freezing, and freeze- other rice distribution pattern reports through
drying. Pages 147-176 in: Rice Utilization, 2001.)
Vol. II. B. S. Luh, Ed. Van Nostrand Reinhold, USDA. 1989. United States Standards for Rice.
New York. U.S. Dep. Agric., Washington, DC.
Lynn, L., Steen, G. J., and Anderson, R. M. 1968. USDA. 2000. Rice Situation and Outlook Year-
New rice oil. Food Technol. 22:1250-1252. book. RCS-2000. Commod. Econ. Div., U.S.
McCaskill, D. R., and Zhang, F. 1999. Use of Dep. Agric., Agric. Res. Serv., Washington,
rice bran oil in foods. Food Technol. 53(2):50- DC.
52. Webb, B. D. 1991. Rice quality and grades.
Mercier, C., Linko, P., and Harper, J. M. 1989. Pages 89-119 in: Rice Utilization, Vol. II. B.
Extrusion Cooking. Am. Assoc Cereal Chem., S. Luh, Ed. Van Nostrand Reinhold, New
St. Paul, MN. York.
Williams, V. R., Wu, W.-T., Tsai, H. Y., and Yoon, S. H., Kim, S. K., Kim, K. H., Kwon, T. W.,
Bates, H. G. 1958. Varietal differences in and Teah, Y. K. 1987. Evaluation of physico-
amylose content of rice starch. J. Agric. Food chemical changes in cooking oil during heating.
Chem. 6:47-48. J. Am. Oil Chem. Soc. 64:870-873.
CHAPTER 17
PREPARATION AND APPLICATIONS

OF RICE FLOUR
An-I Yeh
Graduate Institute of Food Science & Technology
National Taiwan University
Taipei, Taiwan
In the wheat industry, “milling” always produces flour. In the rice industry, the
terminology differs. “Milling” usually refers to the removal of bran for producing
polished rice, and “grinding” broken or polished rice produces rice flour. However,
“milling” has also been used to describe the process of grinding rice. In this chapter,
“polishing” refers to the removal of bran and “milling” to the process of grinding rice
to flour.
The applications of rice flour include traditional Asian foods and baked goods,
many of which have now spread around the world. Rice flour can also be used to
produce gluten-free rice bread for gluten-sensitive individuals and is typically used
in baby foods.
Japan has been called “the land of the rice crop” since ancient times (Sakurai,
1971). In 1987, Japan used 105,000 metric tons (t) of brown rice to produce rice flour
(Hirao, 1990). Although Taiwan is a small island, it consumed about 80,000 t of rice
flour in 1997 (TFSB, 1998). In 1989, about 95% of the world’s rice was produced in
developing countries and 92% of it in Asia (Juliano, 1993). At that time, the
percentages of rice imported by developing countries and Asia were 73 and 76%,
respectively, of the world’s traded rice flour (61,659 t) (Fig. 1) (FAO, 2000). The
data indicate that, in 1989, rice flour was mostly consumed in Asia. The world
annual trading volume increased 85% (from 61,659 to 114,547 t) from 1989 to 1998.
In 10 years, the volume imported by developed countries tripled, and it accounted in
1998 for 45% of the world’s annual trading volume. In the same period, the volume
imported by Europe increased almost four times (from 9,178 to 35,340 t). As an
exporter of rice, the United States increased its exports of rice flour from 1,556 to
6,256 t. The volume imported by Asia or other developing countries fluctuated
slightly from 1989 to 1998. The data imply that the increase in world trading volume
was due to increased consumption in developed countries. So the utilization of rice
flour was broadened from making Oriental products to include the producing of
Western goods.
Thailand, a major exporter of rice flour, offers a price near the world average (Fig.
2) (FAO, 2000). China, the principal rice producer, offers the lowest price for rice
flour in the market. The United States, the fourth exporter of rice flour, offered a
495
price higher than the world average before 1994 but lower than the world average
from 1995 to 1998. Apparently, the difference in price among exporting countries
was narrowed from 1995 to 1998. It appears that the price for rice flour is still good
Fig. 1. Import quantity (t) of rice flour. (Data taken from FAO, 2000)
Fig. 2. Average exporting price (U.S.$/kg) of rice flour. (Data taken from FAO, 2000)
Rice Flour / 497
compared to that for cereal. The FAO cereals price index (1990–1992 = 100)
fluctuated from 1990 to 1995, increased sharply to 180 in mid 1996, and then
declined to below 100 in 2000. The world average price of rice flour showed a steady
growth from 1990 to 2000, which indicates a growing market for rice flour.
PREPARATION
Three methods (wet, semidry, and dry milling) are used to prepare rice flour, as
illustrated in Figure 3. Soaking, milling, drying, and regrinding are involved in wet
milling. Usually rice is soaked in water for more than 4 hr, drained, and ground by a
stone mill with water (three to five times the weight of rice). The key purpose of wet
milling is to prepare flour while causing only minimal damage to the starch. Damage
can be reduced by the cooling and lubricating effects of water. After milling, excess
water is removed by drying, and the flour is then gently reground to prepare the
wet-milled flour. This flour is generally used for products such as Japanese cake, rice
crackers, Taiwanese cake, rice noodles, Filipino rice cake, puto, and Indian
fermented foods such as idli, etc. The treatment of wastewater becomes an issue in
most countries.
In semidry milling, rice is soaked, drained, and ground by using a stamp or pin
mill without adding any water. Because no excess water is used, the concern of
wastewater is eliminated. In general, the applications of semidry milled flour are
similar to those for wet-milled flour.
On the other hand, cleaned (or tempered) rice can be directly ground to different
sizes by various mills, in a process designated as dry milling. Different mills
(including rolling, pounding, shock, stone, and lateral steel) can be used to prepare
rice flour without generating any wastewater, but the starch shows more damage,
which influences the applications of the rice flour. The particle size of rice flour
Soaking Draining Wet Milling

Brown Rice
Milled Rice Semi-dry Drying &
Brokens Milling Grinding
Noodles
Cakes
Rice Flour Crackers
Cleaning Rice Paper
Egg Roll Wrapper
Baked Products (breads, muffins,

crackers, cake, puddings)
Rice
Dry Milling Flour
Puffed & Extrusion-Cooked
(baby foods, breakfast cereals,
snacks)
Air Classification Instant milk

or High-Protein Rice Flour Gruel
Enzyme Digestion Pudding
Fig. 3. Manufacture of rice flours and their applications.

plays a key role in the flour properties. Dry-milled rice flours usually have the same
protein content as does the parent rice and are used for baked products, baby foods,
extrusion-cooked products, and in high-protein flour.
High-Protein Rice Flour
High-protein fractions can be used for nutritional supplements in baby foods or

other products, such as instant milk, gruel, and puddings. These foods provide
needed protein in the diets of young children in rice-eating countries. In the wheat
milling industry, air-classification technology has been successfully used to produce
high-protein flour (Ziegler and Greer, 1971). Combining air classification with
turbo-milling, Houston et al (1964) obtained high-protein rice flour with a 75%
increase in protein content and 8–10% yield. The coarse flour contained the bulk of
the original flour and slightly lower protein.
Another approach is to abrade the outer layers of polished rice kernels to
obtain a 10–15% yield of high-protein flour, with double the protein content of
the original feedstock, and a lower-protein white kernel usable as traditional
polished rice. This process takes advantage of the high concentration of protein
located on the periphery of the rice kernel. A higher concentration of protein was
found in the outer layers of long-grain rice kernels than in those of medium and
short grain (Bean and Nishita, 1985). Compared with that of the parent kernels, the
amylose content of the starch fraction in the high-protein flour didn’t vary. The
pasting characteristics of high-protein flour obtained by abrasion mill were directly
related to the starch content of each fraction, as well as to variety (Houston et al,
1970).
Hansen et al (1981) developed an enzymatic method to produce high-protein
flour. A rice flour slurry (5% rice flour) was pregelatinized by heat and digested by
enzymes. The products (consisting of undigested starch and the intact protein bodies)
were prepared by subsequent treatments, including centrifugation (or filtration),
freeze-drying, and regrinding. The high-protein flour consisted of 25% protein and
32% starch, which is a much higher protein level than that of the starting rice flour
(8.3% protein and 88.7% starch). To reduce the cost of the process, Chen and Chang
(1984) used a 20–30% rice flour slurry as the starting material, and the product was
dried by spray- or drum-dryer. The resulting flour contained 28% protein, which was
slightly higher than that obtained by Hansen et al (1981) and was more economical to
produce because of its higher solids input.
Pregelatinization of rice flour enhances the effectiveness of the enzymatic reac-
tions but is undesirable because the increased viscosity and gelation could cause
difficulties in the processing. Utilization of heat-stable a-amylase can save the step
of pregelatinization before digestion (Brooks and Griffin, 1987). The protein content
can be increased up to 76% in the insoluble fraction by subsequently treating the
solution with other carbohydrate-hydrolyzing enzymes, such as glucoamylase,
cellulase, and hemicellulase after the digestion by heat-stable a-amylase (Shih and
Daigle, 1997). Besides the insoluble fractions, the hydrolyzed products (including
dextrin, maltose, and glucose) can be used for other applications. Glucose can be
further isomerized to make high-fructose rice syrup (Chen and Chang, 1984). The
protein fractions obtained by the enzymatic digestion of rice flour possess the
original subunit structure, solubility, and emulsion activity (Shih and Daigle, 1997).
Thus, use of a-amylase is a good alternative to abrasion milling.
Rice Flour / 499
Brown Rice Flour
Health-conscious consumers are becoming more interested in brown rice flour

because of its nutritional value. This flour is prepared via different milling methods,
using rough rice as the starting material. The presence of bran imparts a different
flavor and chewy texture to products. Unfortunately, brown rice flours have limited
shelf-life stability due to the lipase activity initiated in the bran layers during milling.
The lipid components are enzymatically hydrolyzed, freeing fatty acids and causing
rancidity. The rancidity can be minimized when the enzymes are inactivated. Extru-
sion can be used to stabilize rice bran right after it is separated from the endosperm in
the production of head rice (Randall et al, 1985). The stabilized bran can be ground
to flour and blended with white rice flour to meet consumers’ demands.
The functionality of rice flour depends upon rice variety and preparation methods.
The selection of a specific variety is based on the application. The combination of
variety and preparation methods can produce rice flours for different products, such
as Japanese specialty rice flour.
Japanese Specialty Flours
In Japan, japonica and waxy rice are the two major varieties for preparing
specialty flours. The difference in pretreatment is the key to distinguishing their
properties and applications. When heating is involved in the milling process, the
product is designated as precooked flour. Otherwise, the product is recognized as
native flour. The process for preparing native japonica rice flour is very
straightforward (Fig. 4). Japonica rice is washed, soaked, dried by air to a moisture
content of 18–30%, and then pulverized by a beater, roller, or stamping machine. It is
Rice bun (with red bean)

Dry cake
Herb mochi
Jioshin
flour Leaf-wrapped mochi
(>150 mesh) Senbei
Japonica
rice
Washing, soaking,
draining (18-30% H2O),
grinding Dessert
Snack
Jioy flour Rice noodles
(<200 mesh)
Senbei
(moisture content 10-12%)
Fig. 4. Production and application of native japonica specialty rice flours.
a semidry milling process. After pulverization, flour is further dried to a moisture

content of 10–12% and sieved. Jioshin flour, having an average particle size larger
than 150 mesh, is used for red bean rice bun, dry cake, and various types of mochi.
The fine fraction (<200 mesh), called jioyo flour, is used for Japanese desserts and
rice noodles. Both flours can also be used to prepare rice crackers (senbei). Shironu
flour is a special product produced from the over-polishing of nonwaxy rice used for
sake manufacture. It is relatively rich in protein and is suitable for infant food.
Soaking is an important step in making native waxy rice flour (Fig. 5) and is
generally done at temperatures lower than 10°C. The soaking time varies from two
days to one month, and the soaking water is changed daily to minimize microbial
growth. Pulverization is conducted by using a stone or vertical mill with the addition
of water (twice the amount of rice, by weight) during grinding. A screen of 80–100
mesh is used to sieve the milk, which is further drained by pressing to form a
cakelike dough containing about 45% moisture. The cakelike dough is dried by hot
air at 60–70°C to yield mochi flour, with moisture less than 13%. Mochi flour is used
for red bean mochi, a sesame ball. The flour can be further sieved to produce
different particle sizes. Shiratama flour, consisting of 50% particles of 80–100 mesh
and 50% particles of 200–250 mesh, is generally used in applications similar to those
of mochi flour in the eastern regions of Osaka City (Kanto area). An alternative
method is used to prepare flour in the western region of Osaka city (Kansai area).
The washed waxy rice is dried to about 20% moisture by air (at low temperature),
pulverized to smaller than 120 mesh, and then instantaneously dried by hot air to a
moisture content of 10–12%. The flour, called kyuhi flour, is generally used for
various mochi-related products.
Both japonica and waxy rice are used to prepare precooked flour. Several flours
can be produced as a result of different processing and parching techniques. As
illustrated in Figure 6, washing, soaking, and draining are common procedures for
making japonica precooked flours. Drained rice is pounded, dried, ground, and
Red bean mochi
Tangyuan
Mochi flour
Sesame ball (baked)
Waxy
rice
Washing; soaking (<10oC, 2 days-1 month, renew Red bean mochi
water daily), stone or vertical mill (2-fold water),
80-100 mesh sieve; drying by pressing in a Shiratama flour
filter bag (about 45% H2O); drying by hot air (50% 80-100 mesh +
Tangyuan
o
(60-70 C) (<13%H2O); regrinding 50% 200-250 mesh) Candy
(Kanto area) Small rice ball
Red bean mochi
Air dry (~20% H2O), pulverized (120 mesh), Kyuhi flour Tangyuan
Instantaneous hot air drying (Kansai area) Candy
(10-12%H2O) Small rice ball
Fig. 5. Production and applications of native waxy specialty rice flours.
Rice Flour / 501
parched on a flat pan (using sand or rock salt at 170–200°C) to prepare flour
generally used for baby foods and gravies. Flour can be further sieved to a coarse
fraction called jhonan flour, which is used for Japanese cake (wagashi), and a fine
fraction, called mijin flour, which is used for Japanese dessert. Alternatively, drained
rice is parched on a flat pan by using sand or rock salt at 170–200°C and pulverized
to produce coarse particles with low moisture (3–5%). This flour, called heisou flour,
is used for Japanese dessert. Precooked brown rice flour is prepared from cooked or
gelatinized brown rice and is consumed with milk or hot water.
Washing, soaking, and draining are also common steps in preparing various pre-
cooked waxy specialty rice flours (Fig. 7). Drained waxy rice is steam-cooked,
pounded to mochi dough, parched by a roller drier at 170–200°C, and then
pulverized. Further sieving is used to prepare flours with different particle sizes.
Mijin flour, having particles smaller than 200 mesh, is used for rice cake and
Japanese specialty desserts. The coarse fraction, kanbai flour, having particles of
60–100 mesh, is used for gravies, thickeners, batters for bean-based snack, cakes,
and processed snacks. The mochi dough can alternatively be dried by hot air, ground,
and parched at 170–200°C to yield slightly expanded jhonan flour, which has a
porous structure and a particle size of about 80 mesh. Jhonan flour is used for puffed
dry cake (okoshi), dry cake (rakugan), rice crackers, mochi, and rice buns. The
steam-cooked rice may be dried and crushed into a coarse flour called doumeiji flour,
used for mochi-related products such as leaf-wrapped mochi and rice buns. To
prepare doumeiji flour, the polished waxy rice is usually soaked for one to two days.
Some precooked flour can be prepared without steam cooking. The drained rice is
parched on a flat pan by using sand or rock salt at 170–200°C and is then pulverized
to make lakugan flour, which is used for dry cake. Extending the parching time
produces flour with low moisture (3–5%), called jiyosou flour, which is used for
dessert. Recently pregelatinization has been accomplished by extrusion cooking,
which saves time and labor for preparing mochi dough and simplifies the process for
multipurpose applications. The extruded starches exhibit lower cold-water swelling,
higher solubility, and thinner paste than do the drum-dried samples, due to de-
Baby food
Baby flour
Gravies
Japonica Jhonan Wagashi

rice flour (Japanese cake)
(coarse powder)
Washing, soaking, draining, pounding, drying,
pulverizing, parching
(170-200oC)
Mijin flour Dessert
(fine powder)
Parching (using sand or salt at Heisou flour Dessert

o
170-200 C), pulverized (coarse powder)
(3-5% H2O)
Fig. 6. Manufacture and applications of precooked japonica specialty rice flours.

Cake
Waxy Mijin flour
Snacks
(<200 mesh)
rice
1 Washing, 2 soaking, Gravies
3 draining, 4 steam 5 pounding, 6 parching, Thickeners
o
cooking, (170-200 C) Batters
7 pulverizing Kanbai flour (coating for beans)
(60-100 mesh) Desserts
(such as cake)
Snacks
Batters
Rakugan
Jhonan
Arare
flour
6 Hot-air drying, 7 pulverizing, (80 mesh)
Mochi
o
8 parching (170-200 C) Crackers
Mochi (wrapped
Doumeiji by leaves)
5 Hot-air drying, 6 crushing flour Crackers
(rice soaked for 1-2 days)
Lakugan Dry cake
flour
o Jiyosou flour Japanese snack
4 Parching (using sand or salt at 170-200 C),
(3-5% H2O)
5 pulverizing
Fig. 7. Manufacture and applications of precooked waxy specialty rice flours.
polymerization (Doublier et al, 1986). Delrue and Chamberlin (2000) used a

continuous process to prepare partially gelatinized rice flour with low moisture
content (less than 30%) at low temperature (about 74°C) using steam.
FUNCTIONAL PROPERTIES
Effect of Inherent Properties
Because rice flours are made from polished rice or brokens, their chemical
composition should be the same or similar to that of whole rice (Oryza sativa L.).
The differences among rice flours are governed by inherent cultivar variations,
methods of milling or grinding, and the pretreatments of rice or flour. Rice flour is
identified by the variety or the amylose content, which indicates its inherent
properties. As discussed in previous chapters, rice can be classified traditionally
based upon the original planting area. Indica rice has been grown in India, Sri Lanka,
Pakistan, Bangladesh, Taiwan, southern China, Thailand, Vietnam, Burma, Laos,
Cambodia, etc. Japonica rice is planted in the northern and central regions of China,
Korea, and Japan. Javanica rice is bred in Indonesia. The grain length is a common
standard with which to identify rice type in the United States. Three common types
are long (longer than 6.6 mm), medium (between 5.5 and 6.6 mm), and short (shorter
than 5.5 mm). Sometimes the grain longer than 7.5 mm is called extra long.
Endosperm characteristics are also used for classification. Waxy (or glutinous or
sweet) rice consists of less than 1% amylose, and nonwaxy rice generally has an
Rice Flour / 503
Fig. 8. Amylograph curves for four rice varieties. Short curves, 20% slurries; full curves, 10% slurries.
(Reprinted, with permission, from Nishita and Bean, 1979)
amylose content higher than 10%. In general, indica rice or long-grain rice has a high
amylose content and gelatinization temperature. Japonica rice or short-grain rice
contains an intermediate level of amylose. It appears that the amylose content plays a
key role in the inherent properties of rice flour. Various methods are used to deter-
mine the amylose content in starch. The iodine titration method developed by Juliano
et al (1981) is generally used for rice.
Rice starch granules are the smallest (3–10 µm) among cereal starches. Amylose
content or amylose-amylopectin ratio, gelatinization, and pasting behavior are the
important properties for describing rice starch. Amylography is usually used to ana-
lyze the pasting behavior of rice flour or starch. Figure 8 shows two sets of pasting
curves for four cultivars having contrasting properties (Nishita and Bean, 1979). The
full curves, which include heating, holding, and cooling modes, use 10% rice flour
slurries. The short curves on the left are obtained with 20% slurries. At this high
concentration, a rapid increase in measurable viscosity occurs immediately upon the
onset of granule swelling. The temperature at which a short curve leaves the baseline
gives a fairly reliable estimate of gelatinization temperature. Juliano et al (1985)
estimated gelatinization temperatures from the intercept of short curves with the
20-BU line.
In Figure 8, sample A is Calrose, a California medium-grain with low amylose
content and low gelatinization temperature. Sample C is Labelle, a typical long-grain
rice with both intermediate amylose content and gelatinization. Sample D, Century
Patna, is one of the few atypical U.S. rices. It has long-grain length but low amylose
content and high gelatinization temperature, both characteristics outside those for a
standard U.S. long-grain rice. Sample B, IR-8, is a long-grain rice with high amylose
content but low gelatinization temperature. It is also atypical by U.S. standards. The
pasting curve for IR-8 shows negligible viscosity breakdown during a high-
temperature holding cycle and a marked viscosity increase during the cooling cycle,
indicating that retrogradation is rapid (Bean, 1986). It appears that the grain length
alone is not adequate to represent the properties of rice.
Researchers and industrial users of rice flour have adapted other amylograph
methods. Variations include changes in slurry concentration, starting temperature,
heating rate, and endpoint temperature. They are usually designed to reflect
processing conditions for individual products, and hence they predict functional
performance of the rice flour for a specific application. Sharp (1986) developed a
rapid procedure for determining amylograph viscosity of rice flour. The results have
a linear relationship with the standard method. A Rapid ViscoAnalyser (RVA) has
been used to determine the viscosity of rice flour (Blakeney et al, 1991) using a 3-g
sample instead of the 27- to 28-g sample used in the standard amylograph method.
The test can be completed in 15 min instead of the more than 20 min required for the
standard amylograph procedure. Due to the convenience and small quantity of
sample, the RVA has been widely adopted.
Differential scanning calorimetry (DSC) has been widely used to determine the
temperature and heat of gelatinization. The water content affects the thermal
behavior of rice starch. Biliaderis et al (1986) demonstrated that a single symmetrical
endotherm accounts for the phase transition of starch isolated from rice, IR-8, when
water content is higher than 60% (Fig. 9). As the water content decreases, the size of
the transition endotherm is progressively reduced, with the concomitant
development of a second transition at a higher temperature. The authors also pointed
out that the correlation between DSC data and birefringence endpoint temperature
(BETP) was significant (r = 0.84, P < 0.01). By combining light microscopy with
DSC, Yeh and Li (1996) confirmed that two stages of swelling and
disruption/dissolution of starch granules are involved during gelatinization. The
peak temperature (Tp) of the DSC thermogram is the transition point for the two
stages. The rate of granule rupture reaches maximum at Tp (Fig. 10). When DSC is
used to determine the thermal behavior of starches, the energy transferred from the
surroundings to the system, measured by DSC, can be expressed as a function of
enthalpy and pressure as follows:
DQ = DH – VDP
where DQ is the energy transferred from the surroundings to the system, DH is the
change in enthalpy, V is the volume, and DP is the change in pressure. The change in
energy determined by DSC is DQ. When the system is at constant pressure (i.e., DP =
0), then DQ = DH. However, it is very difficult to ensure that the pressure in the
crucible or sample pan is constant during DSC measurement, particularly when the
water vapor pressure exceeds 1 atm at temperatures higher than 100°C. Therefore, it
is more accurate to describe the endothermic energy of a starch solution measured by
DSC as the “heat” instead of “enthalpy” of gelatinization.
Rice flours and their isolated starch counterparts have been found to exhibit
similar gelatinization characteristics (Chungcharoen and Lund, 1987) in the DSC
thermogram. The presence of sucrose or sodium chloride results in a shift of
gelatinization to higher temperature (Table 1), and heat associated with the
endothermic process decreases. The effect of solutes on gelatinization is more
remarkable at low water content. Among the tested solutes, the surfactant con-
Rice Flour / 505
Fig. 9. Differential scanning colorimetry thermal curves of IR-8 rice at various water contents. Numbers
designate the weight fraction of starch. Starch weight from top to bottom (mg): 1.08, 2.03, 3.22, 4.03, 5.09,
5.41, 4.53, 5.61, and 3.96. Heating rate 10°C/min. All data files were normalized to a constant sample weight
of 5 mg. (Reprinted, with permission, from Biliaderis et al, 1986; ©American Chemical Society)
Fig. 10. Degree of granule rupture of rice starch at different temperatures as observed microscopically.
(Reprinted, with permission, from Yeh and Li, 1996)
Rice Flour / 507
taining 90% glyceryl monostearate has the least effect on the shift in gelatinization
temperature.
The small-amplitude oscillatory rheometer has been used to study the pasting
behavior of rice starch. Figure 11 (Tsai et al, 1997) illustrates the variation of storage
modulus (G) during heating for the starch obtained from japonica (TNu 67), waxy
(TCW 70), and cross-linked waxy (C-TCW 70) rice. At low concentration (10%), G
Fig. 11. Storage modulus (G) measurement of starches from three rice varieties at different
concentrations during heating. (Reprinted, with permission, from Tsai et al, 1997)
Fig. 12. Regression for swelling power, T/Tp, and tan Gmax in log scales. (Reprinted from Li and Yeh,
2001, with permission from Elsevier Science)
increases initially due to the swelling of starch granules and reaches a plateau. At
intermediate concentration (15%), G of japonica rice starch reaches a maximum
(designated as Gmax) at about 77°C. Neither waxy nor cross-linked waxy starches
show the maximal. At high concentration (25%), all the samples exhibit maximal.
The initial increase in G is attributed to the swelling of starch granules that occurs at
low concentration. The decrease in G could be attributed to the melting of remaining
crystallite, which results in swollen granules becoming softer. Li and Yeh (2001)
found that 10 starch samples (20% solids content) from different resources
(including rice, corn, potato, sweet potato, pea, mung bean, etc.) exhibit maximum
G during heating. The loss tangent at Gmax (designated as tan dG max) has been found

to be related to the swelling power of starch granules (Fig. 12) and is expressed as:
Swelling power = 26.12 (T/Tp)14.59(tan dG max)0.5153 (r2 = 0.73).

2
Although the correlation coefficient (r ) is not very high, the data demonstrate that
the thermal and rheological properties are related to swelling power. T/Tp is an
important factor in the effect of temperature on swelling power and appears to be a
driving force. It is much easier to measure swelling power than to determine thermal
and rheological properties. The results show that the swelling power appears to be a
good index for on-line quality control.
Storage Effect
For practical reasons, it is necessary to assess the pasting behavior of rice flour
during storage. Rice flours give higher amylograph viscosities after storage, whether
stored as flours or stored as rice kernels and subsequently ground to flour (Bean,
Rice Flour / 509
Fig. 13. Effect of storage temperature, duration, and moisture content (MC) on amylograph peak viscosity
of rice cultivar Bengal. (Reprinted, with permission, from Perdon et al, 1997)
1986). Both the temperature and duration of temperature significantly affect the
gelatinization of rice flour (Fan and Marks, 1999). Rice stored at high temperature
(38°C) exhibited higher gelatinization temperatures and more gelatinization heat
than did rice stored at low temperatures (4 or 21°C). Storage moisture is another
factor affecting the amylograph viscosity. As illustrated in one study (Fig. 13)
(Perdon et al, 1997), amylograph peak viscosity increased during the first month of
storage. After three months of storage, the peak viscosity started to decline when the
storage moisture was lower than 10.7% but leveled off when moisture was higher
than 12.9%. High storage temperature (37°C) resulted in a larger increase in peak
viscosity than did low storage temperature (3°C).
Previous investigators (Shibuya et al, 1977; Shibuya and Iwasaki, 1982) believed
that starch granule pasting properties do not change during storage. The alteration of
cell wall materials affects the pasting properties of starch. Dhaliwal et al (1991)
pointed out that the increase in amylograph peak viscosity of polished rice with time
of storage of paddy could be due to an increase in free fatty acid content and a
decrease in diastatic activity during storage. Removal of the pericarp and seed coat,
which was accomplished by 1.3% polishing, led to the largest decrease in Tp value
(approximately 5°C) and To values (approximately 1°C) in both the full-fat and par-
tially defatted rice samples compared with the subsequent decrease in these thermal
parameters with further polishing (Champagne et al, 1990). The results demonstrated
that the polishing from 1.3 to 9.1% resulted in approximately 52% decrease in the
lipid-starch weight ratio of rice. The decrease in lipid resulted in a larger decrease in
Tp, as shown in Figure 14.
Fig. 14. Relationship between the decrease in Tp values and the percentages of nonstarch lipids removed from
unpolished (!) and 1.3–9.1% polished (#) kernels. Data points are included for extraction times other than 25
min and 1 hr. Additional points on the curve for unpolished kernels are for 15-min, 2-hr, 4-hr, and 8-hr lipid
extraction times. The lowest point on the curve for polished kernels is for 1.3% polished rice kernels that
were solvent extracted for 15 min. (Reprinted, with permission, from Champagne et al, 1990)
Milling Effect
In addition to the inherent properties and storage history, the milling or grinding
method imposes variation on the functional properties of rice flour.
DRY MILLING
The type of mill or grinder affects the particle size distributions, temperature, and
functional properties of flours. Figure 15 illustrates the particle size distributions of
various flours prepared by different mills (Nishita and Bean, 1982). It appears that
pin mills and hammer mills yield fine particles. Nevertheless, turbo-milled flour has
the finest size among all the flours, according to examination by scanning electron
microscope. This highlights the limitation of sieve-size analyses for such a fine
particle. The finer flours have a higher level of starch damage and do not function
well in yeast-leavened bread. Both the model and design of the grinder affect the
performance, as well as the particle size of the flour. Using three different grinders
for dry milling and analyzing particle size with a laser particle size analyzer, Chen et
al (1999) also reported that turbo-milling yielded the finest flour, with particles finer
than that obtained by Nishita and Bean (1982) (estimated sieve size through which
50% of the sample would pass [PS50] of 62.2 µm vs. 151 µm) (Table 2).
Nevertheless, the hammer-milled flour (PS50 of 197.2 µm) obtained by Chen et al
(1999) was coarser than that (117 µm) reported by Nishita and Bean (1982). Thus, it
is necessary to have basic information on the machine before comparing the
performance of mills or grinders. Both turbo- and hammer mills cause higher
temperatures than do other mills. That is part of reason they yield flour that has more
Rice Flour / 511
Fig. 15. Cumulative particle size distribution by weight for rice flours ground on various mills.
(Reprinted, with permission, from Nishita and Bean, 1982)
TABLE 2
Rice Flour: Median Particle Size by Weight (PS50)a
and Flour Temperatureb
PS50 Flour Temperature
Mill (µm) (°C)
Turbo 151c (62.2) 88
Hammer 117 (197.2) 75
Pin
2× 88 38
1× 111 38
Roller 155 30
Blade 239 50
Burr
Bauer 290 32
Coffee 630 36
a PS50 = estimated sieve size through which 50% of the sample would pass.
b Source: Nishita and Bean (1982). Numbers in parentheses are from Chen et al (1991).
c Determined by scanning electron microscopy to be the finest flour in the study.
damaged starch and does not function well in yeast-leavened bread. Three mills
(turbo-, cyclone, and hammer mill) do not result in a significant difference in
swelling power for waxy rice flour (Fig. 16). In general, the swelling power of rice
flour increases with temperature. Coarse flour exhibits low swelling power at high
temperature. The fine flour from turbo-mills exhibits the highest peak viscosity,
based on amylograph values (Nishita and Bean, 1982) or RVA measurement (Chen
et al, 1999). Coarse flour gives an initial viscosity increase at higher temperature.
Data indicate that fine flours give more thickening power than do coarse flours.
Apparently, the type of mill or grinder affects the particle size as well as the
amylograph pasting characteristics of the resulting flour. Thus, it is necessary to
know milling histories before interpreting the pasting characteristics of rice flour.
WET MILLING
Soaking, adding excess water (three to five times the weight of the rice) during
grinding, and removing the excess water are the three steps that differentiate wet
milling from dry milling. The absorption of water into the rice kernel is a function of
Fig. 16. Effect of milling methods on the swelling power of TCW 70 (A) and TCSW 1 (B) rice flours.
(Reprinted, with permission, from Chen et al, 1999)
Rice Flour / 513
both temperature and time. Using indica rice (TNuS 19) as an example, Yeh et al
(1992) reported that the water content in the rice kernel increased with time and
reached a plateau (a final water content) when the soaking temperature was lower
than 63.5°C. The final water content of the rice kernel was 27% at 48°C and rose to
32% at 63.5°C. The time required to reach the final water content increased with
the raising of temperature. Forty minutes was adequate to reach the final water
content at 48°C, but 80 min was needed at 63.5°C. Soaking at a temperature higher
than 70°C resulted in breakage of the rice kernel due to starch gelatinization. The
product became pregelatinized flour after milling. Table 3 lists the compositions of
rice flours obtained by dry-, wet-, and semidry milling (Lu and Lii, 1989). For both
japonica (TNu 70) and indica (TCS 80) rice, dry-milled flour is highest in protein,
lipid, ash, and reducing sugar. Soaking results in the loss of soluble solids (Medcalf
and Lund, 1985). Thus wet-milled flour is low in protein, lipid, ash, and reducing
sugar. Wet milling yields flour with particles smaller than that produced from
turbo-milled (dry milling) flour for waxy rice flour (Chen et al, 1999). The
wet-milled flour has a lower degree of gelatinization, measured by b-amylase
pullulanase, than that for dry-milled flour, due to the lubricating and cooling effect
of the added water.
Wet-milled flour is superior to dry-milled flour in various practical applications,
particularly for traditional products. Bean et al (1983) observed a noticeable
improvement in bread texture when wet-milled flour was used instead of
dry-milled flour for 100% rice bread. However, the additional cost associated with
redrying, regrinding, and wastewater treatment encourages the seeking of
alternative methods. Hydration of dry-milled rice flour with intensive mixing of
the flour and water has proven to be a viable alternative (Bean, 1986). The
intensive mixing improved the texture of bread and the texture and volume of
white layer cakes (Bean et al, 1983). The effects of mixing are governed by
intensity of mixing, water temperature, and the time and temperature of holding
the hydrated mixture. High-speed mixing is generally used. A long holding time is
needed for a low holding temperature. When the mixture is held at room
temperature, 6–12 hr is required to produce improvement in the texture of rice
bread. As the holding temperature is raised to 60°C, 1–2 hr is sufficient to make the
improvement. Starting with flour hydrated with sufficient water for batter or dough
requirements eliminates the costly removal of excess water and wastewater
treatment. The replacement of wet-milled flour by dry-milled flour is worthy of
further study.
TABLE 3
Chemical Compositions of the Rice Flours from Various Milling Processesa
Protein Lipid Ash Reducing Sugar
Variety Milling (%) (%) (%) (%)
TNu 70 Dry 8.02 0.41 0.45 0.90
Semidry 7.55 0.03 0.15 0.12
Wet 6.67 0.03 0.17 0.15
TCS 10 Dry 7.91 0.27 0.57 0.74
Semidry 7.56 0.08 0.20 0.16
Wet 5.70 0.03 0.22 0.15
a Source: Lu and Lii (1989); used with permission.
SEMIDRY MILLING
Rice is soaked for ½ to 1 hr at room temperature and then centrifuged (2,000 rpm)
for about 1 min before grinding in the process of semidry milling. After soaking, the
moisture content of the rice is approximately 30%, which varies with the rice variety.
There is about 5–10% reduction in moisture content after centrifugation. Hammer or
plate mills have been used for grinding. The pin mill is generally used for commercial
production in Taiwan. After grinding, the flour is redried to about 13% moisture by air
at 40°C. The contents of protein, lipid, ash, and reducing sugar in semidry milled flour
are lower than those in dry-milled flour. The semidry milled flour exhibits low peak
viscosity with delayed initial viscosity increase (as shown by the viscoamylograph)
and has the highest degree of gelatinization (about 40%) among flours prepared by dry,
wet, and semidry milling (Lu and Lii, 1989). During semidry milling, the moisture in
rice is not adequate to cool the heat generated by friction force. Instead, the moisture
from soaking enhances the gelatinization of starch by lowering the gelatinization
temperature. Therefore, semidry-milled flour has a high degree of gelatinization and
exhibits low peak viscosity and delayed temperature for the initial viscosity increase
(in the viscoamylograph). Compared with wet milling, no wastewater is generated in
semidry milling, which saves operating costs. Nevertheless, the related literature is
limited. Semidry milled flours are worthy of further studies to investigate their
applications.
APPLICATIONS
Many rice products exist around the world. The products made either from rice
flour or by the process involving rice milling (or grinding) are discussed below.
Noodles
Rice noodles (Fig. 17) are a popular rice product in Asia. They are also called
bifun (bihon) or vermicelli in Taiwan, China, and east southern Asia and harusame in
Japan. They are usually prepared from high-amylose rice flour. Figure 18 illustrates
Fig. 17. Rice noodles. (Photo by George Chao-Chi Chuang)

Rice Flour / 515
the traditional method of manufacturing rice noodles. Freshly polished rice is pre-
ferred so that rancidity in the product is minimized. Brokens can be used to reduce
the cost. Horizontal stone or concrete mills and vertical steel mills are used to
wet-mill the presoaked rice. Juliano and Sakurai (1985) discussed the function of
each step in the traditional batch-wise process. Microbial growth generally is not
desirable for regular rice noodles. Nevertheless, fermentation is conducted during
long-time soaking for special rice noodles in Thailand. Brokens are soaked for three
days for fermentation, which reduces the pH from 7 to 3.5, with Lactobacillus spp.
and Streptococcus spp. The protein content decreases to 1.6% after one day of
fermentation and to 1.1% after three days (Maneepun, 1987).
In Taiwan, extrusion has been commercially adapted to integrate the actions of
cooking, kneading, and forming. A single-screw extruder is used to partially cook the
feeding flour with appropriate water content (35–40%). The precooked dough is then
kneaded and formed to noodle strands via the second single-screw extruder. The
extruded noodle strands are further steam cooked and then dried in an air oven.
Alternatively, one extruder can be used instead of two extruders. However,
pregelatinized flour or starch is needed as a binder to enhance the strength of the
noodle strands. Khandker et al (1986) pointed out that a single-screw extruder is a
good tool for forming noodles that are given a final cooking in hot water. A
twin-screw extruder has been used to produce rice noodles without adding
pregelatinized flour or starch (Yeh et al, 1991). The manufacture of rice noodles is
then switched from a batch to a continuous process.
Both dry and rehydrated noodles are examined to evaluate the quality, such as
whiteness, translucency, and absence of brokens. The dry noodles are soaked in cold
water until hydrated and soft and then soaked in hot water. The stability of the
rehydrated sample is determined by observing the turbidity of the hot soaking water,
and the sample is manually assessed for surface smoothness and stickiness. Excellent
noodles hydrate with minimum turbidity and surface stickiness. Cooking loss has
been a good index with which to assess the quality of hydrated noodles (Yeh et al,
1991). Low cooking loss is associated with low turbidity. Generally, a cooking loss
lower than 10% is desirable for good quality, clear soup.
Polished rice Dehydration Mixing

Wet &
or milling By pressing
brokens crushing
Steam or water Forming a dough

Forming Kneading
partially cooking (balls or cylinders)
Cooking in Cooling
boiling water in cool water
Cooking Wet Drying Packaging Product

by steam noodles
Fig. 18. Traditional process for manufacturing rice noodles.

Traditionally, rice noodles are made from wet-milled flour, which results in
environmental concerns due to wastewater. In response, some have tried to replace
wet-milled flour with dry-milled flour. Tsao et al (1976) made rice noodles from
dry-milled flour at an extrusion temperature of 99°C. Yeh et al (1991) compared the
product characteristics from wet-milled and dry-milled flours. As discussed previ-
ously, wet-milled flour has a smaller particle size than does dry-milled flour. Under
the light microscope (Fig. 19), dry-milled particles aggregate, but wet-milled parti-
cles disperse individually. The heat generated during dry milling gelatinizes the
surface of the particle, which results in a sticky surface and aggregation. The
Fig. 19. Particles of dry-milled rice flour (A) and wet-milled rice flour (B) under a microscope, showing
differences in aggregation. (Reprinted, with permission, from Yeh et al, 1991)
Rice Flour / 517
dry-milled flour yields rice noodles with a cooking loss as high as 20.9%, which is
much higher than that for the wet-milled product (9.2%). In addition, the surface of
noodles from dry-milled flour is much rougher than is that from wet-milled flour.
The particle size of dry milled flour has been found to affect its suitability for rice
noodle making. A general stamp mill yielded the smallest flour fraction of 100–150
µm, which withstood heating without disintegration (Saito, 1980) and was suitable
for making rice noodles. The rate of water absorption by coarse particles is slower
than that by fine particles, which results in lower degree of gelatinization and higher
die pressure during twin-screw extrusion cooking (Yeh et al, 1992). Flour smaller
than 106 µm was recommended for extrusion cooking of rice noodles with a degree
of starch gelatinization higher than 90%. Using a process similar to traditional
methods, Hemavathy and Bhat (1994) reported that rice noodles prepared from fine
flours (average particle size lower than 115 µm) were too soft and sticky and lacked
firmness, while those from flours of 138 and 165 µm had the desired softness,
smooth surface, and firmness. Coarse flours of 214 and 307 µm yielded undesirably
firm strands with rough surfaces. It appears that the particle size distribution plays a
key role in selecting dry-milled flour.
The amylose content in rice flour has been shown to affect the quality of rice
noodles. High amylose content is necessary to produce a product with fibrillar
structure (Mestres et al, 1988) and to give noodles lower density and more whiteness.
In the Philippines, rice noodles made from intermediate-amylose rice (IR48) are
preferred (Umali, 1981). High-amylose rice with low gelatinization temperature and
hard gel consistency (Taichung no. 1) is preferred for making rice noodles in Taiwan
(Li and Luh, 1980). Milled rice flour with hard gel consistency (35 mm) and high
amylose content (30%) is used in Thailand (Juliano and Sakurai, 1985). In Japan,
rice noodles are made exclusively from indica rice with high amylose content (Saito,
1980). Although satisfactory pasta has been prepared from Italian and Russian rice
with low to intermediate amylose (Resmini et al, 1979; Juliano and Pascual, 1980),
G.-Y. Wu and W. –L. Chen (unpublished data) reported that japonica rice with low
to intermediate amylose content is not suitable for making rice noodles due to
stickiness and low gel consistency.
Surface gelatinization of the extruded noodles improves the stability and texture
upon cooking. The cooked extruded noodles lack the x-ray diffraction crystallinity of
raw rice starch (Chen and Luh, 1980). The average degree of starch gelatinization of
steamed, extruded noodles is reported to be 65–75%, with the surface 98%
gelatinized and core only 55% gelatinized (Juliano and Sakurai, 1985). Electron
micrography of both raw and cooked rice noodles showed a honeycomb-like fibrillar
network of retrograded starch, with protein interwoven in the strands (Resmini et al,
1979; Resmini and Pagani, 1983).
Instant rice noodles are commercially available in Asia. The claim of “instant”
indicates that the dry rice noodles can be rehydrated completely and are ready to
serve in 5 min, most in less than 3 min. The manufacturing process is similar to that
discussed above. In modern plants, a two-extruder setup is utilized to produce the
product. The size or diameter of the product is a critical factor of instant rice noodles.
Yeh et al (1990) conducted a survey and pointed out that there is no significant
difference in the degree of starch gelatinization between instant and regular rice
noodles, but there is significant difference in size. Instant rice noodles are smaller
than 0.68 mm. When the diameter of rice noodles is larger than 0.78 mm, the product
cannot meet the requirement for instant rice noodles.
Fig. 20. Sheeted or flat noodles. (Photo by George Chao-Chi Chuang)
Mitaimu (or bitaibah), one kind of noodle, is eaten as a snack food. A short wet-type
noodle with a large diameter, it is eaten with syrup and ice during the summer. In
Taiwan, mitaimu is also eaten with meat, green onion, or leek. Salt, monosodium
glutamate, and pepper are always added as condiments. The manufacturing of mitaimu
is similar to the process for conventional rice noodles. During the mixing of flour with
water, hot water can be used to gelatinize a portion of starch for binding effects.
Mitaimu generally is shorter than 10 cm and has a diameter of 3–5 mm. It is very
difficult to dry or rehydrate the dried mitaimu; thus, no dried product is commercially
available. As is the case for rice noodles, only indica rice with high amylose content is
used as the raw material. Some starches, such as corn, tapioca, and sweet potato, are
added to adjust the texture of the product. Sweet potato starch is used to increase its
elasticity. Corn or tapioca starch reduces its hardness. Therefore, various products have
been developed to meet consumers’ needs.
Sheeted or flat noodles (Fig. 20) are quite popular in Asia, particularly Thailand
(Siegel et al, 1975), Japan (Imai, 1982), and Taiwan. The starting material can be
either dry-milled or wet-milled flour, but wet-milled is preferred in most areas.
Traditionally, a drum is used to form a milk layer that is then cooked by steam. The
gelatinized sheets are partially dried and cut to noodle strips. In a modern plant, an
extruder is used for cooking and forming. Indica rice with high amylose is
traditionally used. Some Japanese nonwaxy rice (amylose content <20%) has been
successfully used to prepare sheeted or flat noodles (Ishikawa et al, 1980; Saito,
1980). Corn or tapioca starch can be used to adjust the texture of the final products.
Only thin flat rice noodles (thickness <1.5 mm) are dried and packed as commercial
products. The wet-type product is used in most restaurants.
Cakes
Many varieties of rice cakes are eaten in Asia, particularly in Japan, Taiwan, and
China. Rice cakes can be classified into pastry (or unleavened), dry cake, and fer-
mented or leavened products. The pastry, or unleavened, products consist of more
Rice Flour / 519
than 35% moisture, and dry cakes generally have moisture lower than 30%. Mochi is
a typical example of a pastry-type product. Niangao (also called nenkau) is a tradi-
tional rice cake, particularly eaten at the Chinese New Year in Taiwan and south
China. Dumplings of steamed rice (called dango in Japan) and rice pudding (called
uiro in Japan) are other pastry-type products. Dry cake is generally made from
pregelatinized flour. Rakugan is an example. Lunjiaogao, which is yeast-
fermented and steamed to form a unique structure, is an example of a fermented
cake in Taiwan.
PASTRY PRODUCTS
Traditionally, mochi is prepared from waxy milled rice by washing and steaming.
The cooked rice is manually pounded to become the sticky dough known as mochi.
Wooden pestles and mortars can also be used. Mochi used to be a main entrée for
native Taiwanese during special events or festivals. Now, it is a popular snack or
dessert in Japan and Taiwan. In Japan, various mochi products are made from the
specialty rice flour described above. The properties of its products are governed by
different operational conditions during rice flour preparation. The consumption in
Japan increased from 43,000 t in 1983 to 52,300 t in 1986 (T. Iwasaki, unpublished
data). Mochi is usually cut into pieces and eaten right after preparation. Sometime it
is rolled over peanut powder or sprinkled with soy sauce powder for special flavor
and taste. A dry layer (such as seaweed) can be used to wrap mochi, which eliminates
stickiness on the hands and generates a crisp mouthfeel. Some fillings, particularly
red bean paste or green bean paste, are used in mochi to create new products. Dried
mochi is commercially available in Japan and is used with instant noodles. Frozen
mochi filled with ice cream has become a popular dessert recently. Egg white
(powder or liquid) is usually added to prepare frozen mochi dough.
During pounding or kneading, air is incorporated into the dough and forms a
dispersion system consisting of the solid component (mainly amylopectin) and air
bubbles (Horiuchi, 1980a). The free space corresponding to air bubbles results in a
decrease in viscosity as the temperature increases. In relaxation tests, the storage
modulus (G) decreases with time (Fig. 21). The drop in G is more significant at
temperatures ranging from 47 to 57°C (Horiuchi, 1980b). In commercial practice,
wet-milled waxy flour is mixed with water and cooked in a steam kneader. The
dough is consequentially kneaded before cooling to form mochi exhibiting the
properties of commercial products made from waxy rice. Both formulation and
preparation method affect the properties of the final products. With the aid of
rheological measurements, Horiuchi (1980c) pointed out that the optimum solids
content was about 45% for ease of handling of mochi dough. Mochi quality is related
to the molecular properties of amylopectin in the different rice varieties (Palmiano
and Juliano, 1972). Waxy rice flour with a gelatinization temperature of 66–68.5°C
is preferred. Flour with a high gelatinization temperature results in harder mochi
dough after cold storage at 3°C for 24 hr (Yanase, 1982a,b). Nagashima et al (1987)
compared the effects of preparation methods on the properties of mochi. Pounding or
stamping yields a product exhibiting much of the initial rice grain structure. Mochi
prepared with a pounding-type machine shows the greatest shearing strength, and
that made by stamping show the greatest resistance to extension. Mixer-type
machines yield a well-smashed paste showing minimum hardness and the lowest
resistance to extension. Adding waxy cornstarch increases the tenderness, while
added cassava starch results in increased hardness and adhesiveness.
Fig. 21. Relaxation modulus-time relationships of 44.5% shiratama mochi cake at various temperatures
between 32.4 and 57.0°C. (Reprinted, with permission, from Horiuchi, 1980b)
Extrusion cooking has been a good tool to integrate the actions of cooking,
mixing, and kneading. Inaba et al (1988) illustrated that mochi prepared by a
twin-screw extruder has the same quality as that prepared by an electric rice cake
maker made for home use (Toshiba AFC-154). In addition, the extruded products
have a slower rate of retrogradation than those made by an electric rice cake maker
during one month of storage. The repetitive extrusion (up to five times) increased G
during dynamic measurements (Horiuchi, 1987). The mochi dough kneaded by hand
exhibited the lowest G among the different kneading treatments. Chuang and Yeh
(2002) demonstrated that a single-screw extruder can be used to produce mochi with
moisture content up to 55%. The extruded mochi has viscoelastic properties similar
to those of commercial products. High extrusion temperature and high water content
yield a higher degree of starch gelatinization. The increase in specific mechanical
energy results in an increase in both G and the loss modulus (G") (Fig. 22), but
decrease in water soluble index (WSI). The input of mechanical energy enhances the
formation of the network, which reduces the WSI. Thus, kneading is an essential step
on making mochi.
Niangao (also called nenkau) (Fig. 23) is a traditional rice cake eaten at Chinese
New Year and festivals in Taiwan and south China. It has recently become a popular
product for snacking. Wet-milled waxy rice flour is kneaded with water, sweetened
with sugar, transferred into a vessel, and steamed until complete gelatinization
Rice Flour / 521
Fig. 22. G and G" of extrudates increased with the increase of specific mechanical energy. (Reprinted,
with permission, from Chuang and Yeh, 2002)
Fig. 23. Niangao. (Photo by George Chao-Chi Chuang)
occurs. After being cooled to room temperature, niangao is cut into pieces for direct
eating or coated with batter for frying. In addition to sugar, other materials can be
added for new products. Cooked red bean is a popular addition. For unsweetened or
savory-type products, ingredients such as salt, monosodium glutamate, crushed
green onion, and meat may be added instead of sugar before steaming. Short-grain
waxy rice yields a soft product preferred for making niangao. Pregelatinized starch
or flour is added to minimize the precipitation of rice flour during steaming and
ensure the uniformity of product quality.
Luobogao (Fig. 24) is a popular product in Asia, particularly in Taiwan and Hong
Kong. It is a savory product made from indica rice by using procedures similar to
those for niangao. Wet-milled indica rice flour is kneaded with water and cooked for
partial gelatinization. Then, crushed radish is added and mixed thoroughly in the
partially gelatinized slurry, which is transferred into a vessel and steamed until
complete gelatinization. After cooling, the final product can be cut into pieces, which
are stir-fried, cooked with soup, or fried to be eaten separately. Partial pre-
gelatinization is necessary to minimize the precipitation of radish and rice flour and
to ensure uniform texture. Salt, pepper, or monosodium glutamate may be added as
condiments. Due to its good taste and convenience, luobogao has become a popular
breakfast in Taiwan.
In the Philippines, bibingka is a cake product. Wet-milled waxy rice flour is
mixed with coconut milk and sugar, poured into pans lined with banana leaves, and
heated with charcoal below and above. Caramel topping is usually used. Puto
kutsinta is an unleavened cake having texture like a stiff pudding and is prepared
from wet-milled rice flour with white or brown sugar and lye plus flavoring (Ramos
and Reyes, 1981).
A unique unleavened cake, called zhuang-yuan-gao in Taiwan, has elastic texture
similar to that of steamed bread and is used for snacking. In Chinese, zhuang-yuan
means the person who places first in the national examination. Zhuang-yuan-gao is
one of the dishes at a dinner party hosted by the emperor to welcome zhuang-yuan.
Wet-milled rice flour is mixed thoroughly with water and steamed. Rice flour
generally is a mixture of short-grain waxy rice and japonica rice. The use of indica
rice flour with higher amylose content yields products with fluffier texture than does
japonica rice flour. The quantity of water is important for controlling the product
Fig. 24. Luobogao. (Photo by George Chao-Chi Chuang)

Rice Flour / 523
quality. A water content of 37–40% (wb) is generally used. Too much water results
in a doughlike product, which is sticky without elasticity. The product does not cook
completely when there is not enough water. The shape of the product is determined
by various molds, the most popular a mushroom shape. Peanut or coconut powder
mixed with sugar can be used as filling to fit consumers’ preferences. Fine powdered
sugar sometime is mixed with rice flour to impart sweetness to the product.
Zhuang-yuan-gao is a perishable product and retrogrades easily if stored at low
temperature.
Rice pudding, called uiro in Japan, is a perishable product because of its high
moisture content (about 40%). The manufacturing process is quite simple compared
with that for other products. Waxy rice flour is mixed with sugar, cornstarch, water,
and flavoring. The admixture is then placed in a plastic pouch and steamed. Sweet
bean curd, green tea, coffee, cherries, and other foodstuffs can be used to create new
flavors depending upon consumers’ preferences. In China, short-to-medium rice
with low amylose can also be used to prepare puddings.
DRY CAKE
Dry cake is subdivided according to its manufacturing process into pressed
(rakugan in Japan) and puffed (okoshi in Japan) products. The process for making
rakugan is very simple when parched rice flour (waxy or nonwaxy) is used as the
starting material. The precooked rice flour is mixed with sugar and other ingredients
(such as chestnut, buckwheat, and red beans) by continuous stirring until it is sticky.
That mixture is then transferred into a wooden mold and pressed or beaten to make it
congeal. The cake is removed from the mold and dried in a conventional dryer.
Okoshi is a typical Japanese cake made of puffed broken rice mixed with other
ingredients and molded with starch syrup, sugar, and flavoring (Sakurai, 1971).
Bingpi moon cake, a product similar to rakugan, is popular for the mid-fall festival
in Taiwan. Sugar syrup is mixed with the admixture of precooked waxy rice flour and
wheat starch. After thorough agitation and relaxation, the dough is filled with some
foodstuffs (such as red bean, green bean, etc.) and pressed into a mold to form a certain
shape. The cake can be stored at low-temperature or freezing conditions without
significant retrogradation. Ice cream has recently been used to create a new product.
The flour used in a typical formulation consists of precooked waxy rice flour and
wheat starch at a ratio of 4:1. Based on the flour as 100%, a typical formulation is fine
powdered sugar (200%), water (100%), and maltose (30%).
Xue-piang-gao, also a dry cak, is prepared by different processes. Precooked
waxy rice flour is mixed thoroughly with powdered sugar, sugar syrup, and cooking
oil (such as soybean oil). Some toasted black sesame seeds can be added for flavor.
The mixture is pressed into a mold, steamed, cut into strips, and simmered. Some
ingredients, such as chestnut and red beans, can be added to make additional
products. The ratio of sugar to precooked waxy rice flour is generally kept at 0.8–1.1.
Oil is needed to minimize the stickiness and improve the texture. No extra water is
added. The methods of pretreatment of rice flour and the particle size after grinding
affect the product quality. However, the literature on these topics is very limited.
FERMENTED OR LEAVENED CAKE

Fermentation has been incorporated into the manufacture of rice cakes. In the
Philippines, puto is a typical example. Wet-milled rice flour is yeast fermented and
steamed to become a cake. In a comparison of varietal differences, Perez and Juliano
(1988) noted that aged intermediate-amylose rice yielded greater volume than did
rice with lower or higher amylose content and gave optimum softness. Idli and dosai,
two major fermented rice cakes in India, are served typically as breakfast food
(Hesseltine, 1979). The products are soft, spongy, have low bulk density, and possess
sour flavor like that of steamed sourdough bread (Singh et al, 1995). They are made
from a mixture of rice and mung bean (black gram, Phaseolus mungo). Polished rice
(mainly parboiled) and decorticated (dehulled) black gram are washed, soaked in
water, and wet-milled separately to give coarse (about 0.6 mm) rice flour and a
smooth, gelatinous black-gram paste. The flours are mixed with salt (0.8%), and the
thick batter is fermented overnight at room temperature, steamed, and served hot.
Other ingredients, such as cashew nut, ghee, pepper, and ginger, may be added to
enhance the flavor.
Leavening and acidification are two major processes occurring during fermenta-
tion. Steaming gelatinizes the starch for texture. Juliano and Sakurai (1985)
discussed the importance of fermentation and steaming in making idli. Parboiled
rice, which produces soft idli that is not sticky, is more suitable than raw rice. Singh
et al (1995) demonstrated that extruded rice flour can be utilized in making idli. The
suitable extrusion temperature is 125°C when a Brabender laboratory extruder is
used. At a higher level of substitution (30 or 40%) of extruded rice flour, specific
gravity decreases and acidity increases. This could be attributed to fermentation of
the batter induced by the increased quantities of soluble carbohydrates,
enzyme-susceptible denatured proteins, and gelatinized starch, all of which are
readily fermented by the natural microflora.
Dosai is a product similar to idli but usually contains less black gram and is
usually fried, not steamed (van Veen, 1972). It is a pancake made from black gram
and rice flour in a thin batter after fermentation (Susheelamma and Rao, 1979). A
soft and spongy texture and low bulk density are its major textural attributes.
In Taiwan, a unique fermented cake called lunjiaogao is produced. Its appearance
is creamy white with some transparency (Fig. 25), and it exhibits a soft, elastic, and
Fig. 25. Lunjiaogao. (Photo by George Chao-Chi Chuang)

Rice Flour / 525
smooth texture with a sweet-and-sour flavor. Its unique cylindrical structure imparts
a characteristic viscoelasticity and texture to the product. As illustrated in Figure 26,
wet-milled indica rice flour is mixed with a hot water-sugar mixture to partially
gelatinize rice starch and form a paste. The extent of gelatinization of rice starch and
the fluidity of the paste affect the fermentation time and transparency of the product.
Thus, gelatinization is critical in the production process. The optimum ratio of rice
flour (with about 45% moisture) to sugar to water is about 1:1.2–1.4:1.2–1.5. After
mixing and cooling, yeast is added to the slurry, and the mixture is fermented. Small
bubbles floating on the surface indicate that the fermentation is complete. After the
fermented broth is transferred to the vessel for steaming, a 10- to 15-min rest period
can minimize the bubbles on the product surface. Steaming at a high vapor-rising
rate is necessary for making good products. When the vapor-rising rate is too low, no
cylindrical structure is observed and the product surface is concave. A very high
steam-rising rate results in a split surface. The physical force of the upward vapors
appears to contribute to the unique structure. Further investigations are necessary to
understand the mechanism of structure formation. The fermentation has been found
to be the heterofermentative lactic acid type (Chen, 1992). Lactic acid is the major
organic acid generated during fermentation. In addition, malic, succinic, and acetic
acids are detected. Thus, the pH value drops from about 6.0 initially to 3.5 after 14 hr
of fermentation. Starch granules remain intact during fermentation. This indicates
that sugar is the major carbon resource for the growth of yeast. After steaming, no
starch granules are observed with a scanning electron microscope. Lunjiaogao is a
perishable product due to its high moisture content.
In Taiwan, fagao, a leavened cake similar to cupcakes, is eaten at festivals and the
Chinese New Year. In Chinese, one pronunciation of “fa” means getting wealth or
promoted; fagao is symbolic for good fortune. The widely used formulation consists
of indica rice flour (100%), fine sugar (70%), water (90%), and baking powder
(3.5%), which are mixed thoroughly. The mixture is settled for 1 hr and then
transferred to vessels for steaming. Generally, 30 min of steaming is adequate for
making acceptable products, as shown in Figure 27. The split on the surface makes
the product looks like a flower, which is essential for being accepted by consumers.
The variety of rice affects the product characteristics. Indica rice (such as TCN 1 and
TCS 17 in Taiwan), with high amylose content and hard gel consistency, yields prod-
ucts having high specific volume, a fluffy and flaky texture, and a high rate of
Indica rice Wet Pressing for

or draining Crushing
milling
brokens
Steaming Fermentation Cooling Mixing
Yeast
Cooking
Cut Product Packaging

Mixture of
water and sugar
Fig. 26. Process for manufacturing lunjiaogao.

Fig. 27. Fagao. (Photo by George Chao-Chi Chuang)
retrogradation (Huang et al, 1990). Japonica rice (such as TNu 67), with low amylose
content, yields products with a rigid structure and viscous texture. Overall, indica
rice is preferred for making fagao. The product made from wet-milled rice flour is
much superior to that made from dry-milled flour. Brown sugar can be used to
enhance caramelization for coloring. Fruit juice, chocolate or cocoa powder, or nuts
can be added to create products with various flavors.
Malay rice cake (apam) is a similar product. Rice flour is mixed with sugar, water,
and leavening agents (yeast or baking powder) and poured into a vessel. The mixture
is left to ferment for 3–5 hr when yeast is used. After fermentation, the mixture is
steamed for about 25 min. The ingredients affect the characteristics of apam.
Mohamed and Hamid (1998) reported that the optimum amylose-amylopectin ratio
is approximately 1:8. A higher amylopectin content results in sticky, elastic, and
unacceptable cake. Increasing the water content decreases the viscosity of the batter,
thus reducing the gas retention ability and firmness of the cooked batter. For
chemically leavened cake, sugar content correlates linearly with volume expansion
and product tenderness. Nevertheless, a sugar content above 50% of rice flour
decreases the volume and softness of yeast-leavened rice cake. Skimmed milk and
egg whites are found to improve the texture of rice cakes, with an optimum content
of 2–4%. The recommended formula for apam leavened chemically consists of 94
parts (by weight) rice flour, six parts glutinous rice flour, 95 parts water, 24 parts
sugar, 0.8 parts salt, 23 parts margarine, three parts emulsifier, three parts egg white,
and four parts skimmed milk. The formula for yeast-leavened apam is similar, except
it contains less water (92–94 parts) and more sugar (34–42 parts).
Crackers
Rice crackers are a major rice-based snack food. These products include senbei,
arare, xianggao, and putoseko. Senbei and arare are very popular in Japan. In 1983,
Japanese production of senbei and arare totaled 118,000 and 103,000 t, respectively
Rice Flour / 527
TABLE 4
Classification of Rice Crackers by Specific Volumea
Specific Volume
Product Type (mL/g) Name
Light and crispy >4.0 Senbei (niigata, atsuyaki, and sazanami type)
3.5–4.6 Arare (sinagawa and okaki type)
Intermediate 2.5–3.5 Arare (sinagawa and okaki type)
Hard type 2.9–3.5 Senbei (soka type)
2.0–2.5 Arare
a Data from Kurasawai (1988).
(Tani, 1985). Although the market size is smaller than that of chips (including potato
and corn), the consumption in both China and Australia has increased dramatically in
recent years. In general, senbei is made from nonwaxy rice and arare from waxy
rice. The products are very difficult to distinguish by shape. Nevertheless, specific
volume, defined as the volume per gram of product, can be used to classify the
products, as listed in Table 4. A light and crispy cracker generally has a specific
volume higher than 3.5 mL/g. Niigata-type senbei is a crispy cracker. Two types of
arare, sinagawa and okaki, are also crispy crackers. Only waxy rice is used to
manufacture an intermediate type of cracker, which is arare, with specific volume of
2.5–3.5 mL/g. Soka-type senbei has specific volume of 2.9–3.5 mL/g and is a hard
type. Generally, amylopectin expands easily and generates a more porous structure
with high specific volume. Sometimes it can be controlled to yield low specific
volume (2–2.5 mL/g) and a hard-type product. Traditionally, baking expands both
senbei and arare. Frying can be used instead of baking to create a new product, aghe
arare, with a crispier texture. Monaka shell, another example, is made from waxy
rice and contains sweetened red beans.
Figure 28 compares the processes for manufacturing senbei and arare. The
control of moisture content and processing temperature are two key factors. For
making senbei, indica rice is soaked to a 34–38% moisture content and pulverized to
fine particles smaller than 60–80 mesh. The cooked dough is cooled by cooling water
(about 18°C) and dried to form pellets. To make the pellet for arare, mochi dough
with about 40% moisture is prepared as described previously and is quickly cooled to
2–5°C by a quick-freezing method that makes it solidify uniformly. Because waxy
rice consists of almost 100% amylopectin, which does not easily form a gel, a low
temperature is used to enhance gel formation. Primary drying, tempering, and
secondary drying ensure uniform distribution of moisture in the pellet. The pellet is
baked in an oven at 200–260°C and expands to a final product with a moisture
content of about 3%. The expansion behavior of the pellet, which determines the
texture of the final product, is affected by baking temperature, moisture content, and
material properties. Seasoning materials, such as soy sauce powder, sesame seed,
peanut powder, syrup, or honey, can be coated on the surface of the pellet before
baking.
The cooking and kneading steps can be accomplished in an extruder. Chen and
Yeh (2000) demonstrated that a single-screw extruder can be used to cook rice flour
at a moisture content of up to 55%. When waxy rice flour is used to make an
arare-like product, the hardening treatment can be eliminated since the extrudates
form a gel after exiting the die. The extrudates are then dried at 40–70°C to form a
dry pellet, which is expanded by baking or microwaving. The pellet expansion
6(1%(, $5$5(
0RLVWXUH ,VVXHV 0RLVWXUH
DERXW 5RXJKLQGLFDULFH 9DULHW\ RIULFH 5RXJKZD[\ULFH DERXW
'HJUHHRISROLVKLQJ
3ROLVKLQJ 3ROLVKLQJ
SUHIHUUHG
:DVKLQJ
:DVKLQ J6RD NLQJ +\ GUDWLRQ 6RDNLQJ IRUKU
'UDLQLQJ 'UDLQLQJ
5ROOHU0LOOLQJ 3DUWLFOHVL]H
'HJUHHRIVWDUFK 6WHDPFRRNLQJ
6WHDPLQJNQHDGHU
R JHODWLQL]DWLRQ R
&PLQ &PLQ
.QHDGLQ J
R R
&RROLQ JWR &E\ 4XLFNFRROLQ JDW &
DERXW R
LPPHUVLQJLQWRZDWHU & WRKDUGHQIRUGD \V
.QHDGLQJ +RPRJHQL]DWLRQ
)RUPLQJ &XWWLQJ )RUPLQJ &XWWLQJ
3ULPDU\GU\LQJ
R
KRWDLUDW &
&RQWURORIWHPSHUDWXUH 'U \LQJ
DQGWLPHWR\ LHOGXQLIRUP
R
KRWDLUDW &
ZDWHUGLVWULEXWLRQ
7HPSHULQ J
KU
6HFRQGDU\GU\LQJ
R %DNLQJDW
KRWDLUDW & R
&
([SDQVLRQDI IHFWHGE\
WHPSHUDWXUHPRLVWXUH
FRQWHQWPDWHULDOSURSHUWLHV
6HDVRQLQJ
RU FRDWLQJ
3DFND JLQJ
352'8&7
Fig. 28. Manufacturing of senbei and arare. (Adapted from Kurasawai, 1988)
temperature (Te) increases linearly with the glass transition temperature (Tg) of the
pellet (Fig. 29). The data imply that Tg is a good reference by which to understand Te.
The amylose content affects the expansion ratio. Waxy rice flour yields the highest
Rice Flour / 529
Fig. 29. Linear increase in expansion temperature (Te) with glass transition temperature (Tg). (Reprinted,
with permission, from Chen and Yeh, 2000)
expansion ratio among several different feeding materials (Fig. 30). The increase in
amylose content decreases the expansion ratio. The equilibrium moisture of the
pellet is another key factor affecting the expansion behavior. The pellet exhibits the
highest expansion ratio at an equilibrium moisture of 10% (Chen and Yeh, 2001).
Adding dextrin results in lower molecular weight, indicated by the intrinsic
viscosity, and enhances the expansion by reducing the viscosity (Chen and Yeh,
2000). The value of (Te – Tg) is in the range of 20–100°C. It appears that the pellet’s
expansion is related to its glass transition behavior. As the pellet makes the transition
from glassy to rubbery state and becomes flexible due to heating, water is vaporized.
The vapor generates pressure inside the pellet and results in the expansion that occurs
at temperatures higher than Tg.
In addition to senbei and arare, various other rice-based snack foods are common
around the world. In China, xianggao is a product similar to rice crackers. Generally,
it is made from high-amylose rice flour and exhibits hard texture. Low-amylose rice
flour might be used, but the product is less crispy than is that made from
high-amylose rice flour. Puto seko is a similar product in the Philippines. Inter-
mediate- to high-amylose rice flour is generally used as the starting material. The
product is very crispy and breaks easily on handling. Kroepeck is also a popular rice
snack in the Philippines. Rice is soaked in water with lime, wet-milled, and often
mixed with shrimp extract to impart a unique flavor. The dried pellet is puffed by
frying. The process is amenable to continuous manufacturing using an extruder
(Papotto and Cosgriff, 1983). Khao kriap waue is a traditional Thai snack made from
pounded, gelatinized waxy rice, which is partially sun-dried and puffed by roasting
on a direct flame. The expansion ratio has been found to positively correlate with
Fig. 30. Expansion ratio as function of equilibrium moisture and amylose content. (Reprinted, with
permission, from Chen and Yeh, 2001)
amylopectin content. The product made from wet-milled flour is superior to that
from dry-milled flour. The smaller the flour particle size, the better the product
quality (Jomduang and Mohamed, 1994). Malao is a popular rice-based snack food
in Taiwan, particularly eaten during festivals and the Chinese New Year. Wet-milled
waxy rice flour is mixed with hot water and kneaded. Then taro is added into the
dough with kneading. The dough is cooked, kneaded for cooling, cut into sticklike
shapes, then dried to become a pellet, which is expanded by frying. The puffed
product is coated with maltose and dipped into roasted sesame seeds to become
malao.
Baked Products
Baking has been a major method of processing various cereal flours throughout
history. The use of rice flour for baking is probably motivated by its usefulness as a
substitute for wheat in the dietary treatment of wheat-intolerant patients (Dicke et
al, 1953; Sleisenger et al, 1957). Persons with celiac disorder have a high
sensitivity to wheat, manifested in a chronic reaction to certain wheat proteins,
commonly referred to as glutens. Rice flour can be used to produce gluten-free rice
bread for gluten-sensitive individuals. Of the common grains, rice is the safest, as it
rarely troubles anyone. For this reason, it is the first cereal recommended for
infants.
Rice Flour / 531
BREAD OR LEAVENED PRODUCTS

Bread. When mixed, the protein of rice does not develop a film that is capable of
holding fermentation gases. Thus, rice cannot be substituted directly for wheat in a
yeast-leavened product without formula modifications. Various gums and
surfactants have been used in breads made with wheat starch and other cereal flours
and starches (Jongh, 1961; Hart et al, 1970; Christiansen et al, 1974; Kulp et al,
1974). Hydroxypropyl methylcellulose is preferred to permit gas retention in 100%
rice flour bread (Nishita et al, 1976). For a composite flour consisting of 80% rice
flour and 20% potato starch, several combinations of carboxymethylcellulose,
hydroxypropyl methylcellulose, and water result in rice bread with moistness,
cohesiveness, yeast flavor, adhesiveness, aftertaste, top crust color, crumb color, cell
size uniformity, and predominant cell size comparable to those of a reference wheat
bread (Ylimaki et al, 1991). Rice starch predominantly governs the product’s
properties. The amylose-amylopectin ratio, gelatinization temperature, and pasting
behavior are particularly important.
For making 100% rice yeast-leavened bread, rice flour with low amylose content
(<20%) and low gelatinization temperature (<70°C) yields soft-textured crumb with
low loaf volume. Intermediate amylose content and gelatinization temperature yield
sandy and dry crumb (Bean and Nishita, 1985). Ylimaki et al (1991) noted that
medium-grain rice breads met more sensory reference standards than did long-grain
rice breads. In general, the bread made from wet-milled rice flour has a texture
superior to that made from dry-milled flour. In Japan, rice flour with low
gelatinization temperature (<70°C) has been successfully used to replace 60% of the
wheat flour for breadmaking.
These breads tend to stale very rapidly, limiting their distribution. This is a small
industry but very necessary for individuals with celiac disease.
Other Products. In addition to bread, rice flour can also be used in other products.
Japanese rice flour is mixed with wheat flour for making pan-bread (Takano et al,
1979). Untreated rice flour from brown or polished rice showed a typical dilution
effect at 20% substitution. Loaf volume is lower, and farinograph and extensigraph
curves show slightly less strength than 100% wheat flour. A mixture of medium-grain
low-amylose rice flour and waxy rice flour (ratio of 3:1) has been successfully used
to replace wheat flour for making muffins (Stucy Johnson, 1988). The muffin with
added (5–25%) waxy rice flour retained more moisture during baking than the
reference standard muffin (Stucy Johnson, 1990). Sensory evaluation showed no
significant differences among test muffins (with waxy rice flour) and reference
standards regarding tenderness, volume, and flavor characteristics. Appropriate
hydration is a key to improving the performance of rice flour for making
chemically leavened baked goods such as layer cakes, cupcakes, and muffins with
good volume, grain, texture, appearance, and eating quality (Bean et al, 1985;
Perez and Juliano, 1988). Thus, rice flour can be used up to 100% for preparing
baked products.
Flours from short- and medium-grain rice having low amylose content and low
gelatinization temperature are preferred for making layer cake. Long-grain rice flour
gives products with a sandy and dry texture (Bean et al, 1983). A high level of sugar
in a layer cake formula significantly increases the gelatinization temperature of the
starch. This must be considered when optimizing a formula. The ratio of sugar to
total water must be such that starch granules can swell and gelatinize to set the
structure during baking.
The Nepalese of hilly regions traditionally eat shel roti (Tamang et al, 1988). Rice
flour is mixed with water at a ratio of about 1:1. Banana, honey, ghee, and other
spices are then added. The well-mixed batter is allowed to ferment for between 4 (in
summer) and 24 (in winter) hr. The leavened batter is squeezed by hand and
deposited as a ring into hot fat for frying to produce shel roti.
UNLEAVENED PRODUCTS
Wheat is traditionally used to make flat unleavened bread in Pakistan and India;
this chapati or roti is puffed, semi-light, flexible, uniformly round and firm, but not
rough (Hoseney et al, 1981). Rice flour does not make excellent chapati due to the
lack of gluten. Fine rice flour and kneading are necessary to make acceptable
products (Khan and Eggum, 1978). The rice flour with softer gel consistency is
preferred. Juliano and Sakurai (1985) discussed the influence of variety on the
product quality and nutritional value of the final products.
Breakfast Cereal and Baby Foods
Rice breakfast cereals may be divided into two classes: those that require cooking
before serving and those that are ready to eat directly from the package. Rice cereal
that requires cooking is the “farina” type and is generally made from granulated rice,
which is beyond the scope of this chapter. The ready-to-eat rice cereals may be made
from the entire rice grain or from rice flour; they include rice flakes, puffed rice
(puffed by oven, puffing-gun, or extruder), shredded rice cereal, and multigrain
cereals. During the processes, puffing or expansion is a common phenomenon. In
extrusion, the sudden expansion of water vapor as the superheated water exits the die
increases the volume several times. The apparent specific volumes can reach or
exceed those attained by puffing gun, and the process seems to have several advan-
tages over gun puffing (Matz, 1970). Rice flour is mostly used for extrusion-puffed
products and multigrain cereals.
Extruders vary widely in degree of shearing, temperature, and pressure during
operation (Harper, 1981; Hauck, 1981). The particle size of feed material is critical
to the operation, particularly for a single-screw extruder. Fine rice flour results in
agglomeration at the extruder inlet, which causes the extruder to lose its prime
(Mottern et al, 1969). Generally, flours with average particle size ranging from 0.1 to
0.85 mm have been successfully used for extrusion (Spadaro et al, 1971; Molina et
al, 1978; Noguchi et al, 1981). With the advancement of technology, the restriction
on particle size is lessened for new-design extruders. Both moisture content and
temperature affect the degradation of starch molecules and the expansion behavior of
extrudates. During extrusion, the degradation of starch mainly occurs with the
amylopectin fraction (Linko et al, 1981; Mercier, 1981). Increasing barrel tempera-
ture results in more starch degradation and longer residence time. The flow pattern is
changed from a plug flow reactor to a continuous stirred-tank reactor (Yeh et al,
1999). The rice extrudates exhibit optimum expansion at 135 or 150°C and a
pressure of 50 × 105 Pa when a German twin-screw extruder is used. Low-amylose
rice gives higher expansion and a more brittle product than does high-amylose rice
(IRRI, 1985). Waxy rice flour yields extrudates with greater volume expansion and
is more brittle than is nonwaxy rice at 11–15% moisture, 140–150°C, and 123–147
bar (Yanase et al, 1982). However, de Mosqueda et al (1986) pointed out that the
effect of amylose on the expansion ratio was significant only at 135°C. The extrudate
Rice Flour / 533
yields optimum expansion at 15% moisture for brown and yellow polished rice
extruded by a single-screw extruder (Delgado and Victorio, 1996). Thomás et al
(1994) postulated that increasing moisture increases the mobility of the protein
molecules in the starting material, thereby increasing the probability of reactive sites
getting close enough to form cross-links between adjacent molecules and thus
inducing an expandable structure (lower density). High barrel temperature and high
screw speed result in a higher degree of starch gelatinization of rice extrudates, but
lower peak viscosity, hot-paste viscosity, and cold-paste viscosity of ground
extrudates, as determined by the RVA (Guha et al, 1998).
Rice gruel is a traditional weaning food and is also used as a breakfast food in
Taiwan. Rice is cooked with excess water for a longer time than usual. It is
frequently cooked with small fish to enrich its nutritional value and is served as a
main entrée for young children. The cooking is very time-consuming, and extrusion
cooking appears to be an attractive method of conducting the cooking by using rice
flour. The starch molecules degrade during extrusion cooking, which increases the
digestibility of extrudates; this is good for children and senior citizens. In general, the
necessary technology is similar to that for making cereals. Information about the
nutritional value of rice-based products extruded at high or low temperature needs to
be collected to evaluate the performance of extrudates.
Rice Milk and Beverages

Rice milk is a substitute for animal milk and milk powder, particularly for people
who are allergic to the lactose in milk. Sometimes it is recommended by doctors to feed
children having gut problems or diarrhea. Flour obtained from puffed rice or wet
milling of rice is mixed and cooked with water. Sugar and flavoring agents (particu-
larly peanut) are added to adjust the taste of the final product. Agitation is needed to
prevent precipitation during cooking. Rice milk can be served hot or cold. Generally,
the higher the solids content, the higher the nutritional value. However, high solids
content results in high viscosity and pumping difficulty. Hydrolysis of starch by
amylase can increase the solids content without increasing the viscosity of milk
(Mitchell et al, 1988). Lin et al (1988) suggested that a formulation consisting of 3.5%
(wt/vol) brown rice, 7.5% sugar, and 2% peanut oil gives the best sensory score.
In addition to rice milk, rice flour can be used for making other beverages. Boza, a
Turkish traditional beverage made by yeast and lactic acid fermentation of cooked
maize, wheat, and rice flour, is a good example. During fermentation, the pH drops
from 6.1 to 3.5 and the total titratable acidity by means of lactic acid increases from
0.02 to 0.27 mmol/g. Hancio÷lu and Karapinar (1997) reported that 77 isolates of
lactic acid bacteria and 70 yeast isolates were responsible for the fermentation. In
Japan, a new fermented, saccharified beverage using rice flour has been developed
(Tominaga and Sato, 1996). The use of glucoamylase from Rhizopus, commercially
named Gluczyme, yielded less reducing sugar but higher amino acid content and
acidity than did the use of glucoamylase from Aspergillus oryzae. The appropriate
selection of enzyme ensures the development of new products.
Miscellaneous Products
Rice flour has been used in a batter for coating potato strips (Sloan et al, 1991).
The percentage of rice flour in batter was increased from 2–10 to 15–25% by Wu and
Woerman (1997). The use of rice flour provides a desirable tenderness in the finished
products (Horn and Rogols, 2000). In addition, rice flour can reduce about 60% of
the oil uptake during frying, compared with the values for batters from wheat flour
(Shih and Daigle, 1999). Phosphorylated starch and pregelatinized rice flour enhance
both the thickening and oil-reducing capacities of the batter.
Various products are made from rice flour in the Philippines. Inginataan, waxy
rice flour balls, are cooked in sweetened coconut milk together with diced sweet
potato, yams, and sliced, cooked banana and jackfruit and are served either hot or
cold. Palitaw is a flattened, oblong piece of rice dough that is dropped into boiling
water until it is cooked and then is served topped with grated coconut, sugar, and
sesame seed. Buchi is a similar product using wet-milled waxy rice flour. It is a
brown ball enclosing a filling of mashed boiled mung bean or sweet potato, often
with sesame seeds on the surface. Tamale is a nutritious snack consisting of toasted
and ground rice and peanuts, sugar, spices, chicken, or ham, which is cooked until
thick enough to hold its shape. It is then wrapped in banana leaves and steamed for 2 hr
(Ramos and Reyes, 1981). To make espasol, rice flour is cooked in coconut milk
until oily and then sweetened, rolled into flat cylinders, dusted with rice flour, sliced
into segments, and wrapped in white paper (Cordero-Fernando, 1976). Sapin-sapin
is rice mustard made painstakingly with four or five thin layers of different colors.
Carrioca, caramelized balls of glutinous rice stuck three on a bamboo stick, is of
Chinese origin (Juliano and Sakurai, 1985). Sesame ball is a popular dessert in
Taiwan. Wet-milled waxy rice flour is mixed with water to form dough and some
fillings (such as mashed red bean) are added. The dough surface is dusted with
sesame seeds and then the dough is fried for expansion. The product can be served
either hot or cold.
SUMMARY
Rice flour offers potential for conventional and new products. The statistical data
illustrate that the consumption of rice flour has increased in developed countries. The
price of rice flour steadily increased despite the declining cereal price index from
1995 to 2000. This indicates that the market for rice flour is growing.
Wet-milled flour yields products superior to those made from dry-milled flour. It
is still challenging to obtain high-quality flour from dry milling, a process that
eliminates the problem of wastewater associated with wet milling. The impact of the
grinder on the flour properties should be further investigated in order to develop
dry-milled flour suitable for various products. Starch has been recognized to play a
key role in determining the properties of flour. Little is known about protein and lipid
components as they might affect the potential of rice flour for various products.
These would offer a fruitful area of research for developing new insights into the
possible functional contributions of rice flours.
Processes involving cooking, kneading, and forming are used to produce various
products made with rice flour. Extrusion cooking appears to be an attractive method
by which to integrate the actions of cooking, kneading, and forming into one
continuous unit operation. Most of the literature is concerned with expansion by
extrusion at low moisture and high temperature. Nevertheless, most traditional
products are prepared at high moisture, temperatures lower than 100°C, and low
shearing environments. Better understanding of high-moisture extrusion technology
would expand the utilization of extrusion cooking for producing new and traditional
products.
Rice Flour / 535
ACKNOWLEDGMENTS
I would like to express special appreciation to George Chao-Chi Chuang, who took the photos of rice
flour products and helped collect literature; Po-Yuan Chiang, who provided the information on processing
of bingpi moon cake, malao, and zhuang-yuan-gao; and Shin-Pin Kuo, who helped to prepare the figures.
REFERENCES
Bean, M. M. 1986. Rice flour—Its functional flour from broken rice. J. Sci. Food Agric.
variations. Cereal Foods World 31:477-481. 35:1128-1135.
Bean, M. M., and Nishita, K. D. 1985. Rice flour Christiansen, D. D., Gardner, H. W., Warner, K.,
for baking. Chapter 14 in: Rice: Chemistry and Boundy, B. K., and Inglett, G. E. 1974. Xanthan
Technology, 2nd ed. B. O. Juliano, Ed. Am. gum in protein-fortified starch bread. Food
Assoc. Cereal Chem., St. Paul, MN. Technol. 28:23-26, 28-29.
Bean, M. M., Elliston-Hoops, E. A., and Nishita, Chuang, G. C.-C., and Yeh, A.-I. 2002. Effects of
K. D. 1983. Rice flour treatment for product temperature and moisture content on
cake-baking applications. Cereal Chem. viscoelastic properties of glutinous rice extru-
60:445-449. dates. Cereal Chem. 79:36-40.
Bean, M. M., Nishita, K. D., and Elliston-Hoops, Chungcharoen, A., and Lund, D. B. 1987. Influ-
E. A. 1985. Method for preparing improved ence of solutes and water on rice starch
quality rice-containing baked products. U.S. gelatinization. Cereal Chem. 64:240-243.
patent 4,508,736. Cordero-Fernando, G. 1976. The Culinary Culture
Biliaderis, C. G., Perez, C. M., Maurice, T. J., and of the Philippines. Vera-Reyis, Inc., Quezon
Juliano, B. O. 1986. Thermal characterization of City, Philippines.
rice starches: A polymeric approach to phase De Mosqueda, M. B., Perez, C. M., Juliano, B. O.,
transitions of granular starch. J. Agric. Food del Rosario, R. R., and Bechtel, D. B. 1986.
Chem. 34:6-14. Varietal differences in properties of extru-
Blakeney, A. B., Welsh, L. A., and Bannon, D. R. sion-cooked rice flour. Food Chem.
1991. Rapid viscometric analysis of rice flour. 19:173-187.
Int. Rice Newsl. 16(5):11-12. Delgado, L. L., and Victorio, M. G. 1996.
Brooks, J. R., and Griffin, V. K. 1987. Liquefac- Characterisation of brown and milled yellow
tion of rice starch from milled rice flour using rice and development of an expanded snack. J.
heat-stable a-amylase. J. Food Sci. 52:712-714. Sci. Food Agric. 72:148-154.
Champagne, E. T., Marshall, W. E., and Goynes, Delrue, R. M. and Chamberlin, L. W. 2000. Proc-
W. R. 1990. Effect of degree of milling and lipid ess for the production of partially gelatinized
removal on starch gelatinization in the brown rice flour. U.S. patent 6,033,709.
rice kernel. Cereal Chem. 67:570-574. Dhaliwal, Y. S., Sekhon, K. S., and Nagi, H. P. S.
Chen, C.-M., and Yeh, A.-I. 2000. Expansion of 1991. Enzymatic activities and rheological
rice pellet: Examination of glass transition and properties of stored rice. Cereal Chem.
expansion temperature. J. Cereal Sci. 68:18-21.
32:137-145. Dicke, W. K., Weijers, H. A., and van de Kamer, J.
Chen, C.-M., and Yeh, A.-I. 2001. Effect of amy- H. 1953. Coeliac disease II. The presence in
lose content on expansion of extruded rice pel- wheat of a factor having a deleterious effect in
let. Cereal Chem. 78:261-266. case of coeliac disease. Acta Ped. 42:34-42.
Chen, J.-J., Lu, S., and Lii, C.-Y. 1999. Effect of Doublier, J. L., Colonna, P., and Mercier, C. 1986.
milling on the physicochemical characteristics Extrusion cooking and drum drying of wheat
of waxy rice in Taiwan. Cereal Chem. starch. II. Rheological characterization of starch
76:796-799. pastes. Cereal Chem. 63:240-246.
Chen, L.-D. 1992. Changes in physicochemical FAO, 2000. FAOSTAT. Database of the Food and
properties during fermentation of lunjiaogao. Agriculture Organization of the United
Masters thesis, Graduate Institute of Food Sci- Nations.(Online: www.FAO.org)
ence and Technology, National Taiwan Fan, J., and Marks, B. P. 1999. Effect of rough rice
University, Taipei, Taiwan. storage conditions on gelatinization and
Chen, S. Y., and Luh, B. S. 1980. Chemical, retrogradation properties of rice flours. Cereal
physical, and cooking properties of rice vermi- Chem. 76:894-897.
celli. (Abstr.) In: Proc. Rice Technol. Working Guha, M., Ali, S. Z., and Bhattacharya, S. 1998.
Group 18th, Texas A&M University, College Effect of barrel temperature and screw speed on
Station. rapid viscoanalyser pasting behaviour of rice
Chen, W. P., and Chang, Y. C. 1984. Production of extrudates. Int. J. Food Sci. Technol.
high-fructose rice syrup and high-protein rice 33:259-266.
Hancio÷lu, O., and Karapinar, M. 1997. Micro- doughs with a twin screw extruder and their
flora of Boza, a traditional fermented Turkish retrogradation properties. Nippon Shokuhin
beverage. Int. J. Food Microbiol. 35:271-274. Kogyo Gakkaishi 35:15-22.
Hansen, L. P., Hosek, R., Callan, M., and Jones, F. IRRI, 1985. Annual Report for 1984. International
T. 1981. The development of high-protein rice Rice Research. Institute, Los Baños, Laguna,
flour for early children feeding. Food Technol. Philippines.
35:38-42. Ishikawa, S., Egawa, K., and Saito, S. 1980.
Harper, J. M. 1981. Extrusion of Foods. Vols. I and Conditions of rice noodle making. Niigata-ken
II. CRC Press, Boca Raton, FL. Shokuryo Kenkyusho Kenkyu Hokoku
Hart, M. R., Graham, R. P., Gee, M., and Morgan, 17:33-36.
A. I., Jr. 1970. Bread from sorghum and barley Jomduang, S., and Mohamed, S. 1994. Effect of
flours. J. Food Sci. 35:661-665. amylose/amylopectin content, milling methods,
Hauck, B. W. 1981. Control of process variables in particle size, sugar, salt and oil on the puffed
extrusion cooking. Cereal Foods World product characteristics of a traditional Thai
26:170-173. rice-based snack food (khao kriap waue). J. Sci.
Hemavathy, J., and Bhat, K. K. 1994. Effect of Food Agric. 65:85-93.
particle size on viscoamylographic behaviour of Jongh, G. 1961. The formation of dough and bread
rice flour and vermicelli quality. J. Texture structure. I. The ability of starch to form struc-
Stud. 25:469-476. ture, and the improving effect of glyceryl
Hesseltine, C. W. 1979. Some important fermented monostearate. Cereal Chem. 38:140-152.
foods of mid-Asia, the Middle East, and Africa. Juliano, B. O. 1993. Rice in Human Nutrition.
J. Am. Oil Chem. Soc. 56:367-374. Food and Agriculture Organization of the
Hirao, M. 1990. Trend of rice consumption in United Nations, Rome.
Japan. Farming Jpn. 24(1):14-20. Juliano, B. O., and Pascual, C. G. 1980. Quality
Horiuchi, H. 1980a. Specific volume and free characteristics of milled rice grown in different
space relating to mechanical properties of waxy countries. Res. Paper Ser. no. 48. International
rice cake. Agric. Biol. Chem. 44:1231-1235. Rice Research. Institute, Los Baños, Laguna,
Horiuchi, H. 1980b. Stress relaxation of shiratama Philippines.
mochi cake. Nippon Nogeikagaku Kaishi Juliano, B. O., and Sakurai, J. 1985. Miscellaneous
54:247-252. rice products. Pages 569-618 in: Rice:
Horiuchi, H. 1980c. Preparative conditions of Chemistry and Technology, 2nd ed. B. O.
shiratama mochi cake. Nippon Nogeikagaku Juliano, Ed. Am. Assoc. Cereal Chem., St. Paul,
Kaishi 54:241-245. MN.
Horiuchi, H. 1987. Dynamic viscoelasticity of Juliano, B. O., Perez, C. M., Blakeney, A. B.,
shiratama mochi cake. Nippon Nogeikagaku Castillo, T., Kongseree, N., Laignelet, B., Lapis,
Kaishi 61:1417-1423. E. T., Murty, V. U. S., Paule, C. M., and Webb,
Horn, G., and Rogols, S. 2000. French fry potato B. O. 1981. International cooperative testing on
products with improved functionality and the amylose content of milled rice.
process for preparing. U.S. patent 6,080,434. Starch/Stärke 35(5):157-162.
Hoseney, R. C., Varriano-Marston, E., and Dendy, Juliano, B. O., Perez, C. M., Alyoshin, E. P.,
D. A. V. 1981. Sorghum and millets. Pages Romanov, V. B., Bean, M. M., Nishita, K. D.,
71-144 in: Advances in Cereal Science and Blakeney, A. B., Welsh, L. A., Delgado L., L.,
Technology, Vol. 4. Y. Pomeranz, Ed. Am. El Bayâ, A. W., Fossati, G., Kongseree, N.,
Assoc. Cereal Chem., St. Paul, MN. Mendes, F. P., Brilhante, S., Suzuki, H., Tada,
Houston, D. F., Mohammad, A., Wasserman, T., M., and Webb, B. D. 1985. Cooperative test on
and Kester, E. B. 1964. High-protein rice flours. amylography of milled-rice flour for pasting
Cereal Chem. 41:514-523. viscosity and starch gelatinization temperature.
Houston, D. F., Lew, E. J.-L., Chen, L. and Fellers, Starch/Stärke 37(2):40-50.
D. A. 1970. Amylograph viscosities of Khan, M. A., and Eggum, B. O. 1978. Effect of
high-protein rice flour. (Abstr. 75) Cereal Sci. baking on nutritive value of Pakistani bread. J.
Today 15:307. Sci. Food Agric. 29:1069-1075.
Huang, R.-M., Lin, T.-C., Tsiang, J.-M., and Lii, Khandker, A. K., de Mosqueda, M. B., del Rosario,
C.-Y. 1990. The effect of different varieties of R. R., and Juliano, B. B. 1986. Factors affecting
rice and the processing on the qualities of properties of rice noodles prepared with an
leaven-rice cakes. J. Chin. Agric. Chem. Soc. extruder. ASEAN Food J. 2:31-35.
28:267-275. Kulp, K., Hepburn, F. N., and Lehmann, T. A.
Imai, T. 1982. Manufacture of rice noodle. Food 1974. Preparation of bread without gluten.
Sci. Technol. Tokyo 22:50-65. (In Japanese) Bakers Dig. 48:34-37, 58.
Inaba, H., Hatanaka, Y., Iimura, H., Matsumura, Kurasawai, H. 1988. Rice and Rice Processing.
Y., and Mori, T. 1988. Production of rice cake 2nd ed. Kenpakusha Co., Tokyo.
Rice Flour / 537
Li, C. F., and Luh, B. S. 1980. Rice snack foods. prepared by various methods and with various
Pages 690-711 in: Rice: Production and additives. J. Jpn. Soc. Starch Sci. 34:179-185.
Utilization. B. S. Luh, Ed. AVI Publishing Co., Nishita, K. D., and Bean, M. M. 1979. Physico-
Westport, CT. chemical properties of rice in relation to bread.
Li, J.-Y., and Yeh, A.-I. 2001. Relationships Cereal Chem. 56:185-189.
between thermal, rheological characteristics and Nishita, K. D., and Bean, M. M. 1982. Grinding
swelling power for various starches. J. Food methods: Their impact on rice flour properties.
Eng. 50:141-148. Cereal Chem. 59:46-49.
Lin, T.-C., Shao, Y.-Y., and Chiang, W. 1988. Nishita, K. D., Roberts, R. L., Bean, M. M., and
Investigation of the processing and the quality Kennedy, B. M. 1976. Development of a
of rice milk. J. Chin. Agric. Chem. Soc. yeast-leavened rice-bread formula. Cereal
26:130-137. Chem. 53:626-635.
Linko, P., Colonna, P., and Mercier, C. 1981. High Noguchi, A., Kugimiya, W., Hague, Z., and Saio,
temperature, short-time extrusion cooking. K. 1981. Physical and chemical characteristics
Pages 145-235 in: Advances in Cereal Science of extruded rice flour and rice flour fortified
and Technology, Vol. 4. Y. Pomeranz, Ed. Am. with soybean protein isolate. J. Food Sci.
Assoc. Cereal Chem., St. Paul, MN. 47:240-245.
Lu, S., and Lii, C.-Y. 1989. The influences of Palmiano, E. P., and Juliano, B. O. 1972.
various milling processes on the physi- Physicochemical properties of Niigata waxy
cochemical properties of rice flours and the rices. Agric. Biol. Chem. 36:157-159.
rice flake preparation. Food Sci. (Taiwan) Papotto, G., and Cosgriff, M. H. 1983. Advances in
16(1):22-35. the technology of precooked and extruded
Maneepun, S. 1987. Production and consumption cereal based products. Pages 903-908 in: Pro-
of processed rice products in Thailand. Pages gress in Cereal Chemistry and Technology 5B.,
33-40 in: Proc. Int. Symp. Dietary Life of Proc. World Cereal and Bread Congress, 7th.
Rice-Eating Populations. Research Institute for Elsevier Science Publishing Co., Amsterdam.
Product Development, Kyoto, Japan. Perdon, A. A., Marks, B. P., Siebenmorgen, T. J.,
Matz, S. A. 1970. Manufacture of breakfast cere- and Reid, N. B. 1997. Effects of rough rice
als. In: Cereal Technology. S. A. Matz, Ed. AVI storage conditions on the amylograph and
Publishing Co., Westport, CT. cooking properties of medium-grain rice cv.
Medcalf, S. L., and Lund, D. B. 1985. Factors Bengal. Cereal Chem. 74:864-867.
affecting water uptake in milled rice. J. Food Perez, C. M., and Juliano, B. O. 1988. Varietal
Sci. 50:1676-1679. differences in quality characteristics of rice
Mercier, C. 1981. Behaviour of starch in food and layer cakes and fermented cakes. Cereal Chem.
feed processing. (Abstr.) Cereal Foods World 65:40-43.
26:504. Ramos, C. S., and Reyes, C. S. 1981. Gayuma. The
Mestres, C., Colonna, P., and Buleon, A. 1988. prevailing charm of Filippino cookery. Rex
Characteristics of starch networks within rice Book Store, Manila, Philippines.
flour noodles and mungbean starch vermicelli. Randall, J. M., Sayre, R. N., Schultz, W. G., Fong,
J. Food Sci. 53:1809-1812. R. Y., Mossman, A. P., Tribelhorn, R. E., and
Mitchell, C. R., Mitchell, P. R., and Nissenbaum, Saunders, R. M. 1985. Rice bran stabilization by
R. 1988. Nutritional rice milk production. U.S. extrusion cooking for extraction of edible oil. J.
patent 4,744,992. Food Sci. 50:361-364, 368.
Mohamed, S., and Hamid, N. A. 1998. Effects of Resmini, P., and Pagani, M. A. 1983. Ultrastruc-
ingredients on the characteristics of rice cakes. ture studies of pasta. A review. Food Micro-
J. Sci. Food Agric. 76:464-468. struct. 2:1-12, 98.
Molina, M. R., Chauvin, V., Cuevas, R., Elias, L. Resmini, P., Lenner, A., Debernardi, G., and
G., and Bressani, R. 1978. Evaluation of the Gerardo, D. 1979. Technology and Ultrastruc-
Brady crop cooker extruder for the ture of “Pasta” (Noodles) Produced with Rice
manufacture of pre-cooked food items from Flour. Centro Ricerche Soc. Braibanti, Milan,
different cereal/soybean mixtures. Pages 47-52 Italy.
in: Rice Report 1977. S. Barber, H. Mitsuda, Saito, S. 1980. Rice starch and rice powder as
H. S. R. Desikachar, and E. Tortosa, Eds. Inst. materials for food industry. Denpun Kagaku
Agroquim. Technol. Alimentos, Valencia, 27:295-313.
Spain. Sakurai, J. 1971. Rice as an industrial raw material
Mottern, H. H., Vix, H. L. E., and Spadaro, J. J. for the manufacture of processed and
1969. Popping characteristics of rice. Rice J. ready-to-eat rice products. Paper ID/WG 89/15,
70(8):9, 24-31. Interregional Seminar on the Industrial
Nagashima, N., Kawabata, A., and Nakamura, M. Processing of Rice. U.N. Ind. Dev. Org.,
1987. Physicochemical properties of “mochi” Vienna.
Sharp, R. N. 1986. A rapid procedure for deter- Processing Industries. FAO Regional Office for
mining amylographic viscosity of rice flour. Asia and the Pacific, Bangkok.
Cereal Chem. 63:325-326. TFSB, 1998. Taiwan Food Statistics Book. Dept. of
Shibuya, N., and Iwasaki, T. 1982. Effect of the Food, Taiwan Providence Government, Taipei.
enzymatic removal of endosperm cell wall on Thomás, R., Oliveira, J. C., Akdogan, H., and
the gelatinization properties of aged and unaged McCarthy, K. L. 1994. Effect of operating
rice flours. Starch/Stärke 34:300-303. conditions on physical characteristics of
Shibuya, N., Iwasaki, T., and Chikubu, S. 1977. On extruded rice starch. Int. J. Food Sci. Technol.
the changes of rice starch during storage of 29:503-514.
rice (Studies on deterioration of rice during Tominaga, M., and Sato, K. 1996. Lactic acid
storage. Part III). Dempun Koygo Gakkaishi fermentation of saccharified solution from rice
24(3):55-58. flour. J. Food Sci. 61:627-631.
Shih, F. F., and Daigle, K. 1997. Use of enzymes Tsai, M.-L., Li, C.-F., and Lii, C.-Y. 1997. Effects
for the separation of protein from rice flour. of granular structures on the pasting behaviors
Cereal Chem. 74:437-441. of starches. Cereal Chem. 74:750-757.
Shih, F., and Daigle, K. 1999. Oil uptake proper- Tsao, T., Beetner, G., Lorenz, K., and Frey, A.
ties of fried batters from rice flour. J. Agric. 1976. Extrusion processing of instant rice spa-
Food Chem. 47:1611-1615. ghetti. Lebensm. Wiss. Technol. 9:96-98.
Siegel, A., Bhumiratana, A., and Lineback, D. R. Umali, J. G. 1981. Physico-chemical properties of
1975. Development, acceptability, and nutri- milled rice in relation to quality characteristics
tional evaluation of high-protein soy-sup- of noodles (bihon). M.S. thesis, Univ. of
plemented rice noodles for Thai children. Cereal Philippines, Los Baños, Laguna.
Chem. 52:801-812. Van Veen, A. G. 1972. Fermented rice foods.
Singh, N., Bawa, A. S., and Sekhon, K. S. 1995. Pages 428-431 in: Rice: Chemistry and
Quality improvement of idli using extruded rice Technology, 1st ed. D. F. Houston, Ed. Am.
flour. J. Food Qual. 18:193-202. Assoc. Cereal Chem., St. Paul, MN.
Sleisenger, M. H., Rynbergen, H. J., Pert, J. H., and Wu, Y., and Woerman, J. H. 1997. French fry
Almy, T. P. 1957. A wheat, rye, and oat free formulations and method of making. U.S. patent
diet. J. Am. Diet Assoc. 33:1137-1140. 5,648,110.
Sloan, J. L., Middaugh, K. F., and Jacobsen, G. B. Yanase, H., Endo, I., and Chikubu, S. 1982a.
1991. Process for making a starch coated potato Studies on quality and processing suitability of
product. U.S. patent 5,059,435. glutinous rice. II. Processing suitability and
Spadaro, J. J., Mottern, H. H., and Gallo, A. S. sensory evaluation of “mochi” and “arare”
1971. Extrusion of rice with cottonseed and cakes made from various varieties of domestic
peanut flours. Cereal Sci. Today 16:238-240, glutinous rice with respect to storage period.
261. Shokuhin Sogo Kenkyusho Kenkyu Hokoku
Stucy Johnson, F. C. 1988. Utilization of Ameri- 39:1-14.
can-produced rice in muffins for glu- Yanase, H., Endo, I., and Chikubu, S. 1982b.
ten-sensitive individuals. Home Econ. Res. J. Studies on quality and processing suitability of
17:175-183. glutinous rice. IV. Characteristics indicating
Stucy Johnson, F. C. 1990. Characteristics of suitability and evaluation of quality of domestic
muffins containing various levels of waxy rice and imported glutinous rice. Shokuhin Sogo
flour. Cereal Chem. 67:114-118. Kenkyusho Kenkyu Hokoku 40:8-16.
Susheelamma, N. S., and Rao, M. V. L. 1979. Yanase, H., Endo, I., Shibuya, N., and Ohtsubo,
Functional role of the arabinogalactan of black K. 1982c. Product quality and raw materials
gram (Phaseolus mungo) in the texture of of puffed rice by batch-type and ex-
leavened foods (steamed puddings). J. Food Sci. truder-type methods. 1. Selection of raw
44:1309-1312, 1316. materials; glutinous and non-glutinous rice,
Takano, H., Yamagata, J., Hanaki, M., Koganagi, upland and lowland rice, and new and old
Y., and Tanaka, Y. 1979. Bread making rice. Shokuryo Kenkyusho Kenkyu Hokoku
properties of rice flours—Comparison between 41:39-49.
untreated raw rice flour and gelatinized rice Yeh, A.-I., and Li, J.-Y. 1996. A continuous
flour. Shokuryo Kenkyusho Kenkyu Hokoku measurement of swelling of rice starch during
34:35-43. heating. J. Cereal Sci. 28:277-283.
Tamang, J. P., Sarkar, P. K., and Hesseltine, C. W. Yeh, A.-I., Hsiu, W.-H., and Shen, J.-S. 1990. The
1988. Traditional fermented foods and bever- effects of additives on physical properties of
ages of Darjeeling and Sikkim—A review. J. extruded rice noodles. Mem. College Agric.,
Sci. Food Agric. 44:375-385. Natl. Taiwan Univ. 30:104-113.
Tani, T. 1985. Rice processing industries in Japan. Yeh, A.-I., Hsiu, W.-H., and Shen, J.-S. 1991.
In: FAO/UNDP Regional Workshop on Rice Some characteristics in extrusion-cooking of
Rice Flour / 539
rice noodles by a twin-screw extruder. J. Chin. of rice flour. Trans Inst. Chem. Eng. 77 (Part
Agric. Chem. Soc. 29:340-351. C):47-54.
Yeh, A.-I., Hsiu, W.-H., and Shen, J.-S. 1992. Ylimaki, G., Hawrysh, Z. J., Hardin, R. T., and
Moisture diffusion and gelatinization in Thomson, A. B. R. 1991. Response surface
extruded rice noodles. Pages 189-199 in: Food methodology in the development of rice flour
Extrusion Science and Technology. J. L. yeast breads: Sensory evaluation. J. Food Sci.
Kokini, C.-T. Ho, and M. V. Karwe, Eds. Mar- 56:751-755, 759.
cel Dekker, Inc., New York. Ziegler, E., and Greer, E. N. 1971. Principles of
Yeh, A.-I., Wu, T.-Q., and Jaw, Y.-M. 1999. milling. Pages 115-199 in: Wheat: Chemistry
Starch transitions and their influence on flow and Technology, 2nd ed. Y. Pomeranz, Ed. Am.
pattern during single-screw extrusion cooking Assoc. Cereal Chem., St. Paul, MN.
CHAPTER 18
RICE IN BREWING
Kiyoshi Yoshizawa
Tokyo University of Agriculture
Tokyo, Japan
Yutaka Ogawa
Kirin Beer Co. Ltd.
Yokohama, Japan
Rice is widely used in the world for the production of alcoholic beverages. Rice
is the main ingredient, often the sole cereal source, in such beverages as sake in
Japan, shaoshin-chu in China, and miscellaneous beverages in eastern Asia. Rice is
also used as an adjunct in the production of other alcoholic beverages, such as beer.
Beer and sake are the most popular rice beverages and are produced in huge
amounts, about 7 × 106 kL of beer and 106 kL of sake annually. The procedures for
making these two beverages, and the features of rice as an ingredient, are described
in this chapter.
RICE IN SAKE PRODUCTION
Sake, a traditional alcoholic beverage in Japan, is becoming popular in the rest of

the world. The principal raw materials used in sake brewing are rice and water. The
characteristic features of sake brewing are the use of koji—a culture of Aspergillus
oryzae on steamed rice grains—and highly milled rice as the main ingredients.
In 1998, about 41 × 104 tonnes (t) of rice was used for the production of 76 × 104
kL of sake in Japan.
Manufacture of Sake
RICE PROCESSING AND WATER

As shown in Figure 1, the first step in sake brewing is milling (polishing) of rice.
Brown rice, which contains large amounts of proteins, lipids, and minerals, is con-
sidered undesirable for sake brewing. To effectively remove these substances,
which exist mainly in the germ and surface part of the grain, brown rice is highly
milled to remove 25-30% of its weight. Sometimes the rice is milled to below 50%
of the milling ratio (the weight of milled rice divided by that of the original brown
rice times 100). Then the milled white rice is washed, steeped in water, and steamed
for 30-60 min.
541
The water for sake brewing should be colorless, tasteless, odorless, and neutral
or weakly alkaline. It must contain less than 0.02 ppm of iron; only traces of
ammonia, nitrate, and organic substances; and no harmful microorganisms. In gen-
eral, about 25 kL of water is used for 1 t of rice.
PREPARATION OF KOJI
Koji is a culture of A. oryzae that is grown on and within steamed rice grains and
accumulates various enzymes and substances useful for sake brewing. For preparing
koji, a seed-mold culture is made as follows. The A. oryzae strains are cultured on
steamed rice at 34-36°C for five to six days. This process results in abundant spore
formation on the molded rice grains. Aflatoxin-producing strains have not been
found among the Japanese industrial strains of koji mold.
Among about 50 kinds of enzymes found in koji, the most important are amy-
lases and proteases. a-Amylase decomposes and liquefies starch, and glucoamylase
forms glucose and thus regulates yeast growth and fermentation. Acid protease
decomposes proteins to form peptides and helps amylase action, as discussed below.
Carboxypeptidase forms amino acids. Cultural conditions influence the production
of enzymes by the molds. In general, the higher the cultivation temperature (up to
Fig. 1. Brewing process for sake.

Brewing / 543
42°C), the greater the amylase activity. Lower temperature favors development of
protease activity. As cultivation times become longer, especially at the late stage
of koji preparation, more enzymatic activities appear in the koji, as shown in
Figure 2.
The general procedure for making koji is as follows. Steamed rice, cooled to
about 35-40°C, is transferred into the incubation device or room where the tem-
perature and humidity are controlled at a level suitable for growth of the molds.
After inoculation with seed molds, the mixture is incubated for about 40-48 hr, with
occasional mixing, during a controlled temperature rise. Finally, the temperature of
the molded rice rises to 40-42°C, and white mycelia develop to cover and penetrate
into the grains. The resultant molded rice grains, koji, contain sufficient enzymes
and various nutritive substances (e.g., vitamins and lipids) for mashing and yeast
growth and fermentation.
SEED MASH
Seed mash is classified into two types according to the process by which it is
acidified: 1) seed mash acidified by naturally occurring lactic acid bacteria and 2)
seed mash with added lactic acid. In the former type, which has been produced tra-
ditionally, various aerobic bacteria, wild yeasts, lactic acid bacteria, and sake yeasts
grow successively in this order (Fig. 3). In contrast, in the latter type, sake yeast in
pure culture is inoculated at the first step in the seed-mash process. Sugar accumu-
lates to a high content (over 20%) at an early stage in both seed-mash processes.
This, together with the acidic condition, is considered to prevent contamination by
other microorganisms and to facilitate the predominant growth of sake yeast (about
2 × 108 cells per gram).
Fig. 2. Changes in a-amylase and acid protease activities in koji. " = a-Amylase activity, = acid
protease activity.
Fig. 3. Changes in microflora during preparation of the naturally acidified seed mash. A = Nitrate-
reducing bacteria, B = film-forming yeasts, C = wild yeasts, D = lactic acid bacteria (cocci), E = lactic
acid bacteria (rods), F = sake yeasts.
Sake yeasts are taxonomically classified as members of the Saccharomyces cere-

visiae group. However, from the practical point of view, sake yeasts can be differ-
entiated from other types of S. cerevisiae by such additional properties as vitamin
requirement, sugar tolerance, and alcohol tolerance.
The conventional procedure for preparing the lactic-acid-added seed mash,
which is usually used in sake brewing, is as follows. Koji and steamed rice are
mixed with water containing lactic acid (70 mL of lactic acid is usually added to
100 L of water), and a pure culture of sake yeast is inoculated to give 105–106 cells
per gram. The temperature of the mixture is about 20°C or somewhat higher. After
it has stood for a few days with intermittent stirring, the mash is warmed gradually.
Yeast grows and fermentation takes place. After 10-15 days, the seed mash has the
following composition: density about 6° Bp; alcohol, 12-15%; amino acids,
0.15-0.31%; and acids, 0.7-0.8%.
The procedure for making the lactic-acid-added seed mash has many variations;
some breweries use a large amount of compressed sake yeast cells harvested from
an aerobic propagation culture in place of the seed mash.
MAIN MASH FERMENTATION

The main mash is fermented without any special sterilization process. The
weight of white rice used for mashing in one lot is usually 2-7 t, but sometimes it is
more than 10 t. As shown in Table 1, stepwise mashing, covering three steps and an
optional step, is one of the characteristics of the main mash preparation. First,
steamed rice, koji, and water are added to the seed mash, the amount of added
materials being almost twice that of the seed mash. Thus, the acid content and yeast
population in the seed mash are diluted to about one-third, but they still inhibit wild
microorganisms. The temperature of the mixture is about 12°C. After two days, the
yeast population in the seed mash reaches about 108 cells per gram, which is the
same order of magnitude as in the seed mash; then the second addition of materials
(nearly twice the size of the first addition) is made. Since the yeast population and
Brewing / 545
TABLE 1
Proportionsa of Raw Materials Used in a Typical Mash
Total Rice Steamed Rice Koji Water
(kg) (kg) (kg) (L)
Seed mash 140 95 45 155
Additions
1st 280 200 80 250
2nd 530 405 125 635
3rd 890 715 175 1,260
4th (optional) 160 160 … 160
Total 2,000 1,575 425 2,460
a Weights of steamed rice and koji quoted are based on the weight of the original milled rice.
acid content decrease to about one-half, thereby decreasing the inhibition of con-
taminant microorganisms, the temperature is lowered to 9-10°C. The third addition
of materials is done the next day; the residual materials are added at 7-8°C. Yeast
grows, and the fermentation gradually becomes more vigorous. The mash tempera-
ture rises, reaching a maximum of 15-18°C on the sixth to ninth day after the final
addition of materials. The temperature is maintained for several days, after which it
decreases as the fermentation subsides. On the fifteenth to twentieth day after the
final addition of materials, when the alcohol concentration in the mash has reached
17-19% and the fermentation has almost ceased, pure alcohol (a 30-40% solution)
is often added to the mash to adjust the final alcohol concentration to about 20-22%
and give an adequate taste to the final product, sake. Quite often, a final addition of
steamed rice (7-17% of the total amount of rice used) or its materials decomposed
by amylase preparations is made when fermentation has almost ceased to increase
the sugar content of the sake.
In beer brewing, fermentation takes place after filtration of the mash. In contrast,
in the sake mash, sugars liberated from rice grains are fermented successively by
the yeast, and the content of fermentable sugars formed—mainly glucose—regu-
lates the fermentation by yeast. The concentration of fermentable sugars in the mash
reaches a maximum of 7-8% at an early stage and then decreases gradually. This
rather small sugar content—along with the solid matrix of the mash, the proteolipids
contained in koji, and the high alcohol tolerance of sake yeast—are considered to
contribute to the high alcohol production, which can be up to 20%.
FINISHING PROCEDURE
Immediately, or one day, after the addition of alcohol to the mash, the mash is
squeezed to divide it into the filtrate (sake) and the solids, which are used for
making various foodstuffs such as pickles. In general, 2.4 kL of sake, containing
20% alcohol and 200-250 kg of solids, is obtained from 1 t of white rice. The
slightly turbid filtrate is clarified and usually pasteurized (at about 65°C) to kill
yeast and harmful microorganisms (if present), inactivate enzymes, and adjust the
maturation velocity. It is then stored. During storage, sake gradually matures,
deepens color, and becomes mild in taste. In general, maturation takes about three
to eight months. After storage, sake is blended, diluted with water to the appropriate
alcohol content, usually 14-16%, and treated with activated carbon to control its
flavor and taste and adjust its color and clarity. Then the sake is bottled and
pasteurized. Table 2 shows the composition of a typical sake.
TABLE 2
Composition of a Typical Sake
Component Content
Ethanol, %, v/v 15.9
Nonvolatile matters, %, w/v 5.9
Direct reducing sugars, %, w/v 3.49
Acidity, meq/100 mL 1.4
Formol nitrogen, %, w/v 0.0256
Absorbance (430 nm) 0.0146
Rice as a Principal Ingredient
Rice is a principal raw material of sake, and the effect of its physicochemical
properties on the yield and quality of sake has been studied by many research
workers. Whereas the aroma of grapes largely contributes the flavor of wine, rice
has neither a characteristic flavor nor a profound effect on the flavor of sake, which
is affected, instead, by the complicated procedure of sake brewing. Consequently,
rice properties such as ease of processing (e.g., milling and steaming), its high yield
of sugars when its starch is saccharified, and its ability to give well-molded koji
have been considered important in estimating its quality.
Over the last 30 years, many studies have been done on structure and con-
stituents of rice grains as they relate to saccharification, yeast growth, fermen-
tation, and formation of flavor and taste substances in the mash. These results
have revealed that the quality of rice as an ingredient of sake can be estimated
from several physicochemical properties, such as protein content, water absorb-
ability of grains, the weight of 1,000 kernels, moisture content, and occurrence of
white haze (white core) in the center of grains, which are discussed in more detail
below.
Japonica short-grain varieties harvested in Japan have been found to be suitable
for sake brewing. Long-grain varieties form less sugar in saccharification and add
odor to the sake (Kawasaki et al, 1967; Kumagai et al, 1968). Among the many
domestic rice varieties, such varieties as Yamadanishiki are considered specially
suitable for sake brewing and are sold at a high price.
Nature of Rice for Sake Brewing
STRUCTURE OF THE RICE GRAIN

Figure 4 shows scanning electron micrographs of the cross sections of rice
grains used for sake brewing. Yamadanishiki, the most famous rice variety
suitable for sake brewing, shows a radial arrangement of uniformly well-
developed endosperm cells. In contrast is Nihonbare, a popular small-grain
variety in Japan, in which the back side of the grain develops well and makes a
tighter structure than the belly side (Yoshizawa et al, 1978a). Grains such as
Nihonbare are sometimes steamed unequally and thus produce less sugar and
more amino acid in the mash fermentation (Yoshizawa et al, 1979a). In the grain
of Yamadanishiki, endosperm cells are arranged loosely, and the center part is
filled with naked starch granules instead of endosperm cells. This structure
scatters light and causes white core, shinpaku, which is often seen at the center of
large grains (Yoshizawa et al, 1978a; Aramaki et al, 1995). These rice grains
Brewing / 547
Fig. 4. Scanning electron micrographs of the cross sections of Nihonbare (left) and Yamadanishiki
(right) rice grains.
absorb water quickly, steam well, and form a large amount of sugar upon sac-
charification by amylase (Yoshizawa et al, 1979a; Yanagiuchi et al, 1996). Since
the mycelia of A. oryzae grow on the surface of the white-core cavity as well as on
the surface of the grains, these varieties are considered especially suitable as a raw
material for koji (Iemura and Fujita, 1982b).
Many protein granules exist around and within endosperm cells, making up
about 85% of the weight of total proteins contained in the white rice (75%
polishing ratio) (Yoshizawa et al, 1980). Protein granules are divided into two
types, one rich in prolamin (Pb I, less than 30%) and another one rich in glutelin
(Pb II, about 60%) (Kizaki et al, 1991). The latter is easily decomposed into
peptides and amino acids by koji protease. In mash fermentation, a large amount
of a-amylase is absorbed along with protein granules (Shiinoki and Nunokawa,
1980). Acid protease helps saccharification by decomposing protein granules that
have the ability to absorb a-amylase (Shiinoki and Nunokawa, 1980; Yoshizawa
et al, 1980). The steaming process denatures proteins so that they are hardly
decomposed by proteases.
PHYSICOCHEMICAL PROPERTIES
Thousand-grain weight. The weight of 1,000 grains of brown rice, which is a
measure of the size of rice grains, is between 20 and 30 g in Japanese short-grain
varieties. Among them, those weighing over 25 g are generally considered espe-
cially suitable for sake brewing. High correlations exist among such important
properties as the 1,000-grain weight, water absorbability, and others, as mentioned
above.
Specific gravity. The specific gravity of brown rice is between 1.36 and 1.44.
Rice having a white core at the center of grains has a smaller value (Momose,
1979).
Moisture content. In general, the moisture content of brown rice is 14-15%.
Brown rice harvested in the northern districts of Japan often contains more than
15%. During milling, grains with high moisture content often crack and thus reduce
the milling recovery. Grains with low moisture content absorb much water in the
washing and steeping processes and thus contain more water after steaming. The
following equation shows the relationship between the moisture content and the
amount of water absorbed in rice during washing and steeping:
Y = 2.8X + 66
where Y is the percentage of water absorbed in steeped rice and X is the percentage
of moisture content of the rice (Kumagai et al, 1976).
Since it is very important to obtain steamed rice of an appropriate moisture con-
tent to control the processes of sake brewing, and consequently make sake of high
quality, various devices are applied to give the rice grains a suitable moisture con-
tent before they are washed.
COMPOSITION
Carbohydrates. As shown in Table 3, starch is the largest constituent of rice and
accounts for almost all of its carbohydrates. Other minor constituents, such as
sucrose, fructose, and glucose, are found in trace amounts in rice grains (Momose,
1979). The endosperm cell wall of large-grain varieties such as Yamadanishiki is
rich in cellulose as compared with that of small-grain varieties, which contain more
hemicellulose, as shown in Table 4.
In mash fermentation, the amylases in koji decompose starch to form sugars,
mainly glucose, which are successively turned to ethanol by yeast.
Proteins. The protein content (N × 5.95) in brown rice varies from 6 to 8% and is
affected by cultural conditions such as the amount of fertilizer used and changes in
the temperature, as well as by the rice variety. As shown in Figure 5, glutelin is the
main component; the ratio of its content to that of the other proteins increases in the
TABLE 3
Properties of Two Varieties of Brown Rice Grains
Used for Sake Brewing
Yamadanishiki Nihonbare
Weight of 1,000 grains, g 28.01 21.85
Moisture, % 14.8 15.0
Starch, % 70.88 69.53
Total nitrogen, % 1.10 1.34
Crude fat, % 2.28 1.95
Ash, % 1.00 1.06
TABLE 4
Composition (%, w/w) of the Endosperm Cell Wall Fractiona
of Milled Rice Grainsb,c
Cellulose Hemicellulose
Variety Fraction Fraction
Nihonbare 33.7 63.3
Todorokiwase 37.4 59.0
Gohyakumangokud 46.4 50.9
Yamadanishikid 42.4 48.6
a Proteins removed.
b Polishing ratio, 75%.
c Data from Yoshizawa et al (1980).
d Varieties especially suitable for sake brewing.
Brewing / 549
inner part of kernels. Milling reduces the protein content of rice gradually. Rice
milled to 75% of the milling ratio usually contains 5-6% protein. In mash fermen-
tation, acid protease in koji decomposes proteins to form peptides, which are further
decomposed into amino acids by acid carboxypeptidase. Some of the amino acids
are assimilated by yeast and give sake its full and heavy taste. But an excess of
these compounds often gives sake a rough taste, deepens the color, and accelerates
deterioration of its quality. Some amino acids such as leucine, valine, and isoleucine
are assimilated by yeast to form higher alcohols such as isoamyl alcohol, isobutanol,
and 2-methyl butanol, respectively. Among them, isoamyl alcohol is further reacted
with acetyl coenzyme A to form isoamyl acetate, one of the main flavor compo-
nents contributing to the aroma of sake.
As mentioned previously, white rice of high protein content is likely to form less
sugar and more amino acid in mash fermentation. Because consumers prefer sake
with a clear and light taste, various efforts have been made to decrease the protein
content of steamed rice as well as to denature proteins, so that they are minimally
decomposed by proteases in koji during such treatments as steaming for a long time
or at a slightly higher pressure.
Lipids. Brown rice contains about 2% crude fat (extractable with ethyl ether) and
about 0.7% of fat-by-hydrolysis, composed of phospholipids, proteolipids, and other
lipids. Because the crude fat exists mainly in the germ and aleurone layer, milling
Fig. 5. Changes in the composition of proteins in rice grains during milling. s = Glutelin, " = albumin,
! = globulin, r = prolamin, × = amino acids. (Data from Kobuyama et al, 1969)
Fig. 6. Changes in lipid content of rice grains with changes in milling ratio. " = Crude fat, ! = fat-by-
hydrolysis. (Data from Yoshizawa et al, 1973c)
reduces it greatly to below 0.1% in overmilled white rice of 75% milling ratio.
However, fat-by-hydrolysis scarcely decreases, as shown in Figure 6.
Crude fat is composed mainly of fatty acids and their glycerides, of which two
unsaturated fatty acids (linoleic and oleic) and one saturated fatty acid (palmitic
acid) are the main components. With a decreased milling ratio (percentage of the
weight of white rice to that of brown rice), oleic acid decreases more rapidly than
the other two fatty acids, and thus the ratio of the amount of saturated fatty acids to
those of unsaturated ones increases. During the steaming process, some glycerides
decompose to liberated fatty acids, and free fatty acids evaporate and may partly
decompose. Consequently, steamed rice contains about half the amount of crude fat
that it contained before steaming (Ishikawa and Yoshizawa, 1974).
Fatty acids in the medium are rapidly incorporated into yeast cells. These acids
largely affect formation by yeast of aroma esters such as isoamyl acetate during
fermentation (Table 5) (Yoshizawa, 1976). Unsaturated fatty acids incorporated into
yeast cells are located mainly in the cell membrane (Ishikawa and Yoshizawa,
1979), and they suppress isoamyl acetate formation by alcohol-acetyltransferase in
yeast (Ishikawa and Yoshizawa, 1979, 1984; Yoshioka and Hashimoto, 1983).
Under the anaerobic conditions of mash fermentation, yeast usually incorporates
fatty acids into the cell instead of synthesizing them. Consequently, the composition
of fatty acids in the yeast cells and the ratio of the amount of saturated fatty acids to
Brewing / 551
TABLE 5
Ester Formation by Sake Yeast in the Fatty-Acids-Added Mediuma,b
Added Fatty Acid Ethanol Isoamyl Alcohol Ethyl Acetate Isoamyl Acetate
(0.5 mM) (%) (ppm) (ppm) (ppm)
Blank 4.0 141 9.8 1.7
Stearic acid 4.5 145 11.6 2.1
Oleic acid 3.9 125 6.4 0.8
a Sake yeast Kyokai 7 was inoculated (inoculum size, 106 cells per milliliter) in the medium containing
glucose, hydrolyzed casein amino acids, and albumin as an emulsifier and was incubated at 25°C for
three days.
b Data from Yoshizawa (1976).
TABLE 6
Changes in the Amounts of Minerals in Rice Grains with Changes in Polishing Ratioa
Milling Ratio (%)
100b 90 70 50
Ash, % 1.11 0.59 0.20 0.16
Phosphorus, ppm 2,570 1,272 600 432
Potassium, ppm 1,380 600 360 256
Magnesium, ppm 310 104 56 22
Calcium, ppm 135 58 30 23
Iron, ppm 10.2 1.6 0.7 0.7
a Data from Totsuka (1979).
b Brown rice.
that of unsaturated ones reflect the composition and the ratio of fatty acids in the
mash, of which steamed rice and koji are the sole origin. Thus, milling and steaming
procedures make a big contribution to the quality of sake.
Inorganic substances. Brown rice contains about 1% inorganic substances, which
consist mainly of phosphate, potassium, and magnesium. Since these substances
exist in large amounts in the germ and the aleurone layer, they are largely removed
during milling, as shown in Table 6. Because potassium is eluted from grains when
rice is steeped in water, its content in steamed rice is affected by the conditions of
washing and steeping (e.g., the amount of water used and the treating times) as well
as by the potassium content in the white rice used. During storage of rice, some
inorganic substances such as potassium move from the surface of the kernel to the
inner part. Consequently, white rice stored for a long time before milling contains
more potassium than does rice milled before storage (Kubo, 1960; Yoshizawa et al,
1973a).
These three inorganic substances promote the growth of molds and yeasts and
accelerate fermentation; thus, excess amounts of them often negatively affect the
quality of sake.
Estimation of Quality of Rice as an Ingredient
Many studies have been done to determine the qualities of rice suitable for sake
brewing. The results of various experiments (including statistical studies such as
principal component analysis and cluster analysis) show that such properties of
white rice as milling ratio, the weight of 1,000 grains, protein content, potassium
content, and water absorbability contribute greatly to the quality of the sake produced
(Yoshizawa et al, 1973b; Hanamoto et al, 1978; Iemura and Fujita, 1982a). Using
these properties, many rice varieties used for sake brewing were classified into eight
groups (Working Group on Rice for Sake Brewing, 1993). Consequently, several
properties have been selected as being the most important in estimating the quality of
white rice as an ingredient of sake brewing; they are also easy to analyze. These
properties are the weight of 1,000 grains; the contents of moisture, protein, and
potassium; water absorbability in the steeping of rice in water; and the amount of sugars
and amino acids formed in the saccharification of steamed rice by specific enzyme
preparation. Although several other properties, such as lipid content, are important, the
analytical methods are too complicated for routine analysis. To estimate the quality of
white rice, the following equations were proposed (Iemura and Fujita, 1982a,b).
For rice used for mashing,
[(x1 - x1 ) / d1 ] - [(x 2 - x 2 ) / d 2 ]
For rice used for koji making,
[(x1 - x1 ) / d1 ] - [(x 2 - x 2 ) / d 2 ] + [(x 3 - x 3 ) / d 3 ]
where x1 is the weight of 1,000 grains, x1 its mean value, and d1 its standard devia-
tion; x2 is the protein content (N × 5.95), x 2 its mean value, and d2 its standard
deviation; and x3 is the surface area of the grain, including white haze at the center,
x 3 its mean value, and d3 its standard deviation.
Rice Processing
MILLING
In contrast to the use of the whole malted grain in brewing and spirits produc-
tion, brown rice is highly milled before being used in sake brewing. As mentioned
previously, the main purpose of milling is to remove proteins, lipids, and inorganic
substances such as potassium and phosphate, which are in excess in the germ and
the surface part of rice grains and are considered undesirable in sake brewing.
The ratio (percentage by weight) of milled rice to the original brown rice is
called the “milling ratio.” Figure 7 shows changes in the amounts of constituents of
the milled grains with various milling ratios (Yoshizawa and Ishikawa, 1974b). The
contents of crude fat and ash decrease most rapidly, whereas protein content
decreases gradually until the milling ratio reaches 50%, after which it remains prac-
tically constant. But the composition of the proteins changes with various milling
ratios, i.e., the percentage of glutelin, the main constituent, increases while the
others decrease, as mentioned previously. Usually, rice of 65-70% milling ratio is
used for sake brewing. The starch content increases gradually with decreasing
milling ratios.
A type of abrasive roller mill used for milling rice grains is shown in Figure 8.
The roller, made of Carborundum and feldspar, rotates around a vertical axis and
scrapes the surface of the grains. Rice grains, fed from the hopper, are abraded in
the milling chamber and fall to the bottom through the sieve, where rice bran and
germ are removed. The basket conveyor carries rice grains from the bottom to the
hopper. Thus the operation continues until the grains are milled to the required ratio.
The milling efficiency and the shape of the milled rice grains are affected by rota-
Brewing / 553
Fig. 7. Changes in the content of various substances in rice grains with changes in milling ratio. A =
Starch, B = moisture, C = proteins, D = minerals, E = lipids. (Reprinted, with permission, from
Yoshizawa and Ishikawa, 1974b)
Fig. 8. Vertical type of rice mill used in sake brewing. A = Basket conveyor, B = rice hopper, C = rice-
flow adjustment valve, D = milling chamber, E = roller, F = resistance, G = exit, H = sieve, I = bran
reservoir.
tion speeds, resistance at the exit, and the particle size of the Carborundum in the
roller. The speed of milling decreases gradually with a decrease in milling ratio.
The properties of rice, such as the weight of 1,000 grains and the moisture
content, also affect the speed, the milling efficiency, and the shape of the milled
rice grains. With a mill having a roller 56 cm in diameter, the average times for
milling 1.8 t of brown rice are 6 hr for a 75% milling ratio, 10.5 hr for 70%, and
16 hr for 60%.
WASHING AND STEEPING

Rice is washed and steeped in water before steaming. During washing, rice
grains are subjected to a kind of polishing caused mainly from collision of grains in
water. This process removes some of the surface parts of the grain (about 1% of the
total grain weight), which make up most of the suspended solids and account for
most of the biochemical oxygen demand in the waste discharged from sake
breweries (Yoshizawa et al, 1972). To reduce suspended solids and biochemical
oxygen demand, several breweries steep rice without washing it.
During washing, the grains absorb water up to 9-17% of their weights. Among
the various types of washing equipment used (including washing rice grains with air
bubbles in water), those that transport rice grains with water in continuous washing
through a long tube are applied widely.
Washed rice grains are passed into a vat and immediately steeped in water.
During washing and steeping, the grains absorb water to about 25-30% of their
weights, which promotes penetration of heat into the grains easily when they
are steamed and accelerates modification of starch granules and proteins as
Fig. 9. Scanning electron micrograph of the cross section of a steamed rice grain. The endosperm cell
wall is partially broken, and swollen starch granules can be seen.
Brewing / 555
well as destruction of endosperm cell walls (Fig. 9). Absorption of an

appropriate amount of water is very important to the proper and equal steaming
of rice grains, and consequently it greatly affects koji making and mash
fermentation. The velocity of water absorption differs with rice varieties and
milling ratios (Akai, 1963; Saijo et al, 1968). In general, varieties suitable for
sake brewing, such as Yamadanishiki and overmilled rice, absorb water
quickly. The maximum amount of absorbed water is strongly affected by the
moisture content of the rice, as mentioned previously. Usually it takes 3-20 hr
to steep rice. After steeping, the excess water is drained off from the grains for
about 4-8 hr before steaming.
During washing and steeping, sugars and some minerals, such as potassium ions,
are eluted from the grains (Noshiro and Aoki, 1957; Kanoh, 1962; Yoshizawa and
Ishikawa, 1974a), whereas calcium and iron are absorbed into the grains (Horie et
al, 1965; Takase and Murakami, 1965).
STEAMING
During steaming, starch in rice grains changes to the a-form, and protein is
denatured, becoming resistant to enzyme action by koji. At the same time, the
structure of the rice grain, such as the endosperm cell wall, is partly destroyed (Fig.
9) (Yoshizawa et al, 1978b) and the grains are sterilized. It has been reported that
steaming for as short a time as 15-20 min is sufficient to modify the starch and
protein of rice grains that have absorbed sufficient water (Akiyama and Takase,
1958; Akiyama and Yamamoto, 1963); however, it takes more than 45 min to
remove fatty acids effectively from the grains (Yoshizawa et al, 1979b). Addition-
ally, steaming rice for a short time may produce steamed grains with an unmodified
part remaining if the grains absorb insufficient water before steaming (Yoshizawa et
al, 1979a). Consequently, to obtain completely steamed rice, the grains are usually
steamed for 30-60 min.
During steaming, the grains absorb water to about 10% of the weight of the
starting white rice, thus resulting in a total gain of about 35-40% from the begin-
ning of brewing.
In small breweries, steeped rice grains are heaped up in an insulated shallow
tub with a specially designed steam distributor at the center of the base, through
which steam is blown. This is placed on top of a large kettle filled with water,
and steaming takes place for about 30-60 min. The tub is about 160-190 cm in
diameter and 80-90 cm in depth. In many breweries, two types of continuous-
steaming apparatus are adopted, namely a conveyer belt system (Imayasu et al,
1963) and a cylinder system. In the former, steeped rice grains are carried on the
screened conveyer belt and are steamed for about 30 min, with pressure-
controlled steam blown out from steam nozzles placed under the belt. In the latter
system, steeped rice grains pass down from the hopper into the cylindrical vessel
and are treated with steam that comes from nozzles placed at the lower part of the
cylinder.
After steaming, the grain is slightly sticky, somewhat translucent, and strongly
resistant to being crushed between the fingers. The steamed rice is cooled to nearly
40°C for making koji and to about 10°C for preparing mash. The cooling machine
cools steamed rice as it moves on the screened conveyer belt by employing a draft
of air. Then the grains are transferred to the mash tank and the koji-making facili-
ties, often through a pneumatic conveyer.
Alternatives in Rice Processing
STEEPING OF RICE IN WATER WITH ENZYMES

Many studies have been done on the use of enzymes in rice steeping to improve
the quality of steeped rice grains for sake brewing. During steeping, lipase added to
the water decomposes glycerides in rice to liberate fatty acids, which decrease
greatly during steaming, mainly by evaporation. This process is effective in making
sake with light taste and high aroma (Yoshizawa et al, 1976) and has been adopted
by many breweries.
A limited amount of amylase preparation is often used with koji in mash fer-
mentation to supplement saccharification by koji amylases.
DEHYDRATED STEAMED RICE (ALFA-RICE) AND PUFFED RICE

Steamed rice grains deteriorate rather quickly and become resistant to the
enzyme action of koji. The degree of deterioration varies with changes in the mois-
ture content of the steamed rice (Yoshizawa et al, 1974); dehydration of rice grains
immediately after steaming stops deterioration. The dehydrated rice can be stored
for a long time and can be used for mashing without any heat treatment such as
steaming. Various dehydration processes have been studied (Koura et al, 1964;
Nagatani and Nunokawa, 1975), but the dehydration of steamed rice with alcohol
(Kawasaki and Nagatani, 1966; Nakai et al, 1979) is applied in practice. This pro-
cedure removes lipids from the processed rice grains and is thought to reduce dimethyl
sulfide, which gives an off-flavor to the sake. Treatment of rice with methyl bromide,
an insecticide, causes reaction of this reagent with rice proteins to combine methyl
residue with proteins (Kitamoto et al, 1981a; Maekawa et al, 1981). In mash fermen-
tation, these methyl-substituted proteins are decomposed to form methyl methionine,
which divides spontaneously into homoserine and dimethyl sulfide while sake is aged
(Kitamoto et al, 1981b). The sake made from dehydrated rice contains little amino
acids and tastes light (Ohmori et al, 1977; T. Sato et al, 1977).
Processing raw rice grains under high pressure and high-temperature airflow for
a short time (puffing) dehydrates grains, gelatinizes starch, and denatures protein,
leaving it unsusceptible to the protease action of koji (Takahashi et al, 1987;
Takayama, 1992). Puffed rice is used in practice for sake brewing.
SACCHARIFICATION OF RAW RICE

Sake can be produced from raw rice grains with a fairly high savings in the
energy required for steaming (Tanaka et al, 1981). Saccharification requires much
enzyme activity, specifically more than five times the glucoamylase activity exist-
ing in ordinary mash. Since such molds as Aspergillus niger var. awamori and
Rhizopus javanicus grow rapidly on raw rice grains as well as on steamed ones (Fig.
10), koji can be prepared using unsteamed rice grains.
USE OF LIQUEFIED RICE AND RICE BRAN

Addition of liquefied rice bran to the mash as an adjunct is widely adopted by
sake breweries (Ohashi et al, 1978). Rice bran suspended in warm water is
saccharified by enzyme preparations, and the liquid fraction that is separated
from the saccharified mixture by filtration is treated with ion exchangers and
activated carbon, successively, to make a sugar solution. Liquefied rice bran is
added to the mash to adjust the sugar content in sake and increase its yield.
Brewing / 557
Fig. 10. Growth of two mold strains on cooked and uncooked rice grains. " = Rhyzopus javanicus RIB
r
5501 on raw rice grains, ! = same strain incubated on cooked rice grains, = Aspergillus oryzae RIB
128 incubated on raw rice grains, s = same strain incubated on cooked rice grains. (Data from Tanaka
et al, 1981)
Steamed rice is also liquefied by enzyme preparations. The liquefied mixture is

used for mash preparation (Imayasu, 1993).
BREEDING OF NEW RICE VARIETIES

Many efforts have been made to breed new rice varieties suitable for sake brew-
ing. Various rice varieties, including Yamadanishiki and Gohyakumangoku, have
been bred and used widely in many sake breweries. These varieties are desired for
such properties as large grain size and low protein content, together with high crop
yield and strong resistance to diseases and insect damage.
RICE IN BEER PRODUCTION
Raw Materials for Beer Production
The raw materials for beer brewing are malt, hops, water, and adjuncts such as
rice, corn, sorghum, and other kinds of starches and sugars.
Fig. 11. Beer production process. I = mashing stage, II = weighing and grinding of malt, III = prep-
aration of adjunct, IV = mashing stage, V = filtering through lauter tub, VI = boiling of wort, VII =
removal of residues and trub in whirlpool tank, VIII = cooling to yeast-pitching temperature, IX =
fermentation, X = blending and storage.
Figure 11 shows the major stages in beer production and the steps within each
stage.
INGREDIENTS FOR BEER BREWING

The essential ingredient for beer brewing is malt, which is the primary source of
yeast nutrients like carbohydrates and protein, and, at the same time, the sole
enzyme source for mashing. Worldwide, malts are mainly made of barley, with a
Brewing / 559
very small amount from wheat. Two major types of malting barley are used in the
world: two-row, which is mainly used in Europe, Australia, and Asia, and six-row,
used in North and South America.
Adjuncts (nonmalt starch or sugar material) are widely used in many countries
except in Germany, where the Reinheitsgebot or purity law is in effect for beer
brewing for domestic consumption.
The ratio of malt to adjunct must fall within a permissible range if a legal regu-
lation regarding it is in effect. Adjunct ratios are set by each brewer according to the
quality specifications for the final beer and the cost estimation.
MALT
The objectives of malting are 1) to obtain a sufficient amount of enzymes, and 2)
to modify reserved food materials in a grain by controlling barley germination.
Thus, the malt and adjunct substances can be decomposed by enzymes into yeast
nutrients during the mashing process that follows malting.
The malting process has three stages: steeping, germination, and kilning. In the
steeping stage, cleaned, graded barley is steeped in water or subjected to a repeated
combination of steeping, successive draining off, and aeration procedures until the
proper moisture content is obtained to start germination. In the germination stage,
the steeped barley is germinated under controlled conditions; steps are taken to
minimize weight losses, which are caused mainly by respiration. The kilning stage
takes place when the acrospire (plumule) of the germinated barley has reached the
length of two-thirds to three-fourths of the grain. The barley is dried by hot air to
stop both its growth and the enzyme activity and to make the malt—with its char-
acteristic color and aroma—storable.
During germination, the cell walls in endosperm cells are dissolved mainly by
hemicellulases so that various hydrolytic reactions take place. These reactions make
the contents of the endosperm more friable and susceptible to enzymatic hydrolysis
at mashing. The main hydrolytic enzymes formed during germination are amylases,
proteolytic enzymes, hemicellulases, and phosphatases. During the kilning stage, a
reaction between amino acids and sugars (the Maillard reaction) causes nonenzy-
matic browning, in which colored materials called melanoidins are formed.
ADJUNCTS
Adjuncts (usually 20-40% of the total ingredients) are used both to obtain fer-
mentable sugars at a lower cost and because they play a very important and specific
role in giving a characteristic quality to the beer to attract consumers. All-malt beer
is richer in flavor and has more fullness of body and darker color. On the other
hand, beer with adjuncts has a lighter color and taste, so that consumers drink more
because of its less-satiating taste. Also, the beer with adjuncts has higher colloidal
stability due to its lower nitrogen content.
The use of alternative adjuncts has been driven to some extent by the desire to
alter the characteristics of a particular beer—as in the case of wheat malt for Weiss
beer or rice and maize grits to give lighter flavors and/or haze stability—but to a
much larger extent by economic demand (Little, 1994).
Adjuncts in processed form suitable for brewing are classified into two types:
mash adjunct, which is added to the adjunct cooker or the main mash tub, and kettle
adjunct, which is added to the wort kettle. The mash adjuncts are starch itself, either
in its native form or a pregelatinized form; they must be saccharified during the
mashing process. The kettle adjuncts are already converted into sugar form by acids
or enzymes. In the United States, Canada, and Japan, these two types of adjuncts are
derived mainly from corn and rice.
Since each adjunct imparts definite or delicate characteristic tastes to the final
beer, the choice of adjuncts and the relative proportion of each are very important in
relation to beer quality and cost. Because of the subtle differences in beer flavor and
consumer preferences, some major breweries in the United States and Japan prefer
rice as a major adjunct, regardless of its higher cost.
Among several analytical criteria of adjuncts (Vogel, 1950; Coors, 1976; Bradee,
1977), the contents of moisture, carbohydrates, and lipids are most important (Table
7). The most preferable adjunct should have more carbohydrate, which is easily
gelatinized and converted into fermentable sugars, and less lipid. Lipids in an
adjunct may produce rancidity, which will give an unpleasant odor to the wort or
beer and cause starch deterioration.
Manufacture of Beer
As shown in Figure 11, the first stage of beer brewing proper is mashing.
Through a programmed temperature cycle in the mashing stage, the substances in
the milled malt and adjuncts are dissolved or transformed by malt enzymes into
low molecular weight compounds, the so-called “extract,” which yeast can utilize
for fermentation. The two basic mashing methods are infusion and decoction mash-
ing. In the former, which is employed mainly in the United Kingdom and the United
States, the mixed grist and liquor are held in the mash tub and the temperature is
gradually raised without boiling. In the latter, which is employed mainly in Ger-
many and Japan, a part of the mash in the tub is transferred to the mash kettle,
boiled, and then returned to the mash tub to raise the temperature of the whole mash
(De Clerck, 1957).
Before being combined with the main malt mash, cereal adjuncts containing
starch in its native form must be boiled in the adjunct cooker under atmospheric
pressure or pressurized to 0.5-1.0 bar gauge. This causes the starch granules to
TABLE 7
Typical Analysis of Brewing Adjunctsa
Corn Corn Refined Wheat Torrifle
Rice Grits Syrup Starch Starch Barley
Moisture, % 11.0 11.9 … 11.0 11.4 6.0
Extract, % 81.7 79.3 83.4b 92.9 95.1 67.9
Extract dry basis, % 92.7 90.0 … 103.1 105.2 72.2
Total nitrogen, % 0.86 1.47 … 0.06 0.03 2.16
Soluble nitrogen, %c 0.02–0.05 0.10–0.13 … … … …
Lipids, % 0.2 0.8 … 0.04 0.4 1.5
Ash, % 0.3–0.7c 0.3–0.8c 0.19b … … …
Crude fiber, %c 0.3–0.7 0.4–0.8 … … … …
pH 6.4 5.8 5.0b 5.0 5.7 5.9
Gelatinization
temperature, °C 61–78 62–74 … 62–74 52–64 …
a Data from Coors (1976).
b Data from Bradee (1977).
c Data from Vogel (1950).
Brewing / 561
swell, gelatinize, and partially liquefy, which renders them susceptible to attack by
malt diastatic enzymes during mashing. A small amount of ground malt (5-10% of
the total adjuncts) is usually added to the cooker mash to make the starch granules
more fluid (Yoshioka, 1980). The cooker mash is transferred to the mash tub and
mixed with the main malt mash. Thereafter, the combined malt-adjunct mash
undergoes either infusion or decoction mashing. The major degraded products of
mashing are fermentable sugars, oligosaccharides, amino acids, oligopeptides, and
minerals. Among all products, maltose makes up the highest portion, 75-80% of the
total dissolved constituents of the mash. After completion of the mashing (i.e., con-
version of starch), the mash is filtered through the lauter tub to separate wort from
spent grains, which consist mainly of malt hull. The clear wort is boiled with
pressed or processed hops and sometimes with kettle adjuncts in the wort kettle.
The hopped wort is pumped to the whirlpool tank to remove hop residues and
coagulated proteins (trub).
Then the wort is cooled, inoculated with yeast, and transferred to the fermenters,
where the primary fermentation takes place for several days. The fermented wort
(young beer) is transferred into storage tanks and allowed to mature for several weeks.
The matured beer is chilled-proofed, carbonated, and filtered to remove yeast
and turbidity-forming materials. The filtered beer is transferred to the packaging
process.
Although the major products of the fermentation are ethanol and carbon dioxide,
hundreds of minor products are also produced during the fermentation, and these
contribute greatly to the characteristic flavor of beer.
Rice as an Adjunct
Rice is preferred by some brewers because of its lower protein and lipid contents
as compared with those of corn grits. Rice has a neutral aroma and flavor and, when
converted properly, yields a light, clean-tasting beer (Coors, 1976).
Rice is huskless, its aleurone layer is one cell thick, and it responds to gibberellic
acid. Moreover, starch in rice has a complex structure and gelatinization tempera-
ture of 70-80°C. Rice grits and flakes are low in b-D-glucans, pentosans, proteins,
and lipids (Little, 1994).
Rice is the second-most-used adjunct in the United States (Canales, 1979) and in
Japan.
MILLING OF RICE
Rice is usually used for brewing in the form of rice grits or broken rice, which is
obtained as a by-product of rice harvesting and the processing of edible rice. The
rice used for beer brewing consists of almost-pure fragments of endosperm that
contain starch exclusively.
In the U.S. rice-polishing (milling) industry, hull is removed from rough rice,
and the brown rice obtained is then dry-milled to remove the bran-containing aleu-
rone layer and germ. The resulting polished rice (both whole and broken pieces) is
called “mill run.” The objective of milling is to remove the aleurone and germ com-
pletely without causing any damage to the endosperm, which contains starch, so that
the whole grain can be used for edible purposes. However, up to 30% of the grain is
broken during milling. The broken pieces are considered undesirable for home use
due to their appearance and are sold to brewers at a considerably lower price. The
broken pieces are further classified as second heads, brewers’ rice, and screenings.
Brewers’ rice, second heads, and mill run are equally acceptable and differ only in
particle size (Coors, 1976).
In Japan, unlike in the United States, broken rice for beer, miso, and Japanese
rice-cracker production is polished in a line separated from the edible-rice polishing.
The raw materials for the broken rice are largely limited to the broken pieces of
rough rice and brown rice gathered at harvest. The broken pieces are cleaned,
graded, and milled to remove the hull and bran.
QUALITY REQUIREMENTS FOR BREWERS’ RICE

Brewers’ rice should be free from contaminants, such as foreign seeds; soil;
sand; unpleasant odors, among them musty and/or rotten smells; and any trace of
bran residues that are rich in lipids. In addition, brewers’ rice is checked for con-
tamination with insects and/or molds, because the presence of these indicates
improper handling and/or storage, which may give lower quality to the beer.
The particle-size distribution must be within the permissible range set by each
brewery. Breweries are often advised to install a rice mill that can grind rice grits to
a uniform, small granule size. A typical result of screening analysis after grinding is
2% on no. 14 (>1.40 mm), 34% on no. 18 (1.01-1.40 mm), 43% on no. 30
(0.61-1.00), and 21% fines (<0.60) (Bradee, 1977).
As is the case with sake brewing, long-grain U.S. rice is seldom used in beer
brewing because of its gelatinization and viscosity problems in the cooker (Bradee,
1977; Teng et al, 1983). The chemical composition and cooking characteristics of
the rice differ from variety to variety. The growing region also has a significant
influence on the chemical composition and cooking characteristics of rice. For
example, the California short-grain varieties are not always acceptable when grown
in southern states and vice versa (Coors, 1976). In Japan, short-grain varieties are
usually used for beer brewing.
The contents of moisture, extract, and lipids are very important for evaluating the
quality of the brewers’ rice. The analytical data on the extract indicate its brew-
house efficiency; the data on lipids relate to its possible rancidity; and the moisture
content indicates its storage condition. The analytical methods have been officially
established independently by the American Society of Brewing Chemists (ASBC,
1992) and the European Brewery Convention (EBC, 1998). The moisture content is
determined by drying in an oven; the extract, by laboratory infusion mashing; and
the lipids, by extraction with an appropriate organic solvent.
Rice that has been stored for a very long time under unfavorable conditions
sometimes gelatinizes slowly, resulting in a highly viscous cooker mash. Moreover,
this unusual gelatinization often results in a slow and incomplete saccharification of
the combined malt-adjunct mash, run-off problems at the lauter tub, and hence a
lower yield of extract. These problems are caused mainly by oil rancidity.
A high lipid content in the raw materials has been said to have three major kinds
of possible influences on wort and beer characteristics: l) increased yeast growth
and reduced ester formation during fermentation (Ayrääpää and Lindström, 1973;
Anderson and Kirsop, 1974); 2) reduced foam formation and retention of the fin-
ished beer, and 3) off-flavor in beer and gelatinization difficulties during mashing
caused by oil rancidity.
When the adjunct is properly prepared, the greatest part of the lipid content is
derived from the malt, which contains 1.0-2.0% lipids. During wort production, a
Brewing / 563
major portion of the lipids derived from the raw materials is removed with the spent
grains at the lauter tub and with the trub, which mainly consists of coagulated pro-
teins formed during wort boiling (Klopper, 1974). In a study of all-malt brewing on
a laboratory scale, 1.0-2.8% of the lipid in malt was found in wort at the wort kettle
and 0.1-0.7% in the yeast-pitched wort. The lipid content decreased again by more
than 80% during fermentation. The lipid contents of 12 marketed beers were
0.33-0.76 mg/L (Krauss and Holstein, 1972).
Since the composition of lipids in adjuncts is similar to that in malt, the condi-
tions in malt-adjunct brewing are considered to be almost the same as those in all-
malt brewing. With such low concentrations of lipids in the fermented wort or beer,
neither yeast growth nor ester formation is influenced (Klopper et al, 1975). The
foam stability may decrease slightly (Klopper, 1974), but it is practically negligible.
Brewers’ rice usually contains less than 1.5% lipids. In such concentrations, lip-
ids do not affect beer quality unless they have become rancid. In most beer specifi-
cations, a limit of 0.4% of lipid material is given (Pierce, 1987).
Beer off-flavor and starch deterioration due to oil rancidity occur by the expo-
sure of rice grits to the air, particularly under such conditions as high temperatures
and high humidities, which must be avoided during storage of brewers’ rice. The
oil content of rice must not be greater than 0.4%. During storage, fatty acids are
liberated from glycerides and phospholipids. The free fatty acids form a helix
with amylose. Unsaturated fatty acids such as oleic, linoleic, and linolenic acids,
which usually constitute approximately 77% of rice fatty acids (Y. Sato et al,
1977), are oxidized into carbonyl compounds under unfavorable storage condi-
tions. These carbonyl compounds impart unpleasant off-flavors to beer at very
low concentrations, and, at the same time, combine with the protein molecules sur-
rounding the starch granules. These structural changes in the amylose and the pro-
tein molecules depress the water absorption and swelling of starch granules and
cause the unusual gelatinization described previously (Bradee, 1974; Moritaka,
1978). However, with modern storage techniques, oil rancidity is a negligible factor
(Stewart, 1994).
SOME OBSERVATIONS ON RICE IN BEER BREWING

Teng et al (1983) used the gel point measurement to routinely monitor rice qual-
ity and to make predictions on the influence of rice on brewhouse performances.
The gel point is defined as the characteristic temperature at which the mixture of
rice and water in a definite ratio shows a sudden rapid increase in viscosity when
the temperature is raised. Generally, short- and medium-grain rices had lower gel
points than long-grain ones. The rice paste viscosity depended heavily on the gel
point; the higher the gel point, the thicker the rice mash. The influences of the gel
point on total extract yields, wort fermentability, and the flow rate of wort lautering
were more or less significant.
Although brewers’ rice contains 0.8-1.2% total nitrogen, proteolytic enzymes of
malt scarcely degrade proteins at mashing (Canales, 1979), and only 2-3% of rice
nitrogen dissolves into the wort during decoction mashing.
Palamand et al (1970) studied glyoxal in beer, which forms an off-flavor similar
to that of self-soured milk. Of the raw materials examined (barley, malt, rice, corn,
and hops), rice had the highest content of glyoxal. Although its concentrations in
beer brewed with rice were below the threshold value, even at subthreshold con-
centrations, its indirect effects on beer flavor can be significant.
Low-calorie beer is produced by using a debranching enzyme (pullulanase) obtained

from rice. This enzyme reduces the residue extract of the final beer by cleaving a-1,6
linkages of unfermentable limited dextrins to form a-1,4 dextrins. These can be
converted by a-1,4 carbohydrases to sugars that can be fermented by brewers’ yeast. The
enzyme is extracted from rice by warm water or an aqueous buffer system of about pH 6,
and the extract thus obtained is added to mash or to wort before or during fermentation.
When malted rice is used as the enzyme source, a particularly useful mixture of the
enzyme and a-1,4 carbohydrases is obtained (Line et al, 1982).
In Germany, where the purity law is enforced, immunological techniques are
legally employed to detect proteins in beer that are not derived from malt, that is,
those from adjuncts and external proteolytic enzymes. After a protein fraction is
obtained by means of chill precipitation with ethanol or precipitation with saturated
ammonium sulfate, the protein fraction is subjected to double gel diffusion, immuno-
electrophoresis, or countercurrent electrophoresis. The minimum detection levels of
proteins in beer by these three methods are 10-20, 20-30, and 2-3 mg/L, respec-
tively. Each method may be capable of detecting the rice used in brewing at the rate
(weight of rice per unit volume of beer) of roughly 1,000 times the figure of its
detection level (Mebak, 1982).
The feasibility of using a high percentage of rice in beer brewing was evaluated
based on production parameters and wort properties. The experiments were con-
ducted with the amount of rice at 20-80% of total material and with added
enzymes. The results indicated that the amounts of free amino nitrogen and total
soluble nitrogen were reduced by the use of rice. This became a problem when the
rice content was higher than 50%. It was therefore considered feasible to use 50%
rice as an adjunct in beer brewing with added enzymes. Moreover, the effect on
fermentation and beer flavor of a higher ratio of rice (more than 50%) as an adjunct
still requires further study (Vinh et al, 1993).
Beers with approximately 70% adjunct contents are sold in Japan and are called
“happoushu.”
REFERENCES
Akai, T. 1963. On the rice harvested in Hok- 1995. Polishing properties of white-core and non-
kaido. Nippon Jozo Kyokai Zasshi 58:1106- white-core grains fractionated from the same
1109. variety of rice. Seibutsukogaku 73:381-386.
Akiyama, H., and Takase, S. 1958. On the rela- Ayrääpää, T., and Lindström, I. 1973. Influence
tion between the degree of rice steaming and of long-chain fatty acids on the formation of
its enzymatic decomposition. Nippon Jozo esters by brewer’s yeast. Proc. Congr. Eur.
Kyokai Zasshi 53:533-539. Brew. Conv. 14:271-285.
Akiyama, T., and Yamamoto, T. 1963. Effects of Bradee, L. H. 1974. Composition of the grist.
steaming time on the amount of water con- Tech. Q. Master Brew. Assoc. Am. 11:57-62.
densed on the surface of rice grains. Nippon Bradee, L. H. 1977. Adjuncts. Pages 40-61 in:
Jozo Kyokai Zasshi 58:319-325. The Practical Brewer. H. M. Broderick, Ed.
American Society of Brewing Chemists (ASBC). Master Brewers Assoc. Am., Madison, WI.
1992. Methods of Analysis, 8th ed. Adjunct Canales, A. M. 1979. Unmalted grains in
Materials: Cereals-3, Cereals-4, Cereals-5. The brewing. Pages 233-278 in: Brewing Science,
Society, St. Paul, MN. Vol. 1. J. R. A. Pollock, Ed. Academic Press,
Anderson, R. G., and Kirsop, B. M. 1974. The London.
control of volatile ester synthesis during the Coors, J. 1976. Practical experience with differ-
fermentation of wort of high specific gravity. ent adjuncts. Tech. Q. Master Brew. Assoc.
J. Inst. Brew. 80:48-55. Am. 13:117-123.
Aramaki, I., Ogawa, K., Yamamoto, K., Suzuki, De Clerck, J. 1957. Mashing technique. Pages
J., Kanno, M., Kizaki, Y., and Okazaki, N. 270-275 in: A Textbook of Brewing, Vol. 1.
Brewing / 565
(Katherine Barton-Wright, trans.) Chapman & dimethyl sulfide in sake. Agric. Biol. Chem.
Hall, London. 45:1713-1714.
European Brewery Convention (EBC). 1998. Kizaki, Y., Inoue, Y., Okazaki, N., and
Analytica-EBC. Chapter 6, Cereal Adjuncts, Kobayashi, S. 1991. Isolation and determina-
Method 6.2.1, 6.3, 6.4, 6.10. Verlag Hans Carl tion of protein bodies in polished rice
Getränk-Fachverlag, Nürnberg, Germany. endosperm. Nippon Jozo Gakkaishi 86:293-
Hanamoto, H., Nakazawa, E., Soga, E., and 298.
Takemura, S. 1978. On the correlations among Klopper, W. J. 1974. Wort composition, a survey,
various rice properties and classification of Monograph-1, Zeist. European Brewery Con-
rice for sake brewing. Nippon Jozo Kyokai vention, pp. 8-24.
Zasshi 73:115-119. Klopper, W. J., Tuning, B., and Vermeire, H. A.
Horie, S., Kawaski, H., and Nagashima, T. 1965. 1975. Free fatty acids in wort and beer. Proc.
Absorption of calcium onto polished rice and Congr. Eur. Brew. Conv. 15:659-667.
the velocity of rice swelling during steeping. Kobuyama, Y., Namba, Y., Totsuka, A., Ishikawa,
Nippon Jozo Kyokai Zasshi 60:1120-1123. T., Ohba, T., Takeda, M., Shimizu, Y., and
Iemura, Y., and Fujita, E. 1982a. Studies on Miyazawa, S. 1969. Composition of the con-
quality evaluation of rice for sake brewing. I. stituents of domestic and foreign rice varieties.
Characteristics of rice for sake brewing Nippon Jozo Kyokai Zasshi 64:648-652.
harvested in recent years. Nippon Jozo Kyokai Koura, J., Saeki, H., Ohmori, M., Gotanda, A.,
Zasshi 77:255-259. Mori, T., and Sakai, M. 1964. Test of sake
Iemura, Y., and Fujita, E. 1982b. Studies on brewing with dried steamed rice. Rep.
quality evaluation of rice for sake brewing. II. Hiroshima Prefectural Food Res. Inst. 8:37-41.
Evaluation by weight, crude protein and sur- (In Japanese)
face area (including shinpaku) of rice granules. Krauss, G., Zürcher, C., and Holsten, H. 1972.
Nippon Jozo Kyokai Zasshi 77:260-264. Die schaumzerstörende Wirkung einiger
Imayasu, S. 1993. The background of develop- Malzlipide und ihr Schicksale im Verlaufe des
ment and technology in new sake brewing Mälzungs- und Brauprozesses. Monatsschr.
system. Nippon Jozo Kyokai Zasshi 88:499- Brau. 25:113-123.
505. Kubo, S. 1960. Movement of chlorine in the rice
Imayasu, S., Kuriyama, K., and Ando, N. 1963. kernel. V. Inner translocation of a few rice
Studies on continuous rice-steaming of sake elements other than chlorine. Nippon Nogei
brewing. Hakko Kogaku 41:254-261. Kagaku Kaishi 34:689-694.
Ishikawa, T., and Yoshizawa, K. 1974. Studies Kumagai, C., Kuroyanagi, Y., and Noshiro, K.
on changes of lipids in sake brewing. II. 1976. The relation between moisture content
Changes of lipids in the process of washing, and water absorbability of polished rice.
steeping and steaming of rice used for sake Nippon Jozo Kyokai Zasshi 71:718-722.
brewing. Nippon Nogei Kagaku Kaishi Kumagai, E., Ohmachi, T., Kinuyama, Y.,
48:337-341. Kuruma, K., and Akasaka, K. 1968. On sake
Ishikawa, T., and Yoshizawa, K. 1979. Effects of brewing using the imported long-grain rice
cellular fatty acids on the formation of flavor varieties. Nippon Jozo Kyokai Zasshi 63:478-
esters by yeast. Agric. Biol. Chem. 48:45-53. 480.
Ishikawa, T., and Yoshizawa, K. 1984. Effects of Line, W. F., Chaudhary, V. K., Chicoye, E., and
fatty acids on ester-forming enzymes in yeast Mizerak, R. J. 1982. Preparation of debranch-
cells. Nippon Jozo Kyokai Zasshi 79:62-66. ing enzyme for producing a low calorie beer.
Kanoh, S. 1962. The mineral contents of rice and U.S. patent 4355110.
their changes during washing and steeping. Little, B. T. 1994. Alternative cereals for beer
Nippon Nogei Kagaku Kaishi 36:379-380. production. Ferment 7:163-168.
Kawasaki, W., and Nagatani, M. 1966. Sake Maekawa, S., Yokoyama, N., Tabata, N., Shinke,
brewing using alfa-rice (dried steamed rice). R., and Nisira, H. 1981. Factors affecting
Nippon Jozo Kyokai Zasshi 61:642-645. formation of dimethyl sulfide precursor in old
Kawasaki, W., Aso, N., Noshiro, K., and rice grains. Hakko Kogaku Zasshi 59:483-489.
Momiya, N. 1967. On off-flavor of sake pro- Mebak, 1982. Immunologie-Nachweis von
duced of the imported long-grain rice variety. Rohfrucht und Enzymen. Pages 616-623 in:
Nippon Jozo Kyokai Zasshi 62:68-71. Brautechnische Analysenmethoden, Vol. 3. F.
Kitamoto, K., Ohba, T., and Namba, Y. 1981a. Drawert, Ed. Mitteleuropaische Brautech-Analy-
Formation of dimethyl sulfide precursor in rice senkommission, Freising-Weihenstephan, West
by methylbromide fumigation. Nippon Jozo Germany.
Kyokai Zasshi 76:491-494. Momose, H. 1979. Pages 66-67 in: Technology
Kitamoto, K., Ohba, T., and Namba, Y. 1981b. of Sake Brewing. Brewing Society of Japan,
Identification of the heat-stable precursor of Tokyo. (In Japanese)
Moritaka, S. 1987. Quality change of polished Takayama, T. 1992. Sake brewed with polished
rice during storage. Nippon Jozo Kyokai rice treated by heat blast method. Nippon Jozo
Zasshi 73:690-696. Kyokai Zasshi 87:849-857.
Nagatani, M., and Nunokawa, Y. 1975. A new Tanaka, T., Kumagai, C., Okazaki, N., and
method of alfa-rice production. Nippon Jozo Aakiyama, H. 1981. Sake brewing and koji
Kyokai Zasshi 70:345-347. making using unsteamed rice. Nippon Jozo
Nakai, S., Nakazawa, E., Hanamoto, H., Kusama, Kyokai Zasshi 76:665-669.
T., Shimizu, M., Mishima, H., and Takemura, Teng, J., Stubits, M., and Lin, E. 1983. The
S. 1979. Application of dried steamed rice importance of rice variety selection for opti-
using alcohol on sake brewing. Nippon Jozo mum brewhouse operation. Proc. Congr. Eur.
Kyokai Zasshi 74:402-405. Brew. Conv. 19:47-54.
Noshiro, K., and Aoki, M. 1957. On the amount Totsuka, A. 1979. Inorganic substances. Page 2
of potassium eluted from rice grain during in: The Constituents of Fermented Foods. Brew-
rice-steeping in running water. Nippon Jozo ing Society of Japan, Tokyo. (In Japanese)
Kyokai Zasshi 52:1008-1012. Vinh, N. T. T., Viet, N. V., and Mai, N. T. 1993.
Ohashi, M., Yamada, K., Ohiwa, J., and The use of high percentage of rice as an
Kurasawa, M. 1978. Application of rice bran adjunct in beer brewing. Tech. Q. Master
saccharified solution on sake brewing. 1978. Brew. Assoc. Am. 30:42-44.
Rep. for 1977. Saitama Prefectural Res. Inst. Vogel, E. H., Jr. 1950. Rice as brewing adjunct.
Food Technol., pp. 50-57. (In Japanese) Brew. Dig. 25:45-49.
Ohmori, K., Takahashi, Y., Sato, I., Ogawa, J., Working Group on Rice for Sake Brewing. 1993.
Hosokawa, K., and Manome, I. 1977. Pro- Classification of rice varieties for sake brew-
duction of sake using alfa-rice. Rep. Iwate ing using analytical data. Pages 275-277 in:
Prefectural Res. Inst. Brew. Food Technol. Varieties of Rice for Sake Brewing. The
11:14-21. (In Japanese) Working Group, Higashihiroshima, Japan. (In
Palamand, S. R., Nelson, G. D., and Hardwick, Japanese)
W. A. 1970. Further studies on glyoxal and Yanagiuchi, T., Yamamoto, H., Miyazaki, N.,
methylglyoxal in beer. Am. Soc. Brew. Chem. Nagano, T., and Wakai, Y. 1996. Influence of
Proc., pp. 186-191. grain type on physical and chemical properties
Pierce, J. S. 1987. Adjuncts and their effect on of rice for sake brewing. Seibutsukogaku
beer quality. Proc. Congr. Eur. Brew. Conv. 74:97-103.
21:49-60. Yoshioka, K. 1980. Mashing and the role of
Saijo, T., Uchida, T., Frurhashi, M., and enzymes in beer brewing. Nippon Jozo Kyokai
Kawamura, O. 1968. On the influence of the Zasshi 75:841-844.
characteristics of rice and the capacity of water Yoshioka, K., and Hashimoto, N. 1983. Cellular
for steeping upon the water absorption of fatty acid and ester formation by brewer’s
polished rice. Rep. Shizuoka Prefectural Ind. yeast. Agric. Biol. Chem. 47:2287-2294.
Res. Inst. 12:70-74. (In Japanese) Yoshizawa, K. 1976. Effects of higher fatty acids
Sato, T., Saito, K., Tachibana, T., Nakada, T., on the formation of esters by yeasts. Nippon
Ishikawa, T., and Ikemi, M. 1977. Production Nogei Kagaku Kaishi 50:115-119.
of sake using alpha-rice. Rep. Akita Prefec- Yoshizawa, K. 1984. Elucidation of the charac-
tural Inst. Brew. 9:33-37. (In Japanese) teristics of rice for sake brewing. Nippon Jozo
Sato, Y., Ogata, J., Nakazawa, F., and Noguchi, Kyokai Zasshi 79:156-164.
S. 1977. Fatty acid composition of lipids in Yoshizawa, K., and Ishikawa, T. 1974a. Changes
nonglutinous and glutinous rice. Kaseigaku in the amounts of substances eluted from rice
Zasshi 28:403-407. during washing, steeping and steaming pro-
Shiinoki, S., and Nunokawa, Y. 1980. Acid cesses. Nippon Jozo Kyokai Zasshi 69:584-
protease reaction affecting the alpha-amylase 586.
adsorption onto steamed rice. Hakko Kogaku Yoshizawa, K., and Ishikawa, T. 1974b. The
58:355-361. nature of rice used for sake brewing and its
Stewart, G. G. 1994. Adjuncts. Pages 122-123 in: processing. Nippon Jozo Kyokai Zashhi
Handbook of Brewing. W. A. Hardwick, Ed. 69:645-650.
Marcel Dekker, New York. Yoshizawa, K., Ishikawa, T., Unemoto, F., Sato,
Takahashi, K., Matsumoto, E., Tani, K., Sakurai, H., and Noshiro, K. 1972. Studies on rice
K., Shibata, H., and Yoshizawa, K. 1987. washing. Nippon Jozo Kyokai Zasshi 67:645-
Making of sake using puffed low polished 648.
rice. Nippon Jozo Kyokai Zasshi 82:831-835. Yoshizawa, K., Ishikawa, T., Unemoto, F., and
Takase, S., and Murakami, H. 1965. Absorption Noshiro. 1973a. On the content of potassium
of iron onto rice grains during steeping. in various polished rice grains. Nippon Jozo
Nippon Jozo Kyokai Zasshi 60:624-627. Kyokai Zasshi 68:705-707.
Brewing / 567
Yoshizawa, K., Ishikawa, T., and Hamada, Y. in polished rice grains. Nippon Jozo Kyokai
1973b. Rice used for sake making. III. Rela- Zasshi 73:388-391.
tionships among various characteristics of Yoshizawa, K., Ikemi, M., Nakada, T., Iemura,
milled rice. Nippon Jozo Kyokai Zasshi Y., and Momose, H. 1978b. Observation of
68:767-771. steamed rice grains using a scanning electron
Yoshizawa, K., Ishikawa, T., and Noshiro, K. microscope. Nippon Jozo Kyokai Zasshi
1973c. Studies on the changes of lipids in sake 73:737-740.
brewing. I. Changes in fatty acids composition Yoshizawa, K., Momose, H., Ishikawa, T., and
of lipids by the change of the polishing ratio of Iemura, Y. 1979a. Changes in saccharifica-
rice grains. Nippon Nogei Kagaku Kaishi tion and the structure of steamed rice pre-
47:713-717. pared with different steeping and steaming
Yoshizawa, K., Ishikawa, T., Iemura, K., Takeda, times. Nippon Jozo Kyokai Zasshi 74:186-
S., and Fujie, T. 1974. Studies on moisture 189.
contents, saccharification and deterioration of Yoshizawa, K., Ishikawa, T., Iemura, Y., and
steamed rice. Nippon Jozo Kyokai Zasshi Saito, K. 1979b. Studies on the lipids of
69:315-318. steamed rice under various conditions of rice
Yoshizawa, K., Ikemi, M., and Ishikawa, T. steeping in the lipase solution. Nippon Jozo
1976. Application of rice-steeping in lipase Kyokai Zasshi 74:194-197.
solution on sake brewing. Nippon Jozo Kyokai Yoshizawa, K., Momose, H., and Hasuo, T.
Zashi 71:975-978. 1980. The nature of endosperm cell walls of
Yoshizawa, K., Momose, H., Ishikawa, T., and polished rice. Nippon Jozo Kyokai Zasshi
Shimaoka, J. 1978a. Location of the constituents 75:428-431.
CHAPTER 19
RICE BRAN AND OIL
Frank T. Orthoefer
Food Sciences and Technologies, LLC
Germantown, Tennessee
Jennifer Eastman
ACH Food and Nutrition
Memphis, Tennessee
Rice bran is one of three co-products obtained from rice milling (Juliano and
Bechtel, 1985). The other two are milled rice and rice hulls. Rice bran is utilized as
a feed ingredient for its protein and lipid content, and it represents a significant
source of revenue for the miller. To the food industry, rice bran is a source of
dietary fiber, protein, and “friendly” oil; however, only limited quantities are sold
for food applications, partially due to its instability, which leads to off-flavor, and
frequent contamination with rice hulls. Specialty markets relying on a healthful
image have been developed for stabilized bran. Attempts have been made to
develop markets for rice bran oil, a major component of rice bran. High capital
costs for extraction of the oil and limited availability of the stabilized bran have
limited its commercialization. Food and feed utilization for stabilized bran and oil
continue to be investigated.
MILLING
Rice milling began as a simple grinding of the grain with mortar and pestle to
remove the husk and loosen the bran. In today’s modern rice mills, the outcome
is similar but takes place in large-scale, mechanized mills in a series of unit
operations. The milling process begins with paddy rice. Rubber roll dehullers
cleanly remove the husk from the kernel (Juliano, 1985a; Luh, 1991). The
discharge from dehullers includes brown rice, paddy, hulls, bran, dust, broken
rice, and immature grains. Paddy is separated by paddy separators and returned to
the dehullers. Bran, dust, broken rice grains, and other products are separated by
an air stream. Abrasive milling, also called whitening, removes the outer tissues
of the grain, producing “polished” rice and bran. In abrasive whiteners, brown
rice is passed between abrasive rollers and a wire screen or perforated cylinder.
The abrasive roll cuts small bits of bran from the brown rice. A vacuum pulls air
through the machine, removing the loosened bran, which is recovered in a
cyclone.
569
Fig. 1. Relative proportions of the major rice caryopsis components. Hull makes up 20%, bran and
germ 10%, and starchy endosperm 70%. (Reprinted with permission from Orthoefer, 2001)
In friction mills, pressure is applied to the grain as it passes through the clearance
between the screen and mill roller. Bran is loosened and peeled off. Polishing
removes fine bran particles that adhere to the grain. Leather strips may be used to
polish the grains as the shaft revolves. Polish is recovered in cyclone separators.
Bran from modern rice mills is a fine, floury material made up of pericarp, seed
coat, nucellus, aleurone, pulverized germ, some hull fragments, and some endo-
sperm (white rice fragments) (Orthoefer, 2001) (Fig. 1). It makes up approximately 8–
10% of the rough rice weight (Juliano and Bechtel, 1985). The pericarp consists of
several layers of crushed cells about 10 µm thick with a thin cuticle. Next to the
pericarp is a single layer of cells (the seed coat) with a thick cuticle (0.5 µm). Pigments
are located in the pericarp or the seed coat. Nucellular cells have a thick cuticle (0.8
µm). The total thickness of the nucellus is about 2.5 µm. The bond between the seed
coat and nucellus is weak, and the two parts separate easily. The embryo of the rice
kernel, located in the lower abdominal area of the caryopsis, is extremely small
compared with that of other cereals. It is bound on the outside by a single layer of
endosperm cells, the modified aleurone layer, and fibrous cellular remains of the
pericarp, seed coat, and nucellus. The starchy endosperm borders the inner edge of the
embryo. Bran structure is detailed in Chapter 4.
RICE BRAN
Chemical Composition
The composition of rice bran that includes germ is comparable in protein, fiber,
and ash content to that of other cereal brans (Table 1). Oil, phosphorus, and silica
contents are notably higher. Phosphorus, present mainly as phytin phosphorus
(90%), is highest among all the cereal brans. Other minerals present in rice bran
include potassium and magnesium, among others (Table 2). Rice bran is high in B
vitamins and tocopherols but contains little vitamin A and C (Table 3). The
variation reflects varietal differences, degree of milling, and hull contamination.
Vitamins in the bran may be in free or combined forms. For example, 75% of B2
(riboflavin) is esterified, whereas 89% of folic acid, 49% of pantothenic acid, and
89% of niacin are in the bound forms.
Bran and Oil / 571
TABLE 2
Inorganic Constituents of Rice Brana
Constituent Content (ppm, dry basis)
Aluminum 53.5–369
Calcium 140–1,310
Chlorine 510–970
Copper 0.37
Iodine 5
Iron 130–530
Magnesium 8,650–12,300
Manganese 110–877
Mercury 0.3
Phosphorus 14,800–28,680
Potassium 13,650–23,960
Selenium 0.170
Silicon 1,700–16,300
Sodium 0–290
Sulfur 80
Tin 17.6–4.3
Titanium 26
Zinc 80
a Source: Luh et al (1991); used with permission.
TABLE 3
Vitamin Content of Rice Brana
Vitamin Content (ppm, dry basis)
Vitamin A 4.2
Thiamin 10.1–27.9
Riboflavin 1.7–3.4
Niacin (nicotinic acid) 236–590
Pyridoxine 10.3–32.1
Pantothenic acid 27.7–71.3
Biotin 0.16–0.60
Inositol 4,627–9,270
Choline 1,279–1,700
p-Aminobenzoic acid 0.75
Folic acid 0.50–1.46
Vitamin B12 0.005
Vitamin E (tocopherols) 149
The nitrogen content of rice bran is 1–3% (dry basis), most of which is protein
nitrogen. Nonprotein nitrogen accounts for 16% of the total nitrogen and 11% of
rice polish nitrogen (Luh et al, 1991). Free amino acids are glutamic acid (7–31%),
alanine (11–16%), and serine (5–15%). Other nitrogen-containing compounds are
guanine, xanthine, and adenine. The amino acid content of rice bran varies within
wide limits (Table 4) (Juliano, 1985b). The protein content of rice bran plus germ
ranges from 12 to 19% for rice polish.
The major protein fractions in bran are albumin and globulin, whereas in polish
it is glutelin. Prolamin is a minor fraction. The molecular weights of the albumin-
globulin fraction are 2.0 × 105, 0.8 × 105, and 0.3 × 105. Cytochrome C and a blue
protein are the major soluble basic proteins of bran (Morita and Ida, 1968). Blue
protein is a copper-containing glycoprotein that occurs in the aleurone layer.
Bran and Oil / 573
TABLE 4
Amino Acid Composition (g/16 g of N) of Rice Bran and Polish and Other Cereal Bransa
Rice Wheat Rye Oat
Amino Acid Bran Polish Bran Bran Bran
Alanine 6.5–7.0 6.5; 6.6 5.7; 3.8 5.4 5.1
Arginine 8.6–9.1 8.9; 9.0 8.1; 5.9 6.3 6.8
Aspartic acid 10.0–11.0 9.7; 10.7 8.5; 6.1 7.5 8.6
Cystine 2.5–2.8 2.6; 2.8 2.8; 1.9 1.9 2.4
Glutamic acid 14.6–15.0 16.1; 17.6 21.2; 16.8 27.9 21.1
Glycine 5.8–6.2 5.6; 5.7 6.7; 4.6 5.4 5.4
Histidine 2.9–3.7 2.8; 2.9 3.2; 2.5 2.2 2.2
Isoleucine 2.9–4.5 2.9; 4.2 3.4; 2.7 3.7 3.8
Leucine 7.6–8.4 7.2; 8.4 6.8; 5.1 6.8 7.4
Lysine 5.3–6.0 4.6; 5.1 4.5; 3.4 4.1 4.1
Methionine 1.9–2.5 2.4; 3.0 1.7; 1.0 0.4 2.1
Phenylalanine 4.9–5.3 4.6; 5.0 4.3; 3.1 4.6 5.1
Proline 4.6–6.1 4.2; 5.7 6.5; 5.7 4.9 6.2
Serine 5.1–6.0 4.9; 5.9 5.0; 4.3 4.5 4.8
Threonine 4.2–4.6 3.9; 4.4 3.7; 2.9 3.3 3.4
Tryptophan 0.6; 1.4 1.4 1.7 … …
Tyrosine 3.5–3.8 3.8; 4.3 3.3; 2.4 2.7 3.5
Valine 5.4–6.6 4.8; 6.2 3.2; 3.7 5.3 5.5
Ammonia 1.9–7.6 2.2; 6.5 2.1 … 2.5
a Source: Juliano (1985b).
Rice bran and polish protein are higher in lysine than are other cereal brans
(Table 4). The higher albumin content of rice bran protein accounts for some of the
difference. Albumin globulins contributed 60–71% of the extractable protein from
bran and defatted bran. Reported values for the lysine content of bran fractions are
(per 16 g of N) 7.0 g for albumin, 5.7 g for globulin, and 0.9 g for prolamin.
Rice bran is enzyme-active (Barber and de Barber, 1980). Some enzymes
reported include a-amylase, b-amylase, ascorbic acid oxidase, catalase, cytochrome
oxidase, dehydrogenase, deoxyribonuclease, esterase, flavin oxidase, a- and b-
glycosidase, invertase, lecithinase, lipoxygenase, pectinase, peroxidase, phos-
phatase, phytase, protease, and succinate dehydrogenase. The germ and outer layers
of the caryopsis have higher enzyme activities. The amylolytic activities of rice
milling products are (in milligrams of maltose per 10 g) brown rice = 39, milled rice
= 15, bran = 320, polish = 250, and germ = 310. Lipase activity is probably the most
important commercially because it affects the keeping quality and shelf life of rice
bran. Lipoxygenase and peroxidase are also related to rice bran stability.
The equilibrium moisture content of rice bran and polish is lower than those of
brown rice and milled rice because they contain less starch. Defatted bran is more
hygroscopic than is full-fat bran. Stabilization does not change the hygroscopicity of
full-fat bran. The bulk density of bran with oil is 0.2–0.4 g/mL.
Composition of Rice Germ
Rice germ is higher in protein (17.3–26.4%) and fat (16.6–39.8%) than rice bran
but contains less fiber (Luh et al, 1991). Germ accounts for approximately 20% of
rice bran. Removal of germ from bran reduces the fat content of the bran. Rice
germ contains approximately four times as much amino nitrogen as bran. The major
TABLE 5
Vitamin Content of Rice Germa
Constituent (per gram, dry basis) Content (µg/g dry wt)
Vitamin A (carotenes) 1.3
Thiamin 45.3–76.0
Riboflavin 2.7–5.0
Niacin (nicotinic acid) 15.2–99.0
Pyridoxine 15.2–16.0
Pantothenic acid 13.2–30.0
Biotin 0.26–0.58
Inositol 3,725–6,400
Choline 2,031–3,000
p-Aminobenzoic acid 1
Folic acid 0.9–4.3
Vitamin B12 0.001
Vitamin E (tocopherols) 87.3
a Adapted from Juliano (1972).
free amino acids are alanine (14%), aspartic acid (12%), proline (29.5%), and serine
(12%). The protein fractions in rice germ are high in the salt-soluble fraction. A
Japanese rice embryo protein was found to contain 17.4% albumin and 20.0%
globulin (Barber and de Barber, 1980). Embryo globulin consisted of three compo-
nents when analyzed by Sephadex G-200 chromatography—hemoproteins (cyto-
chrome C and peroxidase 556) and chromoproteins (Barber and de Barber, 1980).
Rice germ is high in phosphorus and low in calcium, sodium, and silicon. Sev-
enty-five percent of its phosphorus is present as phytate phosphorus (O’Dell et al,
1972). The vitamin composition and content of rice germ are similar to those of rice
bran (Table 5). Thiamin is higher in rice germ, whereas niacin is higher in rice bran.
Factors Affecting the Composition of Bran
The composition and properties of rice bran vary widely (Saunders, 1990). The
particle size distribution of raw bran is shown in Table 6. The actual particle size
distribution varies significantly with type of milling as well as milling conditions.
The composition of bran also varies as a function of milling degree, as outlined in
Table 7. Generally, a low degree of milling, or so-called “undermilling,” is prac-
ticed for economic or yield reasons. Undermilled rice is oxidatively less stable than
is “well-milled” rice because of residual bran and oil (S. Danforth and F. T.
Orthoefer, unpublished data).
Contaminants in bran such as hulls, endosperm, and weed seeds depend upon the
efficiency of milling (Juliano, 1985b). U.S. rice mill feed is a mixture of the bran
and hull from a single-step milling process. This so-called “huller bran” contains 4–
8% oil. Multipass milling results in reduced bran yields (6%) because less starch
and hulls contaminate the bran stream. Friction mills produce bran that has a higher
oil content than that produced in abrasion mills because of more effective removal
of the aleurone layer. Abrasive mills remove more of the starchy endosperm than do
friction mills. The bran containing the highest amount of oil is from long-grain rice,
and the lowest is from short-grain.
A proportional increase in the concentration of all major components occurs with
oil extraction. Bran produced by a combined solvent-extraction milling process was
Bran and Oil / 575
TABLE 6
Particle-Size Distribution (%) of Raw and Heat-Stabilized Bransa,b
Meshc Particle Size (mm) Raw Bran Moist-Heat Stabilized Bran
18 >1,000 0 0
18–30 1,000–595 2.4 18.6
30–50 595–297 30.0 32.7
50–80 297–177 12.2 18.5
80–100 177–149 8.5 10.8
<100 <149 46.7 19.4
b Japonica rice type; cone mill.
c U.S. standard.
TABLE 7
Variation in Rice Bran Composition as a Function of the Degree of Millinga
Degree of Bran Composition (%)

Milling (%) Protein Fat Fiber Ash NFEb
1st Cone 0–3 17.0 17.7 10.5 9.8 45.0
2nd Cone 3–6 17.6 17.1 10.3 9.4 45.2
3rd Cone 6–9 17.0 16.5 5.7 8.4 52.5
4th Cone 9–10 16.7 14.2 5.7 7.5 55.9
a Courtesy S. Danforth and F. T. Orthoefer.
b Nitrogen-free extract.
reported as higher in protein and lower in fiber than was bran from simple solvent
extraction of oil (Hunnell and Nowlin, 1972).
Varietal difference in bran composition showed a range of 16.2–19.6% oil (IRRI,
1984). Variations also occur in volatile compounds such as low molecular weight
alcohols, aldehydes, and ketones. These carbonyl compounds are likely derived
from oxidation of the unsaturated fatty acids (Juliano, 1985c).
Bran from parboiled rice is higher in oil content than is bran from milled white
rice. Parboiled bran is lower in thiamin and riboflavin. The differences are due to
the degree of milling as well as to conditions used in parboiling. There is less
endosperm contamination of the bran stream in parboiled rice. Parboiled bran con-
tains approximately 2% starch, whereas up to 34% starch may be found in raw rice
bran. Actual contents depend on the extent of milling, particularly above 6% bran
yields. Higher oil levels could be the result of less endosperm contamination, higher
extractability of oil by solvents used in the analysis, and outward movement of fat
within aleurone and germ cells.
Parboiling increases oil yield. Parboiled bran with higher oil content is “oily”
and tends to clog sieves during milling (Juliano 1985b). Parboiled bran tends to
have larger particles than does white rice bran (Table 8).
Antinutritional Factors in Bran
Phytin in bran is in globoids in the aleurone protein bodies as a potas-

sium/magnesium salt (Juliano and Bechtel, 1985). Phytin phosphate groups readily
complex with calcium, zinc, iron, and proteins. Rice bran’s phytin content (Table 1)
is higher than that of wheat, corn, soybean brans, and oat hulls (Thompson and
TABLE 8
Particle Size Distribution Among Brans from Raw and Parboiled Bran of Four Rice Varietiesa
Raw Rice Brans (%, by wt) Parboiled Rice Brans (%, by wt)
Particle Sizeb (mm) Range Mean Range Mean
<0.70 86–94 90 38–82 64
0.70–0.85 6–13 9 11–49 27
0.85–1.00 0 0 2–6 4
>1.00 0 0 0–7 5
a Source: Raghavendra Rao et al (1967); used with permission.
b Sizes correspond to British mesh size of 22 (0.70 mm), 18 (0.85 mm), and 16 (1.00 mm).
TABLE 9
Dietary Fiber (% at 14% moisture) Measured by in Vivo and in Vitro Techniques
and Crude Fiber in Rice Brana
Dietary Fiber
Rice Material In Vivo In Vitro Crude Fiber
Bran 1 41.8 … 12.5
Bran 2 36.2 … 9.5
Bran 3 32.5 31.3 7.6
Bran 4 25.4 27.7 6.8
Bran 5 22.0 … 3.5
Defatted bran 38.1 … 8.4
a Adapted from an unpublished paper by R. M. Saunders, E. Hautala, and E. A. Elliston.
Weber, 1981). Calcium and zinc, as well as phytin phosphorus, are nutritionally
unavailable (Juliano and Bechtel, 1985).
Trypsin inhibitor (TI), an albumin, has been isolated from rice bran (Juliano,
1985b). Calcium is a cofactor for trypsin. The strong calcium binding activity of
rice bran may account for a portion of the observed TI activity. One mole of TI can
inhibit two moles of trypsin. Since other cereal TIs inhibit on a 1:1 ratio, two differ-
ent reactive sites are postulated for rice bran TI (Juliano, 1985b). Most (85–95%) of
the TI is concentrated in the embryo, and 5–10% is in germ-free bran. None is in
milled rice. TI is inactivated by steaming at 100°C for 6 min. Dry heating of bran
for 15–30 min at 100C does not inactivate TI (Juliano, 1985b). Rice bran TI is
different from soybean TI in that it is not water soluble, not readily destroyed by dry
heat, and inhibits other proteases (Luh, 1991).
Hemagglutinins (which are globulins) agglutinate mammalian red blood cells
(Juliano, 1985b). Rice bran hemagglutinins, also called lectins, are similar to wheat
bran hemagglutinins. Rice lectin reacts with wheat germ agglutinin antiserum. It is
structurally different in that it is readily cleaved into smaller polypeptides, does not
dissociate at lower pH, and has fourfold higher specific activity. Lectin is a
glycoprotein containing 27% carbohydrate, mainly glucose, zylose, and arabinose.
Rice bran lectin has a pI of 8.8 and binds 1.8 mL of N-acetyl-glucosamine per
19,000 g of lectin. Others reported a pI of 6.5–6.8. Hemagglutinin activity was
shown to be only in the germ (Barber and de Barber, 1980; Ory et al, 1981).
Heating for 6 min at 100C or for 2 min at 100C in an aqueous extract inactivates
hemagglutinin activity (Ory et al, 1981).
Fiber is sometimes considered “antinutritional” because of its ability to bind
minerals (Rasper, 1979). The dietary fiber of rice bran is equal to about four times
Bran and Oil / 577
Fig. 2. Free fatty acid increase in raw bran during a 135-day storage period. (Reprinted, with
permission, from Orthoefer, 2001)
the crude fiber content (Table 9) and is similar to the neutral detergent fiber content
(Saunders et al, 1982).
Stabilization
The instability of rice bran has long been recognized. Nearly 100 years ago, the
lipase activity of raw rice bran was identified (Brown, 1903). As long as the kernel
is intact, lipase is physically isolated from lipids (Saunders, 1986). Dehulling dis-
turbs the surface, allowing lipase and oil to mingle. Oil in intact bran contains 2–4%
free fatty acids (FFA) (Orthoefer, 1994). Once milled from the kernel, a rapid
increase in the free fatty acid content occurs (Fig. 2).
Lipase activity is highly dependent on storage temperature and humidity
(Orthoefer, 1995a). With rice that is stored under hot, moist conditions, the FFA
content may increase 5–10% per day and about 70% in a month (Orthoefer,
2001). Optimal temperature for lipase activity is 35–40°C. Lower temperatures
reduce the rate of hydrolysis. No lipase activity occurs below freezing temper-
atures.
Rice bran also contains lipoxygenase and peroxidase, both having a negative
impact on the oxidative state of the bran (Table 10). Activity of these enzymes results
in further degradation of the oil contained in bran, as reflected in increased peroxide
value, decreased iodine value, and increased thiobarbituric acid value. Both peroxidase
and lipoxygenase enzymes are inactivated simultaneously with lipase inactivation.
Peroxidase is used as a convenient index of lipase activity. The inactivation tem-
perature for lipases and associated enzymes is dependent on the moisture content.
At 4% moisture, the inactivation temperature for lipoxygenase is 40°C, for lipase it
is 55°C, and for peroxidase it is 70°C (Juliano, 1985b).
Lipase results in hydrolytic rancidity. Little or no change in flavor of the bran

occurs with an increase in FFA (Orthoefer, 1995a). Lipoxygenase activity, however,
increases with the presence of FFA, resulting in oxidative rancidity. The oxidative
deterioration is responsible for the flavor and odor of rancid rice bran.
Methods used to stabilize rice bran include dry extrusion, wet extrusion (expan-
sion), refrigeration, pH modification, and chemical treatment (Orthoefer, 2001).
Dry heat stabilization evolved from high-temperature drying of bran at 100–110°C
(Sayre et al, 1982). Maintenance of low moisture during storage is required to
maintain stability. Some of the methods involve pan roasting, stationary dryers, or
fluid bed dryers. A 20- to 30-min drying time is required. Long heating times and
uniformity of treatment lead to process difficulties resulting in high microbial
counts and insect infestation as well as heat damage to the bran. Prolonged heating
causes dark-colored oil and bran. Residual lipase activity often occurs. Dry-heat
stabilization, in which moisture is retained until completion of the heating phase, is
more effective than is high-temperature drying (Sayre et al, 1982). Pressure is
required to prevent moisture loss until completion of the heating phase.
Development of extrusion methods that utilize retained-moisture heating have
been successful (Saunders et al, 1982). Stabilization within 1 hr after milling is con-
sidered ideal for bran quality. Dry bran (10–14% moisture) is fed to the extruder,
which relies on the generation of frictional heat to obtain temperatures of 130–
150°C (Sayre et al, 1982). Bulk density increases and moisture content decreases to
5–8%. Holding at extrudate temperature for approximately 3 min ensures stabiliza-
tion (Juliano, 1985b). The primary drawbacks to dry extrusion include the follow-
ing: 1) stabilized bran is a powder; 2) dark color may be developed due to high
cooking temperatures; 3) incomplete inactivation of lipases may occur, requiring
maintenance of low moisture levels to prevent an increase in FFA; and 4) the
equipment maintenance cost is high (Orthoefer, 1995a, 2001). Dry extrusion was
found more suitable for stabilizing bran to be used as a food ingredient (Randall et
al, 1985).
Extrusion cooking, as a cost-effective stabilization technique, was pioneered by
the Western Regional Research Laboratory (Sayre et al, 1982; Harper, 1989). Wet
heating is more effective for bran stabilization than is dry heating. Lipase is inacti-
vated in 3 min at 100°C (Sayre et al, 1982). Equipment used includes steam cook-
ers, blanchers, autoclaves, and screw extruders with injected steam and water
(Barber and de Barber, 1980). Extrusion with steam injection and up to 10% added
water reduces the temperature required for lipase inactivation. Temperatures are
TABLE 10
Changes in the Composition of Bran Lipids During Storage of Milyang 23 Rice Bran
at 30°C and 80% RHa
Storage Period (weeks)
Oil Property 0 1 2 3 4 5
Free fatty acids (% as oleic acid) 3.6 33.0 40.3 45.8 61.8 68.2
Peroxide value (meq/kg) 32.8 73.2 96.0 109.3 90.6 91.0
Iodine value (%) 96.8 90.2 85.4 83.2 79.0 74.7
TBAb (mg of malonaldehyde/kg) 0.5 0.8 1.1 0.7 0.7 0.6
a Source: Ryu and Cheigh (1980); used by permission.
b Thiobarbituric acid.
Bran and Oil / 579
reduced to 100–120°C. The product may be held at high temperature (100°C) for
1.5–3.0 min before drying to stable moisture levels. Bran expands as it exits the
extruder, and water flashes to steam (Orthoefer, 2001). Porous pellets, also called
collets, are found to assist in solvent percolation during oil extraction. Fines are
agglomerated as well.
Addition of water or steam to bran during wet extrusion requires drying after
stabilization. Hot air is simply passed through a bed of pellets. This increases the
cost of stabilization, but the lipase inactivation is permanent, with less nutritional
damage to the bran. The recovered oil is lighter in color, with lower refining losses.
The stabilized bran may be stored for extended periods of time, although extraction
should be completed within one month for best-quality crude oil.
Parboiling of rice is also an example of wet heat stabilization. Lipase is com-
pletely inactivated in rough rice by either autoclaving for 3–20 min or by parboiling.
(See Chapter 13.)
Other stabilization methods that have been investigated include refrigeration,
adjustment of pH, and chemical addition. Low temperatures reduce the rate of
hydrolysis by lipase (Orthoefer, 2001). Only the rate is decreased, and lipase activity
resumes once temperatures are increased. However, storing bran under refrigerated
conditions is costly, with limited commercial applications. Adjusting pH to 4.0 by
addition of an acid such as hydrochloric acid reduces lipase activity (Prabhakar and
Venkatesh, 1986). This method is not currently practiced on a commercial basis.
Addition of sodium metabisulfite to bran limits lipase activity. The chemical is added
directly to bran and mixed thoroughly. This process has not been commercially practiced.
Well-stabilized bran has excellent keeping qualities (D. R. McCaskill, S. Danforth,
and F. T. Orthoefer, unpublished data). FFA content remains relatively constant for
both stabilized and parboiled brans (Fig. 3). Oxidative stability as determined by
peroxide value is greater for stabilized rice bran than for parboiled bran (Fig. 4).
Multiple products formulated with stabilized rice bran are currently marketed,
primarily through the Internet. Several animal feeds, particularly horse feeds, claim
Fig. 3. Changes in free fatty acid (FFA) content (%, oil basis) of stabilized and parboiled rice bran
during storage (average of two storage temperatures and two packaging types). (Courtesy S. Danforth
and F. T. Orthoefer)
Fig. 4. Changes in peroxide value (meq/g, oil basis) of stabilized and parboiled rice bran during
storage (average of two storage temperatures and two packaging types). (Courtesy D. R. McCaskill, S.
D. Danforth, and F. T. Orthoefer)
improved performance with the inclusion of stabilized bran in diets. Additionally,

fractionated stabilized brans are promoted for health benefits (Barber and de
Barber, 1980). Dry- as well as wet-fractionation processes are utilized. Extraction
of various components of bran for protein enrichment, as well as for minor compo-
nents, has been attempted, usually with marginal success.
Nutritional Properties of Rice Bran
Rice bran is similar to other cereal brans in that it is deficient in lysine and
threonine (Barber and de Barber, 1980). Protein efficiency ratio (PER) values for
rice bran are 1.7–1.9. Rice bran protein concentrates have higher PER values (1.99–
2.19) and are comparable to the values for soybean and cottonseed protein (Leiner,
1972).
The major rice bran proteins are albumins, globulins, prolamins, and acid-soluble
glutelins (Hamada, 1997). These account for 34, 15, 6, and 11%, respectively, of
total bran proteins. Protein concentrates from bran may be prepared by alkaline
extraction and isoelectric precipitation. Maximum solubility occurs at pH 8.0. Con-
centrates contain 50–71.5% protein (Gnanasambandam and Hettiarachchy, 1995).
In addition to protein, rice bran is a source of oil and complex carbohydrates.
Minor constituents important in nutrition include essential fatty acids, tocopherols,
vitamins B1 and B2, and minerals. Some bran components, such as phytic acid and
TI, are antinutritional factors. Phytic acid is of most concern. The activity of other
antinutritional factors is low (Barber and de Barber, 1980).
The effect of stabilized or parboiled rice bran on serum cholesterol is of most
interest (Kahlon et al, 1989; Panikolaw, 1998). Both rice bran and oat bran lowered
serum cholesterol in hypercholesterolemic hamsters. Hegsted et al (1990) and
Gerhardt and Gallo (Anonymous, 1989) reported that stabilized rice bran and oat
bran lowered serum cholesterol in humans. Rice bran is not a rich source of soluble
fiber, suggesting the active agent in rice bran is different from that in oat bran.
Bran and Oil / 581
Fig. 5. Oryzanol composition.
Defatted rice bran showed little or no effect on serum cholesterol. Cholesterol-

lowering effects were concentrated in the oil fraction (Orthoefer, 2001). Oryzanol
has subsequently been found to be the active cholesterol-lowering agent present in
crude rice bran oil (Sharma and Rukmini, 1987). Oryzanol is the ferulic acid esters
of phytosterols and sterol-like compounds (Fig. 5).
Enzyme-active rice bran has been utilized as a bakery additive for improved
dough formation and baked-product performance. Functional applications of stabi-
lized bran include use as a tortilla additive for quality maintenance and a meat
injection for tenderization. Simply adding it to food products, including breakfast
cereals, enhances the dietary fiber content. Bran may be used for protein recovery
(Hamada, 2000a,b), high-fiber ingredients (Abdul-Hamid and Luan, 2000), or as a
source of oil and of minor components such as oryzanol, a group of sterol-like com-
pounds.
Rice bran may be mixed with feed components to a limited extent. Poultry feeds
can contain up to 40% bran. High-bran diets for dairy cows result in increased milk
production; however, the properties of the milk may be affected. Bran may also be
used in pig feeds at up to 20% of the diet for fattening and up to 80% for pregnant
sows. The highest use levels may affect fat quality and result in reduced weight gain
and feed efficiency.
RICE BRAN OIL
Rice bran is the source for rice bran oil (Barber and de Barber, 1980). Efforts to
extract rice bran oil have been made over the past 100 years (Brown, 1903).
Initially, use of rice bran oil in conventional foods was targeted. More recent efforts
have emphasized the nutritional characteristics of the oil. Rice bran oil, depending
on the methods used for extraction and refining, has been found to reduce serum
cholesterol, a primary risk factor in cardiovascular disease (Orthoefer, 1995a).
Cholesterol lowering has been associated with minor components in the oil. This
low-cost oil has been found to have characteristics similar to those of peanut oil in
frying applications. It may also be used in various food applications, nonfood
products, and animal feeds (Orthoefer, 2001).
Although worldwide production of milled rice is near 400 million tonnes (t),
only about 50% of the rice bran from milling is available for oil production, even
though the oil component represents a potential source of revenue (Orthoefer,
1995b, 2001). The estimated worldwide production of rice bran oil for 1998 was
722.2 thousand tonnes (FAOSTAT, 1998). Major producing countries are shown in
Table 11. India is the leading producer, followed by China and Japan. The total
annual production potential for rice bran oil is approximately 3.2 million tonnes.
The potential U.S. production is 80,000 t, with only 15,900 t, being produced
(Orthoefer, 2001). A practical estimate of potential U.S. production is approxi-
mately 50% of the total due to the scattered production of bran among several rice
millers, with most having insufficient bran production to justify oil extraction.
Although several rice millers have announced intentions to produce rice bran oil,
only one major facility produces significant quantities of it (Orthoefer, 1996;
Orthoefer, 2001).
The processing of rice bran oil presents unique challenges to oil refiners. The oil
is usually high in FFA, waxes, and bran fines, each affecting production efficien-
cies. Careful attention to each processing step is required to obtain reasonable
refining yields and oil quality (Orthoefer, 1995a).
The FFA content of rice bran oil depends on the rice storage conditions, length
of time in storage, time from recovery of bran to stabilization of bran, and storage
of extracted oil (Orthoefer, 2001). The oil from bran recovered from properly stored
rice that was stabilized within 1 hr of milling had an FFA content of about 2.5%.
Typically, FFA in crude oil is <5.5%. Higher FFAs tend to dramatically reduce
refined oil yields. Refining losses of nearly four times the FFA are common.
Waxes from rice bran oil are esters of C:16–C:26 fatty acids with saturated
C:24–C:30 alcohols (Juliano, 1985b; Kao and Luh, 1991). Waxes separated from
crude oils have an iodine value of 111–116, FFA of 2.1–7.3, phosphorus content of
0.01–0.15, and melting point of 75.3–79.9°C. The two broad classes of waxes are
TABLE 11
Production of Rice Bran Oila
Country Million Metric Tons
Bangladesh 1.5
Brazil 1.5
Cambodia 4.6
China 90.0
India 472.7
Indonesia 0.15
Japan 65.0
Korea 11.7
Republic of Korea 9.2
Laos 2.6
Burma 17.6
Nepal 7.6
Pakistan 3.7
Sri Lanka 5.5
Thailand 7.8
Vietnam 7.6
Total 722.2
a Source: Food and Agriculture Organization of the United Nations (FAOSTAT, 1998).
Bran and Oil / 583
1) soft wax, melting at 75°C or less, and 2) hard wax, melting at 80°C or higher.
Soft wax is composed of saturated C:24 and C:30 fatty alcohols, saturated fatty
acids of C:16 and C:26, and n-alkanes of C:21 and C:29 (Yoon and Rhee, 1982).
Other fatty alcohols have also been found. Hard wax is made up of C:24, C:26, and
C:30 saturated fatty alcohols; saturated fatty acids of C:22, C:24, and C:26; and n-
alkanes of C:29 and C:31. The wax content of crude oils depends somewhat on rice
variety. Commonly, 2–4% wax is found, with up to 8% wax in some varieties.
Other variables affecting the wax content of crude rice bran oil include milling
techniques, extraction conditions, and extraction method (Kao and Luh, 1991;
Orthoefer, 1994).
Crude oil extracted at 50°C yields two to three times more wax than does oil
extracted at 20°C (Kao and Luh, 1991; Orthoefer, 1994). Waxes tend to emulsify
with refining agents, increasing oil losses (L. Gingras, unpublished data). Waxes
not removed during refining coat the bleaching earths, requiring higher amounts of
bleach clay. Waxes remaining in the oil crystallize during storage, producing a
cloudy oil, thus reducing consumer appeal.
Wax may be removed from oil by chilling and filtration (Chang et al, 1980). The
technique is simply gradual cooling of oil in settling tanks followed by filtration or
centrifugation. Continuous dewaxing by cooling the oil to less than 20°C followed
by filtration through plate and frame filters has been practiced by Japanese oil refin-
ers (Orthoefer, 1995a). Sodium silicate has been used as an aid for dewaxing
(Kinsey and Hunnell, 1969; Sayre et al, 1985). The recovered wax may be further
purified for use as a substitute for carnauba wax in confectionery polishing.
Excessive bran fines in crude rice bran oil result in high refining losses
(Orthoefer, 1994). Suspended fines are affected by milling method, stabilization
technique, and oil extraction method. Crude oil with a high fines content may settle
in storage tanks and refinery pipelines as well as equipment, leading to poor per-
formance and the need for frequent cleaning (Orthoefer, 2001). Wet stabilization
and hydroclones have been found to be effective in minimizing suspended fines in
crude oil (Sayre et al, 1985; Orthoefer, 1995a).
The pigments in crude rice bran oil are carotenes, xanthophylls, chlorophyll, and
protein degradation products (Orthoefer, 1995a). Chlorophyll can reach levels
greater than 20 ppm. The brown pigments from protein degradation are usually
caused by bran stabilization (Sayre et al, 1985). Rice bran oil from parboiled rice
bran and oil from rice stored under improper conditions also contribute to dark-
colored oil. A finished oil color in the range of 1.5–2.0 red (5¼ in. [13.5 cm] cell).
Lovibond is representative of a fully refined oil.
Oil Extraction
The process for hexane extraction of crude rice bran oil is shown in Figure 6.
Stabilization of bran, in addition to enzyme inactivation, assists in improving
extraction efficiency, reducing fines dispersed in crude oil, and sterilizing the bran
(Sayre et al, 1982, 1985; Orthoefer, 1995a).
Wet extrusion produces a more satisfactory bran for oil extraction. The result of
wet extrusion is a porous cake or pellet that easily allows extraction solvent to per-
colate through the bed, making oil extraction more efficient. Pellets are typically 6–
8 mm in diameter. Binding of fines in the pellet is assisted by starch gelatinization
during heating. Fines are less of a problem since agglomeration has occurred.
Fig. 6. Process for rice oil production.
TABLE 12
Crude Rice Bran Oil Compositiona
Lipid Type Percent
Saponifiable lipids 90–96
Neutral lipids 88–89
Triglycerides 83–86
Diglycerides 3–4
Monoglycerides 6–7
Free fatty acids 2–4
Waxes 6–7
Glycolipids 6–7
Phospholipids 4–5
Unsaponifiable lipids 4.2
Phytosterols 43
Sterolesters 10
Triterpene alcohols 28
Hydrocarbons 18
Tocopherols 1
a Source: Orthoefer (2001), used with permission.
Pressing oil from stabilized bran has been practiced (Juliano, 1983, 1985d; Kao
and Luh, 1991). Bran is initially heated in a stacked deck cooker, then passed
through a screw press to expel oil. Other methods have used steam cooking, drying,
prepressing, and expeller pressing. Stabilization of bran and oil recovery occur as a
single-step process. Both screw and hydraulic presses are generally inefficient for
rice bran, leaving high quantities of oil in the bran (55–65%). The advantages of
mechanical pressing of oil include lower capital cost, increased safety, and the pos-
sibility of a niche product for the health-food industry (Orthoefer, 2001).
Rice bran may be solvent-extracted with hexane using commercial oil-extraction
equipment. Hexane extraction may be by batch, battery, or continuous methods
(Juliano, 1983). Shallow-bed and deep-bed extractors work equally well. Hydro-
clones in series are effective in removing most of the fines from crude oil (Gingras,
2000; Orthoefer, 2001).
Bran and Oil / 585
Fig. 7. Unsaponifiables in common vegetable oils. Saff = safflower, coco = coconut. (Reprinted, with
permission, from Orthoefer, 2001)
Solvents such as acetone, ethyl acetate, isopropyl alcohol, and propane also have
been investigated for oil extraction. These solvents are generally not as cost-
effective as hexane.
The temperature of extraction affects the wax content of crude rice bran oil
(Orthoefer, 1994). The lower the extraction temperature, the less wax is in crude oil.
However, with reduced temperatures, waxes may be segregated based on their melt
point.
The composition of a crude rice bran oil is shown in Table 12. The phospholipids
are phosphatidylcholine, phosphatidylethanolamine, and phosphatidylinositol. Gly-
colipids are the major starch lipids. Palmitic, oleic, linoleic, and linolenic acids are
major fatty acids, and galactose and glucose are major sugars. Glycolipids are asso-
ciated with the protein and endosperm starch.
Sterols in rice bran oil represent the majority of the unsaponifiable content. Rice
oil is particularly high in sterols compared to other oils (Fig. 7). Sterols include free
sterols, esters, sterylglycosides, and acylsterylglycosides. More than 4% of rice oil
may be unsaponifiables. Sitosterol is the most abundant sterol.
Rice Bran Oil Processing
Crude rice bran oil typically contains up to 0.5% bran fines and 0.5–5% wax. To
prevent settling of fines, agitated crude oil storage tanks are required. Heat is also
necessary to prevent crystallization of waxes. Heated transfer lines in cooler cli-
mates are required to provide uniform feedstock for refining.
Because of high FFA, wax, and unsaponifiable content of crude rice bran oil,
dewaxing is often the first processing step. Initial dewaxing may simply be gravity
settling followed by decanting (Chang et al, 1980). The foots (soap stock) recovered
may be added back to the defatted bran, sold as animal feed oil, or further processed
for oil recovery and wax purification. Filtration with decanters is also utilized.
Temperatures of <60°C are used. Additional dewaxing may be used during
degumming or alkali refining (Orthoefer, 2001). It is best to remove the waxes
before neutralization to maximize yields. Dewaxing of refined, bleached oil by
cooling to approximately 5°C, followed by filtration, is necessary for production of

a high-grade, chill-proof oil.
The “gums,” or phospholipids, may be recovered as “rice lecithin.” The phos-
pholipids in rice are similar in composition and function to those of other oil
sources. Production of food-grade lecithin requires prior removal of bran fines and
waxes. Regular water degumming techniques may be used. Temperatures above
80°C are required to prevent crystallization and removal of waxes with the gums. If
food-grade lecithins are not being produced, filtration of bran fines is not required.
Pretreatment with phosphoric or organic acid is necessary to remove non-
hydratable phospholipids. Food-grade surfactants may be added to improve wax
removal (Sah et al, 1983). Degumming at less than 70°C assists removal of most
of the waxes. Wet gums may be added to defatted bran as a method for disposal
(Orthoefer, 2001).
Both alkali refining techniques (neutralization) and physical refining techniques
(steam refining) have been used for FFA removal (Orthoefer, 1995a). With alkali
refining, batch or continuous methods may be utilized. Oil may be pretreated with
phosphoric or organic acid for phospholipid hydration. It is then treated with 16–30
Baumé (Bé) caustic with 20–40% excess. The soaps settle and may be recovered as
“soap stock” or “foots” (Chang et al, 1980).
Continuous refining consists of in-line mixers, heaters, and centrifuges
(Orthoefer, 2001). The combined oil plus alkali is rapidly heated to 55–70°C to
assist breaking of the emulsion of hydrated soap in oil. Higher separation tempera-
tures reduce yields. In instances where neutralization is combined with dewaxing,
separation is performed at 28–32°C. Water washing or postneutralization treatment
with silicates to remove final traces of soaps and phospholipids is the same as for
conventional oils. Micella refining, or refining while still in solvent, may also be
used (Bhattacharya et al, 1986). Higher yields and good quality neutralized oil with
less color are advantages of micella refining. Losses are near the absolute amount
(Reddi et al, 1948). Rice oil micella is often variable, requiring constant adjust-
ments. A 5% FFA crude oil has losses ranging from 12–40% by the cup method.
The source of high refining losses is unknown, but partial esters, oxidized compo-
nents, and waxes are the likely causes.
Titration of rice oil with phenolphthalein as the neutralization endpoint includes
not only the titration of the FFA present but also of the oryzanol compounds. An
alternative indicator for titration utilizes alkali blue for titration of FFAs (Orthoefer,
2001). Oryzanol is incorporated in the soap stock after caustic refining.
Standard methods are used for bleaching, hydrogenating, and deodorizing rice
bran oil. Required bleach clay dosages range from 2 to 10% depending on the oil
quality. Required deodorization temperatures may be higher compared with other
oils: 220–250°C under 3–5 mm Hg vacuum is used, which improves oil stability as
well. Semicontinuous deodorizer units are the most common types used. Storage of
finished oil is the same as for other oils.
Removal of the remaining traces of waxes and stearines is necessary for oils
sold directly to consumers. A 5-hr cold test has been targeted. Without win-
terization, dewaxed rice oil is cloudy or turbid even at room temperature (Chang
et al, 1980). Slow cooling to 15°C over a 12-hr period followed by use of high
filtration capacities are needed to accommodate separation of the crystals from
the viscous oil. Filter aids may be utilized. Cold tests of 5–7 hr are near
maximum.
Bran and Oil / 587
As with all oils, co-products represent a significant revenue stream. Waxes may
be concentrated and refined to compete with other organic waxes. The hard, high-
melting waxes are preferred for most applications.
Soap stock contains fatty acid soaps and, for oil refined with caustic, oryzanol (5–
10%). Soaps may be acidulated for use in animal feeds and the oryzanol concentrated
for other markets (Seetharamaiah and Prabhakar, 1986). Diethyl ether is used for
extraction from soap stock followed by alumina chromatography and recrystallization.
Deodorizer distillate contains tocopherols, tocotrienols, and sterols. Its value is
similar to that of other distillates used for preparation of these compounds.
Steam refining, or physical refining that combines deacidification with deodori-
zation, is an option for yield improvement from high-FFA rice bran oil (Orthoefer,
1995a). Physical refining also maintains the oryzanol in the oil. Dewaxed,
degummed, bleached oil is used for feedstock (Nicolosi et al, 1992).
Refined Rice Bran Oil
A typical specification for finished rice bran oil is shown in Table 13. Except for
color, finished-product specifications are similar to those of other oils. Rice bran oil
has a characteristic nutty, earthy flavor not unlike peanut oil.
The fatty acid composition of rice bran oil is most similar to that of peanut or
groundnut oil (Orthoefer, 2001) (Table 14). The most notable difference is in the
long-chain saturated fatty acid (C20:0 and 22:0) content of peanut oil. Palmitic,
oleic, and linoleic fatty acids make up more than 90% of the fatty acid portion of
the glycerides. As with peanut oil, rice bran oil is most suited for general frying and
cooking applications.
Rice Bran Oil Nutrition
The hypocholesterolemic activity of rice bran is in the oil fraction since the
defatted bran has limited cholesterol-reducing capacity (Table 15). A blind cross-
over study showed that ri ce and oat bran reduced serum cholesterol in humans by
TABLE 13
Product Specifications of Refined, Bleached, and Deodorized Rice Bran Oil
Property Typical Value
Iodine value (Wijs method, g/100 g sample) 99–108
Peroxide value (meq/kg) 1.0 max.
Moisture (%) 0.05 max.
Color (5.25-in. Lovibond red) 5.0 max.
Free fatty acids (% as oleic) 0.05 max.
Flavor/odor 7 minimuma
Chlorophyll (ppb) 75 max.
Saponification value (KOH, mg/g) 180–190
Unsaponifiable matter (%) 3–5
Smoke point (°C) 213
Refractive index 1,470–1,473
Specific gravity (g/mL) 0.916
AOMb (hr) 17.5
a On a scale of 10 = “bland.”
b Active oxygen method.
7% (Hegsted et al, 1990). Results were similar for a free-living intervention study
by Gerhardt and Gallo (Anonymous, 1989). Cholesterol reduction occurred
primarily in low-density lipoprotein (LDL) cholesterol, the “bad” cholesterol
(Kahlon et al, 1992). The hypocholesterolemic activity of the oil fraction was
concentrated in the unsaponifiable matter of rice bran oil (Sharma and Rukmini,
1987) (Table 16). In this study, serum cholesterol reduction also occurred primarily
in LDL cholesterol. Increased fecal excretion of bile acids and neutral sterols
occurred when rice bran oil was fed to rats (Sharma and Rukmini, 1986). Oryzanol,
a major component of rice bran oil unsaponifiables, was found to contribute to the
hypocholesterolemic activity of rice oil in primates (Kahlon et al, 1992). Crude rice
bran oil contains between 1 and 3% oryzanol.
A clinical study with 3.1 g/day of rice bran oil unsaponifiables over a 12-month
period resulted in a 14.1% reduction in total cholesterol and 20.5% reduction in
LDL cholesterol (Table 17) (Watkins et al, 1990). High-density lipoprotein (HDL)
cholesterol rose and triglycerides decreased significantly. Tocotrienols, also present
in rice bran oil, have been reported to reduce serum cholesterol (Qureshi et al, 1991,
2000).
The refining method used in rice bran oil production affects the oryzanol content
of finished oil (F. T. Orthoefer and R. J. Nicolosi, unpublished data). With alkali
refining, most of the oryzanol is removed (Fig. 8), whereas with steam or physical
TABLE 14
Chemical Composition of Rice Bran Oil
Physicochemical Parameters Typical Value
Acid value 1.2
Iodine value 100.0
Saponification value 211.8
Unsaponifiable matter 4.2
Fatty Acid Composition Percent
C14:0 0.6
C16:0 21.5
C18:0 2.9
C18:1 38.4
C18:2 34.4
C18:3 2.2
C20:0 …
C22:0 …
TABLE 15
Effect of Rice and Oat Brans on Serum Cholesterol in Hamstersa
Bran in Dietb Serum Cholesterol (mg/dL)
Cellulose (10%) 395
Rice bran (47.8%) 270
Defatted rice bran (24.7%) 347
Parboiled rice bran (31.8%) 297
Defatted parboiled rice bran (19.6%) 377
Oat bran (53.7%) 289
a Source: Kahlon et al (1992), used with permission.
b Plus 0.5% cholesterol.
Bran and Oil / 589
refining, most of the oryzanol (66%) remains in the oil (Krishna et al, 2001). Physi-
cally refined rice bran oil gave a serum lipid response in hamsters similar to that of the
crude rice bran oil. Various refining methods to preserve oryzanol in the oil have been
TABLE 16
Hypocholesterolemic Activity of Unsaponifiable Matter of Rice Bran Oil in Ratsa
Serum Cholesterolc (mg/dL)
Dietb Total HDL LDL + VLDL
Control (peanut oil) (10%) 374 43 331
Rice bran oil (10%) 228 48 240
Control + 0.2% unsaponifiables 387 48 339
Control + 0.4% unsaponifiables 243 48 195
a Source Sharma and Rukmini (1987), used with permission.
b 1% cholesterol added.
c HDL = high density lipoprotein; LDL = low density lipoprotein; VLDL = very low density lipoprotein.
TABLE 17
Effect of Daily Addition of Rice Bran Unsaponifiables (RBN) on Serum Lipids (mmol/L)
in Hypercholesterolemic Subjectsa
Serum Lipidsb Start 12 Months P
RBN
Cholesterol 6.18 ± 0.33 5.31 ± 0.20 <0.05
LDL cholesterol 4.28 ± 0.37 3.40 ± 0.18 <0.05
HDL cholesterol 0.17 ± 0.02 0.24 ± 0.02 <0.025
Triglyceride/HDL 2.16 ± 0.35 1.21 ± 0.21 <0.05
RBN placebo
Cholesterol 5.70 ± 0.21 6.06 ± 0.32 NSc
LDL cholesterol 3.95 ± 0.18 4.05 ± 0.31 NS
HDL cholesterol 0.21 ± 0.06 0.22 ± 0.01 NS
Triglyceride/HDL 1.54 ± 0.31 1.55 ± 0.20 NS
a Source: Watkins et al (1990), used with permission.
b LDL = low density lipoprotein; HDL = high density lipoprotein.
c Not significant.
Fig. 8. Effect of the refining process on the oryzanol content of rice bran oil. (Reprinted, with
permission, from Reddi et al, 1948)
TABLE 18
Frying Evaluation of Rice Oil (15-day results)
Days to Maximum Levelb
Oil Typea FFA FOS LY LR TPM
Rice (without additives) 3.91 3.74 6 28.0 31.9
Rice (with additives) 5.62 3.46 7 49.6 34.6
Peanut (with additives) 6.87 3.92 8 21.2 35.5
Cottonseed (with additives) 7.22 4.07 7 28.8 37.2
a Specifications: 40 lb (18.2 kg) gas fryers; frying temperature 350°F (177°C); hourly rotation: breaded
chicken, fish, onion rings, French fries; additives: 5 ppm dimethyl polysiloxane antifoam, 200 ppm
tertiary butyl hydroguinone.
b FFA = free fatty acids; FOS = food oil sensor; LY = Lovibond yellow; LR = Lovibond red; TPM =
total polar material.
TABLE 19
Frying Results Using Blends of Rice and Soybean Oils
Total Polar Material (%)
Oil Type 10 Days 15 Days
Rice 21.12 32.78
Peanut 21.07 35.53
Rice/soybean, 50:50 24.11 35.80
Rice/soybean, 25:75 23.25 40.42
TABLE 20
Days at 145°F (62.8°C) Before Rancid Odor Is Detected
Oil Type Days to Detect Rancid Odor
Rice (without additives) 20
Rice (with additives) 25
Peanut (without additives) 14
Cottonseed (with additives) 31
attempted. Sodium carbonate instead of sodium hydroxide has been partially

successful. Adding back unsaponifiables to oil has been patented (Hitotsumatsu and
Takeshita, 1994). Clinical trials have not been performed with high-oryzanol rice bran
oil.
Rice Bran Oil Utilization
Rice bran oil is used in food, feed, and industrial applications. Only high-quality
oil is targeted for foods. Rice bran oil is used primarily as a frying oil in Japan,
where its flavor is preferred over those of alternative oils. The oxidative stability of
rice bran oil is equivalent to those of peanut and cottonseed oils in deep frying
applications (Table 18) (Orthoefer, 2001). Blends of rice bran oil with soybean oil
tend to reduce the rate of increase in total polar material, depending on the content
of rice bran oil in the blend (Table 19).
Rice bran oil for frying of potato chips was evaluated for shelf life using the Schaal
oven test (Table 20). Chips prepared with rice bran oil showed flavor and odor stability at
elevated storage temperatures between those of peanut oil and cottonseed oil.
Bran and Oil / 591
Winterized rice bran oil is suitable for making salad dressing and mayonnaise.
Stearines from winterization may be used to replace other plastic fats in margarines
and shortenings. Hydrogenated rice bran oil is applicable to specialty shortening and
margarine formulations.
Primary nonfood uses of rice bran oil include feed formulations, soaps, and glyc-
erin. Wax from rice bran oil has characteristics similar to carnauba wax with uses in
confectionery, cosmetic, and polish products.
SUMMARY
Rice bran is an underutilized co-product of rice milling. Applications include

foods and feeds, where advantage may be taken of its protein, oil, fiber, and nutri-
tional components. Defatted rice bran is currently commercialized as a feed ingre-
dient. Future work is required to develop processes and markets for the protein,
dietary fiber and minor components. Extraction and refining of rice bran oil capture
a portion of the commercial value of the bran. They provide an oil with perform-
ance properties that can compete with those of widely used oils, and this oil has
further opportunities for use as a nutritional oil.
REFERENCES
Abdul-Hamid, A., and Luan, Y. S. 2000. Func- Hamada, J. S. 2000a. Characterization and func-
tional properties of dietary fiber prepared tional properties of rice bran proteins modified
from defatted rice bran. Food Chem. 68:15- by commercial exo proteases and endo prote-
19. ases. J. Food Sci. 65:305-310.
Anonymous. 1989. Two studies show rice bran Hamada, J. S. 2000b. Ultrafiltration of partially
reduces cholesterol levels. Food Chem. News hydrolyzed rice bran protein to recover value-
31(Nov. 13):40. added products. J. Am. Oil Chem. Soc.
Barber, S., and de Barber, C. B. 1980. Rice bran: 77:779-784.
Chemistry and technology. Pages 790-862 in: Harper, J. M. 1989. Food extruders and their
Rice: Production and Utilization. B. S. Luh, applications. Pages 1-15 in: Extrusion Cook-
Ed. AVI Publishing, Westport, CT. ing. C. Mercier, P. Linko, and J. M. Harper,
Bhattacharya, A. C., Majumdar, S., and Eds. American Association of Cereal Chem-
Bhattacharya, D. K. 1986. Edible quality rice ists, St. Paul, MN.
bran oil from high FFA rice bran oil by Hegsted, M., Windhauser, M., and Lester, S.
micella refining. J. Am. Oil Chem. Soc. 1990. Stabilized rice bran and oat bran lower
63:1189-1191. cholesterol in humans. FASEB 4:368A.
Brown, C. A. 1903. Characteristics of rice oil. J. Hitotsumatsu, H., and Takeshita, S. 1994. U. S.
Am. Chem. Soc. 25:948-954. patent 5,290,579.
Chang, S. C., Saunders, R. M., and Luh, B. S. Hunnell, J. W., and Nowlin, J. F. 1972. Solvent
1980. Rice Oil: Chemistry and technology. extractive rice milling. Pages 201-214 in:
Pages 764-789 in: Rice Production and Utili- Rice, Chemistry and Technology, 1st ed. D. F.
zation. B. S. Luh, Ed. AVI Publishing, Houston, Ed. American Association of Cereal
Westport, CT. Chemists, St. Paul, MN.
FAOSTAT. 1998. Database Collections. Food IRRI. 1984. Annual Report for 1983. Interna-
and Agriculture Organization of the United tional Rice Research Institute, Los Banos,
Nations, Rome. (Website: http://apps.fao. Laguna, Philippines.
org/page/collections) Juliano, B. O. 1972. The rice caryopsis and its
Gnanasambandam, R., and Hettiarachchy, N. S. composition. Pages 16-74 in: Rice: Chemistry
1995. Protein concentrates from unstabilized and Technology, 1st ed. D. F. Houston, Ed.
and stabilized rice bran: Preparation and prop- American Association of Cereal Chemists, St.
erties. J. Food Sci. 60:1066-1069. Paul, MN.
Hamada, J. S. 1997. Characterization of protein Juliano, B. O. 1983. Lipids in rice and rice proc-
fractions of rice bran to devise effective meth- essing. Pages 305-330 in: Lipids in Cereal
ods of protein solubilization. Cereal Chem. Technology. P. J. Barnes, Ed. Academic Press,
74:662-668. New York.
Juliano, B. O. 1985a. Rice hull and rice straw. O’Dell, B. L., deBoland, A. R., and Koirtyohann,
Pages 689-755 in: Rice: Chemistry and Tech- S. R. 1972. Distribution of phytate and nutri-
nology, 2nd ed. B. O. Juliano, Ed. American tionally important elements among the mor-
Association of Cereal Chemists, St. Paul, MN. phological components of cereal grains. J.
Juliano, B. O. 1985b. Rice Bran. Pages 647-687 Agric. Food Chem. 20:718-721.
in: Rice: Chemistry and Technology, 2nd ed. Orthoefer, F. T. 1995a. Rice bran oil. Pages 393-
B. O. Juliano, Ed. American Association of 409 in: Bailey’s Industrial Oil and Fat Prod-
Cereal Chemists, St. Paul, MN. ucts, 5th ed. Y. H. Hui, Ed. Wiley & Sons,
Juliano, B. O. 1985c. Criteria and tests for rice New York.
grain qualities. Pages 443-524 in: Rice: Orthoefer, F. T. 1995b. Rice bran oil: Production
Chemistry and Technology, 2nd ed. B. O. and utilization. Food Technol. Europe 1:140-
Juliano, Ed. American Association of Cereal 144.
Chemists, St. Paul, MN. Orthoefer, F. T. 1996. Rice bran oil: Healthy
Juliano, B. O., and Bechtel, D. B. 1985. The rice lipid source. Food Technol. 50(12):62-64.
grain and its gross composition. Pages 17-57 Orthoefer, F. T. 2001. Rice bran oil: Composi-
in: Rice: Chemistry and Technology, 2nd ed. tion, production, nutrition and utilization.
B. O. Juliano, Ed. Am. Assoc. Cereal Chem., Pages 151-158 in: Proceedings of The World
St. Paul, MN. Conference on Oilseed Processing Utilization.
Kahlon, T. S., Saunders, R. M., Chon, F. I., Chiu, R. F. Wilson, Ed. AOCS Press, Champaign,
M. C., and Betschart, A. A. 1989. Effect of IL.
rice bran and oat bran on plasma cholesterol in Ory, R. L., Bog-Hansen, T. C., and Mod, R. R.
hamsters. Cereal Foods World 34:768. 1981. Properties of hemagglutinins in rice and
Kahlon, T., Saunders, R. M., Sayre, R. M., other cereal grains. Pages 159-168 in: Antinu-
Chow, F., Chiu, M., and Betschart, A. 1992. trients and Natural Toxicants in Foods. R. L.
Cholesterol lowering effects of rice bran and Ory, Ed. Food Nutr. Press, Westport, CT.
rice bran oil fractions in hypercholesterolemic Panikolaw, J. 1998. Rice bran oil gains ground
hamsters. Cereal Chem. 69:485-489. on potential health benefits. Chem. Market.
Kao, C., and Luh, B. S. 1991. Rice oil. Ch. 13 in: Rep. 253(24):10.
Rice Utilization, Vol. II. B. S. Luh, Ed. Van Prabhakar, J. V., and Venkatesh, K. V. L. 1986.
Nostrand Reinhold, New York. A simple chemical method for stabilization of
Kinsey, D. N., and Hunnell, J. L. 1969. U. S. rice bran. J. Am. Oil Chem. Soc. 63:644-646.
patent 3,481,960. Qureshi, A. A., Qureshi, N., Wright, J. J., Shen,
Krishna, A. G. G., Khatoon, S., Shiela, P. M., Z., Kramer, G., Gapor, A., Chong, Y. H.,
Sarmandal, C. V., Indira, T. N., and Mishra, A. DeWitt, G., Ong, A. S. H., Peterson, D. M.,
2001. Effect of refining of crude rice bran oil and Bradlow, B. A. 1991. Lowering of serum
on the retention of oryzanol in the retired oil. cholesterol in hypercholesterolemic humans
J. Am. Oil Chem. Soc. 78:127-131. by tocotrienols (palmvitee). Am. J. Clin. Nutr.
Leiner, I. E. 1972. Nutritional value of food 53:10215-10265.
protein products. Pages 203-277 in: Soybeans: Qureshi, A. A., Huanbiao, M. O., Packer, L., and
Chemistry and Technology, Vol. 1. A. K. Peterson, D. M. 2000. Isolation and identifica-
Smith and S. J. Circle, Eds. AVI Publishing, tion of novel tocotrienols from rice bran with
Westport, CT. hypocholesterolemic, antioxidant and antitu-
Luh, B. S. 1991. Rice Hulls. Ch. 12 in: Rice mor properties. J. Agric. Food Chem. 48:3130-
Utilization, Vol. II. B. S. Luh, Ed. Van 3140.
Nostrand Reinhold, New York. Raghavendra Rao, S. N., Narayana, M. N., and
Luh, B. S., Barber, S., and de Barber, C. B. 1991. Desikachar, H. S. R. 1967. Studies on some
Rice bran: Chemistry and technology. Ch. 14 comparative milling properties of raw and
in: Rice Utilization, Vol. II. B. S. Luh, Ed. parboiled rice. J. Food Sci. Technol. 4:150-
Van Nostrand Reinhold, New York. 155.
Morita, Y., and Ida, S. 1968. Studies on the respi- Randall, J. M., Sayre, R. N., Schultz, W. G.,
ratory enzymes in rice kernel. I. Isolation and Fong, R. Y., Mossman, A. P., Tribelhorn, R.
purification of cytochrome C and peroxidase E., and Saunders, R. M. 1985. Rice bran
556 from rice embryo. Agric. Biol. Chem. stabilization by extrusion cooking for
32:441-447. extraction of edible oil. J. Food Sci. 50:361-
Nicolosi, R. J., Rodgers, E. J., Ausman, L. M., 364, 368.
and Orthoefer, F. T. 1992. Rice bran oil and Rasper, V. F. 1979. Chemical and physical char-
its health benefits. Pages 421-437 in: Rice acteristics of dietary fiber. Pages 93-115 in:
Science and Technology. J. Wadsworth and Dietary Fibers: Chemistry and Nutrition. G. E.
W. Marshall, Eds. Marcel Dekker, New Inglett and S. I. Falkehag, Eds. Academic
York. Press, New York.
Bran and Oil / 593
Reddi, P. B. V., Murti, K. S., and Feugi, R. O. Sayre, R. N., Saunders, R. M., Enochian, R. V.,
1948. Rice bran oil. I. Oil obtained by solvent Schultz, W. G., and Beagle, B. C. 1982.
extraction. J. Am. Oil Chem. Soc. 25:206-211. Review of rice bran stabilization systems with
Ryu, C.-H., and Cheigh, H.-S. 1980. Fractiona- emphasis on extrusion cooking. Cereal Foods
tion of rice bran lipid and storage effects on World 27:317-322.
bran lipid composition. Korean J. Food Sci. Seetharamaiah, G. S., and Prabhakar, J. V. 1986.
Technol. 12:278-284. (In Korean) Oryzanol content of Indian rice bran oil and its
Sah, A., Agrawal, B. K. D., and Shukla, L. S. extraction from soap stock. J. Food Sci. Tech-
1983. A new approach in dewaxing and refin- nol. 23:270-273.
ing rice bran oil. J. Am. Oil Chem. Soc. Sharma, R., and Rukmini, C. 1986. Rice bran and
60:466. hypercholesterolemia in rats. Lipids 21:715-
Saunders, R. M. 1990. The properties of rice bran 720.
as a foodstuff. Cereal Foods World 35:632- Sharma, R., and Rukmini, C. 1987. Hypocholes-
636. terolemic activity of unsaponifiable matter of
Saunders, R. M. 1986. Rice bran: Composition rice bran oil. Indian J. Med. Rev. 85:278-281.
and potential uses. Food Rev. Int. 1:465-495. Thompson, S. A., and Weber, C. W. 1981. Effect
Saunders, R. M., Sayre, R. N., Kratzler, H., and of dietary fiber sources on tissue mineral lev-
Calvert, C. 1982. Stabilized rice bran: Nutri- els in chicks. Poult. Sci. 60:840-845.
tional qualities. Page 63 in: Proc. Rice Watkins, T. R., Geller, M., Kooyenga, D. K., and
Technical Working Group. 19. Texas Agric. Bierenbaum, M. L. 1990. Hypocholestero-
Exp. Stn., Texas A&M Univ., College lemic and antioxidant effect of rice bran oil
Station. nonsaponifiables in hypercholesterolemic
Sayre, R. N., Nayyar, D. K., and Saunders, R. M. subjects. Environ. Nutr. Interact. 3:115-122.
1985. Extraction and refining of edible oil Yoon, S. H., and Rhee, R. S. 1982. Composition
from extrusion—Stabilized rice bran. J. Am. of waxes from crude rice bran oil. J. Am. Oil
Oil Chem. Soc. 62:1040-1043. Chem. Soc. 59:561-563.
CHAPTER 20
NUTRITIONAL PROPERTIES OF RICE

AND RICE BRAN
Wallace Yokoyama
Agricultural Research Service
Western Regional Research Center
Albany, California
The nutritional composition of rice and its fundamental importance as the major
source of caloric energy and protein for most of the population of Asia remain
essentially unchanged since the second edition of Rice: Chemistry and Technology
was published in 1985. However, our comprehension of the structure and organiza-
tion of starch, protein, and other components of rice as related to human nutrition
has been extended to include health promotion and disease prevention in addition to
sustenance. As we begin the twenty-first century, we recognize that obesity and
problems of overnutrition, rather than inadequate nutrition, have become epidemic
in developed countries.
A review of the literature concerning the glycemic properties of rice and their
relationship to diabetes is included in this chapter because diabetes is increasing as
the living standards of rice-consuming Asian countries improve. In the past 10
years, the role of plant sterols in reducing human plasma cholesterol has become
widely accepted. This chapter provides an overview of rice phytonutrients and other
components that contribute to the hypocholesterolemic properties of rice bran and
its oil. These phytonutrients and structural features of the grain may also reduce the
risk of cardiovascular disease, cancer, and diabetes; relevant research findings are
highlighted here.
GLYCEMIC PROPERTIES
Milled Rice
Rice and other cereal grains are consumed primarily for their caloric energy
value, typically 4 kcal/g. Starch, the carbohydrate energy component of grains,
accounts for about 90% of the dry weight of milled rice.
Cooking rice gelatinizes the starch, facilitating its digestion. Digestion begins in
the mouth, where salivary a-amylases hydrolyze starch into maltose and malto-
triose, and it continues in the stomach. Digestion then proceeds in the small intes-
tine, where pancreatic a-amylases aided by glucoamylases hydrolyze maltose to
595
glucose. The glucose is absorbed into the bloodstream and used for energy or stored
as glycogen.
The total amount of rice starch digested and absorbed into the body and the rate
at which this process occurs have important nutritional consequences. In developing
countries, where rice accounts for most of the daily caloric intake, maximum starch
digestibility is necessary for survival. In developed countries, undigested starch may
be considered a form of dietary fiber that contributes to colonic health. In general,
the digestibility of cooked rice starch is complete or nearly complete (Juliano,
1985). Using the breath-hydrogen method, Strocchi and Levitt (1991) found intesti-
nal absorption of rice starch in humans to be complete, compared with 90–95%
absorption of wheat, potato, and corn starches.
However, the rate of digestion can vary. The rates of starch digestion and glu-
cose absorption are of considerable interest because of their relationship to blood
glucose control in people with diabetes and the possibility of delaying noninsulin-
dependent diabetes mellitus (NIDDM), also called adult-onset (type II) diabetes.
The glycemic index (GI) is often used to assign a single numeric value to the
concentration of blood glucose after a carbohydrate meal is consumed (postprandial
blood glucose) (Wolever, 1990). The GI after a 50-g meal of bread or glucose is
scaled to 100 GI units.
Miller et al (1992) found that a high-amylose (28%) rice variety had a low GI
(64), whereas two lower-amylose (20%) varieties had a high GI (83 and 93). Simi-
lar results were reported by Goddard et al (1984), who evaluated waxy (0%) and
low- (14–17%), and high- (23–25%) amylose-containing rice varieties. They
reported a lower glucose response (plasma glucose about 112 mg/dL) at 30 min for
the high-amylose variety compared with that for the 0% amylose variety (plasma
glucose 122 mg/dL). However, even among rice varieties with the same amylose
content (26.7–27%), GI has been found to vary (61, 72, 91) (Panlasigui et al, 1991).
Cooking time and procedure may alter GI. When Panlasigui et al (1991) cooked
the three high-amylose rice varieties (26.7–27.0% amylose) under the same condi-
tions, they had GI values of 61, 72, and 91. However, when cooking conditions
were changed to the minimum cooking time for each variety of rice, no significant
differences were found in GI (75, 78, and 81). Gatti et al (1987) showed that the 60-
min incremental area under the blood glucose curve for baked cooked rice fed to
healthy volunteers was lower than that obtained by feeding boiled rice.
Processing is another factor affecting GI. Several studies (Jenkins et al, 1988;
Casiraghi et al, 1993; Larsen et al, 2000) on healthy subjects and on subjects with
diabetes have shown that parboiling reduces the glycemic response to rice. Larsen et
al (2000) report that three degrees of parboiling—none, traditional (mild), and
severe-pressure parboiling—resulted in GIs of 55, 46, and 39, respectively, with
severe-pressure parboiling significantly different from no parboiling in subjects
with type II diabetes. Similarly, the extrusion of noodles prepared from a high-
amylose variety resulted in a GI that was lower (58) than that of cooked noodles of
milled rice (91), both in healthy subjects and in subjects with type II diabetes
(Panlasigui et al, 1992).
Generally, food form also affects GI. Size reduction of the grain into flour
increases the surface area and accessibility to digestive enzymes, resulting in lower
GI (Crapo and Henry, 1988; Hallfrisch and Behall, 2000).
The digestibility of rice starches can be reduced by food emulsifiers (Guraya et
al, 1997). Five parts of cooked nonwaxy or waxy starch were combined with 30
Nutritional Properties / 597
different emulsifiers. Human a-amylase was used to digest the mixture, and the
released maltose was measured. The greatest decreases in digestibility of nonwaxy
and waxy starches were 30.6 ± 19.5% and 33.7 ± 14.6%, respectively.
Rice Bran
Improvements in blood serum glucose and insulin after consumption of rice-bran

water solubles or rice-bran fiber have been reported (Qureshi, 2002). The subjects,
who were males between the ages of 20 and 65, were recruited from military and
civilian personnel at a military installation and had either type I (juvenile) or II dia-
betes. Subjects consumed 20 g per day of the test materials. Subjects with type I
diabetes showed no differences in glycosylated hemoglobin. Fasting glucose was
reduced from baseline, 9.67 ± 1.36 mmol/L, to 6.89 ± 1.26 mmol/L after 60 days of
consumption of rice bran-water solubles. The rice solubles contained 7.5% protein,
6.0% dietary fiber, 26.5% fat, 270 ppm tocols (mainly tocotrienols), and 2,600 ppm
oryzanol. Rice-bran-fiber concentrates also reduced fasting serum glucose from
baseline, 9.04 ± 1.40 mmol/L, to 7.30 ± 1.50 mmol/L after 60 days. The fiber frac-
tion contained 20.5% protein, 42.0% dietary fiber, 13.5% fat, 30 ppm tocols, and
2,400 ppm oryzanol. In subjects with NIDDM, rice-bran water solubles reduced
glycosylated hemoglobin from 10.69 ± 1.97% to 9.05 ± 1.68% after 60 days. Rice
bran, rice-bran water solubles, and rice-bran fiber reduced fasting glucose by 10, 33,
and 22%, respectively. Insulin was not changed by any of the rice-bran diets.
LIPEMIC PROPERTIES
Milled Rice
The lipemic indices of rice, raggi, and tapioca were determined in human sub-
jects (Kurup, 1993). The equivalent of 400 kcal of rice (115 g), raggi (120 g), or
tapioca (214 g) was served as part of an evening meal for 15 days. On the 16th day,
75 g of oral glucose was administered, and blood samples were collected. The
percentage changes from baseline were summed for total plasma cholesterol, high-
density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol,
and plasma triglyceride. This sum was added to or subtracted from 100 to calculate
the lipemic index. The lipemic index determined by this method was 121 for rice,
55 for raggi, and 57 for tapioca. These results suggest that rice increases undesirable
lipid parameters. However, none of the changes between baseline wheat diets and
the test diets were significant. For example, the largest percentage change in all of
the results was the change in triglycerides in the rice group. The wheat baseline
plasma triglyceride was 156.0 ± 27.5 mg/dL and that for rice was 184.0 ± 18.0
mg/dL. Although the increase was 18%, the difference was not significant.
Additionally, total plasma cholesterol contains both HDL and LDL cholesterol val-
ues, so they were counted twice. HDL cholesterol increases are beneficial and per-
haps should have been subtracted.
Rice Bran
Postprandial lipemia in healthy adults consuming test meals containing 4.2 g of

total dietary fiber from oat bran, rice bran, wheat fiber, or wheat germ resulted in
improvements in some serum lipid parameters (Cara, 1992). The results were based
on the consumption of 70.0–79.4 g of fat, 756 mg of cholesterol, 121.3–139.3 g of

carbohydrates, and 40.6–51.6 g of proteins in the control or test meals by six male
volunteers. Blood was collected every 0.5 hr for 5 hr, and hourly during the sixth
and seventh hours. The area under the D-serum triglyceride curve was significantly
lower for the oat bran, wheat fiber, and wheat germ meals. No differences in serum
insulin were found during the 6-hr period following the test meals. After 6 hr,
subjects consuming the rice bran meal had the lowest serum glucose, which was
significantly different from that of the control.
HYPOCHOLESTEROLEMIC PROPERTIES
Rice Bran
HUMAN STUDIES
In two human studies, consumption of rice bran (60 g for four weeks and 15 g
for three weeks) resulted in 6 and 10% reductions in plasma triglycerides, respec-
tively (Kestin et al, 1990; Sanders and Reddy, 1992). In the first study by Kestin et
al (1990), 24 male subjects with mild hypercholesterolemia were fed approximately
60 g of rice bran for four weeks. The same subjects were also fed about 95 g of oat
bran and 35 g of wheat bran for four weeks each. These intakes of rice, oat, and
wheat brans resulted in the same daily dietary fiber intake of about 12 g per day.
Only oat bran lowered total and LDL plasma cholesterol. The second study by
Sanders and Reddy (1992) compared the consumption of 15 g of wheat bran, 15 g
of rice bran, and 30 g of rice bran for three weeks each. The volunteers were 18
healthy males. Both studies evaluated the effects of dietary fiber in an amount of
rice bran that could be realistically consumed on a daily basis. Although plasma
cholesterol was not affected, even the low amounts of fiber consumed, 3 g per day,
in the Sanders and Reddy study resulted in a significant physiological response
(10% reduction in plasma triglycerides).
At higher levels of rice bran intake and in a diet more typical of the fat intake in
the United States, plasma cholesterol is lowered. The intake of 100 g of rice bran or
oat bran for three weeks by 11 human subjects, 10 male and one female, on diets
containing 37% fat and 300 mg of cholesterol per day, resulted in a 7% reduction of
total plasma cholesterol for both rice and oat brans and 10 and 11% reductions of
LDL cholesterol for rice and oat brans, respectively (Hegsted et al, 1993). The free-
living study included two three-week periods in a crossover design, with two weeks
of a control diet between the bran diets. The diets included 36 g of dietary fiber
from rice bran or 28 g from oat bran per day. The LDL-HDL cholesterol ratio was
also improved.
A more recent study by Gerhardt and Gallo (1998) compared full-fat rice bran to
oat bran in a human study of 13–17 subjects per treatment in a six-week, double-
blind, noncrossover design. The mildly hypercholesterolemic subjects on rice bran
and oat bran reduced their LDL cholesterol by 13.7 ± 2.8 and 17.1 ± 2.4%, respec-
tively, compared with subjects on the control diet. The subjects consumed 84 g of a
bran or rice starch placebo product per day. The dietary fiber contents of the rice
and oat bran products were 19.7 and 8.3 g per 84 g of product, respectively. The a-
tocopherol contents of the rice and oat products were 7.79 and <0.4 mg,
respectively. The magnesium and potassium contents of the 84 g of rice bran
product were 544.3 and 1,276.8 mg, respectively. The latter values indicate a
calculated use of 8.7–12.4 g of rice bran per 84 g of product, based on the range of
magnesium in rice bran (8,650–12,300 ppm) reported by Saunders (1990).
ANIMAL STUDIES
The hypocholesterolemic properties of rice bran have also been studied in pigs,
rats, hamsters, and chickens. The feeding of rice bran or rice bran plus rice oil to 20
adult male pigs was reported by Marsono et al (1993). The rice bran diet contained
163 g of rice bran and 70 g of palm oil per kilogram of diet. Rice-bran oil was sub-
stituted for the palm oil in the rice-bran-plus-rice-bran-oil diet. Rice bran alone did
not reduce plasma cholesterol significantly compared with values for animals fed
white or brown rice (2.08 vs. 2.25 and 2.28 mmol/L, respectively). The diet with
rice bran plus rice oil reduced plasma cholesterol significantly (1.93 mmol/L) com-
pared with values for the animals fed white rice and brown rice.
Rats were fed rice bran or wheat bran in diets containing dietary fiber at 70 g/kg
of diet from the brans, with or without fish oil at 50 g/kg of diet. Six male Wistar
rats per treatment were fed for 10 days. Rats fed rice bran or rice bran with fish oil
had lower total and HDL cholesterol than rats fed the wheat bran diet (Topping et
al, 1990). Total triacylglycerols were lower in rats fed the rice-bran-plus-fish-oil
diet compared with rats fed the wheat bran. The LDL and HDL receptor activities
of the rice bran and rice-bran-plus-fish-oil groups were the same. The LDL receptor
activity of both groups fed rice bran were higher than that of both groups fed wheat
bran. The authors cite the differences between down-regulation of the LDL receptor
by fish oil with wheat bran and the lack of effect with rice bran. Hepatic synthesis
of fatty acids was lower in the rice-bran-plus-fish-oil group compared with that in
both groups fed wheat bran, but hepatic synthesis was higher in the rats fed rice
bran than in the rats in both wheat-bran groups.
The hypocholesterolemic effect of rice bran fed to hamsters is reported in several
studies by Kahlon et al (1992, 1996). In a dose-response study (Kahlon, 1992) of
diets containing 10.9, 21.8, 32.8, or 43.7% full-fat rice bran, 35% defatted rice bran,
and 9% rice bran, which was fed to four-week-old male golden Syrian hamsters, the
43.7% diet lowered total plasma and liver cholesterol. All diets contained 10% fiber
from combinations of rice bran and cellulose, and hypercholesterolemia was
induced by feeding 0.3% dietary cholesterol. In a subsequent study by Kahlon et al
(1996), the feeding of 48% rice bran, either raw or stabilized, did not result in
differences in plasma cholesterol or triglycerides from the values shown by control
animals. The controls were fed 8.1% soy protein and 9–10% peanut oil as
substitutes for the vegetable protein and oil in rice bran. Vegetable proteins are
generally hypocholesterolemic compared with animal sources of protein. The soy
protein in the control diet may have lowered plasma cholesterol in the hamsters fed
the control diets, resulting in lack of difference.
Rice bran was also hypocholesterolemic in single-comb white leghorn chickens
fed diets containing 60% full-fat rice bran and 0.5% cholesterol (Newman et al,
1990). Total cholesterol, triglycerides, HDL cholesterol, and LDL cholesterol were
significantly different from the values for a soy-corn control. In a second
experiment using white leghorn chicks fed 0.5% cholesterol, 18% protein, 14.47%
dietary fiber, and 10.78% lipid from diets containing full-fat rice bran, defatted rice
bran, or corn-soy blend, the LDL cholesterol values were 249, 318, and 275 mg/dL,
respectively. These studies of full and defatted rice bran suggested that the lipid
fraction was responsible for the majority of the hypocholesterolemic property.
Rice Bran Oil
HUMAN STUDIES
In the late 1960s, the Japanese and Indian scientific literature reported feeding
studies in humans and animals of rice bran oil and its effects on blood lipid
parameters. Suzuki and Oshima (1970) reported that blends of rice bran and saf-
flower oils (100:1, 85:15, 70:30, 50:50, and 1:100) fed to 50 young, healthy girls for
seven days resulted in a 26% plasma cholesterol reduction for the 70:30 blend.
Raghuram et al (1989) conducted a free-living study of rice bran oil with patients
with high cholesterol (>225 mg/dL) or high triglycerides (>190 mg/dL). The
experimental group of 12 subjects, nine men and three women, were given rice bran
oil as their cooking oil, and the control group of nine subjects, eight men and one
woman, used their habitual oil. Blood samples were collected at the start and at 15
and 30 days into the study. Edible-oil consumption was estimated from food ques-
tionnaires as 35.2 ± 4.72 and 38.9 ± 5.71 g per day for the rice-bran-oil and habit-
ual-oil groups, respectively. Serum cholesterol was lowered from an initial value of
247.3 ± 10.6 to 204.0 ± 6.6 and 182.7 ± 8.4 after 15 and 30 days, respectively. The
serum cholesterol level of the control group was not changed (244.0 ± 14.0 and
249.6 ± 16.1, initial and at 30 days, respectively). Serum triglycerides of the group
using rice bran cooking oil were also reduced from an initial value of 349.8 ± 42.4
to values of 236.5 ± 31.9 and 212.9 ± 20.0 after 15 and 30 days, respectively.
In another study of 15 middle-aged and elderly subjects ranging in age from 44
to 78 years, with elevated LDL cholesterol ranges of 133–219 mg/dL, Lichtenstein
et al (1994) found no differences in total or LDL cholesterol when the subjects
consumed rice bran, canola, corn, or olive oil. The subjects were placed on a
National Cholesterol Education Panel Step 2 diet consisting of 17% calories from
protein, 53% from carbohydrate, and 30% from fat, along with 80 mg of cholesterol
per 1,000 kcal of diet, for 32-day periods. Rice bran oil contained the highest
amount of saturated fatty acids (22.54%) and the highest amount of palmitic acid
(20.16%).
The hypolipidemic effects of rice bran oil have also been recently reviewed by
Rukmini and Raghuram (1991), Cicero and Gaddi (2001), Sugano et al (1999), and
Sugano and Tsuji (1997).
ANIMAL STUDIES
The effects on plasma lipids of feeding rice bran oil to nonhuman primates have
been studied by Nicolosi et al (1991) and Wilson et al (2000). In the first study
(Nicolosi et al 1991), nine adult male cynomolgus monkeys, ages five to eight
years, were fed diets containing 42% kcal as fat from rice bran oil, butter oil, corn
oil, olive oil, coconut oil, or safflower oil. The monkeys were fed a stabilization diet
for four weeks followed by nine weeks on one of eight test diets. All diets contained
0.1% dietary cholesterol. The control diet contained 100 g of rice bran oil, 81 g of
butter oil, and 32 g of safflower oil. The other diets contained, 168, 29, 37, 96, 40,
74, 27, and 0 g of rice bran oil per kilogram of diet. Total cholesterol, LDL choles-
terol and apo B were shown to be negatively correlated with the percent total calo-
ries from rice bran oil in the diet. In a subsequent study of nine cynomolgus mon-
keys, also ages five to eight years, fed diets containing rice bran oil, canola oil, or
corn oil at 95 g/kg of diet, Wilson et al (2000) evaluated predicted plasma choles-
terol using a Hegsted-type (Hegsted et al, 1965) expected-serum-cholesterol value
calculated from data pooled from 454 experimental periods. The control diet was
the “average American diet” containing 36% of energy as fat with 15% saturated
fatty acids, 15% monounsaturated fatty acids, and 6% polyunsaturated fatty acids.
The control diet contained fat at 177 g/kg of diet, while the test diets contained fat at
143 g/kg of diet. The total and LDL cholesterol values for the control diet were 312
± 131 and 247 ± 145 mg/dL, respectively. The total and LDL cholesterol plasma
concentrations for the monkeys on the rice-bran-oil diet were 237 ± 106 and 172 ±
128, respectively. Monkeys on the corn-oil and canola-oil diets had total and LDL
cholesterol values similar to those of monkeys on the rice-bran-oil diet. A
comparison of the actual decrease in total and LDL cholesterol values with the
predicted values from the quantitative calculation based on the Hegsted-type
equation for the cynomolgus monkey suggested that the actual decrease was greater
than that expected from the fatty acid composition of the rice-bran-oil diets. Other
components of rice bran oil may have contributed to the trend toward lower plasma
cholesterol concentrations.
Rats fed rice bran oil had decreased total serum cholesterol compared with rats
fed groundnut oil (Sharma and Rukmini, 1987). Six male albino rats, 80–90 g, were
assigned to one of five test diets and fed for seven weeks. Two control diets con-
taining either groundnut oil or rice bran oil (100 g/kg) were used. The three test
diets contained 1% cholesterol and either groundnut oil, rice bran oil, or rice bran
oil plus 0.5% oryzanol. Body weight was not different between diet treatments
except for higher weight in the oryzanol-fed group. The animals fed cholesterol had
higher liver weights. The rats fed rice bran oil and cholesterol had lower total cho-
lesterol (134 ± 8.7 vs. 215.7 ± 16.1 mg/dL) than animals fed groundnut oil and
cholesterol. Compared with rats fed rice bran oil and cholesterol, rats fed rice bran
oil, cholesterol, and oryzanol had lower total cholesterol (107.2 ± 4.2 vs. 134.0 ±
8.7 mg/dL). HDL was not different between the cholesterol-fed groups. The cho-
lesterol fraction containing LDL plus very-low-density lipoprotein (VLDL) was
highest in the groundnut cholesterol group, compared with that for the animals fed
rice bran cholesterol or rice bran cholesterol plus oryzanol (190.8 ± 19.6, 115.5 ±
9.5, and 94.9 ± 5.2, respectively). This resulted in an increased LDL-HDL ratio for
the groundnut-cholesterol group compared with that for the two rice-bran-
cholesterol groups (17.93 ± 1.89 vs. 8.91 ± 1.05 and 6.17 ± 0.44, respectively). Free
and esterified cholesterol was elevated by cholesterol feeding in all three test
groups. Esterified hepatic cholesterol was lower in the rice-bran-oil-cholesterol
group compared with the groundnut-oil-cholesterol group. The rice bran oil used
contained 0.1% oryzanol, so using oil at 100 g/kg of diet contributed 0.1 g of
oryzanol. The addition of oryzanol (at 5 g/kg of diet) had less effect on cholesterol
lowering than did rice bran oil, compared with groundnut oil.
ADDITIONAL EFFECTS: INCREASED NEUTRAL STEROL

AND BILE ACID EXCRETION
Total serum cholesterol was lowered in rats fed diets containing 10% groundnut
or rice bran oil for 10 weeks (Sharma and Rukmini, 1986). The diets included 1%
cholesterol or no cholesterol. Rats fed the rice-bran-oil diet containing 1% choles-
terol had total serum cholesterol concentrations of 182 ± 18 mg/dL, vs. 287 ± 19
mg/dL for the groundnut-with-1%-cholesterol control group. Total lowering of
serum cholesterol was not significant between the two oil groups without choles-
terol. Most of the serum cholesterol reduction occurred in the VLDL+LDL fraction.
The VLDL+LDL fraction of the cholesterol-fed groundnut-fed group was 225 ± 20

mg/dL, significantly higher than the same serum fraction of the animals fed rice
bran oil and cholesterol (109 ± 20 mg/dL). Sharma and Rukmini also reported that
neutral sterol excretion was higher in the rice-bran-cholesterol-fed group compared
with the groundnut-cholesterol-fed animals (61.5 ± 5.3 and 26.1 ± 6.4 mg/dL,
respectively). Total bile acid excretion was also higher for the group fed rice-bran
cholesterol compared with the group fed groundnut cholesterol.
PROCESSING OF RICE BRAN OIL: EFFECT ON

CHOLESTEROL-LOWERING PROPERTIES
The effects of processing of rice bran oil on plasma lipids were examined by
Sarkar and Bhattacharyya (1991). Six male Charles Foster rats were fed purified
diets containing 20% fat from crude rice bran oil (4.5% unsaponifiable matter), rice
bran oil 2 (1.1% unsaponifiable matter), or rice bran oil 3 (1.6% unsaponifiable
matter). Compared with either the groundnut control or the crude rice bran oil, rice
bran oils 2 and 3 had lower total serum cholesterol. In the liver, rice bran oil 2 had
lower total cholesterol than the crude rice bran oil. Kahlon et al (1996) found that
unsaponfiables improved the cholesterol-lowering properties of rice bran. A large
body of evidence suggests that sterols and phytochemicals in rice bran with sterol-
like structures are most likely the cholesterol-lowering components. The greater
cholesterol-lowering property of the rice bran oils low in unsaponifiable matter is
contrary to most studies. The most likely explanation is that the sterol composition
of the unsaponifiable matter in the experiment by Sarkar and Bhattacharyya may
have differed between the three bran oils, so that the degree of cholesterol lowering
cannot be attributed to amount of unsaponfiables.
PHYTOSTEROLS, DIMETHYLSTEROLS, TRITERPENE ALCOHOLS,

AND OTHER COMPONENTS
Since the hypocholesterolemic properties of rice bran oil cannot be explained by
its fatty acid composition, attention has been focused on the unsaponifiable fraction
of rice bran oil. This fraction contains sterols, sterol ferulates, stanols, dimethyl-
sterols, triterpene alcohols, and related compounds. Plant sterols have been shown
to interfere with the absorption of cholesterol in ileostomy patients, whose small-
intestinal cholesterol content was higher when sterol and stanol esters were present
in their food (Normen et al, 2000). The acceptance of the cholesterol-lowering
properties of margarines containing stanol esters is due to a year-long study of 153
mildly hypercholesterolemic subjects in Finland (Miettinen, 1995). Subsequent
studies showed that both sterol and stanol esters were similarly effective in mildly
hypercholesterolemic Finnish subjects (Weststrate, 1998; Hallikainen, 2000). These
studies and others resulted in the U.S. Food and Drug Administration allowing
cholesterol-reducing health claims on sterol- and stanol-containing margarines
(Federal Register, 2000). The hypocholesterolemic properties of rice bran, primarily
due to sterol components, have recently been reviewed (Cicero, 2001).
Sharma and Rukmini (1987) reported that rats fed diets containing either 10%
rice bran oil or 10% groundnut oil supplemented with 0.4% rice bran unsaponifi-
ables had lower total and VLDL+LDL serum cholesterol fractions. Eight male
albino rats were fed the test diets for eight weeks. The unsaponifiable fraction was
42% sterols: campesterol, stigmasterol, sitosterol, avanasterols, and cholesterol.
Supplementation of 10% groundnut oil with 0.2% unsaponifiable matter did not
result in changes in plasma lipids. Liver cholesterol and triglycerides were also
lowered by the diets of rice bran oil and groundnut oil supplemented with 0.4%
unsaponifiables.
Plasma lipid concentrations were affected by rice-bran-oil unsaponifiable frac-
tions in hamsters (Kahlon et al, 1996). Hamsters were fed combinations of rice bran
or rice bran supplemented with an equivalent amount of the unsaponifiable matter
found in rice bran so that the total unsaponifiable matter was twice as high as in
typical rice brans. The amount of unsaponifiable matter added was 0.4% of the diet.
The amount of rice bran was 48% of the diet. Rice bran alone did not reduce plasma
cholesterol or triglycerides. Stabilized rice bran or stabilized rice bran combined
with an equivalent amount of unsaponifiable matter present in the same weight of
rice bran significantly decreased total plasma cholesterol compared with that for
cellulose-fed animals.
Oryzanol (a mixture of cycloartenyl ferulate, 24-methylene cycloartanyl ferulate,
campesteryl ferulate, and other ferulates) has been determined to constitute as much
as 0.3% of rice bran (Shin et al, 1997). Oryzanol and tocols in both raw and stabi-
lized rice bran are unstable, and the concentrations drop to less than 50% of starting
concentrations upon ambient storage. The concentration of oryzanol in refined rice
bran oils varies due to processing, as reported in a study of five oils, in which it
ranged from 115 to 787 ppm (Rogers et al, 1993). The variability in the concentra-
tion of oryzanol and other phytochemicals may account for differences in experi-
mental results in studies of the hypocholesterolemic or other properties of rice bran
oils.
A review (Sugano, 1997) of the hypocholesterolemic properties of rice bran
oil and in particular sterols and triterpene alcohols (oryzanol), cites Nakamura’s
(1967) research on oryzanol’s effect on hepatic cholesterol biosynthesis and fecal
excretion of cholesterol metabolites in mice. More recently, Yoshino et al (1989)
studied the effects of 300 mg of g-oryzanol daily in 67 hyperlipidemic human
subjects. In type IIa subjects (initial plasma cholesterol 251 ± 7 mg/dL), plasma
cholesterol was lowered 12% after four weeks, 16% at eight weeks, and 7% at 12
weeks. In type IIb subjects (initial plasma cholesterol 268 ± 8 mg/dL), plasma
cholesterol was lowered 13% at four weeks, 19% at eight weeks, and 11% at 12
weeks. LDL cholesterol was lowered significantly in both types of subjects after
12 months. HDL cholesterol was unchanged in type IIa subjects but increased in
type IIb subjects after 12 months. The authors attribute the decrease in plasma
cholesterol to inhibition of intestinal cholesterol absorption, citing the work of
Mitani et al (1983).
The roles of plant sterols and triterpene alcohols were compared by feeding
human subjects either concentrates from rice bran oil high in plant sterols or sheanut
oil high in triterpenes (Vissers, 2000). The rice bran oil and sheanut oil concentrates
were presented in a sunflower oil-based margarine to 60 subjects for three weeks in
a crossover, double-blind design. The total sterol contents of the rice bran- and
sheanut-based margarines were 7.3 and 0.4%, weight basis, respectively. The triter-
pene content of the sheanut-based margarine was 8.9%, but that of the rice-bran-
based margarine was not reported. The rice bran margarine contained 2.9 g of fer-
ulic acid, and the sheanut margarine contained 2.1 g of cinnamic acid. The subjects
were asked to consume 28.9 g of the margarine daily. The mean daily intake of
plant sterols was 2.1 g from rice bran oil and 2.6 g from sheanut oil. The total serum
and LDL cholesterol values of the subjects consuming the rice-bran-oil margarine
decreased 0.19 and 0.20 mmol/L, respectively. These are 5 and 9% reductions,
respectively. LDL cholesterol was reduced 5% in males in the sheanut-consuming
group, but no significant differences were found when the genders were combined.
Another human study of rice bran sterols, soy sterol esters, and sitostanol esters fed
to 100 normal and mildly hypercholesterolemic subjects resulted in plasma
cholesterol reductions of 8–13% for margarines containing soy sterol esters and
sitostanol esters (Weststrate and Meijer, 1998).
Oryzanol was fed to albino Wistar rats for three, five, and seven weeks, and total
serum cholesterol was reduced by the five- and seven-week feedings
(Seetharamaiah and Chandrasekhara, 1988). Male rats were fed a control diet con-
taining 1% cholesterol and 0.15% bile salts or the test diets containing either 0.2,
0.5, 1.0, or 2.0% oryzanol. After five weeks, the control group had total serum
cholesterol of 311.8 ± 5.5 mg/dL and the total serum cholesterol of the oryzanol-fed
animals was lower and similar to each other, ranging from 224 to 270 mg/dL. After
seven weeks, the control group had total serum cholesterol of 210.4 ± 10.4, and the
oryzanol-fed groups were again lower and similar, ranging from 157 to 160 mg/dL.
Either age or diet resulted in decreased serum cholesterol with length of time of the
study.
Cholesterol absorption and fecal excretion of bile acids were studied in albino
rats (Seetharamaiah and Chandrasekhara, 1990). Groups of 12 rats were fed a con-
trol diet (group 1), a control diet plus 0.5% oryzanol (group 2), a control diet plus
1% cholesterol plus 0.15% bile salts (group 3), or diet 3 plus 0.5% oryzanol (group
4). The animals were fed for seven weeks, and feces were collected the last week.
Rats on diets containing 0.5% cholesterol, groups 3 and 4, had higher fecal
excretion of cholesterol than rats not fed cholesterol-containing diets. Cholic and
deoxycholic acid excretion was higher in the animals fed bile salts compared with
the controls, group 1. The rats in group 4, fed cholesterol, bile salts, and oryzanol,
excreted more cholic and deoxycholic acid than animals fed cholesterol and bile
salts. Cholesterol absorption was lower in the animals from group 4 than the
animals in group 3 (43.8 ± 3.3 and 54.5 ± 2.9%), respectively. The researchers
concluded that oryzanol decreased cholesterol absorption by 20% and increased
fecal excretion of cholesterol by 28% and bile acid excretion by 29%. In contrast to
these results, Shinomiya et al (1983) report that rats fed 0.5 and 2% g-oryzanol had
higher total LDL and VLDL cholesterol than animals fed similar diets without g-
oryzanol.
Hamsters fed 1% oryzanol in a chow diet containing 0.1% cholesterol and 5%
coconut oil resulted in lowered total and non-HDL cholesterol fractions (Rong et
al, 1997). Sixteen male F1B hamsters were assigned to the two diet groups after
2.5 weeks on the hypercholesterolemic diet. Plasma non-HDL cholesterol values
were used to make the assignments. The animals were then fed for 8.5 weeks.
Total plasma cholesterol was 342 ± 32 and 245 ± 6 mg/dL for the control and
oryzanol-fed groups, respectively. The decrease was due to the non-HDL
fraction. Plasma triglyceride was also lower in the oryzanol-fed animals (158 ±
10 mg/dL) compared with the control animals (237 ± 45 mg/dL). Cholesterol
absorption was lower in the oryzanol-fed group (42.9 ± 2.0%) compared with the
control group (57.3 ± 2.1%). The fatty-streak area in the aortas of the oryzanol-
fed animals was lower (5 ± 1 µm2/mm2¼103), compared with the control group (15
± 3 µm2/mm2¼103). Fatty streaks indicate the early stages of atherosclerosis
development.
OTHER COMPONENTS OF RICE
Effects on Cholesterol
TOCOTRIENOLS
A tocotrienol-rich fraction from heated rice bran was fed to hypercholes-
terolemic subjects at the Armed Forces Institute of Pathology, Pakistan (Qureshi,
2002) for 35 days. The tocotrienol-rich fraction included 39.4% g-tocotrienol and
20.9% D-desmethyl plus D-didesmethyl tocotrienols. Ninety subjects with serum
cholesterol levels of >6.5 mmol/L were divided into five groups and given capsules
containing 0, 25, 50, 100, or 200 mg of the tocotrienol-enriched fraction, to be taken
daily. The subjects were placed on an American Heart Association Step-1 diet.
Subjects had total serum cholesterol values of 6.50 ± 0.32, 6.12 ± 0.39, 5.66 ± 0.46,
5.46 ± 0.51, and 5.52 ± 0.47 on the 0-, 25-, 50-, 100-, and 200-mg tocotrienol-
enriched-fraction diets, respectively. Most of the difference was due to decreases in
the LDL fraction. The HDL cholesterol level was not different from that of the
Step-1 control. At least 14 reports on tocotrienol and its hypolipidemic effects have
been published since 1986 by Qureshi and others.
RICE FIBER
Ayano et al (1980) made neutral-detergent-fiber and acid-detergent-fiber
preparations from defatted rice bran. The neutral-detergent-fiber preparation
consisted of 29.3% cellulose, 47.5% hemicellulose, and 16.3% lignin. The acid-
detergent-fiber preparation contained 60.2% cellulose, 5.0% hemicellulose, and
33.4% lignin. Rats were fed 5% dietary fiber. Total serum cholesterol was
lowered by the diets containing neutral detergent fiber (237 ± 51 mg/dL)
compared with the control (318 ± 26 mg/dL). In a subsequent study from the
same laboratory (Aoe et al, 1989), the feeding of 2% rice bran hemicellulose to
rats for nine days was compared with feeding 2% high-methoxylated pectin. The
hemicellulose preparation contained 0.3% cellulose, 72.8% noncellulosic
polysaccharides, and 1.9% lignin. Both the rice bran hemicellulose and the pectin
diets reduced total serum cholesterol (256 ± 13 and 231 ± 11 mg/dL, respec-
tively) compared with the control (316 ± 26 mg/dL). The mechanism of
cholesterol lowering for rice bran appeared to be excretion of acidic steroids,
whereas pectin decreased cholesterol absorption.
Effects on Cancer
Dietary fiber, particularly from wheat bran, has been associated with
decreased risk of colon cancer (Alabaster et al, 1997). Rice bran was compared
with wheat bran, corn bran, and soybean bran in an animal model of colon cancer
(Barnes et al, 1983). F344 rats were fed 20% bran at four weeks of age until the
end of the study, about 11–12 months later. At 8 and 10 weeks of age, the rats
were injected with 135 and 150 mg of 1,2-dimethylhydrazine per kilogram of
body weight. The fiber content of the brans varied largely in the hemicellulose
fraction: wheat 28.3%, rice 19.1%, corn 67.3%, and soybean 9.4%. Soy contained
46.8% cellulose, while the others had less than 9% cellulose. Lignin was less than
5% in all brans. There were no differences in the number of carcinomas,
adenomas, total number of tumors, or tumors in rats bearing tumors, except for
higher numbers in the group fed corn bran. Transit time is thought to play a role
in the exposure of colonic epithelial cells to intraluminal carcinogens. Tomlin and
Read (1988) compared the transit time of rice and wheat brans. Weekly fecal
mass from rice bran was higher than that from wheat bran, but transit time was
not different.
Rice germ fed to F344 rats treated with azoxymethane (AOM) resulted in a
reduced number of aberrant crypt foci (ACF) (Kawabata et al, 1999). Rats were fed
defatted rice germ, defatted rice germ supplemented with 2.5% g-aminobutyric acid
(GABA), or rice germ for five weeks. The numbers of ACF were 93 ± 20, 74 ± 13,
63 ± 19, and 40 ± 5 in rats fed the basal, the defatted rice germ, the GABA-enriched
defatted rice germ, and the rice germ, respectively. The number of ACF in the latter
two groups was significantly lower than that in the rats fed the basal diet. In a fol-
low-up longer-term study in the same report, the authors report the number of neo-
plasms. Most of the tumors occurred in the large intestine. The percentages of ani-
mals with neoplasms were 75, 43, and 29 for the basal, GABA-enriched defatted
rice germ, and rice germ, respectively. Feeding of GABA-enriched defatted rice
germ after AOM injection also resulted in lower numbers of neoplasms (20%). A
measure of cell proliferation, the proliferating cell nuclear antigen assay, showed
significantly lower values for the rice germ and GABA-enriched defatted rice germ
fed after AOM injection.
Rice and other cereal brans are rich in phenolic acids. These acids in rice bran
were tested for their ability to suppress proliferation in three breast and three colo-
nic tumor lines (Hudson et al, 2000). Tricin and genistein were most effective
against breast cell tumor lines MDA MB468 and HBL100 and colon tumor line
HT29. Caffeic acid was also effective against two of the breast and all of the colon
tumor lines. Ethyl acetate extracts of rice bran were most effective against almost
all breast and colon tumor lines, compared with fractions of the extracts or pure
compounds. Ferulic acid and its geranylated ester have also been reported to reduce
the numbers of ACF in F344 rats injected with AOM (Han et al, 2001).
Cycloartenol ferulate, a component of rice bran oryzanol fractions, has been
reported to have anticancer properties in a mouse skin cancer model (Yasukawa et
al, 1998).
CONCLUSIONS
Rice has many potential roles in the twenty-first century. It is and will
continue to be the dominant source of caloric energy and protein for some
populations in Southeast Asia. The rice genome has been mapped, and rice has
been found to be more easily transformed than other cereal grains. Rice is likely
to be the model cereal plant for the introduction of new useful traits such as
higher protein, vitamins, oil, yield, and insect resistance. Another potentially
important role of rice is as a source of nutritional biochemicals. Since the 1980s,
many rice-producing countries have extracted food oil from rice bran. Oil
extraction also provides the opportunity to fractionate lipids into products such as
cholesterol-lowering sterols, oryzanols, fat-soluble vitamins, and other
nutritionally useful components. The solid, defatted residues also contain protein,
phytic acid, fiber, and antioxidants. Research will continue to improve the nutri-
tional properties of rice, which will, in turn, improve human health and prevent
diet-related diseases.
REFERENCES
Alabaster, O., Tang, Z., and Shivapurkar, N. in vitro digestibility, complexing index, and
1997. Inhibition by wheat bran cereals of viscosity. Cereal Chem. 74:561-565.
the development of aberrant crypt foci and Hallfrisch, J., and Behall, K. M. 2000. Mecha-
colon tumours. Food Chem. Toxicol. nisms of the effects of grains on insulin and
35:517-522. glucose responses. J. Am. Coll. Nutr. 19:320S-
Aoe, S., Ohta, F., and Ayano, Y. 1989. Effect of 325S.
rice bran hemicellulose on the cholesterol Hallikainen, M. A., Sarkkinen, E. S., Gylling, H.,
metabolism in rats. Nippon Eiyo Shokuryo Erkkila, A. T., and Uusitupa, M. I. 2000.
Gakkaishi 42:55-61. Comparison of the effects of plant sterol ester
Ayano, Y., Ohta, F., Watanabe, Y., and Mita, K. and plant stanol ester-enriched margarines in
1980. “Dietary fiber” fractions in defatted rice lowering serum cholesterol concentrations in
bran and their hypocholesterolemic effect in hypercholesterolaemic subjects on a low-fat
cholesterol-fed rats. J. Nutr. Food (Jpn.) diet. Eur. J. Clin. Nutr. 54:715-725.
33:283-291. Han, B. S., Park, C. B., Takasuka, N., Naito, A.,
Barnes, D. S., Clapp, N. K., Scott, D. A., Oberst, Sekine, K., Nomura, E., Taniguchi, H., Tusno,
D. L., and Berry, S. G. 1983. Effects of wheat, T., and Tsuda, H. 2001. A ferulic acid deriva-
rice, corn, and soybean bran on 1,2-dimethyl- tive, ethyl e-(4-geranyloxy-3-methoxy-phenyl)-
hydrazine-induced large bowel tumorigenesis 2-propenoate, as a new candidate chemo-
in F344 rats. Nutr. Cancer 5:1-9. preventive agent for colon carcinogesis in the
Cara, L., Dubois, C., Borel, P., Armand, M., rat. Jpn. J. Cancer Res. 92:404-409.
Senft, M., Portugal, H., Pauli, A. M., Bernard, Hegsted, D. M., McGandy, R. B., Myers, M. L.,
P. M., and Lairon, D. 1992. Effects of oat and Stare, F. J. 1965. Quantitative effects of
bran, rice bran, wheat fiber, and wheat germ dietary fat on serum cholesterol in man. Am. J.
on postprandial lipemia in healthy adults. Am. Clin. Nutr. 17:281-295.
J. Clin. Nutr. 55:81-88. Hegsted, M., Windhauser, M. M., Morris, K., and
Casiraghi, M. C., Brighenti, F., Pellegrini, N., Lester, S. B. 1993. Stabilized rice bran and oat
Leopardi, E., and Testolin, G. 1993. Effect of bran lower cholesterol in humans. Nutr. Res.
processing on rice starch digestibility evalu- 13:387-398.
ated by in vivo and in vitro methods. J. Cereal Hudson, E. A., Dinh, P. A., Kokubun, T.,
Sci. 17:147-156. Simmonds, M. S. J., and Gescher, A. 2000.
Cicero, A. F., and Gaddi, A. 2001. Rice bran oil Characterization of potentially chemopreven-
and gamma-oryzanol in the treatment of tive phenols in extracts of brown rice that
hyperlipoproteinaemias and other conditions. inhibit the growth of human breast and colon
Phytother. Res. 15:277-289. cancer lines. Cancer Epidemiol. Biomark.
Crapo, P. A., and Henry, R. R. 1988. Postpran- Prev. 9:1163-1170.
dial metabolic responses to the influence of Jenkins, D. J., Wolever, T. M., and Jenkins, A. L.
food form. Am. J. Clin. Nutr. 48:560-564. 1988. Starchy foods and glycemic index.
Federal Register. 2000. Plant sterol/stanol esters Diabetes Care 11:149-159.
and coronary heart disease; Interim final rule. Juliano, B. O. 1985. Polysaccharides, protein,
Dep. Health and Human Serv., Food and Drug and lipids of rice. Pages 59-174 in: Rice:
Admin., 21 CFR Part 101, Food Labeling: Chemistry and Technology, 2nd ed. B. O.
Health Claims. Sept. 8, 2000. 65(175):54685- Juliano, Ed. Am. Assoc. Cereal Chem., St.
54739. Paul, MN.
Gatti, E., Testolin, G., Noe, D., Brighenti, F., Kahlon, T. S., Chow, F. I. Sayre, R. N., and
Buzzetti, G. P., Porrino, M., and Sirtori, C. R. Betschart, A. A. 1992. Cholesterol-lowering in
1987. Plasma glucose and insulin responses to hamsters fed rice bran at various levels, defat-
carbohydrate food (rice) with different thermal ted rice bran and rice bran oil. J. Nutr.
processing. Ann. Nutr. Metab. 31:296-303. 122:513-519.
Gerhardt, A. L., and Gallo, N. B. 1998. Full-fat Kahlon, T. S., Chow, F. I., Chiu, M. M., Hudson,
rice bran and oat bran similarly reduce hyper- C. A., and Sayre, R. N. 1996. Cholesterol-
cholesterolemia in humans. J. Nutr. 128:865- lowering by rice bran and rice bran oil
869. unsaponifiable matter in hamsters. Cereal
Goddard, M. S., Young, G., and Marcus, R. Chem. 73:69-74.
1984. The effect of amylose content on insulin Kawabata, K., Tanaka, T., Murakami, T., Okada,
and glucose responses to ingested rice. Am. J. T., Murai, H., Yamamoto, T., Hara, A.,
Clin. Nutr. 39:388-392. Shimizu, M., Yamada, Y., Matsunaga, K.,
Guraya, H. S., Kadan, R. S., and Champagne, E. Kuno, T., Yoshimi, N., Sugie, S., and Mori, H.
T. 1997. Effect of rice starch-lipid complex on 1999. Dietary prevention of azoxymethane-
induced colon carcinogenesis with rice-germ Panlasigui, L. N., Thompson, L. U., Juliano, B.
in F344 rats. Carcinogenesis 20:2109-2015. O., Perez, C. M., Yiu, S. H., and Greenberg, G.
Kestin, M., Moss, R., Clifton, P. M., and Nestel, R. 1991. Rice varieties with similar amylose
P. J. 1990. Comparative effects of three cereal content differ in starch digestibility and gly-
brans on plasma lipids, blood pressure, and cemic response in humans. Am. J. Clin. Nutr.
glucose metabolism in mildly hypercholes- 54:871-877.
terolemic men. Am. J. Clin. Nutr. 52:661-666. Panlasigui, L. N., Thompson, L. U., Juliano, B.
Kurup, P. G., and Krishnamurthy, S. 1993. Gly- O., Perez, C. M., Jenkins, D. J., and Yiu, S. K.
cemic response and lipemic index of rice, 1992. Extruded rice noodles: Starch digesti-
raggi and tapioca as compared to wheat diet in bility and glycemic response of healthy and
human. Ind. J. Exp. Biol. 31:291-293. diabetic subjects with different habitual diets.
Larsen, H. N., Rasmussen, O. W., Rasmussen, P. Nutr. Res. 12:1195-1204.
H., Alstrup, K. K., Biswas, S. K., Tetens, I., Qureshi, A. A., Sami, S. A., Salser, W. A., and
Thilsted, S. H., and Hermansen, K. 2000. Gly- Khan, F. A. 2002. Dose-dependent suppression
caemic index of parboiled rice depends on the of serum cholesterol by tocotrienol-rich fraction
severity of processing: Study in type 2 diabetic (TRF25) of rice bran in hypercholesterolemic
subjects. Eur. J. Clin. Nutr. 54:380-385. humans. Atherosclerosis 161:199-207.
Lichtenstein, A. H., Ausman, L. M., Carrasco, Raghuram, T. C., Brahmaji Rao, U., and
W., Gualtieri, L. J., Jenner, J. L., Ordovas, J. Rukmini, C. 1989. Studies on hypolipidemic
M., Nicolosi, R. J., Goldin, B. R., and effects of dietary rice bran oil in human sub-
Schaefer, E. J. 1994. Rice bran oil consump- jects. Nutr. Rep. Int. 39:889-895.
tion and plasma lipid levels in moderately Rogers, E. J., Rice, S. M., Nicolosi, R. J.,
hypercholesterolemic humans. Arterioscler. Carpenter, D. R., McClelland, C. A. and
Thromb. 14:549-556. Romanczyk, L. J., Jr. 1993. Identification and
Marsono, Y., Illman, R. J., Clarke, J. M., quantitation of g-oryzanol components and
Trimble, R. P., and Topping, D. L. 1993. simultaneous assessment of tocols in rice bran
Plasma lipids and large bowel volatile fatty oil. J. Am. Oil Chem. Soc. 70:301-307.
acids in pigs fed on white rice, brown rice and Rong, N., Ausman, L. M., and Nicolosi, R. J.
rice bran. Br. J. Nutr. 70:503-513. 1997. Oryzanol decreases cholesterol absorp-
Miettinen, T. A., Puska, P., Gylling, H., tion and aortic fatty streaks in hamsters. Lipids
Vanhanen, H., and Vartiainen, E. 1995. 32:303-309.
Reduction of serum cholesterol with Rukmini, C., and Raghuram, T. C. 1991. Nutri-
sitostanol-ester margarine in a mildly hyper- tional and biochemical aspects of the hypo-
cholesterolemic population. New Engl. J. lipidemic action of rice bran oil: A review. J.
Med. 333:1308-1312. Am. Coll. Nutr. 10:593-601.
Miller, J. B., Pang, E., and Bramall, I. 1992. Sanders, T. A., and Reddy, S. 1992. The influ-
Rice: A high or low glycemic index food? Am. ence of rice bran on plasma lipids and lipo-
J. Clin. Nutr. 56:1034-1036. proteins in human volunteers. Eur. J. Clin.
Mitani, K., Kido, Y., Shimizu, S., and Morita, S. Nutr. 46:167-172.
1983. Effect of g-oryzanol on the metabolism Sarkar, S., and Bhattacharyya, D. K. 1991.
of cholesterol in hypercholesterolemic rats. J. Nutrition of rice bran oil in relation to its
Jpn. Atheroscler. Soc. 11:36. purification. J. Am. Oil Chem. Soc. 63:956-962.
Nakamura, H. 1967. Effect of g-oryzanol on Saunders, R. M. 1990. The properties of rice bran
hepatic cholesterol biosynthesis and fecal as a foodstuff. Cereal Foods World 35:632-636.
excretion of cholesterol metabolites. Radio- Seetharamaiah, G. S., and Chandrasekhara, N.
isotopes 15:371-374. 1988. Hypocholesterolemic activity of oryza-
Newman, R. K., Betschart, A. A., and Hofer, P. J. nol in rats. Nutr. Rep. Int. 38:927-935.
1990. Serum cholesterol in leghorn chicks fed Seetharamaiah, G. S., and Chandrasekhara, N.
full-fat or defatted rice bran. FASEB Proc. 1990. Effect of oryzanol on cholesterol
4:A925. absorption and biliary and fecal bile acids in
Nicolosi, R. J., Ausman, L. M., and Hegsted, D. rats. Indian J. Med. Res. (B) 92:471-475.
M. 1991. Rice bran oil lowers serum total and Sharma, R. D., and Rukmini, C. 1986. Rice bran
low density lipoprotein cholesterol and apo B oil and hypocholesterolemia in rats. Lipids
levels in nonhuman primates. Atherosclerosis 21:715-717.
88:133-142. Sharma, R. D., and Rukmini, C. 1987. Hypocho-
Normen, L., Dutta, P., Lia, A., and Andersson, H. lesterolemic activity of unsaponifiable matter
2000. Soy sterol esters and beta-sitostanol of rice bran oil. Ind. J. Med. Res. 85:278-281.
ester as inhibitors of cholesterol absorption in Shin, T.-S., Godber, J. S., Martin, D. E., and
human small bowel. Am. J. Clin. Nutr. Wells, J. H. 1997. Hydrolytic stability and
71:908-913. changes in E vitamers and oryzanol of
extruded rice bran during storage. J. Food Sci. Vissers, M. N., Aock, P. L., Meijer, G. W., and
62:704-708, 728. Katan, M. B. 2000. Effect of plant sterols from
Shinomiya, M., Morisaki, N., Matsuoka, N., rice bran oil and triterpene alcohols from
Izumi, S., Saito, Y., Kumagai, A., Mitani, K., sheanut oil on serum lipoprotein concentra-
and Morita, S. 1983. Effects of gamma- tions in humans. Am. J. Clin. Nutr. 72:1510-
oryzanol on lipid metabolism in rats fed high- 1515.
cholesterol diet. Tohoku J. Exp. Med. Weststrate, J. A., and Meijer, G. W. 1998. Plant
141:191-197. sterol-enriched margarines and reduction of
Strocchi, A., and Levitt, M. D. 1991. Measure- plasma total and LDL-cholesterol concentra-
ment of starch absorption in humans. Can. J. tions in normocholesteroleaemic and mildly
Physiol. Pharmacol. 69:108-110. hypercholesterolaemic subjects. Eur. J. Clin.
Sugano, M., Koba, K., and Tsuji, E. 1999. Health Nutr. 52:334-343.
benefits of rice bran oil. Anticancer Res. Wilson, T. A., Ausman, L. M., Lawton, C. W.,
19:3651-3657. Hegsted, D. M., and Nicolosi, R. J. 2000.
Sugano, M., and Tsuji, E. 1997. Rice bran oil and Comparative cholesterol lowering properties
cholesterol metabolism. J. Nutr. 127:521S- of vegetable oils: Beyond fatty acids. J. Am.
524S. Coll. Nutr. 19:601-607.
Suzuki, S., and Oshima, S. 1970. Influence of Wolever, T. M. 1990. The glycemic index. World
blending of edible fats and oils on human Rev. Nutr. Diet. 62:120-185.
serum cholesterol level. Jpn. J. Nutr. 28:3-6. Yasukawa, K., Akihsa, T., Kimura, Y., Tamura,
Tomlin, J., and Read, N. W. 1988. Comparison of T., and Takido, M. 1998. Inhibitory effect of
the effects on colonic function caused by cycloartenol ferlate, a component of rice bran,
feeding rice bran and wheat bran. Eur. J. Clin. on tumor promotion in two-stage carcinogene-
Nutr. 42:857-861. sis in mouse skin. Biol. Pharm. Bull. 21:1072-
Topping, D. L., Illman, R. J., Roach, P. D., 1076.
Trimble, R. P., Kambouris, A., and Nestel, P. Yoshino, G., Kazumi, T., Amano, M., Tateiwa,
J. 1990. Modulation of the hypolipidemic M., Yamasaki, T., Takashima, S., Iwai, M.,
effect of fish oils by dietary fiber in rats: Hatanaka, H., and Baba, S. 1989. Effects of
Studies with rice and wheat bran. J. Nutr. gamma-oryzanol on hyperlipidemic subjects.
120:325-330. Curr. Ther. Res. 45:543-552.
CHAPTER 21
UTILIZATION OF RICE HULL AND RICE STRAW

AS ADSORBENTS
Wayne E. Marshall
In the 1985 edition of Rice Chemistry and Technology, Juliano (Juliano, 1985)
wrote a comprehensive chapter on the structure, composition, properties, and uses
of rice hull and rice straw entitled, appropriately enough, “Rice Hull and Rice
Straw.” My objective in this “update” of that chapter is to present a narrower
focus, zeroing in on the utilization of rice hull and straw as value-added
products—specifically, value-added products in the form of adsorbents. In-
formation on the structure, composition, and properties of hull and straw was
thoroughly covered by Juliano up to 1985. Little new information in these areas
has emerged since then compared to the number of literature reports in the
burgeoning field of utilization. This chapter takes the utilization route because
this is an area of hull and straw chemistry and technology that underwent perhaps
the greatest change during the 1990s. This area involves technology and resultant
adsorbent products that may be useful in a variety of environmental and industrial
applications.
The total rice production in the United States for crop year 2002 was 21.1
billion pounds (9.6 billion kilograms) of rough rice (NASS, 2003). Therefore,
approximately 4 billion pounds (1.8 billion kilograms) of rice hulls and 26
billion pounds (11.8 billion kilograms) of rice straw were potentially produced.
Thus, a major issue facing the rice processing industry is what to do with this
prodigious quantity of co-products. A large quantity of hull is burned by the
rice miller to produce steam for processing, particularly parboiling (e.g.,
Producers Rice Mill, Inc./Riceland Foods, Stuttgart, AR) or for production of
electricity (e.g., Agrilectric Power Partners Ltd., Lake Charles, LA). Hull
burning results in rice hull ash, a by-product that consists of varying amounts
of silica and carbon and minor amounts of alkali oxides (K2O, Na2O) and alkali
earth oxides (MgO, CaO), depending on the amount of air used for hull
burning. However, the net result is both unburned rice hull and rice hull ash for
either disposal or utilization.
The same picture emerges for rice straw. Since rice straw is a field residue found
on individual rice farms, rather than a processing residue collected at a single loca-
tion, as is rice hull, its collection and transportation to a central incineration facility
611
is more expensive. Most rice straw is burned in the field. However, one state, Cali-
fornia, has enacted legislation to restrict open burning, which contributes to air pol-
lution in the rice-growing areas. In fact, California has the largest biomass-to-
energy power plant in the world at Williams, CA. A mixture of hulls and straw is
burned to create electricity; Oxford Energy Company has contracted with growers
and millers to supply the feedstock.
Rice hull, hull ash, and straw need to be utilized, especially for the production
of value-added products. With a bit of ingenuity on the part of scientists and
technologists, a number of different uses have been envisioned and studied for
these co-products (USAID, 1989). Figure 1 shows the diversity of the products
that have been developed in the United States from hull and straw. The figure
does not include co-product conversion to chemicals, such as ethanol, lignin, and
methane. These products can be considered value-added, but the market is very
small when the feedstocks are hull and straw. Most of the products depicted in
Figure 1 are of low value; the highest-value products are listed under the
adsorbent category.
Since the objective of this chapter is to describe value-added products of rice hull
and straw, preference is given to their use as adsorbent material in the following
forms: char, activated carbon, ash, and unmodified hull and straw, or these co-
products modified by chemical treatment.
Rice Hull Rice Straw

Particle board Fiber board
Poultry bedding Packaging material
Brick making Mushroom growing
Fiberboard Cattle feed
Package cushioning Paper pulp
Adsorbent Adsorbent
Char Char
Activated carbon Activated carbon
Modified hull Modified straw
Energy feedstock Energy feedstock
Electricity Electricity
Steam Steam
Ash Ash
Adsorbent Adsorbent
Steel lading Steel lading
Silicates Silicates
Mortar cement Mortar cement
Fig. 1. Products generated from the utilization of rice hull and rice straw.
Hull and Straw as Absorbents / 613
UTILIZATION OF RICE HULLS AND STRAW
As a Char or an Activated Carbon
Lignocellulosic materials, such as rice hull and straw, can be converted to a car-
bonized form, or char, at a temperature below 800°C in a continuous stream of inert
gas, usually nitrogen (N2) (Bansal et al, 1988). Carbonization volatilizes many of
the nonelemental-carbon (C) constituents in the starting material, including oxygen
(O2), nitrogen, and hydrogen (H2). Some elemental carbon is also volatilized, but
the overall effect of carbonization is the enrichment of the char in both elemental
carbon and elemental silicon (Si). Physically, chars are macroporous particles with
rudimentary surface areas. They are not particularly good adsorbents because they
lack significant surface area.
Activated carbons, on the other hand, are excellent adsorbents and represent one
of the most important types of adsorbents used in residential, municipal, industrial,
and general environmental applications. They are produced by activation of chars to
produce large surface areas, which impart high adsorptivity to these materials. The
high adsorptivity is the result of a highly developed micro-, meso-, and macropor-
ous structure, with most (80–90%) of the surface area resulting from the carbon’s
microporous structure. Some of these structural elements can be observed in Figure
2. Because of its pores of different sizes, activated carbon demonstrates the unique
ability to adsorb substances of different molecular sizes from air or water.
Structurally, a char consists of more or less disordered elementary graphitic
crystallites, with a poorly developed porous structure. Subsequent activation dra-
matically increases the surface area and creates a highly developed pore structure.
The activation process can be either physical or chemical (Bansal et al, 1988).
Physical activation is a process in which chars of rice hull or straw are heated to
temperatures generally between 750 and 900°C in the presence of suitable gases
such as steam (H2O), carbon dioxide (CO2), air, or any mixture of these gases. The
gases react with carbonaceous material in the char structure to enhance the forma-
tion of surface area in a process known as “burn-off.” The spaces between the ele-
mentary graphitic crystallites become cleared of the less organized, loosely bound
Fig. 2. Scanning electron micrograph of a granular activated carbon made from rice hull.
material, called “tar.” The resulting channels through the graphitic regions, the
spaces between the elementary crystallites, together with fissures within and paral-
lel to the graphitic planes, constitute the porous structure (Rodriquez-Reinoso and
Solano, 1989).
Chemical activation occurs when rice hull or straw is impregnated with a
“chemical” activating agent, such as phosphoric acid (H3PO4), sulfuric acid
(H2SO4), potassium or sodium hydroxide (KOH or NaOH), or zinc chloride (ZnCl2).
Chemical activation occurs at lower temperatures than physical activation, with
temperatures between 400 and 600°C commonly reported.
The process of surface-area formation is different for chemical activation than
for physical activation. The common feature of “chemical” activating agents is that
they are dehydrating agents that inhibit the formation of carbonaceous material in
and between pores (Bansal et al, 1988). Carbons produced by chemical activation
generally have higher surface areas and higher yields than carbons developed by
physical activation. However, the activating agent should be removed from the car-
bon before it is used.
Because of their highly developed pore structures and excellent adsorptivity,
activated carbons are usually the adsorbents of choice for a wide range of industrial
and residential applications, including municipal and industrial wastewater treat-
ment, residential tap-water purification, industrial air purification, food and phar-
maceutical purification, and solvent recovery. Commercial activated carbons are
currently made primarily from coal, wood, coconut shell, and peat, but not from rice
hull or straw. The demand for activated carbon in the United States is forecast to
increase about 3% per year to about 450 million pounds (204 million kilograms) by
2006 (Anonymous, 2002). The market value is anticipated to rise about 3.5% per
year to about $330 million (U.S. dollars) by 2006. Carbon cost is variable, with the
higher-value carbons selling in bulk for $1.50 or more per pound. In contrast, the
other adsorbent market (namely the ion-exchange market) is currently valued at
about $330 million, with demand presently at about 165 million pounds (74.8
million kilograms). The cost of resin is also variable, with bulk prices generally in
the range of $1.50 to $10 per pound.
AS A CHAR
Little published evidence shows rice hull or straw chars being used for adsorption
studies. As determined by the Brunauer-Emmett-Teller equation, the surface areas
(SBET) of chars are generally low, and a low surface area of less than 50 m2/g normally
precludes significant adsorption of inorganic or organic compounds.
One of the earliest references describing the use of a char from rice co-products
to adsorb inorganic and organic contaminants from water is contained in a U.S. pat-
ent (Harrison, 1987). Rice hull char was used to remove ammonia, biological oxy-
gen demand, and total organic carbon from a sample of wastewater. The char was
most effective at removing ammonia; at least 93% of the original ammonia in the
sample was adsorbed.
Additionally, Imagawa et al (2000) demonstrated good adsorption of three chlo-
rinated hydrocarbons (tetrachloroethylene, trichloroethylene, and 1,1,1-trichloro-
ethane) with a rice hull char of low surface area (44 m2/g). Adsorption was only
about 5% lower than that of a commercial, coal-based, granular activated carbon of
higher surface area (384 m2/g), but it was about 65% higher than noncarbonized rice
bran and more than 90% higher than noncarbonized rice hull and straw.
Further research is required before the adsorptive benefits of rice hull and rice
straw chars can be ascertained. Since the adsorption of organic constituents found in
water and wastewater is normally dependent on the surface area of the adsorbent,
the lower-surface-area chars would be at a disadvantage compared to activated car-
bons made from the same material. However, chars have the advantages of higher
yield and lower production costs than activated carbons. Future research must
examine the balance of adsorptive properties versus product cost before a fair
assessment of char utilization can be made.
AS AN ACTIVATED CARBON
In a number of studies, rice hull and straw have undergone both physical activa-
tion (with steam, carbon dioxide, or air) or chemical activation (with phosphoric
acid, zinc chloride, or sodium/potassium hydroxide) to produce activated carbons
with a broad range of surface areas suitable for the adsorption of a wide variety of
organic compounds and metal ions.
Sugar Decolorization. Ahmedna et al (1997) used rice hull and rice straw with a
sugarcane molasses binder to produce activated carbons by carbon dioxide activa-
tion. These carbons were evaluated for their ability to decolorize solutions of raw
sugar. Decolorization potential was measured against that of a commercial (refer-
ence) decolorizing carbon. Both rice hull- and straw-based carbons were better
decolorizing carbons than carbons made from pecan shells and soybean hulls. In a
dilute sugar solution of 10° Brix, the rice hull carbon and the commercial carbon
showed the same decolorizing potential. However, in a more concentrated 60° Brix
sugar solution, the rice hull carbon achieved about 85% of the decolorizing potential
of the reference carbon.
This study was followed by a report (Pendyal et al, 1999) in which rice hull and
straw were mixed in combinations with four different binders, then activated with
carbon dioxide. All experimental carbons were compared to two commercial, sugar-
decolorizing carbons for their ability to remove molasses color or decolorize a 60°
Brix sugar solution. A rice straw-based carbon made with a corn syrup binder did
well in the molasses-color-removal test by removing 56% of total color, while the
reference carbons adsorbed 54 and 66% of total color. In the decolorization experi-
ment, a rice hull carbon with sugarcane molasses as binder removed 45% of total
color compared to 67 and 72% color removal for the reference carbons.
Another study of the ability of rice hull- and straw-based carbons to decolorize
raw sugar solutions and remove molasses color was presented by Ahmedna et al
(2000). In this study, a rice hull-based carbon made with a coal tar binder and acti-
vated with carbon dioxide removed 15% molasses color, compared to 28 and 33%
for two commercial carbons used as reference, and had an efficiency of 12% for
raw sugar decolorization compared to 15 and 18% for the reference carbons. How-
ever, low surface area (120 m2/g) for this carbon, compared to that of the reference
carbons (>900 m2/g), placed it at a distinct disadvantage.
More research is needed on developing hull- and straw-based carbons for sugar
decolorization. The foregoing results showed that these carbons demonstrate prom-
ise as sugar decolorants, but their surface areas require substantial improvement.
Steam activation rather than carbon-dioxide activation should be evaluated, as
should higher activation temperatures and longer activation times. Under these sets
of conditions, carbon yield would diminish but sugar decolorization should
improve, perhaps to the level of the reference carbons.
Removal of Low Molecular Weight Organic Compounds. Carbons produced by

physical activation of hulls and straw have been used to remove a variety of differ-
ent low molecular weight organic substances from solution. Johns et al (1998)
determined the adsorptive properties of carbon-dioxide- or steam-activated rice
straw carbons in the presence of a suite of six organic molecules that differed in
polarity from polar methanol to nonpolar toluene. Both carbon-dioxide- and steam-
activated, straw-based carbons showed between 13 and 28% less adsorption than
four commercial carbons.
The rice straw carbons examined by Johns et al (1998) had surface areas of <400
m2/g, while the reference carbons had surface areas >700 m2/g. The importance of
surface area for adsorption of organic molecules was demonstrated by Kim and
Choi (1998), who used steam-activated rice hull carbons with high iodine numbers
(>1,200 mg/g); iodine number usually mimics surface area. These authors removed
more than 80% of the original phenol in solution with the hull-based carbon, while a
commercial carbon, Calgon F-400, removed about 65% of the phenol. Carbons with
high surface areas generally are preferred for adsorption of organic compounds,
particularly for nonpolar or low-polarity compounds like phenol.
Removal of High Molecular Weight Dyes. Two recent publications (Kannan and
Meenakshisundaram, 2002; Malik, 2003) reported dye binding by rice hull and
straw carbons prepared by steam activation. They displayed adsorption capacities
toward a neutral dye, Congo Red, that were 0.34 mmol/g (hull) and 0.58 mmol/g
(straw), compared to an adsorption capacity of 0.71 mmol/g for a commercial
carbon (Kannan and Meenakshisundaram, 2002). Surface areas were not given for
these carbons, but they may have played a role in explaining the differences in dye
binding. Malik (2003) used an acidic dye, Acid Yellow 36, as adsorbate and a
steam-activated rice hull carbon as adsorbent. Considerably less dye was adsorbed
by the rice hull carbon (0.23 mmol/g) compared to the amount adsorbed by a carbon
made from sawdust (wood) (0.49 mmol/g). In this case, both the surface area (516
m2/g) and surface basicity (2.34 meq/g) of the wood-based carbon were greater than
the surface area (272 m2/g) and surface basicity (0.45 meq/g) of the hull-based
carbon. Enhancement of both properties could readily contribute to increased
adsorption of the acidic dye.
Since surface area features prominently in adsorption of organic molecules, the
development of activated carbons with high surface areas would be beneficial.
Usmani et al (1994) used chemical activation (zinc chloride) to achieve high surface
areas (>1,000 m2/g) and high cationic dye (methylene blue) adsorption profiles
(>0.94 mmol/g) in rice hull carbons. Khalil (1996) employed another chemical
activation method (phosphoric acid) to create rice hull carbons with an SBET
exceeding 870 m2/g and methylene blue adsorption capacities of about 0.50 mmol/g.
Yet another chemical activation, using potassium hydroxide as the activating agent,
was performed on rice straw (Oh and Park, 2002). These authors used a two-stage
process, pre-carbonization in an atmosphere of N2 gas at temperatures at or above
700°C, followed by activation with potassium hydroxide at 900°C. The resulting
rice straw carbons had impressive SBET values of up to 2,410 m2/g and methylene
blue adsorption values that exceeded 2.50 mmol/g. Guo et al (2003) employed
potassium hydroxide activation and also sodium hydroxide activation to develop
very high surface areas in rice hull carbons. Some sodium-hydroxide-activated rice
hulls had SBET values that exceeded 2,700 m2/g. The authors determined adsorption
of the basic dye, malachite green, to be about 2.0 mmol/g and found that dye
adsorption was higher for the sodium-hydroxide-activated carbon than for the
potassium-hydroxide-activated carbon.
The above reports clearly demonstrate that high surface areas are achievable in
carbons made from rice co-products and that these high surface areas lead to high
adsorption capacities for organic dyes. These surface areas are produced from
chemical activation as opposed to physical activation. Chemical activation produces
carbons with high microporosity, usually greater than 90%, while physical activa-
tion results in carbons with lower microporosity, usually about 60–70% (Johns et al,
1999). Since surface area results primarily from the amount or percentage of
micropores in the carbon, the highest surface areas would arise from chemical
activation. Carbons with very high surface areas (>2,000 m2/g) are excellent
adsorbents of small to medium-sized organic and inorganic molecules and serve a
useful purpose in that regard. Rice hull and straw carbons produced by chemical
activation might be able to replace or complement carbons made from traditional
sources (such as coal, coconut shell, and wood) for adsorption of a range of organic
molecules.
Removal of Cations (Metal Ions). While the surface area of an activated carbon
may predict the carbon’s ability to adsorb organic compounds from solution, this is
not necessarily the case when a carbon is called upon to adsorb metals (metal ions)
from solution. Youssef and Mostafa (1992) activated rice hull by either physical
activation (steam) or chemical activation (zinc chloride) and found some copper ion
(Cu2+) adsorption (up to 0.46 mmol/g) and surface acidity (up to 0.27 mmol/g) for
the zinc-chloride-treated carbons but no Cu2+ adsorption or surface acidity for the
steam-activated carbons. However, when the authors introduced more surface acid-
ity to both types of carbons by treatment with nitric acid, they found a large increase
in metal ion adsorption (up to 1.42 mmol/g) and surface acidity (up to 0.92
mmol/g). The authors concluded that the chemistry of the carbon surface rather than
the surface area dictated Cu2+ adsorption. The acidic sites on the carbon surface
were the centers of adsorption for the metal ion.
To further make the point that the carbon surface requires modification to
improve metal binding, Mostafa (1994) significantly increased adsorption of palla-
dium ion (Pd2+) on a steam-activated, rice hull carbon by treatment with hydroflu-
oric acid (HF) and/or 2-hydroxyimino-3-(2-hydrazonopyridyl)-butane (HHB), a
known chelating agent for metal ions. The author determined that exposing the car-
bon to both HF and HHB resulted in material with the highest Pd2+ binding (0.57
mmol/g) compared to carbon with no treatment, which adsorbed no measurable
metal ion. In further research on carbon surface modification, Mostafa (1997) pre-
pared rice hull carbons by both physical (steam) and chemical (zinc chloride) acti-
vation. The carbon surface was modified by treatment with sulfuric acid. Rice hull
carbons that were chemically activated and oxidized with sulfuric acid had the
greatest adsorption capacities for mercury (Hg2+), lead (Pb2+), and cadmium (Cd2+)
ions, 0.12, 0.12 and 0.16 mmol/g, respectively.
Teker et al (1997) used a rice hull carbon produced by zinc-chloride activation
but no additional treatment to adsorb cobalt ion (Co2+) from solution. They observed
very little binding, as the highest adsorption capacity for Co2+ was 0.03 mmol/g.
Teker et al (1999) extended these observations to determine the binding capacities
of zinc-chloride-activated rice hull carbon for Cu2+ or Cd2+ in solution. Adsorption
capacities were determined to be 0.06 and 0.01 mmol/g for copper ion and cadmium
ion, respectively. The low adsorption for all three metal ions was likely due to low
surface charge, since Mostafa (1997) had previously shown that rice hull carbons
prepared by zinc chloride activation without surface treatment adsorbed very little
cadmium ion.
Johns et al (1998) prepared rice straw carbons by steam activation and carbon-
dioxide activation and oxidized the carbon surfaces with either an oxygen-nitrogen gas
mixture composed of 38% oxygen or by a solution of the oxidizing agent ammonium
persulfate ([NH4]2S2O8). Gas oxidation improved Cu2+ binding performance for both
steam and carbon-dioxide-activated carbons, but chemical oxidation did not.
Adsorption of copper ion was about 0.23 mmol/g for both carbons, while the untreated
rice straw carbons adsorbed about 0.14 mmol/g. The rice straw carbon that had been
oxidized by the gas mixture was compared with commercial carbons made from either
2+
coal, coconut shell, or peat for Cu binding. No significant difference in metal
adsorption among the commercial carbons and experimental carbon was observed.
However, when this rice straw carbon was compared to commercial carbons using a
suite of five metal ions (cadmium, copper, nickel, lead, and zinc) in solution, the rice
straw carbon adsorbed more total metal ions than any of the commercial carbons.
Removal of Anions. While several studies have examined the binding of metal
ions or cations by rice hull or straw activated carbons, very few studies have inves-
tigated the binding of anions to these carbons. An early study by CiCi and Keles
(1990) determined the binding of chromium ion (Cr6+) as sodium dichromate
(Na2Cr2O7) to rice hull carbons prepared by zinc-chloride activation. The authors
found a maximum Cr6+ adsorption of 0.38 mmol/g. Further recent work on this
subject was undertaken by Guo et al (2002), who used a carbon activated with
potassium hydroxide to achieve an adsorption capacity of 0.67 mmol/g for Cr6+.
A summary of the effectiveness of rice hull- and rice straw-based activated car-
bons for adsorption of metal ions is presented in Table 1. In some cases, adsorption
values are given as adsorption capacity, calculated from adsorption isotherms.
Adsorption capacities are good estimates of the potential of the carbon for binding a
particular metal ion under a given set of experimental conditions. Other values
TABLE 1
Adsorption of Metal Ions by Activated Carbons from Rice Hull and Rice Straw
By- Metal Ion
Product Activation Surface Adsorption
Metal Ion Precursor Type Modification (mmol/g) Reference
Cadmium2+ Hull Zinc chloride None 0.01a Teker et al (1999)
Hull Zinc chloride Sulfuric acid 0.16a Mostafa (1994)
Chromium6+ Hull Zinc chloride None 0.38b CiCi and Keles (1990)
Hull KOH None 0.67a Guo et al (2002)
Cobalt2+ Hull Zinc chloride None 0.03a Teker et al (1997)
Copper2+ Hull Zinc chloride None 0.06a Teker et al (1999)
Hull Zinc chloride Nitric acid 1.42a Youssef and Mostafa (1992)
Straw CO2 38% O2 0.23b Johns et al (1998)
Straw Steam 38% O2 0.23b Johns et al (1998)
Lead2+ Hull Zinc chloride Sulfuric acid 0.12a Mostafa (1994)
Mercury2+ Hull Zinc chloride Sulfuric acid 0.12a Mostafa (1994)
Palladium2+ Hull Steam HF and HHBc 0.57a Mostafa (1994)
a Obtained from adsorption isotherms.
b Obtained at a single metal ion concentration.
c HF = hydrofluoric acid; HHB = 2-hydroxyimino-3-(2-hydrazonopyridyl)-butane.
given in Table 1 are adsorption efficiencies, obtained from metal-ion solutions at a

given concentration rather than from a range of metal concentrations used to con-
struct adsorption isotherms. These values are, in some cases, rough estimates of
adsorption capacity.
A good indicator of how well rice hull and rice straw carbons bind metal ions
would be a comparison between the experimental carbons and representative com-
mercial or reference carbons. Unfortunately, of all of the studies listed in Table 1,
only the Johns et al (1998) study gave this comparison. These authors compared
their rice straw carbon results with those of six different reference carbons that var-
ied in Cu2+ adsorption from 0.19 to 0.27 mmol/g. In this author’s experience, based
on metal-ion binding by numerous commercial carbons, adsorption values of >1.00
mmol/g of carbon would be considered excellent, values between 0.50 and 1.00
mmol/g would be considered good, and adsorption between 0.25 and 0.49 mmol/g
would be judged fair. Using these criteria, only one study (Youssef and Mostafa,
1992) produced a carbon with excellent metal-ion binding, and two studies devel-
oped carbons possessing good metal-ion binding.
Youssef and Mostafa (1992) showed that excellent adsorption of metal ions to a
rice co-product-based activated carbon is possible. More research is needed to
develop a simple combination of activation and oxidation or a simple oxidation pro-
cedure following activation to permit consistent achievement of excellent metal-ion
binding without the use of strong and potentially dangerous oxidants such as nitric
acid. Recently, two patents have appeared (Toles et al, 2000; Johns et al, 2003) that
address this issue. Both describe methods of activation and oxidation that produce
carbons specifically targeted for the adsorption of metal ions. Utilization of these
activation-oxidation methods in the author’s laboratory has resulted in Cu2+ adsorp-
tion values between 1.30 and 1.50 mmol/g for carbons made from either white oak
wood or oat hulls (unpublished observations). The patented procedures are simple,
straightforward, and require only air as the oxidizing agent, rather than the mineral
acids (nitric, hydrofluoric, sulfuric) listed in Table 1. The patented methods hold
promise for the development of rice hull- and straw-based carbons with excellent
metal-ion-adsorbing properties and should be a priority for investigators seeking to
commercialize activated carbons made from rice co-products.
As Noncarbonized Material
Thus far in this chapter, adsorption of organic and inorganic species by hull and
straw chars and activated carbons has been documented. While activated carbons,
and chars to a lesser extent, have been shown to possess adsorptive properties
toward a variety of adsorbates, they are somewhat expensive to manufacture. Chars,
and particularly carbons, necessitate energy-consuming processes that require
expensive equipment, high capital investment, and a final product in low yield.
Moreover, in the current regulatory environment, char and carbon manufacture calls
for strict control of pollutants that emanate from the heating of char or carbon feed-
stock, which adds additional expense.
With this in mind, several laboratories have investigated noncarbonized, unmodi-
fied hull and straw as the adsorbent material. If investigators succeed in finding
good to excellent adsorption of different organic and inorganic substances by these
co-products and determine that they are as efficient as commercial activated car-
bons, the adsorbent market could be flooded with low-cost products.
Rice hulls and rice straw are lignocellulosic materials that contain cellulose, lig-
nin, hemicellulose, and protein as the major polymeric constituents. A typical
analysis is given in Table 2. The major polymer is cellulose for both co-products;
rice straw has significant hemicellulose and low lignin content.
In addition, hulls and straw contain 13–22% silica, in the amorphous form rather
than in the crystalline form commonly found in quartz (sand). All of the above
polymers and the silica are capable of interacting with the inorganic and organic
components of various water and wastewater sources. However, predicting the
degree or extent of interaction with hull or straw is difficult because these polymers
and silica are compartmentalized within the hull or straw particle. In addition, inter-
action would depend on the rate of diffusion of the molecules to the adsorption
sites, the polarity of the adsorbing surface, and the polarity of the adsorbate. For
ionic species, the extent of interaction could be enhanced if the surface of the hull or
straw were modified to increase either the positive or negative surface charge.
In the examples given below, both unmodified, noncarbonized hull and straw
have been used to adsorb organic compounds or metal ions. In contrast, some
investigators, perhaps frustrated with the lack of significant binding of anions and
cations to unmodified hull and straw, have resorted to various chemical modifica-
tion strategies to increase ion attachment. These studies are also presented.
UNMODIFIED HULL AND STRAW

Dye Adsorption. McKay et al (1986) made a comprehensive study of dye binding
by rice hulls, among other low-cost adsorbents. The authors examined adsorption of
acidic, basic, disperse, and direct dyes to rice hull and found significant binding
only with the basic, or cationic, dyes methylene blue and Safranine. Adsorption
capacities were 0.22 and 0.09 mmol of dye per gram for Safranine and methylene
blue, respectively. These results indicate the presence of some negative charge on
the surface of the hull and basically no positive charge, since no adsorption of acidic
or anionic dyes occurred.
Nawar and Doma (1989) found results similar to those of McKay et al (1986)
using rice hull as an adsorbent for the basic dye Sandocryl Orange or the acidic dye
Lanasyn Black. They observed a significantly greater adsorption capacity for
Sandocryl Orange (0.36 mmol/g) than for Lanasyn Black (0.14 mmol/g) and attrib-
uted the results to interactions between cationic centers on the basic dye and acidic
sites, mostly carboxylate groups on the hull fibers, probably originating from the
hemicellulose moieties.
The adsorptive properties of rice hull were also used to remove chemical oxygen
demand (COD) from textile effluent (Meyer et al, 1992). A number of low-cost,
waste materials were investigated. Barbecue charcoal and rice hull were shown to
TABLE 2
Content of the Major Polymeric Constituents of Rice Hull and Rice Strawa
Polymer Content (g/100 g, dry wt. basis)
By-Product Cellulose Lignin Hemicellulose Protein
Rice hull 34.8 17.2 6.0 3.2
Rice straw 43.3 5.4 25.1 5.6
a Wartelle and Marshall, unpublished data.
adsorb between 65 and 67% of the original effluent color or approximately 65–67
mg of COD per gram of hull or charcoal, respectively, which was the most color
adsorbed by any of the various materials tested.
Rice straw has also been evaluated for the removal of the cationic dye methylene
blue (Deo and Ali, 1996). About 0.20 mmol of dye per gram was removed from
solution under the conditions set forth in the study.
In the above studies, the greatest adsorption took place with cationic dyes, indi-
cating that hull and straw may have an anionic surface. Although surface charge
was not measured in the cited studies, the surface charge appears to be inherently
low for these co-products.
Adsorption of Organic Compounds. Besides dyes, other organic molecules, such
as phenol, have been adsorbed by rice hull (Munaf et al, 1997). These investigators
found that the adsorption capacity for phenol was 0.45 mmol/g. The authors also
used rice hull to remove phenol from industrial and hospital wastewaters, but they
found a much lower amount removed (59.9% industrial and 37.5% hospital) com-
pared with amounts removed by laboratory-prepared solutions (95.5%). Since the
wastewater effluents had many uncharacterized organic substances, these unknown
materials could have been competing for adsorption sites.
Cation and Anion Adsorption. While the literature reported above demonstrates
the use of unmodified hull and straw in the removal of dyes and phenol from solu-
tion, rice co-products have also been used in the sequestration of cations and anions.
In this case, both unmodified and modified rice hulls and straw have been deployed.
As ion-exchange resins, these products would be expected to compete with com-
mercial ion-exchange resins made from both natural (cellulose) and synthetic (sty-
rene-divinylbenzene) materials. In this market, in contrast with the activated-carbon
market, different benchmarks of adsorption activity apply. These benchmarks are
given as adsorption capacities for the cation or anion with the greatest binding and
are based on the author’s experience with numerous commercial products. They are
excellent = 1.50+ mmol of ion adsorbed per gram of resin; good = 1.00–1.49
mmol/g; fair = 0.25–0.99 mmol/g; poor = <0.25 mmol/g.
Kumar and Dara (1979) presented early work on the adsorption of Cu2+ by both
hull and straw. They found adsorption to be poor, as only 0.06 and 0.04 mmol were
bound per gram of hull and straw, respectively.
Roy et al (1993) determined adsorption capacities for Pb2+ and Cd2+ using rice
hull. The capacities appeared to be rather low, as only 0.05 and 0.19 mmol of Pb2+
and Cd2+, respectively, were bound to the hull. Marshall and Champagne (1995)
utilized rice straw to adsorb Zn2+. They calculated an adsorption capacity of only
0.06 mmol/g for the metal ion. Mishra et al (1997) also determined an adsorption
capacity for Zn2+ binding, using rice hull, and obtained a value of 0.30 mmol/g. This
large difference could be due to the by-product used (straw vs. hull) and/or the dif-
ferent pH values used. Metal-ion binding is sensitive to pH, as would be expected
for ionic or charged species in solution. Marshall and Champagne (1995) deter-
mined binding at pH 4.5–5.0, while Mishra et al (1997) kept the pH at 6.2. With
metal ions, binding is normally enhanced as the pH increases because of less com-
petition with protons (H+) for anionic adsorption sites on the adsorbent.
In a series of publications (Khalid et al, 1998a,b; 1999a,b; 2000), radiotracer
methodology was employed to examine metal-ion binding to rice hull. The
researchers treated the rice hull with different mineral acids at various concentra-
tions before deciding to use 0.01M nitric acid in all subsequent binding experiments.
These authors optimized reaction conditions, such as amount of adsorbent, concen-

tration of adsorbate, temperature, time, and pH. Metal ion adsorption obtained
under optimal conditions was 0.06, 0.02, 0.05, 0.04, and 9.4 × 10–5 mmol/g for Cd2+,
Pb2+, mercury (Hg2+), chromium (Cr3+), and antimony (a mixture of Sb3+ and Sb5+),
respectively. In all of their studies, the authors speculated that metal-ion binding
was occurring on the surface of the silica (SiO2) contained within the hull rather
than with one or more of the polymeric constituents, such as hemicellulose or pro-
tein, which would offer anionic functional groups for adsorption purposes. Their
adsorption model envisioned the hydrolysis of SiO2 with acid to form an –SiOH or
silanol group. The silanol group, a weak acid, would dissociate at the pH values
used (6–8) in their experiments to form anionic –SiO–, which would bind with cati-
onic metal ions. Because of the poor binding observed, the data suggest that there
are few exposed –SiO– groups available for interaction.
MODIFIED RICE HULL AND STRAW

The results thus far presented strongly indicate that unmodified rice hull and
straw are poor metal-ion adsorbents. Thus, the lack of sufficient anionic binding
sites on the hull or straw surface would require that the surface be modified to
obtain enhanced adsorption.
Adsorption of Cations. Marshall et al (1993) treated rice hulls with 5% NaOH
and exposed them to individual solutions of Cr3+, Co2+, Cu2+, nickel ion (Ni2+), or
Zn2+. The resulting binding was <0.01 mmol/g for all metal ions. Therefore,
modifying the hull surface with base had little, if any, effect on metal-ion
adsorption.
To improve the negative or anionic surface charge on rice hull, Suemitsu et al
(1986) adsorbed two anionic dyes, Procion Red and Procion Yellow, on the surface
of the hull. Adsorbates used were both metal cations (Cu2+, Ni2+, Zn2+, Cd2+, Pb2+,
and Hg2+) and a metal anion (Cr6+ as dichromate). The dye-coated hull removed
more cations than the unmodified hull, but anion removal was similar in both cases.
Adsorption of metal ions ranged from 0.06 mmol/g for Cr6+ to 0.12 mmol/g for
Cu2+. Although steps were taken to enhance metal-ion binding, only a modest
improvement was observed.
Low et al (2000a) used surface adsorption of ethylenediaminetetraacetic acid
(EDTA), a well-known metal-ion chelator, to improve metal-ion binding to rice
hull. Significant improvement was noted for the EDTA-treated hull compared to an
NaOH-treated control. Maximum sorption capacities were 0.14, 0.18, 0.15, and 0.14
mmol/g for Cu2+, Cr3+, Ni2+, and Pb2+, respectively, using EDTA-treated material.
The foregoing reports described resins that showed poor binding of metal ions. A
method was needed to substantially increase the anionic surface charge on the sur-
face of the rice co-product. In recognition of this problem, Wartelle and Marshall
(2000) and Wong et al (2003) attempted to covalently attach polycarboxylic acids to
the hull surface. These methods take advantage of a reactive primary alcohol –OH
group in glucose, which is abundant in the glucose polymer cellulose. Through the
application of heat, an ester linkage was formed between the reactive –OH group
and reactive acid anhydride in either citric acid (CA) (Wartelle and Marshall, 2000)
or tartaric acid (TA) (Wong et al, 2003). Adsorption of Cu2+ was 0.76 and 0.46
mmol/g for CA and TA, respectively. In addition, Wong et al (2003) determined the
sorption capacity of Pb2+ to be 0.52 mmol/g after TA modification. These values
represent a significant increase over previous values found in the literature.
Adsorption of Anions. In addition to cation or metal-ion binding, there is a strong

need to develop inexpensive anion-exchange resins for use in environmental reme-
diation. Some serious environmental challenges involve the removal of toxic anions,
+
such as arsenate (As5+), chromate (Cr6 ), and selenate (Se6+)—or the ubiquitous
5+
anion, phosphate (P ), which has been implicated in “blooms” of toxic algae. These
troublesome pollutants are difficult to remove unless rice hull and straw can be
endowed with a significant number of cationic sites, sufficient to achieve reasonable
adsorption.
One approach is to apply a single-step quaternization procedure. This method
involves the use of the quaternizing reagent N-(3 chloro-2-hydroxypropyl)
trimethylammonium chloride (CHMAC). This reagent forms an ether linkage with
the primary alcohol –OH groups in the glucose moiety of cellulose. CHMAC also
reacts to a limited extent with other reactive groups in lignin, hemicellulose, and
protein. Several recent studies (Low et al, 1997, 2000b; Lee et al, 1999) applied the
quaternization method to rice hulls and used the quaternized product to remove Cr6+
(as dichromate), As5+ (as arsenate), and P5+ (as phosphate) from solution in three
separate studies. Their adsorption capacities were 0.62, 0.25, and 0.31 mmol/g for
Cr6+, As5+, and P5+, respectively. L. H. Wartelle and W. E. Marshall (unpublished
data) recently applied the single-step quaternization procedure to both rice hull and
straw and observed P5+ adsorption values of 0.46 and 0.54 mmol/g for hull and
straw, respectively.
In contrast to the single-step method of quaternizing rice co-products, two-step
(Orlando et al, 2003) and three-step (Orlando et al, 2002a,b) quaternization methods
have been employed to increase the anion-adsorption capacity of rice hull. The two-
step method takes advantage of a reaction between alcohol –OH groups in the lignin
polymer and thionylchloride (SOCl2) to produce a chloride that reacts with
dimethylamine to add a cationic nitrogen anion-binding site. With rice hulls as sub-
–
strate, the maximum adsorption was only 0.10 mmol of nitrate ion (NO3 ) bound per
gram. The authors felt that the relatively low lignin content of rice hulls contributed
to the lack of significant nitrate adsorption. In the three-step reaction, the authors
utilized an initial reaction of pyridine and dimethylformamide with the primary
alcohol –OH group on glucose contained within the cellulose polymer. Epichloro-
hydrin was then added to create a reactive site for the addition of dimethylamine,
which carried the cationic or reactive nitrogen. This method proved to be successful
–
for the adsorption of NO3 , and in separate studies, adsorption maximums of 1.20
mmol/g (Orlando et al, 2002a) and 1.30 mmol/g (Orlando et al, 2002b) were
achieved.
The data from this section are summarized in Tables 3 and 4. Table 3 clearly
shows that unmodified hull and straw are poor adsorbers of metal ions. There
appears to be only one modification strategy that produced hull resins in the “fair”
category of metal-ion binding. Reaction of the hull surface with heat in the presence
of a polycarboxylic acid (citric or tartaric) resulted in a major improvement in the
uptake of metal ions. Table 4 demonstrates that a single-step quaternization with
CHMAC allows for much higher anion binding than adsorption of anionic dyes on
the surface of rice hull. A three-step quaternization imparted to rice hull greater
adsorption for nitrate ion than for any other cation or anion given in the tables.
If rice hull and straw are to be considered value-added cation- or anion-exchange
resins able to compete effectively with commercial resins, they must undergo sur-
face modification to dramatically increase their surface cationic or anionic charge.
TABLE 3
Adsorption of Different Cations by Unmodified and Modified Rice Hull and Straw
By- Ion Adsorption
Cation Product Modification (mmol/g) Reference
Antimony2+ Hull None <0.01b Khalid et al (2000)
Cadmium2+ Hull None 0.19a Roy et al (1993)
Hull None 0.06b Khalid et al (1998a)
Hull Anionic dyes 0.12b Suemitsu et al (1986)
Chromium3+ Hull None 0.04b Khalid et al (1999b)
Hull 5% NaOH <0.01b Marshall et al (1993)
Hull EDTA 0.18a Low et al (2000a)
Cobalt2+ Hull 5% NaOH <0.01b Marshall et al (1993)
Copper2+ Hull None 0.06b Kumar and Dara (1979)
Straw None 0.04b Kumar and Dara (1979)
Hull Citric acid 0.76b Wartelle and Marshall
(2000)
Hull Tartaric acid 0.46a Wong et al (2003)
Lead2+ Hull None 0.05a Roy et al (1993)
Hull Tartaric acid 0.52a Wong et al (2003)
Mercury2+ Hull None 0.05b Khalid et al (1999a)
Nickel2+ Hull 5% NaOH <0.01b Marshall et al (1993)
Zinc2+ Straw None 0.06a Marshall and Champagne
(1995)
Hull None <0.01a Mishra et al (1997)
Hull Anionic dyes <0.01b Suemitsu et al (1986)
a Obtained from adsorption isotherms.
b Obtained at a single metal ion concentration.
TABLE 4
Adsorption of Various Anions by Modified Rice Hull and Straw
Ion Adsorption
Anion By-Product Modification (mmol/g) Reference
Chromium6+ Hull CHMACa 0.62b Low et al (1997)
Hull Anionic dyes 0.06c Suemitsu et al (1986)
Arsenic5+ Hull CHMAC 0.25b Lee et al (1999)
Phosphorus5+ Hull CHMAC 0.31b Low et al (2000b)
Hull CHMAC 0.46c Wartelle and Marshalld
Straw CHMAC 0.54c Wartelle and Marshalld
Nitrate1– Hull DMAe 1.20b Orlando et al (2002a)
Hull DMA 1.30b Orlando et al (2002b)
Hull DMA 0.10b Orlando et al (2003)
a N-(3 chloro-2-hydroxypropyl) trimethylammonium chloride.
b Obtained from adsorption isotherms.
c Obtained at a single metal ion concentration.
d Unpublished data.
e Dimethylamine.
Adsorption capacities of 0.50 mmol/g or greater could compete with some of the
low-end commercial resins. Tables 3 and 4 show that this adsorption value was
exceeded in only six out of 37 cases and only after chemical modification. Further
research is required to produce a competitive ion-exchange resin from hull or straw.
Investigators should begin their quest by starting with proven modification reagents
(such as polycarboxylic acids, CHMAC, or dimethylamine) and refining or altering
these modification strategies.
As Rice Hull Ash
As was noted at the beginning of this chapter, approximately 4 billion pounds

(1.8 billion kilograms) of rice hull and 26 billion pounds (11.8 billion kilograms) of
rice straw were produced in the United States in crop year 2002. A significant
portion of the rice hull and lesser amounts of rice straw are burned in boilers.
Energy gleaned from these co-products can be harnessed to manufacture steam for
powering turbines and generating electricity or parboiling rice. Either way, the
controlled burning of hull and/or straw produces a residue, namely, rice hull or rice
straw ash. Ash is a by-product that consists of about 5–30% carbon and about 70–
95% silica, usually with <1% alkali oxides and alkali earth oxides. The variation in
carbon content is due to the amount of air present during burning. Low amounts of
air in contact with the hulls or straw produce ash with a high carbon content. On the
other hand, if an ash with a high silica content is desired, large volumes of air in
close contact with the hull or straw particles must be maintained during combustion.
Ash is a low-cost combustion by-product that contains two adsorptive compo-
nents, a carbon-rich char and silica. Since ash is predominantly silica, its adsorptive
properties are largely attributable to this component.
PURIFICATION OF FOOD OIL

Several components in crude vegetable oils require removal before the oil can be
sold to the consumer. They consist of free fatty acids (FFAs), which can impart
rancid flavor to the oil if not removed; color constituents, such as lutein and caro-
tene, that impart a yellow or yellow-orange hue to the oil that consumers find
objectionable; and phospholipids, which can be hydrolyzed and oxidized to create
rancidity.
Normally, the oil refining industry uses silicate-based bleaching clays or
“bleaching earths” composed mainly of the mineral montmorillonite or amorphous
synthetic silica adsorbents. Hull or straw ash is composed of silicate in the form of
trymidite or cristobalite. However, considering the quantity of rice hull or straw ash
produced, and the fact that ash has a high silica content, this material should be
investigated as a value-added, potentially less-expensive substitute for the commer-
cial bleaching clays.
Before use, bleaching clay is “activated” by exposure to a mineral acid that
enhances its adsorptive properties. Proctor and Palaniappan (1989) determined that
acid activation of alkaline (unactivated) rice hull ash improved its adsorption of the
yellow pigment soy oil lutein to the extent that there was little difference in lutein
adsorption between the acid ash and a commercial bleaching earth. However, acid
activation of alkaline rice hull ash inhibited FFA binding to the ash, which differed
from the results found with lutein adsorption (Proctor and Palaniappan, 1990). Since
the FFA portion of the soybean oil was adsorbed with the lutein, the mode of action
of binding may be different for the two fractions. On the other hand, exposure of
rice hull ash to water before use can “deactivate” the ash and significantly reduce
lutein adsorption (Palaniappan and Proctor, 1990). Another important component of
crude vegetable oil, phospholipid, was found to be adsorbed about equally by either
acid or alkaline rice hull ash, but the presence of isopropanol enhanced phospholipid
binding to both types of ash, with adsorption to acid-activated ash slightly more
enhanced than absorption to the alkaline material (Proctor et al, 1992). In a
comprehensive study of rice hull ash adsorbent performance, Proctor et al (1995)
compared the bleaching performance of low-carbon ash (silica), high-carbon ash,
silica hydrogel, and a commercial bleaching clay. Their abilities to adsorb pigment
(lutein), FFAs, phospholipids, and peroxides were determined under commercial soy-
oil bleaching conditions. Rice hull silica excelled in the removal of phospholipids,
with adsorption greater than that of the bleaching clay. However, bleaching clay was
considerably better than all other adsorbents in removal of peroxides and lutein. None
of the materials was able to significantly reduce FFA levels. The authors concluded
that the low surface area of the rice hull silica hindered adsorption of pigments and
peroxides and believed that a product with much higher surface area would make this
material more competitive with commercial bleaching clays.
Rice hull ash was also investigated for its ability to remove carotene from palm
oil (Liew et al, 1993). Three types of ash (untreated, acid-activated, and unwashed
acid-treated) were used. Unwashed, acid-treated ash was found to be the best
adsorbent of the three and was nearly as effective as Filtrol 105, a commercial
bleaching clay. The authors believed that carotene removal was a result of surface
interactions between the adsorbed mineral acid, sulfuric acid, and the carotene.
A series of recent investigations has centered on the removal of impurities from
sesame oil by rice hull ash. Toro-Vazquez et al (2000) evaluated several adsorbents
(including rice hull ash, two synthetic silicas, and a carbon made from pine wood)
for their ability to remove FFAs and caroteniods. The carbon had significantly
higher FFA adsorption than the other adsorbents. However, none of the adsorbents
exhibited a significant advantage in carotenoid removal. H.-K. Chen and collabora-
tors have been particularly interested in improving the bleaching efficiency (BE) of
rice hull ash in a sesame oil system by careful analysis of how the ash was prepared
(Chang et al, 2001a), effect of activation conditions (Lin et al, 2001), and an
analysis of ash structure (Chang et al, 2001b). Using rice hulls that had been
combusted under specific time and temperature conditions and activating the ash
with sulfuric acid under specific time and temperature conditions, Chang et al
(2001a) observed a BE of 46.2%, which was still lower than the 52.3% BE achieved
by an activated, commercial bleaching clay. Lin et al (2001) found that the ratio of
rice hull ash to sesame oil had the largest effect on BE of any of the activation
condition variables. At a hull-oil ratio of 0.05 g/g, a BE of more than 60% was
determined. This value surpassed the BE for a commercial bleaching clay. How-
ever, the clay-oil ratio was lower than 0.05 g/g. In addition, high hull-oil ratios
increase the amount of residual oil on the adsorbent, thus increasing production
costs for refining the oil. A compromise adsorbent-oil ratio is necessary to balance
the factors of ash required, oil adsorbed, and BE.
Pore size and surface area of rice hull ash were found to play a role in BE. Chang
et al (2001b) observed maximum BE at a surface area of about 200 m2/g and an
average pore diameter of 50 Å. Unfortunately, the maximum value for the BE was
only 33.7%.
Although the investigations described in this section failed to provide a rice hull
ash that was equal to or better than a commercial bleaching clay in removal of
vegetable oil impurities, several of the ash products came close. One must carefully
compare the economics of rice hull ash production vs. the manufacturing or sales cost
of potentially competitive bleaching clays. Rice hull ash is a renewable commodity
present in large quantities as a by-product from energy or steam generation, while
bleaching clays must be dug from specific sites in the United States and is a non-
renewable resource. Although BE may be somewhat lower with hull ash, the potential
difference in cost may be sufficient to drive the market toward ash, especially in
countries that have plentiful rice hulls but must import bleaching clay.
DYE AND METAL ION ADSORPTION

Like its rice hull and rice hull carbon counterparts, rice hull ash has been used to
adsorb acid dyes and metal ions from solution. Chou et al (2001) employed rice hull
ash to adsorb the acidic dye Congo Red from solution and determined an adsorption
capacity of 0.24 mmol/g. The authors also combined hull ash with kaolin and starch
and observed an 88% decrease in the ability of the ash to adsorb the dye. They con-
cluded that the dye was interacting with silanol groups (SiOH) on the ash surface
and that the additives kaolin and starch reduced the number of silanol groups avail-
able to the dye by simple dilution.
Rice hull ash has been observed to remove Hg2+ (Tiwari et al, 1995) and gold
(Au1+) contained within a gold-thiourea complex (Nakbanpote et al, 2000). Tiwari
et al (1995) used a low-carbon (10%) ash. After optimization of experimental con-
ditions of contact time, metal ion concentration, particle size, and pH, an adsorption
capacity of 0.33 mmol/g of ash was calculated. Nakbanpote et al (2000) found
that a high-carbon ash (38%), compared to low-carbon ashes, was a better
adsorbent for Au1+. The authors also compared their high-carbon ash to a com-
mercial activated carbon and found maximum gold adsorption to be 0.11 and 0.17
mmol/g of ash or carbon, respectively. They postulated that the mechanism of
Au1+ adsorption on the high-carbon ash was attributable to contributions from
silanol groups on the silica portion of the ash and various carbonyl groups on the
carbon portion.
In the limited literature on dye and metal ion adsorption by rice hull ash, only
low levels of adsorption were observed. Perhaps, as with activated carbon and non-
carbonized rice hull and straw, the best approach to increased adsorption is through
surface modification. In this case, though, it will be primarily surface functional
groups on silica that will be modified. However, ash modification may be a unique
approach, as the author is not aware of any literature that describes surface modifi-
cation of ash from rice co-products by chemical or physical means.
SUMMARY
As stated at the outset, the objective of this review was to highlight the utiliza-
tion of the rice co-products, hull and straw, as value-added products. From this
broad classification, the review narrowed the focus to the utilization of hull and
straw, in various forms, as adsorbent material. The narrowed focus was brought
about by the high value and environmental impact of commercial adsorbents, such
as activated carbon and ion-exchange resins, in numerous residential, industrial, and
environmental applications.
The use of agricultural co-products, in general, and rice co-products, in particu-

lar, for adsorbent production is a very active research endeavor that is taking place
in laboratories worldwide. Many scientists believe that the world pollution level is
increasing with each passing year. Effective adsorbents are available to remediate at
least part of the pollution, but they can be cost prohibitive, especially to small busi-
nesses and to emerging industrial nations, where air and water pollution are
endemic. In countries, like the United States, with a strong agricultural base, bil-
lions of pounds of commodity co-products are generated. For the most part, they
have little or no value, and their disposal can be problematic. Many of the emerging
industrial nations have strong agricultural bases and are in an excellent position to
utilize their co-products as adsorbent material.
The studies highlighted in this review have demonstrated that rice hull and straw
and their by-products, hull and straw ash, have little adsorptive capacity. When they
were converted to the carbon form, their adsorptive capacity increased for primarily
organic adsorbates, but sequestration of metal ions was not significantly improved.
However, methods were identified to increase the adsorptive characteristics of the
adsorbents to the point at which, in some cases, they could be considered competi-
tive with commercial products. If hull- or straw-based adsorbents can be developed
to effectively compete in the marketplace, they should be less expensive to manu-
facture, based solely on the low cost of the starting material.
Currently, adsorbent markets in the United States are mature, but there are no
hull- and straw-based adsorbents in those markets. Current research on the utiliza-
tion of hull, straw, and ash as adsorbents is at a point where economic analyses
should be carried out on processes that yield the best adsorbents. The production
costs generated should be matched to production costs for commercial carbons, ion-
exchange resins, and bleaching clays for a specific application. If commercial pro-
duction costs are not available, then comparisons should be made with known mar-
ket costs. Favorable cost comparisons would generate discussions with potential
customers to determine market size and precise characteristics of the product
desired for a specific consumer application. In the final analysis, good to excellent
adsorbents have been and can be developed using various modification strategies.
However, it will take entrepreneurship on the part of scientists working with
manufacturers to carry the products to the marketplace. This can be accom-
plished, and the rewards will be beneficial to the rice grower, processor, and
commercial sector.
REFERENCES
Ahmedna, M., Johns, M. M., Clarke, S. J., Bansal, R. C., Donnet, J.-B., and Stoeckli, F.
Marshall, W. E., and Rao, R. M. 1997. Poten- 1988. Manufacture of active carbons. Pages 1-
tial of agricultural by-product-based activated 26 in: Active Carbon. Marcel Dekker, Inc.,
carbons for use in raw sugar decolorization. J. New York.
Sci. Food Agric. 75:117-124. Chang, Y. Y., Lin, C. I., and Chen, H. K. 2001a.
Ahmedna, M., Marshall, W. E., and Rao, R. M. Effect of acid activation on sesame oil
2000. Production of granular activated carbons bleaching by rice hull ash. J. Chem. Eng. Jpn.
from select agricultural by-products and 34:1-6.
evaluation of their physical, chemical and Chang, Y. Y., Lin, C. I., and Chen, H. K. 2001b. Rice
adsorption properties. Bioresour. Technol. hull ash structure and bleaching performance
71:113-123. produced by ashing at various times and
Anonymous. 2002. Activated Carbon. Report temperatures. J. Am. Oil Chem. Soc. 78:657-660.
1593. The Freedonia Group, Inc., Cleveland, Chou, K. S., Tsai, J. C., and Lo, C. T. 2001. The
OH. adsorption of congo red and vacuum pump oil
by rice hull ash. Bioresour. Technol. 78:217- Khalil, L. B. 1996. Adsorption characteristics of
219. activated carbon obtained from rice husks by
CiCi, M., and Keles, E. 1990. Removal of chro- treatment with phosphoric acid. Adsorpt. Sci.
mium (VI) ions from solutions with active Technol. 13:317-325.
carbon prepared from rice hulls. Pak. J. Ind. Kim, K. S., and Choi, H. C. 1998. Characteristics
Res. 33:347-350. of adsorption of rice-hull activated carbon.
Deo, N., and Ali, M. 1996. Dye removal from Water Sci. Technol. 38:95-101.
aqueous solutions by adsorption on a low cost Kumar, P., and Dara, S. S. 1979. Studies on
material. Inst. Eng. (India) Environ. Eng. Div. binding of copper ions by some natural poly-
76:48-51. meric materials. Chem. Era (Dec.):20-23
Guo, Y., Qi, J., Yang, S., Yu, K., Wang, Z., and Lee, C. K., Low, K. S., Liew, S. C., and Choo, C.
Xu, H. 2002. Adsorption of Cr(VI) on micro- S. 1999. Removal of arsenic (V) from aqueous
and mesoporous rice husk-based active car- solution by quaternized rice husk. Environ.
bon. Mater. Chem. Phys. 78:132-137. Technol. 20:971-978.
Guo, Y., Yang, S., Fu, W., Qi, J., Li, R., Wang, Liew, K. Y., Yee, A. H., and Nordin, M. R. 1993.
Z., and Xu, H. 2003. Adsorption of malachite Adsorption of carotene from palm oil by acid-
green on micro- and mesoporous rice husk- treated rice hull ash. J. Am. Oil Chem. Soc.
based active carbon. Dyes Pigm. 56:219-229. 70:539-541.
Harrison, G. C. 1987. Biological filtration proc- Lin, S. P., Lin, C. I., and Chen, H. K. 2001. Ses-
ess. U.S. patent 4,676,907. ame oil bleaching with activated rice hull ash.
Imagawa, A., Seto, R., and Nagaosa, Y. 2000. J. Chem. Eng. Jpn. 34:737-742.
Adsorption of chlorinated hydrocarbons from Low, K. S., Lee, C. K., and Ng, A. Y. 1997.
air and aqueous solutions by carbonized rice Chromium (VI) sorption on quaternized rice
husk. Carbon 38:628-630. hulls. J. Environ. Sci. Health A32:1849-
Johns, M. M., Marshall, W. E., and Toles, C. A. 1860.
1998. Agricultural by-products as granular Low, K. S., Lee, C. K., Wong, S. Y., and Tang,
activated carbons for adsorbing dissolved met- P. L. 2000a. Metal sorption enhancement of
als and organics. J. Chem. Technol. Biotech- rice hull through chemical modification. Envi-
nol. 71:131-140. ron. Technol. 21:1239-1244.
Johns, M. M., Toles, C. A., and Marshall, W. E. Low, K. S., Lee, C. K., and Wong, S. M. 2000b.
2003. Activated carbons from low-density Sorption enhancement of phosphorus by
agricultural waste. U.S. patent 6,537,947. chemically modified rice hull. J. Environ. Sci.
Juliano, B. O. 1985. Rice hull and rice straw. Health A35:253-261.
Pages 689-755 in: Rice: Chemistry and Tech- Malik, P. K. 2003. Use of activated carbons pre-
nology, 2nd ed. B. O. Juliano, Ed. Am. Assoc. pared from sawdust and rice-husk for adsorp-
Cereal Chem., St. Paul, MN. tion of acid dyes: A case study of acid yellow
Kannan, N., and Meenakshisundaram, M. 2002. 36. Dyes Pigm. 56:239-249.
Adsorption of congo red on various activated Marshall, W. E., and Champagne, E. T. 1995.
carbons: A comparative study. Water Air Soil Agricultural byproducts as adsorbents for
Pollut. 138:289-305. metal ions in laboratory prepared solutions and
Khalid, N., Rahman, A., Ahmad, S., Kiani, S. N., in manufacturing wastewater. J. Environ. Sci.
and Ahmed, J. 1998a. Adsorption of cadmium Health A30:241-261.
from aqueous solutions on rice husk. Radio- Marshall, W. E., Champagne, E. T., and Evans,
chim. Acta 83:157-162. W. J. 1993. Use of rice milling by-products
Khalid, N., Ahmad, S., Kiani, S. N., and Ahmed, (hull and bran) to remove metal ions from
J. 1998b. Removal of lead from aqueous solu- aqueous solution. J. Environ. Sci. Health
tions using rice husk. Sep. Sci. Technol. A28:1977-1992.
33:2349-2362. McKay, G., Ramprasad, G., and Mowli, P. P.
Khalid, N., Ahmad, S., Kiani, S. N., and Ahmed, 1986. Equilibrium studies for the adsorption
J. 1999a. Removal of mercury from aqueous of dyestuffs from aqueous solutions by low-
solutions by adsorption to rice husks. Sep. Sci. cost materials. Water Air Soil Pollut. 29:273-
Technol. 34:3139-3153. 283.
Khalid, N., Rahman, A., Ahmad, S., Toheed, A., Meyer, V., Carlsson, F. H. H., and Oellermann,
and Ahmed, J. 1999b. Adsorption behavior of R. A. 1992. Decolourization of textile effluent
rice husk for the decontamination of chro- using a low cost natural adsorbent material.
mium from industrial effluents. J. Radioanal. Water Sci. Technol. 26:1205-1211.
Nucl. Chem. 240:775-781. Mishra, S. P., Tiwari, D., and Dubey, R. S. 1997.
Khalid, N., Ahmad, S., Toheed, A., and Ahmed, The uptake behaviour of rice (jaya) husk in the
J. 2000. Potential of rice husks for antimony removal of zn(II) ions—A radiotracer study.
removal. Appl. Radiat. Isot. 52:31-38. Appl. Radiat. Isot. 48:877-882.
Mostafa, M. R. 1994. Adsorption of palladium Proctor, A., Clark, P. K., and Parker, C. A. 1995.
ions by modified carbons from rice husk. Pak. Rice hull ash adsorbent performance under
J. Sci. Ind. Res. 37:96-99. commercial soy oil bleaching conditions. J.
Mostafa, M. R. 1997. Adsorption of mercury, Am. Oil Chem. Soc. 72:459-462.
lead, and cadmium ions on modified activated Rodriquez-Reinoso, R. F., and Solano, A. L.
carbons. Adsorpt. Sci. Technol. 15:551-557. 1989. Microporous structure of activated car-
Munaf, E., Zein, R., Kurniadi, R., and Kurniadi, bons as revealed by adsorption methods.
I. 1997. The use of rice husk for removal of Chapter 1 in: Chemistry and Physics of Car-
phenol from wastewater as studied using 4- bon, Vol. 21. P. A. Thrower, Ed. Marcel Dek-
aminoantipyrine spectrophotometric method. ker, Inc., New York.
Environ. Technol. 18:355-358. Roy, D., Greenlaw, P. N., and Shane, B. S. 1993.
Nakbanpote, W., Thiravetyan, P., and Adsorption of heavy metals by green algae
Kalambaheti, C. 2000. Preconcentration of and ground rice hulls. J. Environ. Sci. Health
gold by rice husk ash. Miner. Eng. 13:391- A28:37-50.
400. Suemitsu, R., Uenishi, R., Akashi, I., and
NASS. 2003. Grain and feed. Page I-16 in: Agri- Nakano, M. 1986. The use of dyestuff-treated
cultural Statistics 2003. U.S. Dep. Agric.-Natl. rice hulls for removal of heavy metals from
Agric. Statistics Serv. U.S. Govt. Printing waste water. J. Appl. Polym. Sci. 31:75-83.
Office, Washington, DC. Teker, M., Saltaba, Ö., and emamoXlu, M. 1997.
Nawar, S. S., and Doma, H. S. 1989. Removal Adsorption of cobalt by activated carbon from
of dyes from effluents using low-cost agri- rice hulls. J. Environ. Sci. Health A32:2077-
cultural co-products. Sci. Total Environ. 2086.
79:271-279. Teker, M., emamoXlu, M., and Saltaba, Ö. 1999.
Oh, G. H., and Park, C. R. 2002. Preparation and Adsorption of copper and cadmium ions by
characteristics of rice-straw-based porous car- activated carbon from rice hulls. Turk. J.
bons with high adsorption capacity. Fuel Chem. 23:185-191.
81:327-336. Tiwari, D. P., Singh, D. K., and Saksena, D. N.
Orlando, U. S., Baes, A. U., Nishijima, W., and 1995. Hg(II) adsorption from aqueous solu-
Okada, M. 2002a. Preparation of agricultural tions using rice-husk ash. J. Environ. Eng.
residue anion exchangers and its nitrate maxi- 121:479-481.
mum adsorption capacity. Chemosphere Toles, C. A., Marshall, W. E., and Johns, M. M.
48:1041-1046. 2000. Activated nutshell carbons from agri-
Orlando, U. S., Baes, A. U., Nishijima, W., and cultural waste. U.S. patent 6,033,573.
Okada, M. 2002b. A new procedure to pro- Toro-Vazquez, J. F., Mendez-Montealvo, G., and
duce lignocellulosic anion exchangers from Charó-Alonso, M. A. 2000. Adsorption effi-
agricultural waste material. Bioresour. Tech- ciency of selected adsorbents in sesame oil
nol. 83:195-198. miscellas. J. Food Lipids 7:151-162.
Orlando, U. S., Okuda, T., Baes, A. U., USAID. 1989. Rationale for commodity focus.
Nishijima, W., and Okada, M. 2003. Chemical Pages 3-9 in: Biomass Energy Systems and
properties of anion-exchangers prepared from Technology Project. The A.I.D. Approach:
waste natural materials. React. Funct. Polym. Using Agricultural and Forestry Wastes for the
55:311-318. Production of Energy in Support of Rural
Palaniappan, S., and Proctor, A. 1990. Competi- Development. (Bioenergy Systems Rep.)
tive adsorption of lutein from soy oil onto rice Office of Energy-U.S. Agency for Interna-
hull ash. J. Am. Oil Chem. Soc. 67:572-575. tional Development, Washington, DC.
Pendyal, B., Johns, M. M., Marshall, W. E., Usmani, T. H., Ahmad, T. W., and Yousufzai, A.
Ahmedna, M., and Rao, R. M. 1999. Removal H. K. 1994. Preparation and liquid-phase char-
of sugar colorants by granular activated car- acterization of granular activated-carbon from
bons made from binders and agricultural by- rice husk. Bioresour. Technol. 48:31-35.
products. Bioresour. Technol. 69:45-51. Wartelle, L. H., and Marshall, W. E. 2000.
Proctor, A., and Palaniappan, S. 1989. Soy oil Citric acid modified agricultural by-products
lutein adsorption by rice hull ash. J. Am. Oil as copper ion adsorbents. Adv. Environ. Res.
Chem. Soc. 66:1618-1621. 4:1-7.
Proctor, A., and Palaniappan, S. 1990. Adsorp- Wong, K. K., Lee, C. K., Low, K. S., and Haron,
tion of soy oil free fatty acids by rice hull ash. M. J. 2003. Removal of Cu and Pb by tartaric
J. Am. Oil Chem. Soc. 67:15-17. acid modified rice husk from aqueous solu-
Proctor, A., Tan, L. C., and Palaniappan, S. 1992. tions. Chemosphere 50:23-28.
Phospholipid adsorption onto rice hull ash Youssef, A. M., and Mostafa, M. R. 1992.
from soy oil miscellas. J. Am. Oil Chem. Soc. Removal of copper ions by modified activated
69:1049-1050. carbons. Indian J. Technol. 30:413-416.
INDEX
2-Acetyl-1-pyroline, 44, 94, 96, profile, 447 Arrhenius equation, 204
452, 458 recrystallization of, 117 Aspergillus spp., 541, 542, 547,
Activated carbon, from rice complexation with lipids, 55, 556
activation methods, 613–614 60, 62, 63, 451 Atmosphere, modified, in
hull and straw, 614, 615–619 content storage, 280–281
use in brewing, 545, 556 breeding criterion, 43–44
Adenosine 5-diphosphatase and cultural practices, 97, 98 Baby foods, 475, 488–489, 498,
(ADP), 118 determination, 441–443 532
ADP-glucose, 118, 119–120 and hydration during Backcross breeding, 36
ADP-glucose pyrophos- cooking, 424, 426 Basmati rice, 334, 335, 417
phorylase, 118–120, 123 insoluble, 442, 443, 444 Beer, 489
Aeration, in storage, 249, 253, in parboiled rice, 331, 360, adjuncts in, 559–560
254–255, 274–275 384, 386 rice as, 561–564
airflow, 197, 256, 257 in rice noodles, 517 all-malt, 563
Aging, effect on rice properties, and solids loss in cooking, analytical methods for, 562
275–277 427–428 glyoxal in, 563
Agrobacterium tumefaciens, 34 variation in, 122 ingredients, 557–560
Albumin, 145, 146 and water-to-rice ratio, 439, low-calorie, 564
Aleurone, 83, 86, 169 440 malting, 559
Alkali spreading value, 43, 45 in cultivars, various, 66, 67, manufacture of, 560–561
Amino acids 69, 70, 71, 72 off-flavor, 563
in milled rice, 145, 154 molecular weight distribution, Beetles, 278–279
in Osborne protein fractions, of chains, 115 Bin drying, 256–261
145–148 in products from parboiled Birds, and storage, 281–282
in parboiled rice, 365 rice, 64, 68 Birefringence, 111, 428
processing, effects on, 154–155 quality types, 443–444, 445, Blue protein, 572
in rice bran, 572, 574 447 “Boil-in-the-bag” rice, 409, 410
in rice for brewing, 549 and starch lipid content, 165 Bran (kernel part)
Amorphous regions, of starch structure, 109, 114–115, 448, content in kernel, 149
granule, 112–113 450 lipids in, 170
Amylase, 556 synthesis, 123–125, 129 in parboiled rice, 374
alpha, 498, 543, 547 Amylose extender, 148 removal, 293–295
beta, 113 Amylose-amylopectin ratio, 113 Bran (milling fraction) (see Rice
Amylopectin, 383–384, 390, Anemia, 407 bran)
393 (see also Preamylo- Angle of repose, 197–198, 371 Branching enzymes
pectin) Angoumois grain moth, 278 BEI, 128–129
chains, 113, 124, 125, 442 Animal feed, 480, 579–580, 581 BEII, 129–131
distribution of, 114, 444– (see also Pet foods) Breeding (see also Varieties)
445, 448 Anther culture, 36–38 bulk, 36
elongation of, 125–128 Antioxidants history of, in United States, 33
profile, 445–447 in kernel, 90 lines, development, 34–39
recrystallization of, 117 and parboiled rice, 367, 372, programs, public, 49, 51, 55,
clusters, 112 374 61, 64, 66
structure, 109, 113–114, 447, Antisense technology, 125 for “quality” attributes, 417–
448, 450 Arborio rice, 44, 64, 471, 476 418
synthesis, 123, 131–132, 133 Aroma (see also Odor) selection methods, 36–38, 39–
Amylose factors affecting, 307, 313 45
apparent, 442, 455, 456 volatile compounds in, 94, 452– of “super rice,” 22
branches, 129 453 techniques, 33–34
chains, 114 Aromatic rices, 5, 443, 475 (see Brewers’ rice, 476, 489
elongation of, 120–125 also Specialty rices) quality requirements, 562–563
molecular weight trade, 14 Brewing
distribution of, 115, 117 varieties, 69–70, 72 amount brewed, 541
631
of beer (see Beer) effect of rice bran on, 580–581, Diabetes mellitus, 596, 597
enzymes in, 542–543, 547, 556, 597–599 Dietary fiber, 576, 581
564 effect of rice bran oil on, 581, Diets, rice in
with liquefied rice, 556–557 587–588, 600–604 in Asia, 18, 153
milling of rice for, 541, 552, Climate, and grain composition, diversification, 18, 20
554, 555, 561–562 98 nutritional properties, 155–156
milling ratio, 541, 550, 552, Coefficients Diffusion, equation, 204
554 of film heat transfer, 200 Digestibility of rice, and
and molds, 542, 547, 556 of friction, 196 processing, 154, 155
rice varieties used in, 546, 548, of hygroscopic expansion, 196 Discoloration, of grain
557, 562 of thermal expansion, 195–196 and parboiling, 32, 368–370
of sake (see Sake) Color, of grain, 408, 410 (see by pests, 31
steaming, 555 also Discoloration) from wet holding, 302, 303–
steeping, 554–555 Composition 304
yeast in, 544–545, 550 of bran, 91–93 Dry matter loss, 302
Brown plant hopper, 34 of brown rice, 95–96 Dryers
Brown rice, 290, 473 factors affecting, 96–100 adjustments to, 253–255
composition, 95–96 of hull, 88–91 column, 215, 216
carbohydrates, 548 of milled rice, 93–94 combination system, 255
lipids, 168, 170, 171, 176, of rough rice, 88 concurrent-flow, 252
181, 549–551 Consumption, of rice continuous-flow, 215, 216,
minerals, 551 per capita, 19–20 250–252, 347
proteins, 143, 145–148, 548– in various regions, 18–19 depth of rice in, 257
549 worldwide, 18 multipass, 252
dimensions, 193 Cooking, of rice Drying (see also Pre-drying
grain volume, change with ease of, and parboiling, 338 period)
hydration, 339–340 and glycemic index, 596 delayed, 305
parboiling of, 358–359 methods, for instrumental dielectric, 261–262
physicochemical properties, testing, 438–441 duration, 256, 308
547–548 and textures, 331 and fissuring, 201, 307, 346–
production of, steps, 291–293 Cooking quality 347 (see also Fissuring)
Bulgur wheat, and parboiling, breeding for, 43–44 fluid-bed, 261
329–330 and chalky kernels, 65 identity preservation, 248–249
Burkholderia glumae, 32, 41 effects of rice aging on, 275– infrared, 255, 261
277 microwave energy for, 262
Cadra cautella, 280 of noodles, 515, 517 milling quality, effects on,
Calcium, fortification with, 412 Cooking time, 424, 427 307–310
California growing region, 61– optimal, 424–425 of parboiled rice, 346–347
63, 71–73 Cracking (see also Fissuring) regions, in grain, 204
Cancer, reduced risk, and rice and glass transition research needed, 264
bran, 605–606 temperature, 201, 347 of rice bran, 579
Canned rice, 355–356, 392 and thermal expansion, 195– of rough rice
varieties for, 67, 69–70 196 alternative methods, 261–262
Carbohydrates time of, 346–347 bin drying, 256–261
in kernel, 28–29, 88 Cracks, 195 (see also Fissures) commercial drying, 248–255
in rice for brewing, 548 Cross-pollination, 35–36 drying models, 245–248
Carbonyl compounds, 452–453 Cryptolestes spp., 279 drying process, 238–345
Celiac disorder, 530 Crystallinity, of starch granule, early studies, 224–229
Ceramides, in lipids, 176–178 111–112 in the field, 211, 229–236
Cercospora janseana, 32 Cultivars (see also Varieties) and harvest moisture
Cereal grains other than rice Asian, differences from U.S. content, 236–238
comparison with rice cultivars, 27, 28 high-temperature, 246, 248,
brans, 570–571, 575 identification of, 458–460 254–255, 262
production, 1–2 semidwarf, 31 low-temperature, 248
starch molecules, 110, 113 superior-processing, 417, 428, in sacks, 261
yields, 1–2 431–432, 451 sensory quality, effects on,
studies of starch synthesis, U.S. cultivars, breeding, 33 310–311
126, 128, 130, 131–132 Cultural practices, effects on in storage, 274, 277
Chalkiness, 32, 64–65, 423 grain quality, 31–32 Dryland rice, 22 (see also
Chars, from rice hull and straw, kernel composition, 96–97 Upland rice)
613, 614–615 lipids, 184
Chatti process, for parboiled Ecosystems, for rice growing,
rice, 335 Degree of milling, 419–421 2–3
Cholesterol, in blood Dehydrated steamed rice, 556 Elasticity, modulus, 209–210
effect of milled rice on, 597 Density, and moisture content, Elongation ratio, 435
effect of phytonutrients on, 339–340 Embryo, structure, 82–83
605–606 Dewaxing, of rice bran oil, 585 Endosperm, 83–88, 168, 170
Index / 633
Enrichment (see also Fissuring (see also Cracking) espasol, 534

Fortification) cultivars resistant to, 64 fagao, 525–526
coated-kernel, 409 after drying, 242–245, 246– harusame, 514
definition, 405 247, 263 idli, 524
powder, 408–409 and drying conditions, 246– inginataan, 534
regulations concerning, 405– 247, 263 khao kriap wanu, 529–530
408 and fertilization, 234–235 kroepeck, 529
testing of, 410–411 and glass transition lunjiaogao, 519, 524–525
Enzymes temperature, 217, 248 luobogao, 522
in brewing, 542, 547, 556, 564 and grain development, 230 malao, 530
and parboiling, 359, 367, 374 history of research on, 224– mitaimu (bitaibah), 528
in rice bran, 573 229 mochi, 500, 501, 519–520
of starch synthesis, 119–133 and moisture monaka, 527
Equilibrium moisture content adsorption, 211–217, 247 niangao (nenkau), 519, 520–
and cracking of grains, 325 content in the field, 211, 522
during drying, 308 248, 229–236, 263 okoshi, 501, 523
for forms of rice, 202 gradients, 194, 239, 242, palitaw, 534
for individual kernels, 202 245–247 puto, 523, 529
and milling quality, 225 at harvest, 31, 230–234, rakugan, 501, 525, 529
and moisture adsorption and 236–238 sapin-sapin, 534
readsorption, 229–230 readsorption, 224, 227, 229, senbei, 499, 526–527
for parboiled rice, 372, 376– 235, 244–245, 263 shel roti, 532
380, 381, 383, 384, 390 partial cracks, 240, 242 tamale, 534
of rice bran and polish, 573 potential for, 214, 216 uiro, 519, 523
Expansion, coefficients of, 195– and relative humidity, 321, wagashi, 501
196 323, 324, 325 xue-piang-gao, 523
Extraction tensile stresses, 240, 247 zhuang-yuan-gao, 522–523
of lipids, 164, 175 time of, 230, 241 products, Asian types (see also
of protein components, 150 Flaked rice, 66 Products and co-products)
Extrusion Flavor, factors affecting, 307, baby foods, 533
and oil extraction, 583 310, 313, 320–321 beverages, 533
for stabilization of rice bran, Flood-prone rice, 3, 22 breakfast cereals, 532–533
578 Flour, rice cakes, 518–526
Asian countries, use in, 495 crackers, 500, 526–530
False smut, 41 composite, 531 dry cake, 523
Fans, use in drying, 256, 258– extrusion of, 499, 515, 520, fermented, 519, 523–526,
259 524, 527–529, 532–533 533
Fat (see also Lipids; Oil, rice functional properties, leavened products, 523–526,
bran) influences on 531–532
changes during parboiling inherent properties, 502–508 miscellaneous, 533–534
(steaming), 361–363, 372 milling, 510–514 noodles, 514–518
Fatty acids storage, 508–509 pastry products, 519–523
and brewing, 550–551 high-amylose, 514, 529 pudding, 519, 523
in ceramides, 177 from parboiled rice, 361, 376, rice milk, 533
in developing grain, 184, 185 378 snacks, 534
in nonstarch lipids, 168 particle size, effects, 510–512, unleavened products, 532
in protein-body lipids, 173 516–517, 532 for wheat intolerance, 530
in rice bran, 577 precooked, 499, 500–501 protein in, 152, 498, 513
in rice bran oil, 171, 582, 585, preparation methods types
586, 587 dry milling, 497–498, 510– brown, 479, 499
in starch lipids, 165, 166, 167 512 high-protein, 498
in stored grain, 186 semidry milling, 497, 514 pregelatinized, 479, 485, 498
in waxes, 182 wet milling, 497, 512–513 specialty (Japanese), 499–
Feed (see Animal feed) price, 495–497 502
Fertilization of plant, 29 production, 495 waxy (see Waxy rice)
Fertilizer treatments, 31, 97 products, Asian (see also Folate, 406–407
Ferulate esters, 174 (see also Products and co-products) Folic acid, fortification with,
Oryzanol) apam, 526 405, 406–408, 410
Fiber, 605 arare, 526–527 Foots (see Soap stock)
dietary, 573, 576–577, 581, bibingka, 522 Fortification (see also
597, 599 bifun (bihon), 514 Enrichment)
Field moisture content, 215 bingpi, 523 definition, 405
Fissures boza, 533 with folic acid, 405, 406–408,
and broken kernels, 321–325, buchi, 534 410
419 carrioca, 534 guidelines, 406
“sun-cracks,” 225, 264 dango, 519 in various countries, 411–413
and tensile strength, 206 dosai, 524 Fragrant rice (see Aromatic rices)
Friction, coefficients of, 196 Grain (see also Kernel; Seed) economic analysis, need for,
Frying, rice bran oil for, 587, broken, 561–562 627, 628
590 composition for food oil, purification,
of milling fractions, 88–96 625–627
Gel consistency, determination of rough rice, 88 modified material, 622–625
of, 431 development, 30, 236 for organic compound
Gelatinization forms of, 476–480 removal, 616, 621
of rice noodles, 517 hygroscopic properties, 201– surface area, 614, 616, 617,
of starch, 116 205 620, 623, 625
in rice flour, 503–507 market classes, 416–417 unmodified material, 620–
Gelatinization temperature mass properties, 195–198 622
and amylopectin chains, 448 properties after parboiling, composition, 88–91
of cultivars, various, 66, 67, 367–370 as fuel, 259, 611, 612, 625
69, 70, 71 quality (see Quality, of grain) lipids in, 168
determination of, methods, structure, 77–88, 570 production, annual, 611
428–431 thermal properties, 199–201 role in hydration of paddy,
and genetic markers, 457–458 translucency, 27, 50, 58, 61 341, 345
and hydration during cooking, types, 473–476 structure, 79
425, 426 weight, normalized, 423 value-added products from,
for parboiled rice, 339, 344 Granule-bound starch synthase, 612
and rice flours, 503, 504, 513, 110, 114 Huller bran, 574
514 and amylose content, 30, 121 Hybrid rice, 38–39
and starch crystallinity, 116 and amylose-chain elongation, Hydration
General Agreement on Tariffs 115, 120–125 during cooking, 423–427
and Trade, 12 gene for, 32–33 equations for, 340, 341, 343
Genetic diversity in rice, 148 marker for, 45 of paddy, 338–345
Genetic improvement (see Green (undried) rice, 249 Hygroscopic conductivity and
Breeding) diffusivity, 204–205
Genetic markers Handling, of rice after harvest, Hypoallergenicity, and rice
for cultivar identification, 249, 260–261 products, 157, 479, 481, 488,
458–460 Happoushu, 564 489
for quality trait identification, Hardness of kernel Hypolipidemic effects, of rice
455–458 effect of parboiling, 367 components, 599, 600–604
Genetic modification, 34 and storage effects, 312–313
Genotype, effect on grain Harvest moisture content, 237 Indica rice, 4
composition, 99 Harvesting, time of, and milling amylopectin chains, 114, 126–
Germ quality, 63, 236–238, 301–302 127
composition, 573–574 Head rice, 63, 476 and amylose content, 32, 120–
defatted, nutritional effects, Head-rice yield, 64, 65, 232– 121, 122
606 233, 236–237, 418, 419 and fatty acid content, 184
lipids in, 168 factors affecting, 232, 242, and parboiling, 336
Germ plasm, variation in, 416, 276, 305, 307–310 protein digestibility, 155
417, 422 and rate of drying, 242, 254–255 trade, 14
Glass transition, 116, 248 Health benefits Iodine number, 168
Glass transition temperature, of oryzanol fraction of lipids, Insects
116, 200–201, 248 91–92, 175, 480 management of, 31, 280–281
and cracking, 201, 217 of rice bran, 92, 480 and parboiled rice, 374
definition, 196 of rice bran oil, 92, 175, 481, storage, 277–280
and fissuring, 217, 248 581, 587–590 Instant
and pellet expansion of sterol lipids, 178, 180 noodles, 517
temperature, 528 Heating (see Thermal parboiled rice, 338
and tempering, 248, 253 processing) “super rice,” breeding of, 22
and thermal conductivity, 200 Holding, before milling, 305, 307 Iron
Globoids, 83, 84 Hot spots, in grain mass, 302 deficiency, 413
Globulin, 146–147 Hull (husk) fortification with, 405, 408,
Glutelin, 145, 147–148, 547, as adsorbents 411, 412
548 activated carbons, 613, 615– “Iron rice,” 27
Glutinous rice, 5, 18 (see also 629 Irradiation, for insect control,
Waxy rice) ash, 625–627 281
trade, 14 for cation and anion Irrigated rice, 3
Glycemic index, 596 removal, 617–619, 621– Isoamylase, 113, 131, 132
Glycolipids, 165, 166, 171, 585 622, 623–625, 627
Glycosyl glycerides, 176 chars, 213, 214–215 Japonica rice , 4–5, 37–38
Golden Rice, 22, 23 for decolorization of sugar, amylopectin chains, 114, 126–
Government policies 615 127
and rice production, 4, 8, 9, 10 for dye removal, 616–617, and amylose content, 32, 120–
and trade, 12, 14, 17 620–621, 627 121, 122
Index / 635
and brewing, 546 Lissorhoptrus oryzophilus, 43 removal of broken grains,

and fatty acid content, 184 Lodging, of plant, 30 296–300
and parboiling, 336 Lysine removal of husk, 291–293
protein digestibility, 155 in mutant proteins, 148–149 yield, 31, 32, 63, 293, 294, 300
for specialty flour, 499, 500 in Osborne protein fractions, Milling machines
temperate vs. tropical, 50 145–148 graders and sorters
trade in, 14 processing, effects on, 154– for density, 286–288, 292–
Jasmine-style rice, 417 155 293
Javanica rice, 37–38, 50, 184 in rice bran, 573 electronic, 11–16
for impurities, 285–286
Kernel (see also Grain; Seed) Maltooligosaccharides, 123–124, for length, 285–286, 296–
breakage, 293–294, 295, 296, 133 297
321, 324–325, 331, 419 Market classes, U.S., 49–50, 65– for thickness, 289, 290–291
chalkiness, 64–65 66 husker, 291–292
classifications, 192 Market structure, of rice trade, laboratory, 418–419
defects, 223 14 metal separation devices, 288–
dimensions, 193–195 Megaloblastic anemia, 407 289, 298, 300
discoloration, 31, 32, 302, Methyl bromide, 280 polisher, 295
303–304 Micronutrients, deficiency of, roller mill, 552
extruded, 410 413 vertical and horizontal models,
fissuring, 64, 321–325 Microorganisms (microflora) 295–296
immature, 289 action during parboiling whiteners, 294–295, 296
length-width ratio, 421 processes, 333, 350, 360 Milling quality
and moisture entry, 202, 204 control in storage, 249, 277 analysis of, 418–421
properties management of disease pests, drying, effects on, 307–310
mechanical, 206–218 31–32 and equilibrium moisture
physical, 191–205 Microwave parboiling, 347, 354 content, 225
surface area per unit weight, Milled rice, 290, 474 and harvesting, 63, 236–238,
422, 425 amount worldwide, 582 301–303
calculation, 422 composition, 93–94 of southern varieties, 63–64
weakness, 64, 419 glycemic properties, 595–597 Milling ratio, 541, 550, 552, 554
Kernel smut, 32, 41 lipemic properties, 597 Minerals
Koji, for sake making, 542–543, production steps, 293–295 and climate, 98
549, 555, 556 proteins in, 144, 145 in kernel, 90–91, 93
storage, 273 in milling fractions, 89, 96,
Label, regulations, 407, 409, 459 Milling 551, 552
Lasioderma serricorne, 279 in brewing process, 549, 552, in parboiled rice, 365–366
Leaf blight, resistance, 34 554, 555, 561–562 in rice bran, 570
Lecithin, 586 cleaning, 284–291 and steeping, 555
Lectins, 576 control, 300 Moisture
Length of grain (see Grain, definition, 283, 495 adsorption, consequences, 211–
market classes; Grain, types) degree of 217
Lesser grain borer, 278 and functionality, 313–314, movement into grain, 202, 204
Levees, for planting, 25–26 319–320 and storage
Lipase, 92, 573, 577, 578 and kernel thickness, 314, of flour, 509
Lipids (see also Fat; Oil, rice 316, 317, 318, 319 of grain, 273, 274, 275, 277
bran) measurement of, 419, 420– Moisture content (see also
in beer, 560, 562, 563 421 Equilibrium moisture
in brown rice, 549 and sensory quality, 320– content)
categories of, 163, 164–173 321 basis, wet or dry, equations,
composition, factors affecting, delays in, effects, 305, 307 238–239
183–186 factors affecting, 283–284 and breaking or fissuring, 237,
content, 88–89, 91, 165, 168 to flour, 495–498, 510–514 324, 325
relationship to other and grain breakage, 293–294, after drying, 217, 242–245,
parameters, 451–452 295, 296, 321–325, 331 263
extraction of, 164, 175 material balance, 300 during drying, 215–216, 307–
glycosyl glycerides, 176 of parboiled rice 308
neutral, 165, 166, 167 and breakage, 370–371 and grain thickness, 314
oryzanol fraction, 173–176 and immature kernels, 289 of grains in the field, 214–215,
in purified starch, 166–168 reduced losses, 363, 365, 230–236
sphingolipids, 176–178 366–367 at harvest, 215, 274, 557
sterols, 178–180 and protein content, 154, 318, and head-rice yields, 64, 232,
surface, 420–421 552 305
tocopherols and tocotrienols, removal of bran, 293–296 optimum, 236–238
180–182 abrasive and friction and physical properties
waxes, 182–183 methods, 293–295, 552, friction coefficients, 196
Lipoxygenase, 92, 577, 578 569, 574 grain density, 195
and respiration of grain, 303 Oil, rice bran, 481 (see also Fat, nutritional benefits, 99, 332,
during storage, 312–313 Lipids) 363–367, 374
and thermal expansion, 195– blends, 590 origin of process, 329–331
196 bran fines in, 583 process steps, 338–347
variability in, 214, 215, 230 composition, 585 of rice bran, 575, 579
Moisture gradients extraction, 574–575, 583–585 systems, 333–336, 348–359
and fissuring, 194 health benefits, 92, 175, 481, Particle size, 510–512, 516–517,
during soaking of paddy, 343– 581, 587–590 532, 562
344 cholesterol reduction, Pasting characteristics
Molecular marker technology, studies, 600–604 of rice cultivars, measurement,
33, 44–45 pigments in, 583 431–435
Moths, in stored grain, 279– processing, 582, 585–587 of rice starch, 117, 447–448,
280 effects, 175–176, 181–182 507–508
Mutants production, 582 and storage, 276, 433, 509
amylose extender, 130–131 refined, 587 Peck, 31
and protein digestibility, 155 shelf life, 590 Pedigree breeding, 36, 38
in rice and other cereals, 148– stabilization of, 582, 584 Pentosans, 96
149 uses of, 590–591 Pericarp, 79–80
waxy gene, 120, 130 waxes from, 582–583 Pernicious anemia, 407
Mutation breeding, 34 yield, 575 Pests (see Birds; Insects;
Oryza Microorganisms; Rodents)
Narrow brown leaf spot, 32, 41 glaberrima, 1 Pet foods, 480, 489 (see also
Near-infrared sativa, 1, 77 Animal feed)
for end-use quality analysis, Oryzaephilus surinamensis, 279 Phenolic acids, 606
453–455 Oryzanol Phosphine, 280
to measure components, 72 composition, 173–175 Phospholipids
Neural tube defects, 405, 406– extraction of, 175–176, 587, in nonstarch lipids, 171, 173
407, 408 588–589 in purified starch, 167
Neutral lipids, 165, 166, 167, health benefits of, 91–92, 175, in rice bran oil, 585
168 480, 581, 588, 603, 604 in rice starch lipids, 165, 166
Niacin Oryzenin, 311 Phosphorus, 570
fortification with, 405, 408, Osborne’s classification of Phytates, 93
411 proteins, 144–148 Phytin (phytic acid), 575–576,
in rice bran, 574 Oxidative rancidity, 372 580
Nitrogen, effect on grain Phytoglycogen, 131, 132
composition, 96, 97 Paddy (see also Rough rice) Pigmented rice, 154
Nodes, on plant, 27 cleaning, 284–289 Plant
Nonstarch lipids grain volume, change with breeding (see Breeding)
in brown rice, 96 hydration, 339–340 cultural management of, 31–32
as contaminants in starch husking, 291–292 environmental factors
lipids, 165, 167 structure, 77 affecting, 32–33
distribution of, in grain, 91, Panicle growth, 27–30
168, 170 development, 28, 29 harvest, 30–31
extraction, 164 grain location on, effects, 98–99 planting, 25–26
fatty-acid composition, 170 Panicle blight, 32, 43 Plodia interpunctella, 280
glycolipids of, 171 Parboiled rice, 476 Polished rice, 495, 569, 570 (see
phospholipids of, 171 angle of repose, 198, 371 also White rice)
from protein bodies, 173 changes during processing, Polishing, 554, 561 (see also
Nonwaxy rice, lipids in, 165, 359–367, 596 Milling, removal of bran)
166, 167, 168, 171, 184 physicochemical properties, Population, effects, 12, 20, 22
North American Free Trade 376–380 Porosity, of rough and brown
Agreement, 12 products, 329, 392–394 rice, 195
Nutraceutical, oryzanol as, qualities of, 367–375 Potassium, 551
175 starch in, 360, 380–384 Preamylopectin, 132
Nutritional properties tests of, 386–392 Pre-drying period, 301–307
of milled and brown rice, 154, trade in, 334–338 Pressure-parboiling, 355, 367,
595–597 variety, effects of, 384–386 380, 392, 396
of phytonutrients in rice, 605– Parboiling Processing
606 advantages and disadvantages, and glycemic index, 596
of rice bran, 580–581, 597– 375 grain content, effects on, 99–
599 and composition of rice, 99–100 100, 185
of rice bran oil, 587–590, 600– definition, 329 nutritional effects, 154–155,
604 future of, 66–68 596
industry, in various countries, of parboiled rice
Odor, 93, 350, 360 334–338 of protein products, 149–152
Oebalus pugnax, 31 modernization, 332, 333, 352– of rice bran, 149–150, 175
Off-flavor, 92, 452–453 354 of rice bran oil, 582, 585–587
Index / 637
effects, 175–176, 181–182, changes during parboiling differential scanning

602 (steaming), 360 calorimeter, 429
rice types for, 417 composition, 143–149 genetic markers, 455–460
of sake, 541, 553–557 content, 91, 143, 425–426 image analysis, 419, 421, 423
vitamins, changes during pro- and hydration during Instron tester, 437
cessing, 363–364 cooking, 425–426 laboratory mills, 418–419
Production relationship to other light meters, 423
diversification in, 23 parameters, 451–452 optical, 421
of hulls, 611 and cultural practices, 98 near-infrared spectroscopy,
in major rice-growing areas, 5– fractions, 144–148, 572 420–421, 453–455
12 functional properties, 152 randomly amplified
of milled rice, 582 in germ, 573–574 polymorphic DNA
and population, 20, 22 granules, 547 analysis, 458–459
of rice bran oil, 582 in milled rice, 94, 549, 552 Ranghino test, 427
of straw, 611 mutants (see Mutants) Rapid ViscoAnalyser, 428–
U.S., 611 nutritional properties, 153–157 429, 431–433
worldwide, 1, 3–5, 143 products separation techniques, 450
Products and co-products, 18 processing, 149–152 texture analyzer, 438
Asian (see Flour, products, utilization, 152–153 Tensipresser, 438–439
Asian; Flour, products, in rice bran, 572–573, 580 Viscoamylograph, 431
Asian types) varieties of rice, identification Viscoelastograph, 437
beer (see Beer) of, 148 physical properties, 418–423
bran (see Rice bran) Protein bodies, 83, 84, 88, 144, preferences in various
brown rice (see Brown rice) 155 countries, 417
crisped rice, 478, 486, 488 Protein efficiency ratio, 481–482, research needs, 460–461
ethnic foods, 482 580 research programs, 415–416
flour, 479–480 (see also Flour, Protein isolate, 150–151 Quality, of grain
products) Puffed rice, 355–356, 392–393, analyses (see Quality, for end
health benefits of (see Health 556 uses)
benefits) cakes, 72 breeding for, 43–45
high-protein isolates, 151–152, Pullulanase, 113, 131, 132 factors influencing
153, 481–482 Pyricularia grisea, 31–32, 41 cultural management, 31–32
from hull (see Hull) drying, 223, 252, 307–311
and hypoallergenicity, 157, Quality, for end uses environment, 32–33
479, 481, 488, 489 analyses milling, 313–325
individually quick-frozen rice, amylopectin characteristics, storage, 275–282, 311–313
477–478 443–451 wet holding, 5–7
meal, 479–480 amylose characteristics, 443– for milling (see Milling quality)
oil (see Oil, rice bran) 451 sensory, 307, 310–311, 312–
parboiled rice (see Parboiled amylose content, 441–443 313, 320–321
rice) aroma, 452–453 Quality, of protein, 155–156
puffed rice, 478 cooking methods for, 438–441 Quick-cooking rice
processed foods elongation ratio, 435 brown, 476–477
baby foods (see Baby foods) gel consistency, 431 parboiled, 338
breakfast cereals, 417, 484– gelatinization temperature, varieties for, 67, 69–70
485 428–431 white, 476–477
dairy alternatives, 490 grain color, 423
desserts, 18, 488 hydration, 423–427 Rainfed lowland rice, 3
energy bars, 482, 485 kernel dimensions, 421–423 Rancidity
entrées, 487 lipid content, 451–452 in rice bran, 149, 578
market trends, 482–484 milling quality, 418–421 volatiles causing, 452–453
meat and meat analogs, 489– pasting characteristics, 431– Ratoon crop, 31, 32, 43
490 435 Raw rice, nutritional properties,
packaged mixes, 486–487 protein content, 451–452 154
pet foods, 480, 489 solids loss, 427–428 Ready-to-eat rice, 338
snacks, 485–486 texture, cooked, 436–438 Recurrent selection, 36
soups, 487–488 biochemical properties, 441–453 Red rice, 26, 30, 41
quick-cooking rice, 476–477 for brewing, 562–563 Refining, of rice bran oil, 586–
starch (see Starch) for cooking (see Cooking 587, 588–589
from straw (see Straw) quality) Residual breakage, 314, 321
syrup, 480 functional properties, 423–441 Resistance to pests and diseases,
white rice (see White rice) instrumentation and techniques breeding for, 42–43
Prolamin, 144, 147, 547 alkali spreading value, 429– Respiration, of grain
Protein, 481–482 (see also 430 effects of, 302, 303–304
Storage protein) amylograph, 428–429, 431 factors affecting, 303
breeding for, 22 amylose-iodine complex, 441– in storage, 249
in brown rice, 95, 548 442 control of, 273–274
Retrogradation, of starch, 117– sack drying of, 262 extraction of, 164
118 structure, 81–88 formation of, 110–111
Rhizoctonia solani, 31, 41 Seed coat, 82 Starch surface lipids, 163–163,
Rhizopus javanicus, 556 Seed mold, for brewing sake, 165
Riboflavin 542–543 Starchy endosperm, structure,
deficiency, 413 Seedling vigor, breeding for, 41 86–88
fortification with, 405, 410, Semidwarf cultivars, 31, 41 Steaming, of parboiled rice,
411 Sheath blight, 31, 36, 41, 42 345–346, 360–363
in rice bran, 575 Shrinkage, of kernel, 194, 195 Stearines, 591
Rice blast disease, 31, 34, 41 Silica, 88, 620, 622, 626, 627 Sterol lipids, 91–92, 178–180,
Rice bran, 480 Sitophilus spp., 278 585
composition, 91–93, 313, 570 Sitotroga cerealella, 278 Stickiness
antinutritional factors, 575– Soaking and amylose content, 66, 71,
577 changes during, 359–360 387, 417, 443
chemical composition, 570– conditions, for parboiled rice, of cooked rice
575 338–345, 389 parboiling, effects of, 371,
factors affecting, 574–575 before milling, 500 372, 386
fiber concentrates, 597 Soap stock, 173, 174, 585, 586, and protein content, 152
milling, 569–570 587 measurement of, 436, 437, 438
undermilling, 574 Soil, effect on grain compo- of noodles, 517
nutritional properties, 580–581 sition, 98 Storage (see also Holding,
glycemic, 597 Solar energy, for bin drying, 259 before milling)
hypocholesterolemic, 598– Specialty rices, 417, 458 (see effects on
599 also Aromatic rices) functional properties, 311–
lipemic, 597–598 breeding for, 44, 45 312, 508–509
oil from (see Oil, rice bran) Specific gravity, 547 lipid content, 185–186
processing, 149–150, 175 Specific heat, 199 quality factors, 275–277
products made from, 579–580, Spherosomes, 163 sensory properties, 312–313
581, 590–591 Sphingolipids, 176–178 management during, 273–275
stabilization, 499, 577–580, Spikelet, 78–79 packing conditions, and
582, 583, 584 Spina bifida, 406, 408 airflow, 197
Rice stink bug, 31 Stabilization, of rice bran, 499, of parboiled rice, 372, 374
Rice Technical Working Group, 577–580, 582, 583, 584 protection during
55 Stackburned rice (see Yellow from insects, 277–281
Rice water weevil, 43 rice) from microflora, 277
Rodents, and storage, 281–282 Stalk, moisture content, 302 from rodents and birds, 281–
Rough rice (see also Paddy) Starch, 481, 561 282
composition, 88 amount produced, 109 of rough rice, 269–282
dimensions, 193 content in kernel, 92–93, 152, of sake, 545
drying (see Drying, of rough 553 stability of enriched/fortified
rice) damage, 497, 510, 511 rice during, 409, 410, 412
proteins in, 143 digestion, 595–596 structures
effects of parboiling, 360, farm-scale, 270–272, 275
Saccharomyces cerevisiae, 544 380–384, 390–392, 575 large-scale, 272, 274–275
Sake functional properties, 115–118 warehouses, 273
manufacture of and glass transition, 248 of undried rice, 248–249
fermentation of mash, 544– granules, 86, 109–115, 447– Storage modulus, of flours, 507–
545 448, 503 508
finishing, 545 nonfood uses, 481 Storage protein, 86, 88
koji making, 542–543 nonrice, in rice products, 518 Straw
seed mash, 543–545 and proteins, interactions, 152 as absorbents
rice as ingredient, 546 purified, lipids in, 166–168 activated carbons, 613, 615–
composition needed, 548– synthesis, enzymes of, 118– 619
551 133 for cation and anion
grain structure needed, 546– branching, 128–131 removal, 618–619, 621,
547 debranching, 131–133 623–625
physicochemical properties disproportionating, 133 chars, 613, 614–615
needed, 547–549 granule-bound synthase (see for decolorization of sugar,
processing of, 541, 553–557 Granule-bound starch 615
quality, estimation of, 551– synthase) for dye removal, 616–617,
552 other synthases, 125–128, 621, 627
Sand-heating parboiling process 457–458 economic analysis, need for,
(sella), 334, 335 waxy-rice, 481, 488, 597 627, 628
Saponification number, 168 Starch lipids as modified material, 622–
Seed (see also Grain; Kernel) bound and unbound, 165, 168 625
planting, 26 in brown rice, 96 for organic compound
production, 29 composition of, 165–166 removal, 616
Index / 639
surface area, 614, 616, 617, fortification with, 405, 408, 411 superior-processing, 66, 69
620, 623, 625 in parboiled rice, 332, 363–364, waxy (see Waxy rice)
as unmodified material, 620– 366 Vitamins
622 in rice bran, 574 A, 22, 413
burning of, 611–612 Thousand-kernel weight, 547 B-12, 407
value-added products from, Tillers, on plant, 27–28 changes during processing,
612 Tilletia barclayana, 32 363–364
Strength, compressive, 208–209 Tissue culture techniques, 34, deficiency, 407, 413
Stress cracks (see Fissures) 35 in kernel, 91, 93
Structure Tocopherols, 180–182, 570 in milling fractions, 89, 96
of hull, 79 Tocotrienols, 180–182, 588, in rice bran, 570, 574
of paddy rice, 77–79 605 Volatile compounds in milled
of pericarp, 79–80 Toughness, modulus of, 210 rice, 94
of seed, 81–88 Trade Volume
of starch, 109–115 exporting countries, 14–16 expansion, and rice aging,
Subaleurone, 86 importing countries, 16–17 427
Sugars, in parboiled rice, 365 liberalization, effects, 12, 23 and weight, equations, 217
Sun checking, 63 in parboiled rice, 334–338
Sun drying, of rice, 346–347 population effects, 12 Wastewater, 497
Surface lipids, 420–421 world, 12–17 Water
Sushi rice, 417 Transgenic rice, 22–23 absorption
Sweet rice (see Glutinous rice) Tribolium castaneum, 279 during cooking, 423–427
Swelling number, 423 Trypsin inhibitor, 576 effects of drying and storage,
Swelling power, of flours, 508, Turning, of grain in storage, 274 311, 312
511 Types of rice, 4–5, 14, 502–503 by paddy, 305, 307, 340–341
Swelling ratio, 424 ratio, 423
Ultra Rice, 410, 412 in brewing, 542, 548, 554–
Temperature United Nations Population 555, 556
and brewing, 542–543, 544– Division, projections, 22 in cooking tests, 432, 439,
545 Upland rice, 3 440
and drying of grain, 310, 311 Uruguay Round Agreement on and glass transition
with aeration, 254–255 Agriculture, 12 temperature, 116
in dryers, 250, 251, 254– Ustilaginoidea virens, 32 milling, use in, 295
255, 256 movement into kernel, 202,
in storage, 249 Value-added products (see 204
gelatinization, 116 Products and co-products) movement into starch granule,
glass transition, 116 Varieties (see also Cultivars) 113
during grain development, 32, aromatic, 69–70, 72, 475 and rice production, 3, 22
33 for brewing, 546, 548, 557, uptake, 423, 425, 426–427
of postharvest grain, 303, 304 562 apparent, 426
and rice bran stabilization, 577 from California and solids loss, 427–428
for soaking of paddy, 339, breeding programs, 55, 61 true, equation for, 427
342, 344 leading varieties, 61, 63 Waxes, 182–183, 582–583, 587,
and storage naming system, 61 591
of flour, 509 origins, 51, 55 Waxy rice, 474–475
of grain, 273, 277, 281, 313, quality characteristics, 71–73 for Asian products, 502, 518,
323, 324 high-amylose, 69, 517, 532, 519, 522, 527
Tempering 596 flour from, 479–480, 485, 500,
in drying process, 252–253 hybrid, 38–39 519, 527, 528–529, 532
to reduce fissuring, 201, 217, impact scores, 60 gene for, 120–122
310 long-grain, 51, 56–57, 60, 61, for Japanese specialty flours,
Tenebroides mauritanicus, 278– 65, 67, 72, 474 499–500
279 low-amylose, 73, 596 lipids in, 165, 183–184
Texture, cooked medium-grain, 51, 56, 57, 60, starches, freeze-thaw stability,
measurement of, instrumental, 63, 65, 67, 474 481, 488
436–438 proteins, use for identification, use in processed foods, 417
near-infrared analysis, 454 148 Weevils, in storage, 278
of parboiled rice, 331, 372, semidwarf, 51, 60–61 Weight and volume, equations,
375, 381, 386 short-grain, 51, 61, 65, 474 218
and rice type, 331 from southern region Wheat granules, comparisons
Thermal conductivity and diffu- breeding programs, 51, 64, with rice, 110
sivity, 199–200 66 White belly, 343, 348, 355, 423
Thermal processing, nutritional leading varieties, 55–57, 60– White core, 423, 556, 557
effects, 154–155, 596 61 White rice, 474, 541, 549
Thiamin origins, 50–51 quality estimation, 551–552
in brown rice, 96 quality characteristics, 63– Whiteness determination, 421
deficiency, 413 67, 69–70 Whitening, 294–295, 569
Winterization, of oil, 586 Yields, 1–2, 42 and moisture content at

World Trade Organization, 12, field, 237 harvest, 232
17 head-rice, 64, 65, 225, 232–233, and rate of drying, 242, 254–
236–237, 418, 419 255
Yellow (stackburned) rice and chalky kernels, 65 variation in, 223
nutritional changes, 154 factors affecting, 232, 242, milling, and nitrogen, 31
from postharvest holding, 277, 275–276, 305, 307–310
302, 303–304 and harvesting, 64 Zinc deficiency, 98

Champagne, Elaine T - Rice - Chemistry and Technology-Amer Assn of Cereal Chemists (2006) PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Champagne, Elaine T - Rice - Chemistry and Technology-Amer Assn of Cereal Chemists (2006) PDF

Uploaded by

Copyright:

Available Formats

Cover photographs courtesy of U.S. Department of Agriculture, Agricultural

Reference in this publication to a trademark, proprietary product,

Library of Congress Catalog Card Number: 2003115879

© 1972,1985, 2004 by the American Association of Cereal Chemists, Inc.

All rights reserved.

Printed in the United States of America on acid-free paper.

American Association of Cereal Chemists, Inc.

Donald B. Bechtel, U.S. Department of Agriculture–Agricultural Research Service,

Rice, as one of the two major food cereals, provides—together with a

1. Production and Utilization of Rice. NATHAN W. CHILDS, 1

2. The Rice Plant: Growth, Development, and Genetic Improvement.

3. Rice Varieties. K. A. K. MOLDENHAUER, J. H. GIBBONS,

5. Starch. MELISSA FITZGERALD, 109

6. Rice Proteins. FREDERICK F. SHIH, 143

7. Rice Lipids. J. SAMUEL GODBER and BIENVENIDO O. JULIANO,

9. Rough-Rice Drying—Moisture Adsorption and Desorption. OTTO R.

11. Rice Milling. NORRIS BOND, 283

12. Impact of Drying, Storage, and Milling on Rice Quality and

13. Parboiling of Rice. KSHIROD R. BHATTACHARYA, 329

14. Enrichment and Fortification of Rice. DIANE WRIGHT HOFFPAUER,

15. Rice End-Use Quality Analysis. C. J. BERGMAN, K. R.

16. Value-Added Rice Products. HEIDI C. WILKINSON

17. Preparation and Applications of Rice Flour. AN-I YEH, 495

18. Rice in Brewing. KIYOSHI YOSHIZAWA and YUTAKA OGAWA, 541

19. Rice Bran and Oil. FRANK T. ORTHOEFER and JENNIFER

20. Nutritional Properties of Rice and Rice Bran. WALLACE

21. Utilization of Rice Hull and Rice Straw as Adsorbents. WAYNE E.

PRODUCTION AND UTILIZATION OF RICE

RICE COMPARED WITH OTHER CEREAL GRAINS

The ecosystems within which rice is grown are characterized by elevation,

RAINFED LOWLAND RICE

WORLD RICE PRODUCTION

PRODUCTION IN MAJOR RICE-GROWING REGIONS

China and Northeast Asia

Europe and the Middle East

third. Uzbekistan is the second-largest rice-producing country in the region;

WORLD RICE TRADE

GLOBAL RICE CONSUMPTION

Regional and Country Comparisons

the second-largest rice-consuming continent, accounting for almost 4% of global

RICE IN THE TWENTY-FIRST CENTURY

POPULATION AND PRODUCTION

food labeling regulations. In addition, the EU is finalizing its traceability and

THE RICE PLANT: GROWTH, DEVELOPMENT,

Anna Myers McClung

GROWTH AND DEVELOPMENT OF THE RICE PLANT

Planting, Germination, and Emergence

Vegetative Growth Stage

FACTORS INFLUENCING GRAIN QUALITY

GENETIC IMPROVEMENT OF RICE

U.S. Breeding History

Fundamental to the success of rice-improvement programs is access to germ

Population Structure and Breeding Line Development

The first step in initiating a breeding program is to generate plant populations

Breeding Goals and Selection Methods

BREEDING GOALS FOR AGRONOMIC TRAITS

BREEDING GOALS FOR GRAIN QUALITY TRAITS

still very expensive relative to traditional phenotypic screening methods. In addi-

volume markets include the Southern Long-Grain, Southern Medium-Grain, and

VARIETY ORIGIN AND CLASSIFICATION

Fig. 1. Southern varieties, 1929–1984.

Fig. 2. Southern varieties, 1988–2001.

The short-grain varieties, predominantly Caloro and Colusa, occupied essentially

Adenosine 5-Diphosphatase Glucose Pyrophosphorylase

Fig. 4. Starch synthesis pathway. Adenosine 5-diphosphatase (ADP) glucose pyrophosphorylase

Debranching enzymes catalyze the removal of branches by hydrolyzing a-(16)