Indigenous Fermented Foods of Southeast Asia - J. David Owens

Indigenous
Fermented Foods
of
Southeast Asia
Edited by
J. David Owens
Indigenous
Fermented Foods
of
Southeast Asia
FERMENTED FOODS AND BEVERAGES SERIES
Series Editors
M.J.R. Nout and Prabir K. Sarkar
Indigenous Fermented Foods of Southeast Asia (2014)

Editor: J. David Owens
Cocoa and Coffee Fermentations (2014)
Editors: Rosane F. Schwan and Graham H. Fleet
Handbook of Indigenous Foods Involving Alkaline Fermentation (2014)
Editors: Prabir K. Sarkar and M.J.R. Nout
Solid State Fermentation for Foods and Beverages (2013)
Editors: Jian Chen and Yang Zhu
Valorization of Food Processing By-Products (2013)
Editor: M. Chandrasekaran
Indigenous
Fermented Foods
of
Southeast Asia
Edited by
J. David Owens
Boca Raton London New York
CRC Press is an imprint of the

Taylor & Francis Group, an informa business
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2015 by Taylor & Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works

Version Date: 20141029
International Standard Book Number-13: 978-1-4398-4481-6 (eBook - PDF)
This book contains information obtained from authentic and highly regarded sources. Reasonable
efforts have been made to publish reliable data and information, but the author and publisher cannot
assume responsibility for the validity of all materials or the consequences of their use. The authors and
publishers have attempted to trace the copyright holders of all material reproduced in this publication
and apologize to copyright holders if permission to publish in this form has not been obtained. If any
copyright material has not been acknowledged please write and let us know so we may rectify in any
future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information stor-
age or retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copy-
right.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222
Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that pro-
vides licenses and registration for a variety of users. For organizations that have been granted a photo-
copy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Contents
S e r i e s P r e fa c e vii
A c k n o w l e d g e m e n t s ix
I n t r o d u c t i o n xi
E d i t o r xxi
C o n t r i b u t o r s xxiii
C h a p t e r 1 Te m p e and R e l at e d P r o d u c t s 1
J . DAV I D OW E N S , M A RY A S T U T I A N D
K A P T I R A H AY U K U S WA N T O
C h a p t e r 2 S ta r t e r C u lt u r e s 109
L I L I S N U R A I D A A N D WA R AW U T K R U S O N G
C h a p t e r 3 S w e e t, S o u r , A l c o h o l i c S o l i d S u b s t r at e
F u n g a l F e r m e n tat i o n s 137
L I L I S N U R A I DA A N D J . DAV I D OW E N S
C h a p t e r 4 A l c o h o l i c B e v e r a g e s 157
N G O T H I P H U O N G D U N G , WA R AW U T K RU S O N G
A N D K A P T I R A H AY U K U S WA N T O
C h a p t e r 5 L a c t i c V e g e ta b l e and F r u i t F e r m e n tat i o n s 185

L I L I S N U R A I DA , J . DAV I D OW E N S , J U N A I DA H
A B U B A K A R A N D K A P T I R A H AY U K U S WA N T O
v
vi C o n t en t s
C h a p t e r 6 L a c t i c F e r m e n t e d R i c e N o o d l e s 211
R E N U P I N T H O N G A N D J . DAV I D OW E N S
C h a p t e r 7 L a c t i c F e r m e n tat i o n s o f F i s h a n d
F i s h e r y P r o d u c t s 257
L E O N A R DA S . M E N D O Z A
C h a p t e r 8 L a c t i c M e at F e r m e n tat i o n 313
WO N N O P V I S E S S A N G UA N , V E T H AC H A I
P L E N G V I D H YA , N I PA C H O K E S A J J AWAT E E
A N D J U N A I DA H A B U B A K A R
C h a p t e r 9 S oya S au c e 359
S A R D J O N O AT M O KO
C h a p t e r 10 B a c i llu s F e r m e n tat i o n s 373

J. DAV I D OW ENS A N D K A P T I R A H AY U K US WA N T O
Series Preface
Natural fermentation precedes human history, and since ancient

times humans have been controlling the fermentation process.
Fermentation, the anaerobic way of life, has attained a wider mean-
ing in biotransformations resulting in a wide variety of fermented
foods and beverages.
Fermented products made with uncontrolled natural fermentations
or with defined starter cultures, achieve their characteristic flavour,
taste, consistency, and nutritional properties through the combined
effects of microbial assimilation and metabolite production, as well as
from enzyme activities derived from food ingredients.
Fermented foods and beverages span a wide diverse range of starchy
root crops, cereals, pulses, vegetables, nuts and fruits, as well as ani-
mal products such as meat, fish, seafood, and dairy.
The science of chemical, microbiological, and technological factors
and changes associated with manufacture, quality, and safety are pro-
gressing and are aimed at achieving higher levels of control of quality,
safety, and profitability of food manufacture.
Both producers and consumers benefit from scientific, technologi-
cal, and consumer-oriented research. Small-scale production needs to
be better controlled and safeguarded. Traditional products need to be
characterised and described to establish, maintain, and protect their
authenticity. Medium- and large-scale food fermentation requires
vii
viii Serie s P refac e
selected, tailor-made, or improved processes that provide sustain-

able solutions for the future conservation of energy and water, and
responsible utilisation of resources and disposal of by-products in the
environment.
The scope of the CRC book series Fermented Foods and Beverages
includes (i) Globally known foods and beverages of plant and animal
origin (such as dairy, meat, fish, vegetables, cereals, root crops, soy-
beans, legumes, pickles, cocoa and coffee, wines, beers, spirits, starter
cultures, and probiotic cultures); their manufacture, chemical and
microbiological composition, processing, compositional and func-
tional modifications taking place as a result of microbial and enzymic
effects; their safety, legislation, development of novel products, and
opportunities for industrialisation. (ii) Indigenous commodities from
Africa, Asia (South, East, and South-East), Europe, Latin America,
and Middle East; their traditional and industrialised processes and
their contribution to livelihood. (iii) Several aspects of general interest
such as valorisation of food-processing by-products, biotechnology,
engineering of solid-state processes, modern chemical and biological
analytical approaches (genomics, transcriptomics, metabolomics, and
other -omics), safety, health, and consumer perception.
The fifth book, born in the series, is titled Indigenous Fermented
Foods of Southeast Asia. This treatise, edited by Dr. J. David Owens,
deals with the indigenous fermented foods of Thailand, Vietnam,
Indonesia, Malaysia, and the Philippines. The region is known for
its large diversity of fermented foods and the editor has made a great
effort to represent this diversity at the microbiological and ingredient
levels. Thanks to his network of scientists from the Southeast Asian
countries he was able to bring together much of the current knowl-
edge and state of the art. We are convinced that this compendium
will serve as an essential reference for all those scientists as well as
consumers who are looking for the latest knowledge and inspiration.
Acknowledgements
I would like to express my appreciation to all the authors for the work
they have put into their contributions and to Ellen Owens for doing
a final proofreading and for making many helpful suggestions. I also
thank the Department of Food and Nutritional Sciences, University
of Reading for providing me with office space and access to comput-
ing and library facilities.
ix
Introduction
J. D av i d O w e n s
Food fermentations are noted for the creation of a multiplicity of
aromas, flavours and textures from a single starting material and
Southeast Asian fermented foods are no exception in creating a diver-
sity of products from soya beans, rice, cassava as well as from various
waste products of the tofu, peanut oil, tapioca and coconut industries
(Table I.1). In addition, Asia, including Southeast Asia, is noted for
its much wider utilisation of fungi in food fermentations than is the
case in Western countries.
Despite the long history of food fermentations in Southeast Asia,
they have received relatively little attention from the indigenous sci-
entific establishment and, for many products, little has been published
over the past 30 years. Even where research has occurred, in some
countries, there seems to be a predilection to report findings at con-
ferences and in reports that are not widely disseminated rather than
as publications in peer-reviewed international scientific journals. This
hampers research progress and does not provide encouragement to oth-
ers to undertake research in the area. Consequently, many of the foods
remain as artisanal products produced by small-scale backyard pro-
ducers, in contrast to the situation in Japan or, increasingly, in China.
While, in some cases, it is deliberate government policy to support
small producers because of the employment they provide, this does not
xi
x ii In t r o d u c ti o n
preclude upgrading the technology to produce a safer and more consis-

tent product, which would also help to ensure the continued existence
of the products and their producers. As Kiuchi and Watanabe (2004)
noted, ‘Modern technology has transformed natto (Bacillus-fermented
soya beans) from a locally distributed seasonal product of uncertain
quality and safety to a consistent, nationally distributed, high-quality
product that can be safely enjoyed year-round’. There are no reasons
why the production of many Southeast Asian indigenous fermented
foods should not be put on to a similarly sound footing.
It is unfortunate that little research has been carried out on indig-
enous Southeast Asian fermented foods, as many of the fermentations
offer fascinating ecosystems that are ripe for investigation by newer
chemical and molecular biological techniques. Additionally, research-
ers with direct access to indigenous producers have opportunities not
easily available to researchers in other parts of the world. Also, not
to be overlooked is the possible contribution to the success of local
manufacturers and national economies. Seen in this light, indigenous
food fermentations may offer more opportunities than following the
latest fashion, be it prebiotics, probiotics or using molecular biological
techniques but without addressing significant questions.
Partly because little work has been done on many of the indigenous
fermented foods since the 1970s and, in particular, since the 1977
Bangkok conference (Steinkraus 1983, 1996) and partly because there
are relatively few researchers active in fermented food research in
the region and who are available to write reviews, this book does not
offer as comprehensive a cover of products as provided by Steinkraus’
books. Nevertheless, it includes chapters dealing with examples of
all the major categories of fermented foods (Table I.1). The produc-
tion, microbiology, biochemistry, nutritional value and dietary roles
of a wide variety of indigenous fermented foods of Southeast Asia are
described. The emphasis is on the microbiological and biochemical
processes in the fermentations and on the factors that influence the
development of the characteristic microfloras and chemical changes
induced. The classification of products is based on their microbial
ecology (i.e. the predominant microbes involved; Table I.1). The ratio-
nale for this being that traditional fermentations represent solutions
In t r o d u c ti o n x iii
to the problem of how to obtain a desired microbial flora and product

outcome under non-aseptic conditions and in the presence of an ini-
tially highly diverse microflora. Understanding how this is achieved is
essential in establishing reliable and safe processes.
One of the aims of a review is to detect deficiencies in knowledge or
understanding in order to aid researchers in identifying areas for future
research. The contributions here have all attempted to address a series
of basic questions, including the following: (i) What are the dominant/
desired microbes and what factors in the processing and environment
select for them? (ii) What other microbes are commonly present? (iii)
What compounds are utilised as major carbon and energy sources and,
in particular, what are the sources of fermentable carbohydrates? Since
rice is often an ingredient, it has frequently been assumed that rice
serves as the main source of fermentable carbohydrates but Paludan-
Müller et al. (1999) showed that in Thai low-salt fish fermentations
garlic fructans are a major source of fermentable carbohydrate. (iv)
What are the main biochemical activities and chemical changes in the
fermentation? (v) What is the true yield of product per kg of initial raw
materials? Yield has a large impact on the economics of a process and
it is important to understand what causes losses of materials and how
these may be minimised. (vi) What possible hazards may be associated
with a product and how may they be minimised or eliminated?
In the cases of many indigenous Southeast Asian fermented foods,
even these quite basic questions have not been answered unambigu-
ously and there are, therefore, many opportunities to undertake good
research in answering them. The aim in every case should be to have
a sufficiently good understanding of a process to be able to ensure the
production of a consistently high-quality and safe product. If this is
not done, the danger is that many of these traditional products will be
lost and/or displaced by imported variants. If this book plays any role
in encouraging such research activities and in promoting the develop-
ment of indigenous Southeast Asian fermented foods, then I will be
happy that it has served its purpose.
xiv
Table I.1 Some Indigenous Fermented Foods Produced in Southeast Asia

CATEGORY AND SE ASIAN
NAME OF APPEARANCE AND MODE OF COUNTRIES
PRODUCT MAIN METABOLIC ACTIVITIES SUBSTRATE CONSUMPTION WHERE MADE REFERENCE
1. AEROBIC MOULD FERMENTATIONS
Tempe Mould growth (Rhizopus spp.) Cooked, dehulled soya beans Cotyledons bound into a solid cake Indonesia Chapter 1
Fried; used in soups, curries
Oncom merah Mould growth (Neurospora intermedia) Waste materials from As tempe but pink Indonesia Chapter 1
(red oncom) production of soya bean
curd, tapioca
Oncom hitam Mould growth (Rhizopus spp.) Wastes from production of As tempe but grey-black Indonesia Chapter 1
(black oncom) peanut oil, tapioca, coconut
milk
In t r o d u c ti o n
Dage Mould growth (Mucor spp.) Wastes from production of As tempe but grey Indonesia Chapter 1
peanut oil, coconut milk,
cassava starch
Koji Mould growth (Aspergillus oryzae); Whole soya beans or defatted Beans/grains covered with fungal All Fukishima
production of amylase and protoases soya bean meal + wheat mycelium. Used to make soya sauce (2004)
grains and other products
Red rice Mould growth (Monascus purpureus); Rice Red rice grains. Food colouring agent ? Steinkraus
(ang kak) production of red pigment (1996)
Mould-ripened Surface growth of mould(s); Soya bean curd, salt, rice Cubes covered with fungal mycelium Thailand Han et al.
soya bean protein → peptides, amino acids wine, sugar, spices in brine–ethanol–wine–sugar–spices (2001)
curd (sufu) solution. Relish
2. STARTER CULTURES
Ragi tape Growth of moulds and yeasts Rice flour, spices, coconut Dried balls or discs containing moulds Indonesia, Chapter 2
water or sugar cane juice and yeasts. Starter for tape, rice other
wines and so on countries
Loog paeng Growth of moulds and yeasts Rice flour + herbs Dried balls or discs containing moulds Thailand Chapter 2
and yeasts. Starter for tape, rice
wines and so on
Ragi tempe Mould growth Cassava solid waste or rice Dry white-grey powder. Starter for Indonesia Chapter 2
(laru) flour tempe
Ragi tempe Mould growth Cooked soya beans inter- Dried leaves. Starter for tempe Indonesia Chapter 2
(usar) leaved with hibiscus leaves
3. SWEET, SOUR, ALCOHOLIC SOLID-SUBSTRATE FUNGAL FERMENTATIONS

Tape ketan Starch → sugars, ethanol, lactic acid Cooked glutinous rice Sweet, slightly sour, slightly alcoholic All Chapter 3
(Indonesian) cooked rice dessert
Tape singkong Starch → sugars, ethanol, lactic acid Peeled, steamed cassava Sweet, slightly sour, slightly alcoholic, Indonesia Chapter 3
(tape ketela) roots cassava roots with whitish fungal

growth. Dessert
Brem Starch → sugars Tape ketan liquid Sweet, solid rectangular or circular Indonesia Steinkraus
cakes (1996)
4. ALCOHOLIC DRINKS
Rice wine Starch → sugars → ethanol Rice Alcoholic drink, usually distilled or Vietnam Chapter 4
fortified
Satoh Starch → sugars → ethanol Rice Non-distilled alcoholic drink Thailand Chapter 4
Cui Starch → sugars → ethanol Molasses Distilled alcoholic drink Indonesia Chapter 4
xv
continued
xvi
Table I.1 (continued) Some Indigenous Fermented Foods Produced in Southeast Asia
Brem bali Starch → sugars → ethanol Tape ketan Sweet, alcoholic beverage Indonesia
Various Starch → sugars → ethanol Alcoholic drinks All
5. LACTIC ACID BACTERIAL VEGETABLE AND FRUIT FERMENTATIONS

Various Sucrose/glucose/fructose → lactate, Various vegetables Fermented vegetables All Steinkraus
acetate, ethanol, CO2 (1996)
Budu pakis Sucrose/glucose/fructose → lactate, Fern fronds, salt Sour, slightly salty fern fronds, Brunei Chapter 5
acetate, ethanol, CO2 consumed as a side dish
Growol Sucrose/glucose/fructose → lactate, Cassava root Sour dough. Alternative to rice Indonesia Chapter 5
acetate, ethanol, CO2
Tempoyak Sucrose/glucose/fructose → lactate, Durian flesh Fried with spices. Condiment or Indonesia, Chapter 5
acetate, ethanol, CO2 appetiser Malaysia
6. LACTIC ACID BACTERIAL CEREAL FERMENTATIONS

Kanomjeen Starch → sugars → lactate, acetate, Rice Fermented rice noodles eaten with Thailand Chapter 6
ethanol soups and curries
7. LACTIC ACID BACTERIAL FISH AND SEA FOOD FERMENTATIONS

Burong isda Starch → sugars → lactate, acetate, Fish, cooked glutinous rice, Cooked and consumed with rice and Philippines Chapter 7
ethanol salt vegetables
Tinapayan Starch → sugars → lactate, acetate, Dried, salted mudfish fillets, Dry, fried flakes, consumed as a meat Philippines Chapter 7
ethanol fermented rice, spices substitute
Balao balao Starch → sugars → lactate, acetate, Fresh water shrimps, cooked Sautéed in oil and spices and served Philippines Chapter 7
ethanol rice, salt as a condiment
Pla-som Garlic fructans → fructose; Minced fish, ground boiled Consumed raw or cooked on its own or Thailand Paludan-
starch → sugars → lactate, acetate, rice, garlic, salt as part of a main meal Müller et al.
ethanol, CO2 (1999)
8. LACTIC ACID BACTERIAL MEAT FERMENTATIONS

Nham Starch/garlic Pork, garlic, cooked rice, salt, Pork sausage, eaten raw or cooked on Thailand Chapter 7
oligosaccharides → sugars → lactic acid spices, nitrite its own or as part of a main meal
Belutak Sugar → lactic acid Cow or buffalo small Slightly acidic, salty and chewy Brunei Chapter 7
intestines, meat, salt, sugar fermented sausage. Fried with
chillies and onions, with rice
9. LACTIC ACID BACTERIAL DAIRY FERMENTATIONS

Dadih Lactose → lactate, ethanol, CO2 Unheated buffalo milk Yoghurt-like Indonesia Akuzawa et al.
(2011)
Tairu Lactose → lactate Heated cow milk Yoghurt-like drink, and used in cooking Malaysia Steinkraus
(1996)
10. MIXED LACTIC ACID BACTERIAL AND YEAST FERMENTATIONS

Soya sauce Proteins → amino acids; Koji (soya bean ± wheat) Pale brown liquid with umami and All Fukishima
sugars → lactate, acetate, ethanol, CO2 salty taste. Flavouring agent (2004)
Kecap Proteins → amino acids; Koji, palm sugar Dark brown, sweet liquid. Flavouring Indonesia Chapter 9
sugars → lactate, acetate, ethanol, CO2 agent
Tauco Proteins → amino acids; Cooked, dehulled soya beans, Brown paste used in soups and side Indonesia Steinkraus
sugars → lactate, acetate, ethanol, CO2 palm sugar, salt dishes (1996)
continued
x vii
Table I.1 (continued) Some Indigenous Fermented Foods Produced in Southeast Asia
x viii

11. BACILLUS FERMENTATIONS

Thua nua Protein → amino acids → cells, NH4, CO2 Soya bean Brown paste or dried disks; flavouring Thailand Chapter 8
agent
Cabuk Protein → amino acids → cells, NH4, CO2 Sesame seed press cake Black-brown, slightly sticky balls. Indonesia Chapter 8
Flavouring agent or snack
Semayu Protein → amino acids → cells, NH4, CO2 Coconut residue after Black-brown, slightly sticky balls. Indonesia Steinkraus
extraction of milk Flavouring agent or snack (1996)
12. ACETIC ACID BACTERIAL FERMENTATIONS

Vinegar Ethanol → acetic acid Rice wine Sour liquid. Condiment Probably all Steinkraus
(1996)
Nata de coco Sugars → cellulose Coconut water Translucent jelly-like cubes in sugar Philippines, Steinkraus
solution. Dessert Thailand (1996)

Nata de pinea Sugars → cellulose Dilute mashed pineapple Translucent jelly-like cubes in sugar Philippines, Steinkraus
solution. Dessert Thailand (1996)
In t r o d u c ti o n xix
References
Akuzawa, R., T. Miura and I.S. Surono. 2011. Asian fermented milks. In
Encyclopedia of Dairy Sciences, Volume 2, second ed., edited by J.W.
Fuquay, P.F. Fox and P.L.H. McSweeney. Amsterdam: Elsevier and
Academic Press.
Fukishima, D. 2004. Industrialization of fermented soy sauce production cen-
tering around Japanese shoyu. In Industrialization of Indigenous Fermented
Foods, second ed., edited by K.H. Steinkraus, pp. 1–98. New York: Marcel
Dekker.
Han, B.-Z., F.M. Rombouts and M.J.R. Nout. 2001. A Chinese fermented
soybean food. International Journal of Food Microbiology 65:1–10.
Kiuchi, K. and S. Watanabe. 2004. Industrialization of Japanese natto. In
Industrialization of Indigenous Fermented Foods, second ed., edited by K.H.
Steinkraus, pp. 193–245. New York: Marcel Dekker.
Paludan-Müller, C., H.H. Huss and L. Gram. 1999. Characterization of lactic
acid bacteria isolated from a Thai low-salt fermented fish product and the
role of garlic as a substrate for fermentation. International Journal of Food
Microbiology 46:219–229.
Steinkraus, K.H. ed. 1983 and 1996. Handbook of Indigenous Fermented Foods,
first and second ed. New York: Marcel Dekker.
Editor
J. David Owens graduated in microbiology and chemistry from

the University of Bristol, UK in 1962 and earned his PhD at the
University of Reading, UK, in 1966 for research on soil and herb-
age coryneform bacteria. After postdoctoral research at the Johns
Hopkins University, Baltimore, Maryland on the ecology of bacteria
in an estuarine bay in Jamaica, he took up a lectureship at the West
of Scotland Agricultural College and research on the biological treat-
ment of farm wastes. In 1973 he moved to the Science University of
Malaysia, Penang, Malaysia and worked on marine pollution prob-
lems and various aspects of food microbiology and local fermented
foods. In 1976 he joined the University of Sydney, Australia and ini-
tiated research on the evolutionary ecology of methylamine-utilis-
ing bacteria. After a year at the Australian National University he
returned to Reading in 1980. At Reading he expanded his interests
in indigenous fermented foods and the physiology of lactic acid bac-
teria and conducted research on yoghurt, Philippine fermented fish/
rice mixtures, Mexican fermented maize dough, Indonesian tempe,
Nepalese and African Bacillus fermentations, Papua New Guinean
fermented taro/coconut gruel, Tanzanian cassava fungal fermenta-
tion and Zambian fermented maize beverage. He also investigated
the mechanisms of conductance changes in microbial cultures and
xxi
x x ii Ed it o r
invented the Indirect Method for the conductimetric assay of micro-

bial populations. Although formally retired since 2005, he contin-
ues to do some teaching at the Department of Food and Nutritional
Sciences, University of Reading.
Contributors
Mary Astuti Nipa Chokesajjawatee

Faculty of Agricultural Food Biotechnology Research
Technology Unit
Gadjah Mada University National Center for Genetic
Yogyakarta, Indonesia Engineering and Biotechnology
Klong Luang, Pathum Thani,
Sardjono Atmoko Thailand
Department of Food and
Agricultural Product Ngo Thi Phuong Dung
Technology Biotechnology Research and
Gadjah Mada University Development Institute
Yogyakarta, Indonesia Can Tho University
Can Tho City, Vietnam
Junaidah Abu Bakar
Universiti Brunei Warawut Krusong
Darussalam (Formerly) King Mongkut’s Institute of
Bandar Seri Begawan Technology Ladkrabang
Brunei Bangkok, Thailand
x x iii
x xiv C o n t ribu t o rs
Kapti Rahayu Kuswanto Renu Pinthong

Faculty of Agricultural Department of Food
Technology Science and Technology
Gadjah Mada University (Formerly)
Yogyakarta, Indonesia Chiang Mai University
Chiang Mai, Thailand
Leonarda S. Mendoza
Institute of Fish Processing Vethachai Plengvidhya
Technology (Formerly) Food Biotechnology
University of the Philippines in Research Unit
the Visayas National Center for
Miag-ao, Iloilio, Philippines Genetic Engineering and
Biotechnology
Lilis Nuraida Klong Luang, Pathum Thani,
Department of Food Science Thailand
and Technology
Bogor Agricultural University Wonnop Visessanguan
Bogor, Indonesia Food Biotechnology
Research Unit
J. David Owens National Center for
Department of Food and Genetic Engineering and
Nutritional Sciences Biotechnology
(Formerly) Klong Luang, Pathum Thani,
University of Reading Thailand
Reading, United Kingdom
1
Tempe and R el ated
P roducts
J . DAV I D O W E N S , M A R Y A S T U T I
A N D K A P T I R A H AY U K U S WA N T O
Contents
1.1 Tempe 3
1.1.1 Description of Product 3
1.1.2 History of Tempe 5
1.1.2.1 Origin of Usar Tempe Inoculum 5
1.1.2.2 Soya Beans in Indonesia 7
1.1.2.3 Javanese and Tempe 8
1.1.3 Places and Scale of Production, How Tempe Is
Consumed and Its Role in the Diet 9
1.1.3.1 Distribution of Tempe Producers 9
1.1.3.2 Consumption of Tempe in Indonesia 10
1.1.4 Traditional and Current Production Methods 13
1.1.4.1 Soaking 16
1.1.4.2 Cooking 17
1.1.4.3 Dehulling 18
1.1.4.4 Washing 18
1.1.4.5 Draining and Cooling 18
1.1.4.6 Inoculation 19
1.1.4.7 Packaging 20
1.1.4.8 Incubation 24
1.1.4.9 Characteristics of Fermenting Tempe 25
1.1.5 Microbiology of Tempe Fermentation 26
1.1.5.1 Characteristics of Tempe Moulds 27
1.1.5.2 Nutritional Requirements 30
1.1.5.3 Responses to Environmental Conditions 36
1.1.6 Characteristics of the Substrate 40
1.1.6.1 Soya Beans 40
1.1.7 Substrate Changes during Processing and
Fermentation 42
1
2 Indigenous Fermented Foods of Southeast Asia
1.1.7.1 Changes during Hydration of Beans 44

1.1.7.2 Natural Lactic Acid Bacterial Fermentation 46
1.1.7.3 Acidification of Soya Beans by Soaking
and/or Cooking in Acid Solutions 51
1.1.7.4 Dehulling 51
1.1.7.5 Cooking 52
1.1.7.6 Overall Losses of Dry Matter Prior to the
Mould Fermentation 53
1.1.7.7 Changes during Mould Fermentation 54
1.1.8 Chemical Changes during Tempe Fermentation 55
1.1.8.1 Changes in Major Chemical Components
during Tempe Fermentation 57
1.1.8.2 Changes in Minor Chemical Components
in Bacteria-Free Tempe 63
1.1.8.3 Changes in Tempe Containing Bacteria 63
1.1.9 Nutritional Value 66
1.1.9.1 Protein Quality 67
1.1.9.2 Lipids 68
1.1.9.3 Carbohydrates 68
1.1.9.4 Minerals 68
1.1.9.5 Antioxidants in Tempe 69
1.1.9.6 Functional Food Attributes 70
1.1.10 Safety Considerations 70
1.1.10.1 Growth of Pathogenic Bacteria in Soak
Water 71
1.1.10.2 Growth of Pathogens and/or Production
of Microbial Toxins in Tempe 71
1.1.10.3 Production of Mycotoxins by Tempe Moulds 73
1.1.10.4 Bongkek Poisoning 74
1.1.11 Future Prospects and Research Needs 76
1.2 Tempe from Other Pulses 77
1.2.1 Lamtoro (Leucaena leucocephala) Tempe 77
1.2.1.1 Preparation of Lamtoro Tempe 78
1.2.2 Velvet Bean Tempe (Tempe Benguk) 78
1.2.2.1 Preparation of Velvet Bean Tempe 78
1.2.3 Sword Bean Tempe (Tempe Koro) 79
1.2.3.1 Preparation of Sword Bean Tempe 79
1.2.4 Pigeon Pea Tempe (Tempe Gude) 80
T em p e a n d Rel at ed P r o d u c t s 3
1.2.4.1 Preparation of Pigeon Pea Tempe 80

1.2.5 Lablab Bean Tempe (Tempe Koro Wedus or
Tempe Kacang Komak) 80
1.2.5.1 Preparation of Lablab Bean Tempe 80
1.2.6 Tofu Waste Tempe (Tempe Gembus) 81
1.2.7 Tofu and Peanut Waste Tempe (Tempe Menjes or
Tempe Enjes) 81
1.3 Indonesian Oncom (Fermented Food Processing
By-Products) 81
1.3.2 Places of Production, How Consumed and
Role in Diet 83
1.3.4 Microbiology 85
1.3.5 Biochemical Changes 86
1.3.8 Industrialisation 89
1.3.9 Future Work and Prospects 90
1.4 Indonesian Dage 90
1.4.2 Places of Production 91
1.4.5 Characteristics of the Substrates and Changes
during Processing/Fermentation 93
References 94
1.1 Tempe
J. David Owens and Mary Astuti
1.1.1 Description of Product
Tempe is a natural product made from soya beans or other pulses that
are dehulled, hydrated, cooked and inoculated with Rhizopus spp.
moulds, without the addition of salt or other ingredients. The mould
Figure 1.1 Tempe, showing cooked soya bean kernels tightly bound together by mould mycelium.
(a) Indonesian tempe; (b) laboratory-made tempe with location of aeration hole delineated by a ring
of black fungal spores. (Courtesy of J.D. Owens.)
growth binds the cooked bean kernels into a solid cake covered by a
matt of white mycelium on its surface and without yellow spots or
evidence of black sporulation. The cake is sufficiently firm, such that it
does not disintegrate when cut with a knife (Figure 1.1). Tempe has a
slightly beany flavour and an aroma characteristic of mould mycelium
and boiled soya beans. Tempe is a traditional product from Indonesia,
especially Java, and is now widespread over the world.
The word tempe, pronounced ‘tempay’ (tĕmpā, U.S. Dictionary
transcription) in Indonesian and, based on the 1996 international
agreement in Bangkok, the English spelling is tempe rather than
tempeh (Anonymous 1996).
Different kinds of raw materials may be used to prepare tempe
and, commonly, the name of the product includes reference to its
raw material. Thus, tempe kedele is soya bean (Glycine max), yellow or
black, tempe and the term tempe alone usually refers to soya bean
tempe. Black soya bean tempe may be referred to as tempe kedele hitam
(Figure 1.2). Among other pulses that may be used to prepare tempe
are velvet bean (Mucuna pruriens; tempe benguk; Handayani 1997),
sword bean (Canavalia gladiata; tempe koro; Figure 1.3), winged bean
(Psophocarpus tetragonolobus; tempe kecipir), pigeon pea (Cajanus cajan;
tempe gude), lablab bean (Lablab purpureus; tempe kacang komak or
tempe koro wedus) and Leucaena leucocephala bean (tempe lamtoro). In
addition, tempe may be made from the waste left over from tofu man-
ufacture, generally called tempe gembus (Subchan and Rukmi 2007)
or a mixture of tofu waste and waste of defatted peanut, tempe menjes.
The most popular and widely consumed is tempe made from soya
bean. Tempe lamtoro, from Leucaena leucocephala bean, is only found
Figure 1.2 Black soya bean tempe (tempe hitam). (Courtesy of M. Astuti.)
in areas with unfertile land, such as in Wonogiri (southeast Central

Java province) and Gunungkidul (southeast Jogyakarta province) dis-
tricts. Tempe koro and tempe benguk can only be found in certain
areas of Central Java, East Java and Yogyakarta province.
1.1.2 History of Tempe
1.1.2.1 Origin of Usar Tempe Inoculum The question of where tempe

inoculum or usar came from was raised by Widagdo in an article
Figure 1.3 White sword bean (Canavalia gladiata) tempe (tempe koro). (Courtesy of M. Astuti.)
about how to make tempe in Majalah Guru Desa, the village teacher
magazine, published in 1915. None of the many usar producers to
whom he asked knew of its origin, but all agreed that it had been
handed down for many generations. The raw materials for usar prepa-
ration are black soya beans and leaves of Hibiscus tiliaceus. The soya
beans are first boiled well, allowed to cool, dehulled and then soaked
overnight in water. Half of the soakwater is removed and is re-used
to boil the next batch of soya beans. The cooked, dehulled kernels are
inoculated with usar from a previous batch, using two to four usar
leaves per kilogram wet weight of soya beans. Next, new clean hibis-
cus leaves are selected. They must not be washed or wiped as this
would damage the fine hairs that are important for growth of the
mould. Banana leaves are then cut into small pieces (5–6 cm2), small
holes are punched in them with a sharp twig and they are then placed
on the hibiscus leaves (hairy undersurface uppermost). A thin layer of
the inoculated soya bean kernels is then spread over the banana leaves
and covered with more hibiscus leaves, their hairy undersurface in
contact with the soya beans. Layers of beans and leaves are built up
in this way until 10 or so layers have been created. The pile of leaves
is then rolled and tied with rice straw. This roll is placed in a gunny
bag or basket and left for 2–3 days at ambient temperature. The usar
is ready when the white fungal mycelium appears on the surface of
the banana leaves and out through the holes. At this stage the roll is
opened up. The hibiscus leaves, which will have a layer of sporulating
mycelium on their hairy under surface, are taken and dried for 2 days
in a well-ventilated place. When black spores appear over the surface
of the leaves the drying is stopped and the usar is ready to be sold in
the market (see Chapter 2, Figure 1.3d).
Almost 100 years after Widagdo had speculated on the origins of
usar, Ogawa et al. (2004) noted that Rhizopus oryzae was commonly
present on fresh leaves of Hibiscus tiliaceus. This led to the suggestion
that the origin of usar and tempe lay in the ‘accidental’ discovery that
cooked soya bean kernels wrapped in hibiscus leaves became a solid
mass that could be utilised in a variety of ways.
Usar exists in the market and some tempe producers use it as a
source of mould. Modern tempe inoculum is a dried powder made by
growing a mixture of Rhizopus oligosporus and R. oryzae on a substrate
of rice or cassava powder.
The use of an inoculum to ‘steer’ a fermentation in the desired

direction can be a very important factor in determining the outcome
of that fermentation. Indonesian tempe is primarily a mould fermen-
tation of non-salted soya beans in contrast to the non-salted soya bean
fermentations found in Japan, Thailand, Korea, China, Myanmar and
Bhutan, where the causative organism is Bacillus subtilis (see Chapter
10). The Bacillus fermented products, such as Japanese natto and Thai
thua nao, are quite different from tempe. They are greyish or brownish
in colour with a sticky texture, a strong flavour and a musty, ammo-
niacal smell. The traditional Bacillus fermentations are not inoculated,
suggesting that one factor promoting a tempe mould fermentation
is inoculation. Undoubtedly, other environmental factors are also
involved in obtaining mould fermentation rather than a Bacillus one.
1.1.2.2 Soya Beans in Indonesia Soya beans originated in north China

and from there have spread around the world. The word for soya
bean in Indonesian is kedelai, from the Javanese kadele. Among old
Javanese manuscripts, the word kadele is first recorded in the Serat Sri
Tanjung manuscript (Prijono 1930), believed to have been written in
the twelfth or thirteenth century. The word for soya bean is also found
in the Serat Centhini manuscript, written in 1814 (Kamajaya 1986).
Soya beans are not only consumed as tempe but also have a role in the
Javanese marriage ceremony, where the groom gives presents of food,
including rice, maize, black soya beans and long beans, as a symbol of
his responsibility for the welfare of the family. Black soya beans also
appear in the dish, nasi udug (rice cooked with coconut milk), and
their use has not been replaced by yellow soya beans. Black soya beans
are always used as the raw material for making usar. Thus, black soya
beans continue to have a great significance in Javanese culture.
Rumphius (1747) described a type of soya bean plant, Phaseolus
niger, which the Javanese called kadele, the Chinese authau, and the
Dutch zwarte boontjes (black beans). His description of the plant is
as follows: ‘The soybean plant resembles a small shrub with pointed
leaves, yellow flowers and seeds in pods. In Java and Bali, soybeans
are widely cultivated and harvested by pulling up the plant with its
pods. To store them, the leaves are first washed, then, 8–10 plants
are tied together and hung up in a bunch. When they are required
for eating, the entire plant is put into boiling water, or the beans are
separated from the pods and boiled. Eating the boiled beans alone is
not frequently done as they are quite hard and have a bitter flavour.
The soybean plant grows in Java and Bali but is seldom found in
Ambon’.
Black soya beans were cultivated by the Javanese and mainly sold to
the Chinese, who ground them into flour from which they manufac-
tured laksa or tautsjian (a type of flat noodle). In addition, soya beans
were processed by heating, removing the black husks and milling the
beans into flour for tofu production. Rumphius (1747) does not make
any mention of the use of soya beans in tempe, possibly because he did
not observe soya bean processing in rural areas. A magazine published
in the 1900s stated that Javanese people made tempe from black soya
beans and Chinese people made tempe from yellow soya beans (M.
Astuti, unpublished data). Black soya beans were cheaper than yellow
soya beans. The Chinese used imported yellow soya beans for mak-
ing tempe and tofu. Tempe made from black soya beans is very rarely
found in the market nowadays and almost all tempe is made from
yellow soya beans. Tempe made from black soya beans is now priced
higher than that made from yellow soya beans.
Before Indonesian independence in 1945, black soya bean was
the dominant type cultivated and studied in Indonesia (Anonymous
1996). Black soya beans were mainly used as raw material for fer-
mented soya sauce and were made into tempe. However, preferences
shifted to yellow soya beans when they began to be imported from
China and America and the black traditional variety has almost dis-
appeared (Agranoff 2001).
1.1.2.3 Javanese and Tempe The earliest written record of the word
tempe, comes from the 1814 Serat Centhini manuscript (Kamajaya
1986). The manuscript includes a description of the journey of Mas
Cebolang when he traveled between Prambanan temple and Pajang,
via Tembayat in the Klaten sub-district of Central Java province. Here,
Cebolang was served a lunch, described in its entirety, that included
a dish of tempe in coconut milk and a tempe sauce made from over-
fermented tempe (cf. section on Indonesian dage below; Gandjar and
Hermana 1972). At that time, tempe made from soya beans was an
ordinary food found only in rural areas but tempe is currently popular,
found everywhere, not only in rural areas but also in the cities. Rather
surprisingly, Raffles (1817 and 1830) makes no mention of tempe.
The hypothesis that tempe originated in Java is supported by the
fact that tempe can be found in every corner of the island, with varia-
tions only in terms of the type of substrate used to manufacture it.
The production and consumption of tempe are integral to the Javanese
lifestyle and the bond between tempe and the Javanese is so strong
that the two have become inseparable. Wherever there are Javanese,
there is sure to be a source of tempe too.
The spread of tempe outside Java began with the migration of the
Javanese to other regions, both within Indonesia as well as abroad.
Within Indonesia, many transmigrants settled in Lampung, Sumatra,
taking with them their tempe technology. Interestingly, although
these Javanese passed on their knowledge of tempe-making to
natives of Lampung, the failure of the industry to develop among
the Lampungese illustrates the special affinity that only the Javanese
seem to have for this food. Trade links between Java and China as
well as with Western nations have existed over several hundreds of
years. The Javanese migrated to places such as Malaysia, Thailand,
Surinam and the Netherlands and took with them their traditions as
well as their knowledge of tempe production. As a result, tempe man-
ufacture can now be found in Malaysia, Thailand, Surinam and the
Netherlands and, more recently, has been taken up in many western
countries, including the United States, Japan, Australia and Europe.
1.1.3 Places and Scale of Production, How Tempe Is

Consumed and Its Role in the Diet
1.1.3.1 Distribution of Tempe Producers Indonesia comprises 33 prov-

inces, 399 districts and 98 municipals, and tempe factories are found
in all of them, with the greatest number being in Java. Java com-
prises six provinces and is home to 241 million people (National
Population and Family Planning Board 2013). Tempe factories exist
from Sabang, Aceh Special District in western Sumatra to the eastern
most parts of Papua. In Indonesia there are more than 100,000 tempe
producers with 10,000 tempe producers in Yogyakarta province alone
(Anonymous 2010).
The amount of soya beans processed into tempe varies between

5 kg and 2 tonnes per day in factories classified as micro and small
enterprises. These manufacturers are protected by law and are highly
promoted by the government because of the high level of employment
they provide. Tempe production in these small factories is mostly
carried out by hand, using simple instruments and machines rather
than complex ones. Producers with capacities of 1 tonne per day
may use some stainless-steel equipment but smaller producers tend
to use cheaper aluminium items. Machine packaging is rarely used.
Tempe production workers and tempe makers are usually members
of the Indonesia Tempe Makers Cooperative Enterprises (Koperasi
Perajin Tempe Indonesia/KOPTI). KOPTI estimated that in 2012 the
tempe industries contributed about USD700 million to the Indonesian
economy.
The increasing demand for tempe, due to economic growth and the
increasing population, has led to increases in tempe production and
greater importation of soya beans to supplement locally grown soya
beans.
1.1.3.2 Consumption of Tempe in Indonesia Generally, tempe is con-

sumed as a source of protein in the Indonesian diet. Historically,
tempe was viewed as a protein source of the lower classes and ani-
mal protein was categorised as high-class food. Although tempe was
considered a home food and was rarely served in large restaurants
or high-class hotels, people of all economic classes and of all ages
consumed tempe and it was a favourite traditional food, served as a
daily side dish in every household, especially in Java island. Currently,
tempe has gained wider acceptance and is increasingly available in
supermarkets and high-class restaurants.
Tempe is rarely consumed fresh but is incorporated into other
dishes. It is suggested that there are more than 100 tempe recipes in
Indonesia, with a wide diversity of tastes and appearances. Generally,
dishes made from tempe can be classified as side dishes, snack foods or
foreign-adapted foods. Tempe can be sliced or cut into cubes as either
the main or as an additional ingredient of side dishes. It can be stir
fried or seasoned and boiled, alone or with vegetables. Tempe dishes
include oseng-oseng tempe (stir-fried tempe), besengek (whole tempe
cooked with spices in thick coconut milk), sambel goreng tempe (sliced
Figure 1.4 Tempe restaurant dish, mendoan (tempe dipped in thin flour batter and fried).
(Courtesy of J.D. Owens.)
tempe cooked with chili and spices in coconut milk), sambel kering
tempe (sliced tempe and peanuts dry-fried and then cooked in oil with
spices, chili, tamarind and coconut sugar), tempe bacem (savory, dark
brown tempe made by boiling with spices and coconut sugar followed
by frying), tempe penyet (fried tempe pressed flat onto the frying pan
surface) among others. Most of these are Javanese cuisine specialties.
Snack foods made from tempe include mendoan (tempe is dipped in
thin flour dough and then fried; Figure 1.4) and keripik tempe (crunchy
tempe crisps; Figure 1.5). There are also foreign-adapted products,
such as tempe nuggets and tempe burgers.
Deep-fat frying, which takes only 3–5 min, is the most popular
cooking method for serving tempe at a meal or as a snack food (Figure
1.6). The temperature for frying should be about 180°C. During frying,
Figure 1.5 Tempe crisps. (Courtesy of J.D. Owens.)

Figure 1.6 Deep-fat-fried tempe. (Courtesy of J.D. Owens.)
the colour changes from white to a golden brown but, if the tempera-
ture is too high, an unpleasant, dark brown colour is produced.
Tempe is generally consumed along with staple foods, such as
rice, corn or cassava. Although tempe has a high protein content,
it is estimated that it supplies less than 20% of total protein in the
Indonesian diet (Table 1.1). Cereals, which are consumed in much
greater quantities, supply around 60% (Central Bureau of Statistics
2012). In Yogyakarta province, the frequency of tempe consumption
is ~50 times a month. This means that, on average, people consume
tempe almost twice a day (M. Astuti, unpublished data). Although
the quality of tempe protein is lower than that of animal protein, mix-
ing rice and tempe in the ratio 7:3 improves the quality of protein
(Astuti 1992).
Table 1.1 Consumption of Protein Food Sources in Indonesia in 2012

FOOD CONSUMPTION (KG WET WT CAPITA−1 Y−1)
Fresh fish and shrimp 13.5
Chicken 9.3
Tempe 7.0
Tofu 7.0
Broiler, local chicken meat 4.0
Beef, buffalo meat 0.36
Source: Central Bureau of Statistics. 2012. Food Balance Sheet. Jakarta: Central Bureau of
Statistics.
In Indonesia, tempe is one of the cheapest sources of protein. Tempe

prices in 2013 (USD) per kg were as follows: tempe, 1.0; chicken eggs,
1.75; and beef, 14.5. On a per kg protein basis it is as follows: tempe,
5.1; fresh water fish, 7; sea water fish, 8; tofu,11; egg, 14; and beef, 97.
1.1.4 Traditional and Current Production Methods
There are many variations in the methods for making tempe in

Indonesia, although they all share the same basic process which may
be used for most types of beans. Only the preparation of tempe gembus
(from tofu waste) and bungkil (residue from pulses, especially peanut,
after extraction of oil) bean tempe are a little different.
The traditional method described by Widagdo (1915) is as fol-
lows: ‘First of all, the whole soya beans are boiled until cooked, then
allowed to cool. When cool, the hulls are removed by treading under-
foot several times (Figure 1.7) until they float easily to the surface
of the water (Figure 1.8). The hulls can then be easily separated and
discarded, more water being added to clean the soya beans. The beans
are then left to soak overnight until the soak water begins to smell bad
(acid smell). At this stage they are boiled until soft, drained and left to
cool on a bamboo or woven leaf mat. When cool they are mixed with
‘usar’. The way this is done is by tearing up the usar into tiny pieces
Figure 1.7 Dehulling cooked soya beans by treading, Bogor 1995. (Courtesy of J.D. Owens.)
Figure 1.8 Separating hulls by flotation, Bogor 1995. (Courtesy of J.D. Owens.)
and placing these on a clay pot which is heated over a fire, without
letting the usar burn. When the leaf fragments are fully dried, they
are agitated until the mould becomes detached. The mould is then
mixed with the cooled soya beans. The beans are wrapped in leaves
and fermented in a gunny sack for 48 hours until they become tempe.’
The traditional method, as described by Widagdo, is still used in
current tempe preparation (Table 1.2, procedure A), but the removal
of the soya bean coat is done differently. In current methods, tempe
producers use a soya bean dehuller to separate the testa and cotyledons
of the soya beans (Figure 1.9). Use of tempe inoculum in the form of
usar is now very limited and producers have changed to using pow-
dered tempe inoculum (Figure 1.10). Perforated plastic bags are now
used to contain the fermenting soya beans and have largely replaced
natural wrapping materials, such as banana or teak leaves. For the
fermentation, only very small-scale tempe producers still use a gunny
sack and the bigger manufacturers use a fermentation rack made from
bamboo, wood or metal (Figure 1.11).
Basic tempe processing involves two stages, preparation and fer-
mentation, and the variations occur only in the preparatory stage
(Saono et al. 1986; Steinkraus 1996; Shurtleff and Aoyagi 1985). The
preparation stage changes the raw, hulled soya beans into cooked,
dehulled kernels/cotyledons that provide a suitable substrate for
growth of the mould. A survey of the methods of tempe preparation
Table 1.2 Some Traditional Methods for Preparing Tempe

PROCEDURE
STEP A B C D E F G
1 Clean Clean Clean Clean Clean Clean Clean
2 Boil Boilb Boil Soaka Soak Soak Dehull
3 Cool Soak Soak Dehull Boil Boil Wash
4 Dehull Dehull Dehull Wash Wash Cool Boil
5 Wash Wash Wash Boil Boil Dehull Drain
6 Soak Boil Drain Drain Cool Wash Cool
7 Boil Drain Inoculate Cool Dehull Soak Inoculate
8 Drain Cool Wrap Inoculate Mash Drain Wrap
9 Cool Inoculatec Incubate Wrap Drain Drain Incubate
10 Inoculate Wrap Incubate Inoculate Cool
11 Wrap Incubate Wrap Inoculate
12 Incubate Incubate Wrap
13 Incubate
Source: Modified from Saono, S., R.R. Hull and B. Dhamcharee. 1986. A Concise Handbook of
Indigenous Fermented Food in the Asia Countries. Jakarta, Indonesia: The Indonesian
Institute of Science.
a If only once, 60–120 min, if twice, first 30 min, second 90–120 min.
b 10 h to overnight.
c 36–48 h at room temperature (26–30°C).
in Indonesia conducted by Saono et al. (1986) revealed seven variant

processes (Table 1.2). It can be seen that there are between 9 and
13 stages in the whole process. However, the procedures described
by Saono omit the cleaning of the beans. Cleaning is important to
remove materials, such as stones, metal and other impurities.
Figure 1.9 Mechanical dehulling machine. (Courtesy of M. Astuti.)

Figure 1.10 Powdered ragi tempe inoculum. (Courtesy of M. Astuti.)
Although details of preparation methods vary from one place

to another and from one producer to another, essentially, they are
all similar. It involves three basic steps: soaking, dehulling and
cooking.
1.1.4.1 Soaking Some producers use a slow method which includes

two periods of soaking. Others use a rapid method which only
includes one soaking. The soaking stage allows full hydration of the
Figure 1.11 Bags of tempe on incubation rack. (Courtesy of M. Astuti.)

Figure 1.12 Soaking soya beans. (Courtesy of M. Astuti.)
soya beans, softens the beans and improves their texture for eating
(Figure 1.12).
1.1.4.2 Cooking Cooking can be done by steaming or by boiling

(Figure 1.13). It serves to eliminate contaminant vegetative bacteria
and fungi from the beans, to soften the beans and to inactivate tryp-
sin inhibitor and other anti-nutritional factors. Cooking in water is
easier than steaming, but loss of water-soluble nutrients is higher with
boiling than with steaming. The temperature of steaming is slightly
higher than cooking in excess water. The soya beans may be boiled
once or twice, depending on which system is preferred. If they are
Figure 1.13 Boiling soya beans in small-scale industry. (Courtesy of M. Astuti.)

boiled twice, the first boil is often short, about 30 min, and the sec-
ond boil then lasts about 60–90 min (Figure 1.13). However, there is
much variation in boiling times and Efriwati et al. (2013) observed a
process in which the first boiling was for 2–3 h and the second was
for ~2 h.
1.1.4.3 Dehulling Dehulling soya beans is now generally performed

using a mechanical dehuller machine as this is somewhat more
hygienic and economical for those who process in excess of 50 kg soya
beans per day (Figure 1.9). Dehulling may be performed with a simple
roller mill, adjusting the distance between rollers to somewhat less
than the size of the soaked soya beans. The pairs of cotyledons are
slightly pressed from each other by rotating cylinders and at the same
time the hulls detached. The hulls can then be separated from the
kernels by floating them off in running water. However, an equipment
that separates all of the skins is not available and so around 10 − 15%
of seed coats of the beans remain in the tempe. Nurrahman et al.
(2011) suggested that tempe is still acceptable with around 15% of
the seed coat with the cotyledons. If the seed coat content was higher
than 20%, the taste of the tempe was not accepted by consumers.
It is also possible to dehull the dry beans (Steinkraus 1996) and
dry dehulling of soya beans, using a mechanical dehuller, is done in
the Tempe Murni factory in Yogyakarta province. However, tempe
produced from dry dehulled soya beans can have a lower quality than
tempe prepared from wet dehulled soya beans (M. Astuti, unpub-
lished data).
When tempe producers use black soya beans as raw material some
of the seed coat is visible on the surface of the tempe as black specks,
but if yellow soya beans are used the seed coat is not easily seen. This
is one reason why consumers, and producers, prefer yellow soya bean
tempe to black soya bean tempe.
1.1.4.4 Washing Washing with clean water removes dirt attached to

the beans and in the soaking water and replaces the acidic soaking
water with clean water so that the beans do not become too sour.
1.1.4.5 Draining and Cooling The cooked kernels are drained to

remove excess water and cooled, to obtain material that is dry on
the surface and sufficiently cool to allow fungal growth. Spreading

the hot kernels in thin layers allows evaporative cooling to achieve
this (Figure 1.14). A suitable temperature for growth of the mould is
30°C, close to ambient temperature in Indonesia. The cooked kernels
are not sterile and may contain bacteria, both contaminants and bac-
terial spores that survive the cooking, but in the environment offered
by the kernels the growth of bacteria is slow and growth of the mould
dominates.
The kernels need to be drained thoroughly since excessive water,
and especially surface water can promote bacterial growth. Usually,
the moisture content of the kernels is around 50–60%. If the moisture
content is higher than 60%, bacterial growth in the free water on the
kernels is liable to affect growth of the mould.
1.1.4.6 Inoculation Cooled beans at approximately room temperature

can be inoculated with starter and homogenised. Some tempe manu-
facturers still use usar. One usar leaf is used for 2–4 kg of cooled soya
beans. The amount of usar used is influenced by the environment.
If the room contains a lot of tempe mould spores, only one piece of
usar is needed per 10 kg of kernels. Most tempe producers use tempe
inoculum in the form of a powder. The powder consists of ground rice
and R. oligosporus and R. oryzae and is mixed with the cooked and
cooled kernels (Figure 1.14). Usually, 1 g of inoculum powder is used
per 1 kg of cooled beans. In the rainy season, the amount of inoculum
Figure 1.14 Draining, cooling and inoculating cooked cotyledons. (Courtesy of M. Astuti.)
powder used is a little higher than in the dry season, possibly because
of the lower ambient temperatures.
1.1.4.7 Packaging It is important to pack the inoculated kernels prop-

erly in order to obtain a final product in which the white mycelium has
developed abundantly and bound the soya bean kernels into a compact
cake. To achieve a good quality tempe, an equilibrium should be main-
tained between keeping the kernels moist and restricting contact of
the beans with air. Small-scale producers pack containers by hand but
automatic packaging machines are now used by large-scale producers
(Figures 1.15 and 1.16). Packaging materials used are as follows.
1.1.4.7.1 Trunk of Banana Trees Banana tree trunk (called debog)

has a hollow center that allows it to be used as a container. The trunk
is cut into desired lengths. Normally, the trunk is cut lengthwise into
strips which are then tied together in pairs and filled with inoculated
kernels (Figure 1.17a). Alternatively, sections of trunk can be packed
directly (Figure 1.17b).
1.1.4.7.2 Banana Leaves Banana leaf used to be the most common

packaging material used by tempe producers. Fibres in banana leaves
allow the leaf to be moderately flexible, such that it can be folded,
but not easily torn when it is wilted. Banana leaf became a favoured
packaging material as it was easily obtained at a low price and was
Figure 1.15 Hand packing of inoculated cotyledons into plastic bags. (Courtesy of M. Astuti.)
Figure 1.16 Machine for packing inoculated cotyledons in plastic tubes. (Courtesy of M. Astuti.)
favoured by consumers, who perceived that it provided a desirable

aroma. The leaves commonly used as packaging material for tempe are
from a variety of banana referred to as klutuk, as its leaves are less stiff
than those of some other varieties and are, therefore, easily folded to
make packets. Banana leaf allows even air diffusion to the fermenting
tempe and, hence, produces a high-quality product (Suprapti 2003).
Banana leaves are cut into pieces and wiped to clean the sur-
face. The inoculated kernels are placed on the under surface of the
leaf, which has a pale green-whitish colour, and are then wrapped
Figure 1.17 (a) Banana stem sections. (b) Inoculated cotyledons packed in banana stem sections.
(Courtesy of M. Astuti.)
Figure 1.18 Tempe packed in banana leaf. (Courtesy of M. Astuti.)
by folding the banana leaf to form a cylindrical, oblong, cuboidal or

pyramidal packet (Figure 1.18). This leaves the attractive shiny green
upper surface of the leaf on the outside of the package. During the
rainy season, some producers pack tempe in banana leaves combined
with paper. The paper helps keep the fermenting kernels warm to pro-
mote the growth of the mould.
Although much tempe is now packed in plastic bags, some con-
sumers prefer tempe in banana leaf and one finds tempe fermented in
banana leaf wrapped in plastic in supermarkets (Figure 1.19).
1.1.4.7.3 Teak (Tectona grandis) Leaves Teak leaves are widely used
to wrap tempe in the areas of Yogyakarta, Central Java and East Java.
Teak leaves are much smaller than banana leaves and they are used
Figure 1.19 Tempe packed in banana leaf and plastic. (Courtesy of J.D. Owens.)
Figure 1.20 Tempe packed in teak leaves. (Courtesy of M. Astuti.)
to prepare small-sized tempe packages, with less variety in shape

than with banana leaves due to the limitations imposed by their size.
Otherwise, the procedure of folding and tying the leaves around the
inoculated kernels with teak leaves is similar to that with banana
leaves, although in the case of teak leaves it is the upper leaf surface
that is in contact with the kernels (Figure 1.20).
1.1.4.7.4 Plastic Bags The use of plastic packaging to wrap tempe

is now widespread and highly developed in Indonesia (Figure 1.21).
Plastic bags have several advantages, including being cheap, readily
Figure 1.21 Inoculated cotyledons packed in plastic bags. (Courtesy of M. Astuti.)

Figure 1.22 Hole piercer for making holes in plastic bags, Bogor 1995. (Courtesy of J.D. Owens.)
available in different shapes, sizes and thicknesses, easily formed,

transparent and devoid of contaminants that may be on plant leaves.
Many modern tempe producers use plastic as their packaging mate-
rial both due to its practical benefits and because brand advertising
and product information can be placed on the bag. Plastic packaging
also offers benefits to consumers since the physical condition of the
product can be seen without opening the packaging, unlike the case
with products packed in leaves. Plastic bags are filled with inoculated
kernels (Figure 1.16) and sealed using a heat sealer.
Plastics used to wrap tempe are polyethylene or polypropylene.
The main concern with wrapping tempe in plastic is the availability
of oxygen to support the growth of the mould during fermentation.
Plastics have a lower permeability to oxygen than leaves and hence
the exchange of oxygen, water vapour and heat with the exterior is
restricted. Therefore, the plastic needs to be perforated to allow gas-
eous exchange between the fermenting mass and the exterior. Initially,
plastic bags were perforated by hand (Figure 1.22) but now ready-
perforated bags are invariably used. The sizes and distribution of the
perforations need to be designed to provide optimal conditions for the
growth of fungi and the production of high-quality tempe. Typically,
holes are 1–2 mm in diameter.
1.1.4.8 Incubation The innoculated and packaged kernels are placed

on racks made of bamboo, wood or stainless steel and incubated at
room temperature (~30°C) for 36–48 h (Figure 1.11).
1.1.4.9 Characteristics of Fermenting Tempe During fermentation the

mould grows profusely and binds the loose soya bean kernels into a
solid cake that may be cut with a knife without any loss of structure.
The pale yellow kernels become completely covered with mould myce-
lium so that the mature tempe has a bright white surface appearance,
much like a brie or camembert cheese (Figure 1.1). After more than
48 h incubation the tempe starts to deteriorate and acquires yellowish
surface spots and an ammoniacal smell (Table 1.3).
1.1.4.9.1 Shelf Life of Tempe Tempe is a perishable product which

only lasts for a few days when stored at room temperature (Kasmidjo
1990). After about 72 h at room temperature the quality of tempe
decreases due to physical, sensorial and chemical changes. Shurtleff
and Aoyagi (1985) noted that tempe kept for 2–3 days beyond the
optimal incubation time of 24–48 h at 24–29°C became dark brown
in colour and developed a smell rather like that of mature camembert
cheese mixed with a faint ammoniacal odour but still retained a good
texture. After 3–5 days, the colour was an intense dark brown, there
was a strong ammoniacal odour and the texture softened.
Refrigerated storage of fresh tempe keeps it only for a short time,
up to about 5 days. Freezing fresh tempe extends the shelf life up to
several months and, hence, allows time for distribution and display.
Table 1.3 Characteristics of Fermenting Tempe

FERMENTATION SURFACE
TIME (H) COLOUR FLAVOUR TEXTURE APPEARANCE
0 Pale yellow Strong beany Loose soya bean Shiny cotyledons
flavour cotyledons
24 Greyish Less beany Cotyledons loosely Partial cover with
white flavour attached to each other mould mycelium
48 White Strong tempe Mould mycelium Entirely covered by
flavour throughout cotyledons white mould
mass, binding mycelium
cotyledons into a firm
cake
72 Yellowish Ammoniacal Mycelium deteriorating Moderately covered
white smell and cotyledons mass with mould
becoming less firm mycelium,
yellowish spots
and moist areas
1.1.5 Microbiology of Tempe Fermentation
The traditional tempe-making process involves two fermentations:

an initial lactic acid bacterial fermentation during the soaking of the
beans and a subsequent mould fermentation when the cooked beans
are bound into a solid cake. The soaking stage involves a relatively
diverse microbial flora and is dealt with later. The mould fermentation
is the essential component of the tempe process and correct designa-
tion of the mould(s) involved is important, not least since some closely
related moulds are agents of human infections or produce mycotoxins.
The mould fermentation is essentially a fungal sold-substrate
fermentation but, in the absence of special procedures, invariably
includes substantial numbers of bacteria and commonly also yeasts.
The bacterial population in tempe commonly reaches 107–1010 cfu g−1
(Efriwati et al. 2013; Mulyowidarso et al. 1990; Nout et al. 1987a;
Samson et al. 1987; Sudarmadji and Markakis 1978). This population,
apart from potentially pathogenic species, has received rather little
study but the predominant bacteria include Bacillus species, lactic acid
bacteria, Enterobacteriaceae and Staphylococcus species (Efriwati et al.
2013; Mulyowidarso et al. 1990; Samson et al. 1987). Such popula-
tions are comparable to levels occurring in many bacterially fermented
foods and are certainly high enough to cause observable chemical
changes. However, these concentrations equate (taking one bacterial
cell as 10−13 g dry matter) to only 0.001 to 1.0 g dry biomass (kg wet
tempe)−1, which is much less than the estimated ~70 g dry fungal bio-
mass (kg fresh tempe)−1 (see later). Hence, it is evident that the major
chemical changes observed must be due to the activities of the mould,
but changes in minor components cannot unequivocally be attributed
to activity of the mould unless tempe free of bacteria is made.
Samson et al. (1987) found yeast populations of 105–108 cfu g−1 in
a majority of commercial tempe samples in the Netherlands, while
Efriwati et al. (2013) reported populations of 106.8 and 109.7 cfu g−1
in tempe manufactured by two small producers in Bogor, Indonesia.
A population of 109.7 cfu g−1 appears to be extraordinarily high as,
depending on the sizes of the yeast cells, it implies that the volume
of yeast biomass could represent 17 to 33% of the tempe (for cells 4
or 5 μm in diameter). A population of 108 yeast cells (~5 μm diam-
eter spheres) per g fresh tempe would constitute a dry biomass of
~6.5 × 10−1 g (kg fresh tempe)−1 compared with an estimated mould

biomass of ~70 g dry biomass (kg fresh tempe)−1. Again, it is evident
that major chemical changes observed must be due to the activities
of the mould but some contribution from yeasts to minor chemical
changes cannot be excluded.
1.1.5.1 Characteristics of Tempe Moulds

1.1.5.1.1 Taxonomy The moulds in the tempe fermentation
are Zygomycota fungi of the order Mucorales. The Zygomycota
(Zygomycetes) are characterised by the presence of non-septate
hyphae, asexual reproduction by means of non-motile sporagiospores
formed in uni- or multi-spored sporangia or merosporangia. Sexual
reproduction involves the fusion of gametangia to produce zygospores
(Benny 2009). The Mucorales typically reproduce asexually by spo-
rangiospores formed in a sack-like sporangium borne on an aerial
stalk or sporangiophore. Sexual reproduction is by zygospores formed
on opposed or apposed suspensors, though they are rarely seen as two
strains of opposite mating types are required.
Hesseltine et al. (1963a) examined a large number of moulds iso-
lated from Indonesian tempe samples and established that the fungi in
the tempe fermentation are primarily Rhizopus species and especially
R. microsporus Tiegh. var. oligosporus (Saito) Schipper and Stalpers.
Other species isolated included R. oryzae Went and Prinsen Geerligs,
R. stolonifer Ehrenberg and R. arrhizus Fisher. Hesseltine et al. (1963a)
showed that, under aseptic conditions, any of these species may be used
to make perfectly good tempe. The strains used in commercial tempe
manufacturing are normally strains of R. microsporus var. oligosporus,
including Northern Regional Research Laboratory (NRRL) 2710
(=Centraalbureau voor Schimmelcultures [CBS] 338.62 = American
Type Culture Collection [ATCC] 22959) and NRRL 5905 (=CBS
112586).
Schipper and Stalpers (1984) commented that R. microsporus var.
oligosporus ‘shows features of stunted (prematurely blocked) develop-
ment’ and Jennessen et al. (2008) noted that it had a high proportion
(10–31%) of large and irregular sporangiospores and was significantly
different from other, natural Rhizopus taxa in this respect. They con-
cluded that R. oligosporus has a defect in the spore formation process
and suggested that this may be related to the domesticated nature of

R. oligosporus. Pitt and Hocking (1997) also noted that ‘R. oligosporus
appears to be a domesticated fungus. It has rarely, if ever, been reliably
isolated from sources other than fermented foods.’
The taxonomy and identification of Mucorales fungi is primarily
based on morphological features. The genus Rhizopus (Schipper 1984)
has sporangiophores mostly formed on stolons and opposite rhizoids,
either single or more often in clusters, unbranched, occasionally divided
near the top, bearing multispored, terminal sporangia. Sporangia are
globose, distinctly columellate, apophytate (i.e. having a swelling at
the top of sporangiophore beneath the sporangium), greyish to brown-
ish at maturity. Sporangiospores are (sub)-globose to ellipsoidal and
angular. Zygospores covered with spines or warts, formed in aerial
mycelium between non-ornamented, isogamous, opposite suspensors.
Rhizopus differs from Mucor by having stolons and rhizoids.
Schipper (1984) recognised three groups, R. stolonifer group, R.
microsporus group and R. oryzae, within the genus. More recent stud-
ies (Liou et al. 2007), based on the analysis of large subunit rDNA
partial sequences, supported the recognition of the R. oryzae (equiva-
lent to R. arrhizus of Zheng et al. 2007) and R. microsporus groups but
divided Schipper’s R. stolonifer group into R. lycococcus and R. stoloni-
fer group. The R. microsporus group (Liou et al. 2007; Schipper and
Stalpers 1984) is characterised by simple rhizoids, sporangiophores
mostly up to 0.5 (rarely up to 1) mm long, sporangia up to 100 μm in
diameter, temperature maximum for growth usually 45°C or more.
Four R. microsporus varieties were recognised by Schipper and
Stalpers (1984), R. microsporus var. microsporus, R. microsporus var.
rhizopodiformis, R. microsporus var. oligosporus and R. microsporus var.
chinensis. R. microsporus var. oligosporus is described (on malt extract
agar): colony pale yellowish brown to grey; rhizoids simple, subhy-
aline; sporangiophores on stolons, up to 300 μm in length, 15 μm
wide, brownish, in groups of 1 −3; sporangia blackish, up to 80 (−100)
μm diameter; columellae (sub)-globose to subglobose-conical, mouse
grey; sporangiospores (sub-) globose, up to 9 μm diameter, heter-
ogenous, larger extremes irregular; zygospores unknown; restricted
growth at 45°C; good growth and sporulation at 40°C. R. microsporus
var. oligosporus is commonly referred to as simply R. oligosporus and is
so referred to elsewhere in this chapter.
Jennessen et al. (2008) made an extensive study, using low-temper-

ature scanning electron microscopy, of the shape, size and ornamen-
tation pattern of sporangiospores of fungi of the R. microsporus group
and showed that R. oligosporus could be clearly differentiated from
other R. microsporus varieties.
While two of the Rhizopus species groups of Schipper (1984) were
supported by the molecular studies of Liou et al. (2007) they were not
supported by the extensive, morphological, physiological, genetic and
molecular studies of Zheng et al. (2007) and Liu et al. (2007). Zheng
et al. (2007) identified 10 Rhizopus species, with R. arrhizus having
three varieties and R. microsporus having six, namely azygosporus, chi-
nensis, microsporus, oligosporus, rhizopodiformis and tuberosus. Growth of
R. microsporus var. oligosporus NRRL 2710 was identified as R. arrhizus
var. arrhizus but R. microsporus var. oligosporus CBS 338.62, which is
purportedly the same strain (Centraalbureau voor Schimmelcultures
n.d.), was confirmed to be R. microsporus var. oligosporus.
Liu et al. (2008) investigated delimitation of the Rhizopus varieties
based on the intergenic spacers of ribosomal RNA gene (IGS rDNA)
sequences. They concluded that ‘Morphologically R. microsporus has
formed six stable and obviously differentiable varieties (Zheng et al.
2007) but phylogeneically only three lineages were developed, also that
these morphological varieties are not consistent with the phylogenetic
lineages.’ Similarly, Walther et al. (2013) found that the described
morphological varieties are not supported genetically. rDNA internal
transcribed spacer (ITS) sequences identified one clade that included
strains representing the varieties microsporus, chinensis and oligosporus.
The remaining R. microsporus strains, which had been assigned to the
morphological varieties azygosporus, chinesis, oligosporus, rhizopodifor-
mis and tuberosus, had identical ITS sequences.
In the following sections the original names have been retained, not
least because some of the strains have not been subjected to taxonomic
re-examination and, therefore, their current status is not known with
certainty.
1.1.5.1.2 Genomic Information To date, no complete Rhizopus spe-

cies genome has been sequenced and the only genetic information
is one complete mitochondrial DNA sequence for a R. oryzae (Seif
et al. 2005) strain and some sequence data for 28s rDNA, internal
transcribed spacer (ITS) rDNA, elongation factor DNA, beta tubu-

line DNA and the pyrG gene for some strains (Centraalbureau
voor Schimmelcultures n.d.; Liou et al. 2007; Liu et al. 2007, 2008;
Walther 2013).
1.1.5.2 Nutritional Requirements Tempe moulds are relatively nutri-

tionally versatile and able to utilise a wide range of compounds as sole
sources of carbon and energy (Table 1.4). All grow well in mineral
salts media with glucose as sole carbon and energy source and ammo-
nium as sole nitrogen source and do not require any vitamins or other
growth factors (Graffham et al. 1995).
Table 1.4 Substrates Used as Sole Source of Carbon and Energy by Rhizopus Species
SUBSTRATE USED AS SOLE SOURCE OF CARBON AND ENERGYa
R. OLIGOSPORUS R. ORYZAE R. STOLONIFER
NRRL NRRL NRRL NRRL NRRL
SUBSTRATES 2710 5905 OTHER 1526 3563 OTHER A-2293 OTHER
Acetate + e + e +e + e +e
L-lactate −e −e −e −e −e −e
D-lactate −e −e −e −e −e −e −e
Ethanol −e −e +e +e +e
Glycerol −bde −e −b −e +e −b +e, −b −b
Erythritol −b −b −b −b −b
L-arabinose −bc +b
D-arabinose −be −e −e −e +b −e
D-ribose −b +b +b +b
Xylose +bcd +b +b +b
Myo-inositol −e −e −e −e −e −e
Fructose +cdef +e +ef +e +e +ef +e +f
Galactose +bcdef +e +bef +e +e +bef +be +bf
Glucose +bcdef +e +bef +e +e +bef +be +bf
Mannitol +cd
L-rhamnose −e −e −e −e −e −e
L-sorbose +bc +b +b
Cellobiose +b + b +b +b +b
Lactose −bcd −b −b −b −b
Maltose +cde, +e +e +e −e −be −b
−b
Melibiose −cef −e −bef −e −e +f, −b −be +f, −b
Sucrose −bcde −e −e +e −e −b, −be
Table 1.4 (Continued) Substrates Used as Sole Source of Carbon and Energy by Rhizopus
Species
SUBSTRATE USED AS SOLE SOURCE OF CARBON AND ENERGYa
R. OLIGOSPORUS R. ORYZAE R. STOLONIFER
NRRL NRRL NRRL NRRL NRRL
SUBSTRATES 2710 5905 OTHER 1526 3563 OTHER A-2293 OTHER
Trehalose + b + b + b + b +b
Raffinose −bcef − e −bef − e − e +f −be +f, −b
Stachyose −cef −e −ef −e −e +f −e +f
Arabinogalactan −e −e −e −e −e −e
Inulin −b −b −b −b −b
Methylcellulose −e − e −e − e − e −e
Pectin −e −e +e +e +e
Starch +bcd +b +b −b −b
Xylan −e − e −e − e − e −e
Palmitic acid +e +e +e +e +e
Stearic acid −e −e −e −e −e −e
Oleic acid +e +e +e +e +e
Linoleic acid +e +e +e +e +e
α-Linolenic acid NB NB NB NB −e
γ-Linolenic acid +e +e +e +e +e
Soya bean oil +bcd +b +b +b +b
Sunflower oil +f + e + e + e +e
Glycine −c
DL-α-alanine −c
β-alanine −c
L-serine −c
L-glutamate +e +e +e +e +e
DL-threonine −c
L-leucine −c
DL-isoleucine −c
DL-norleucine −c
L-arginine ±c
DL-phenylalanine −c
Gelatin +e +e +e +e +e
a + , growth; –, no growth; ±, dry matter yield 18% of yield on glucose; blank, no data.
b Hesseltine et al. (1963a).
c Sorenson and Hesseltine (1966).
d Nahas (1988).
e Graffham et al. (1995).
f Rehms and Barz (1995).
1.1.5.2.1 Carbon and Energy Sources The three Rhizopus species

commonly found in tempe are largely similar in the range of com-
pounds utilised but there are some differences (Table 1.4). None uti-
lised lactate, the major acid produced during the natural lactic acid
bacterial fermentation during the soaking of the beans, which con-
fers the initial acidic pH on the beans. R. oligosporus does not utilise
ethanol whereas R. oryzae and R. stolonifer do. Glycerol is generally
not used. All utilise hexose sugars, glucose, fructose and galactose,
the constituent sugars of the major soya bean oligosaccharides. Of
the major oligosaccharides in soya beans, neither sucrose, raffinose,
stachyose nor the hydrolysis product, melibiose (Figure 1.23), are
used by R. oligosporus or by most strains of R. oryzae or R. stolonifer.
However, Rehms and Barz (1995) described strains of R. oryzae and
R. stolonifer that grew on these compounds and they suggested that
the use of such strains would allow the production of tempe devoid
of flatulence-producing oligosaccharides and confirmed that this was
possible in practice.
The utilisation of the oligosaccharides present in soya beans
in substantial amounts, namely stachyose, raffinose and sucrose,
depends upon the possession of the enzymes α-galactosidase and/
or β-fructosidase (Figure 1.23). Rehms and Barz (1995) showed that
Rhizopus species could be divided into three groups depending on
whether they possessed none, β-fructosidase only or both of these
enzymes. R. oligosporus NRRL 2710 lacks both enzymes and so is
unable to use any of the oligosaccharides. The second group possessed
1 1 2
Galactose – Galactose – Glucose – Fructose
Sucrose
Melibiose
Mannotriose
Raffinose
Stachyose
Figure 1.23 Soya bean oligosaccharides, hydrolytic enzymes and degradation products.
Hydrolytic enzymes: 1,α-galactosidase(s); 2,β-fructosidase. (Adapted from Rehms, H. and W. Barz.
1995. Applied Microbiology and Biotechnology 44 (1/2):47–52.)
β-fructosidase only and utilised sucrose, stachyose with release of

mannotriose and raffinose with release of melibiose. The third group
produced both enzymes and utilised all the oligosaccharides with
only transient accumulation of intermediates.
Of the amino acids tested, relatively few serve as sole sources of
carbon and energy (Table 1.4).
1.1.5.2.2 Nitrogen Sources R. oligosporus NRRL 2710 utilises

ammonium but not nitrate as sole nitrogen source (Table 1.5).
However, Casey and Walsh (2004) grew R. oligosporus ATCC 22959
(NRRL 2710) in a liquid corn starch/glucose medium with nitrate as
sole nitrogen source. R. oligosporus NRRL 2710 also utilises a wide
range of amino acids, including many that are not utilised as sole
carbon and energy sources.
1.1.5.2.3 Phosphorus Source Strains of R. oligosporus, R. oryzae and

R. stolonifer all utilised phosphate in a chemically defined medium but
no strain grew with phytic acid (myo-inostitol hexakisphosphate) as
sole source of phosphate (Graffham et al. 1995). Wang et al. (1980)
also failed to detect phytase activity in three Rhizopus cultures grown
in a glucose-peptone-salt (without phosphate) medium. However,
Rhizopus strains do produce extracellular phytase when grown in rice
flour (Sutardi and Buckle 1988; Wang et al. 1980) or in corn flour
Table 1.5 Compounds Used as Sole Nitrogen Sources by R. oligosporus NRRL2710

SUBSTRATE GROWTHa SUBSTRATE GROWTHa SUBSTRATE GROWTHa
NaNO3 − b L-cysteine ± b L-proline +b
NH4NO3 +bc DL-aspartic acid +bc L-arginine +bc
NH4Cl +bc L-glutamic acid +bd DL-lysine ±b
CH3CONH2 +b DL-threonine ±b L-histidine ±b
Urea +c L-leucine +b DL-phenylalanine ±b
Glycine +bc DL-isoleucine −b DL-tryptophan −b
DL-α-alanine +b DL-norleucine ±b Gelatin +d
β-alanine ±b DL-valine ±b
L-serine +b DL-methionine ±b
a + , good growth; −, no growth; ±, dry matter yield 14 to 37% of yield on NH4+.
b Sorenson and Hesseltine (1966).
c Nahas (1988).
d Graffham et al. (1995).
(Casey and Walsh 2004) media. Sudarmardji and Markakis (1977)

and Sutardi and Buckle (1985a) also detected phytase activity in
tempe but, as the tempes were not produced aseptically, they will have
included bacteria and it is not possible to unequivocally attribute the
activity to the tempe mould.
The purified phytase enzymes exhibited relatively high thermo-
stability with maximum activities at 55°C (Sutardi and Buckle 1988)
or 65°C (Casey and Walsh 2004) and wide pH tolerance, leading
to the suggestion that the R. oligosporus ATCC 22959 (NRRL
2710) enzyme may have commercial application in animal feeds.
Casey and Walsh (2004) unequivocally demonstrated the extracel-
lular hydrolysis of phytate in an agar plate assay by R. oligospo-
rus ATCC 22959 (NRRL 2710) and other unspecified Rhizopus
strains but only after relatively long periods of incubation. Wang
et al. (1980) also noted that long incubation periods were required
to obtain good phytase activities and this raises the possibility that
the observed activities were due to phosphohydrolases released by
autolysis of old mycelium rather than to true secreted phytases.
Certainly, the distinction of true myo-inositol-hexakisphosphate
3-phosphohydrolase from other acid phosphatases is not straight-
forward (Mitchell et al. 1997).
1.1.5.2.4 Extracellular Hydolases In common with many fungi,

Rhizopus species produce a variety of extracellular hydolases (Table
1.6). Noteworthy differences between the species are amylase pro-
duction by R. oligosporus and R. oryzae but not by R. stolonifer, the
production of pectinase by R. oryzae and R. stolonifer but not by R.
oligosporus (although Sarrette et al. (1992) detected production of
polygalacturonase by R. oligosporus NRRL 5905) and the absence of
protease production by some strains of R. stolonifer. Since all three
species may be used to make perfectly acceptable tempe (Hesseltine
et al. 1963a) this suggests that none of these activities are essential in
tempe fermentation.
There have been some detailed studies on Rhizopus polygalac-
turonases and proteases but little on other extracellular hydrolases.
An endo-polygalacturonase from a soil, R. stolonifer was purified
and characterised by Manachini et al. (1987). It had an optimum
pH of pH 5.0 at 45°C and was highly specific for non-methylated
Table 1.6 Extracellular Hydrolases Produced by Rhizopus Species

GROWTH ON
POLYMER OR
R. OLIGOSPORUSa R. ORYZAE NRRL R. STOLONIFER
PRODUCTION OF
NRRL STRAIN NUMBER STRAIN NO. NRRL STRAIN NO.
EXTRACELLULAR
ENZYME 2710 5905 OTHER 1526 3563 OTHER A-2293 OTHER
Inulin −b −b −b −b −b
Pectinase −bf −ef −bf +bf +f +b +bf +b
Starch +b +b +b −b −b
Amylase Lb,+f +f Lb,+f +f +f +b −bf −b
Xylanase +e
Gelatin hydolysis +b +b +b −b −b
Acid protease +c +c +c −c
Lipase +d
Polygalacturonase +e
Exocellulase −e
Endocellulase +e
Arabinase +e
a + , growth or enzyme production; −, no growth or enzyme production; L, amylase activity detected
only in old cultures; blank, no data.
b Hesseltine et al. (1963a).
c Wang and Hesseltine (1965).
d Nahas (1988).
e Sarrette et al. (1992).
f Graffham et al. (1995).
polygalacturonic acid. Yoshida et al. (2004) cloned and sequenced

an exo-polygalacturonase gene from a plant pathogenic R. oryzae.
Mertens et al. (2008) identified 18 putative polygalacturonase genes
in the genome of R. oryzae strain 99-880, with only two of the
genes being identical. Of the 17 genes, 15 were shown to produce
active enzymes, with 12 encoding for endo-galacturonases and three
encoding for exo-galacturonases. Phylogentic analysis indicated
that the genes form a distinct monophyletic group among fungal
polygalacturonases.
Wang and Hessseltine (1965) showed that R. oligosporus produced
at least two proteolytic systems with pH optima of 3.0 and 5.5 and
Baumann and Bisping (1995) suggested that cell wall-bound pro-
teases were primarily responsible for the proteolysis. A number of
Rhizopus aspartic proteases have now been studied in detail and infor-
mation is available on gene sequences, amino acid sequences, cleavage
specificity and crystal structure (Chen, C.C. et al. 2009; Delaney
et al. 1987; Farley and Ikasari 1992; Horiuchi et al. 1988). Farley and
Ikasari (1992) showed that secretion of aspartic proteases in R. oli-
gosporus is repressed by ammonium, sulphate and arginine and was
not induced by the presence of protein.
1.1.5.3 Responses to Environmental Conditions As with all non-aseptic

food fermentations, promoting the growth of the desired Rhizopus
mould in the tempe fermentation depends upon creating condi-
tions in the substrate that allow the desired organism(s) to dominate
the fermentation. The use of a starter culture, no doubt, assists that
domination but still the environment offered needs to ensure that
the desired organism(s) grow at least as fast as potential competitors.
Historically, these conditions have been discovered by a process of
trial and error and current practice represents solutions to the prob-
lem of how to ensure that the desired outcome is achieved. But, if
traditional processes are to be put on to a more reliable and scientific
basis or transferred to other substrates, then a good understanding
is required of the interactions between the fungus, the environment
offered by the substrate, and potential competitors. One component
of this understanding is how the growth of the fungus is affected by
the environment offered by the substrate. A noteworthy example is of
how some small, presently unknown, differences in the environment
allows R. oligosporus to dominate cooked dehulled soya beans to make
tempe while the residue from soya milk manufacture is dominated by
Neurospora sitophila to make oncom.
1.1.5.3.1 Temperature R. oligosporus is characterised by a maxi-

mum growth temperature of 45°C or higher and some strains are able
to grow at 50°C (Schipper and Stalpers 1984). R. oryzae (R. arrhizus
var. arrhizus) and R. stolonifer have lower maximum growth tempera-
tures of 40–45°C and 30–33°C, respectively. Given the low maxi-
mum growth temperature for R. stolonifer, it is surprising that strains
have been isolated from tempe, where internal temperatures may
exceed 40°C during the fermentation. Most grow at 15°C, usually
with reduced or poor sporulation (Schipper and Stalpers 1984). Pitt
and Hocking (1997) state that R. stolonifer spores barely germinated at
5°C but otherwise there appears to be little information on minimum

growth temperatures.
1.1.5.3.2 pH Value R. oligosporus NRRL 2710 grows over the

pH range 3.0–9.0 (Graham et al. 1976; Sparringa et al. 2002), but
actual minimum and maximum pH values for growth have not been
reported. Sparringa et al. (2002) noted that the hyphal extension rate
of NRRL 2710 was very sensitive to pH value and declined rapidly
on either side of pH 5.5 at 30°C, 37°C and 42°C. This report is in
contrast to the observation of Graham et al. (1976) who noted that
the rate of biomass production was similar at pH 3.0, 4.0 and 5.0 but
lower at 6.0 and 7.0.
1.1.5.3.3 Water Activity R. oligosporus NRRL 2710 and 5905 both

grow best under high water activities (~1.00–0.99) (Sarrette et al.
1992; Sparringa et al. 2002) and growth rate was reduced progres-
sively at lower water activities down to 0.96. Growth at further lower
water activities was not evaluated. Hocking and Miscamble (1995)
examined the effects of water activity on the growth of some non-
tempe Rhizopus strains. The observed minimum water activities for
growth were 0.84 for R. stolonifer, 0.88 for R. oryzae and 0.90 for
R. microsporus. Maximum hyphal extension rates occurred at a water
activity of 0.995 for all the strains and were progressively lower at
lower water activities. Sporulation tended not to occur at the mini-
mum water activity for growth but generally did occur at water activi-
ties 0.02 units higher.
1.1.5.3.4 Carbon Dioxide The hyphal extension rate of R. oligospo-

rus NRRL 2710 was reduced in the presence of 12.5% and 25% CO2
compared with that in the presence of 0.03% (Sparringa et al. 2002).
The organism appeared to be particularly sensitive to CO2 at pH 7.5,
when growth was predicted to be inhibited at 15% CO2 at 33°C and
aw 0.96. Seaby et al. (1988) failed to observe any effects of 10%, 15% or
20% CO2 on hyphal extension rate and suggested that the extension
rate was stimulated by 2.5% and 5% CO2. De Reu et al. (1995) noted
that while the hyphal extension rate of R. oligosporus NRRL 5905 in
the presence of 0 or 35% CO2 was reduced to ~65% of the rate in 5%
CO2, biomass production was practically inhibited, being only ~10%

of that produced in 5% CO2.
Generally, CO2 has little effect on sporulation (Seaby et al. 1988;
Sparringa et al. 2002) but at 42°C, pH 7.5 and water activity ~1.0 or
0.96 sporulation of R. oligosporus NRRL 2710 was inhibited in the
presence of 25% CO2, whereas it occurred under the same conditions
in the presence of air (0.03% CO2). Of all the conditions examined by
Sparringa et al. (2002) these were the only ones to inhibit sporulation.
1.1.5.3.5 Oxygen All the tempe moulds do, of course, grow aer-

obically. R. oligosporus NRRL 2710, 5905 and A-10457 and R. sto-
lonifer NRRL A-2296 are obligate aerobes and are not able to grow
anaerobically (Hesseltine et al. 1985; Graffham et al. 1995). However,
the R. oryzae strains, NRRL 3513, 1526 and IMI 215407, exam-
ined by Graffham et al. (1995) grew anaerobically as did five uniden-
tified Rhizopus species tested by Hesseltine et al. (1985). R. oryzae
has recently been divided into lactic acid producers (R. oryzae) and
fumarate–malate producers (R. delemar) (Abe et al. 2007; Kito et al.
2009), although the latter might equally accurately be described as
ethanol producers. It is not clear whether or not all of the R. ory-
zae-delemar group can grow anaerobically, although it is established
that Amylomyces rouxii, which comprises domesticated variants of R.
oryzae/delemar, is able to grow anaerobically (Graffham et al. 1995;
Hesseltine et al. 1985).
1.1.5.3.6 Antimicrobial Compounds Hesseltine et al. (1963a) found

that if they prepared tempe following procedures that avoided dis-
carding any soak water, with the aim of reducing dry matter losses,
the mould grew poorly, sporulated throughout the cake and yielded
an unacceptable tempe with a disagreeable odour, whereas tempe pre-
pared by traditional methods, that include discarding soak water, sup-
ported good mould growth and yielded a normal white cake. They
attributed this to the presence of unidentified antimicrobial com-
pounds in the soya bean grits that are normally removed by leaching
and discarding the leachate. No further investigations appear to have
been made on this phenomenon.
The other antimicrobial compound that may occur in tempe, espe-
cially as it ages, is ammonia (Steinkraus 1996). Growth of R. oligosporus
NRRL 2710 was unaffected by NH+4 concentrations up to 0.3 mol L −1

but was slower in the presence of 0.24 and 0.84 mmol L −1 NH3 and
was inhibited by 1.3 mmol L −1 (Sparringa and Owens 1999a). In aging
tempe, the pH rises and it is quite possible for these concentrations
of NH3 to be reached, leading to the suggestion that the cessation of
growth in tempe and autolysis could be due to the accumulation of
toxic levels of NH3.
1.1.5.3.7 Interactions among Environmental Factors In the tempe fer-

mentation process the mould is, of course, exposed to all the com-
ponents of the environment simultaneously and it is its response to
the overall environment that determines growth rate. It is commonly
observed that the temperature supporting the fastest growth is lower
when other environmental conditions are unfavourable than when
they are at optimal values and this phenomenon was clearly dem-
onstrated by R. oligosporus NRRL 2710. The optimum temperature
was lowered from ≥42°C at pH 5.5 and 7.5 to 36–37°C at pH 3.5
(Sparringa et al. 2002).
1.1.5.3.8 Relative Growth Rates Relative growth rates of fungi

are commonly estimated by measuring the rates of hyphal extension
on solid media. While this is simple to do, it is recognised that the
observed rates may not accurately reflect the rate of increase in bio-
mass because the density of the mycelium may differ under different
conditions. Many zygomycete fungi are characterised by very high
growth rates and Rhizopus species are typical in this respect. The
predicted maximum hyphal extension rates for R. oligosporus NRRL
2710 in a glucose-salts medium were 1.7 mm h−1 at 42°C, pH 5.85,
aw ~1.00 and CO2 0.03%, although the highest rate actually observed
was ~1.3 mm h−1 (37°C, pH 5.5, aw 0.98 and CO2 0.03%, and 42°C,
pH 5.5, aw 0.98 and CO2 12.5%). De Reu et al. (1995) observed hyphal
extension rates of R. oligosporus NRRL 5905 up to 1.6 mm h−1 at 30°C
on a malt extract-soya peptone medium. R. stolonifer exhibited the
fastest mycelia extension rate, at 2 mm h−1, that Pitt and Hocking
(1997) had recorded for any (food-related) fungus at 25°C. For com-
parison, Ryan et al. (1943) observed a maximum extension rate of
5.2 mm h−1 for Neurospora crasa on a chemically defined sucrose-salts
medium at 35°C.
1.1.6 Characteristics of the Substrate
1.1.6.1 Soya Beans Soya bean seeds comprise an outer protective

seed coat or hull, an embryonic stem and root, the hypocotyl, and
two cotyledons. The proximate compositions of the different parts
are shown in Table 1.7. The cotyledons are the primary storage organ
in the seed. The cotyledon cells have a cell wall, membranes and
cytoplasm but lose most organelles, including the nucleus, mitochon-
dria and endoplasmic reticulum during maturation. Instead, they
contain large protein bodies and many small oil bodies. The protein
bodies are typically 8–10 μm in diameter, with a range of 2–20 μm,
and are bound by a single membrane (Murphy 2008). The oil bod-
ies, also referred to as pherosomes or oleosomes, are lipid-containing
vesicles, relatively homogenous in size with diameters of 0.2–0.5 μm
(Liu 1997).
Soya beans have the highest protein content of the cultivated
legumes at ~40% dry matter, compared with 20–30% in other legumes
and 8–15% in commercial cereals. The major proteins in soya beans
are the storage proteins, glycinin and β-conglycinin, which constitute
65–85% of the seed proteins (Murphy 2008).
Soya beans also have high oil content of ~20% dry matter, sec-
ond among the cultivated legumes only to peanut with ~48%. Most
legumes have far lower oil contents of 1–4% (Liu 1997). The triacylg-
lycerols incorporate mostly unsaturated fatty acids (Table 1.8), includ-
ing good amounts of the essential fatty acids linoleic and linolenic.
Table 1.7 Proximate Composition of Soya Bean Seeds and Their Component Parts for Six U.S.
Varieties
CHEMICAL COMPOSITION (% DRY MATTER)
PERCENT OF
WHOLE SEED PROTEIN
DRY MATTER (N × 6.25) LIPID CARBOHYDRATE ASH
Whole seed 100 40 21 34 4.9
Hull 8 8.8 1 86 4.3
Hypocotyl 2 41 11 43 4.4
Cotyledon 90 43 23 29 5.0
Source: Kawamura, S. 1967. Kagawa Daigaku Nogakubu Gakuzyutu Hokoku (Technical Bulletin of
the Faculty of Agriculture, Kagawa University) 18:118–131. (In Japanese). Cited by Wolf and
Cowan (1971) and Liu (1997).
Table 1.8 Typical Fatty Acid Composition of Soya Bean Oil

FATTY ACID (RELATIVE%)
MYRISTIC PALMITIC PALMITOLEIC STEARIC OLEIC LINOLEIC LINOLENIC ARACHIDIC BEHENIC
C14:0 C16:0 C16:1 C18:0 C18:1 C18:2 C18:3 C20:0 C22:0
0.1 11.0 0.1 4.0 23.4 53.2 7.8 0.3 0.1
Source: Adapted from Liu, K. 1997. Soybeans: Chemistry, Technology and Utilization. New York: Chapman &
Hall.
Soya beans contain a variety of anti-nutritional factors, including

trypsin inhibitors, lectins, lipoxygenase, oligosaccharides and phytate,
and the fates of these components during processing and fermentation
is of considerable importance. The trypsin inhibitors, which bind to
serine proteases and reduce the rate of substrate cleavage, are proteins
and include a Kunitz trypsin inhibitor (specificity directed primarily
towards trypsin) and a Bowman–Bink inhibitor (inhibits trypsin and
chymotrypsin via independent reaction sites). Feeding raw soya bean
meal to rats or chickens leads to irreversible pancreatic blistering, exces-
sive secretion of pancreatic juice, increased excretion of nitrogen in fae-
ces and poor growth rates. In vitro, legume inhibitors generally inhibit
human trypsin but soya Kunitz inhibitor, though not the Bowman–
Birk inhibitor, is completely inactivated by human gastric juice. The
Kunitz inhibitor is relatively heat labile while the Bowman–Birk inhibi-
tor is more stable. Nevertheless, normal heating/cooking procedures are
adequate to eliminate the growth-suppressing effects of raw soya meal
(Belitz et al. 2009; Liu 1997; Richardson and Hyslop 1985).
Lectins are sugar-binding proteins that commonly exhibit toxic
effects in animals. However, soya lectins are inactivated during dry
heating or cooking (Belitz et al. 2009).
The major carbohydrates in soya bean cotyledons are sucrose (2.5–
8.2% dry matter), raffinose (0.1–0.9%) and stachyose (1.4–4.1%) (Liu
1997). Humans lack α-galactosidase and, hence, raffinose and stachy-
ose are not digested and reach the colon intact. In the colon they
are fermented by bacteria leading to the production of carbon dioxide
and hydrogen and possibly uncomfortable flatulence. However, these
oligosaccharides may also function as prebiotic compounds and pro-
mote the growth of intestinal bifidobacteria. Hence, there has long
been considerable interest in the fate of these oligosaccharides during
tempe fermentation.
1.1.6.1.1 Phytic Acid Soya bean seeds contain 1–1.5% phytate (ino-
sitol hexaphosphate), representing 51–57% of the total phosphorus.
Phytic acid can form extremely insoluble salts with divalent metal ions
which can pass through the digestive tract unabsorbed (Tannenmaum
et al. 1985).
1.1.6.1.2 Lipoxygenase Lipoxygenase (linoleic acid oxygen oxido-

reductase) catalyses the oxidation of some unsaturated fatty acids to
monohydroperoxides, which are implicated in undesirable beany fla-
vours. Fortunately, they are very sensitive to heat and are inactivated
during normal cooking (Liu 1997).
1.1.6.1.3 Functional Food Components Food components that do not

make a significant nutrient contribution but which may confer some
beneficial effect(s) are referred to as ‘functional foods’. Soya beans
include two classes of compounds that may lay claim to functional
food attributes, the oligosaccharides mentioned above and isofla-
vones. While flavonoid compounds (polyphenols possessing two ben-
zyl rings joined by a three carbon bridge which may or may not be
closed into a pyran ring) occur in many plants, isoflavone flavonoids
occur in only a few plant families and the concentrations in soya beans
are among the highest in any seeds (Liu 1997). The isoflavones are
mainly present as malonylglucosides and glucosides with only small
amounts of the aglycones, daidzien, genistein and glycitein (Table
1.9). Glycitein and its derivatives occur in the hypocotyls only. The
isoflavones are antioxidants and have been postulated to confer health
benefits. However, according to Belitz et al. (2009), unambiguous evi-
dence to support these claims is lacking.
1.1.7 Substrate Changes during Processing and Fermentation
Tempe manufacture involves four main steps:

1. Soaking the beans in water, with or without the occurrence of
a natural lactic acid bacterial fermentation
2. Dehulling, which may be done with dry beans or after
hydration
3. Cooking
4. Mould fermentation
Table 1.9 Isoflavones Content of Soya Bean Seeds

COMPOUND AND CONCENTRATION (MG [KG DRY MATTER]−1)a
ISOFLAVONE
CATEGORY MEAN RANGE MEAN RANGE MEAN RANGE
Aglycone Daidzein Genistein Glycitein
7 tr – 38 17 7–45 19 tr – 21
B-glucoside Daidzin Genistin Glycitin
204 80–780 318 136–806 69 50–97
Acetylglucoside 6″-O-acetyl- 6″-O-acetyl- 6″-O-acetyl-
daidzin genistin glycitin
1 tr – 12 1 tr – 4 33 tr – 41
Malonylglucoside 6″-O-malonyl- 6″-O-malonyl- 6″-O-malonyl-
daidzin genistin glycitin
414 222–752 1061 670–1558 105 60–183
Total 626 1397 226
Source: Adapted from Liu, K. 1997. Soybeans: Chemistry, Technology and Utilization. New York: Chapman &
Hall.
a Means and ranges for three Japanese and four American varieties.
To have a complete understanding of the overall process one would

like to know:
1. What each processing step contributes to the properties of the

final tempe.
2. What the major substrates utilised by the mould are and what
the major products of its metabolism are?
3. What changes occur in minor, but potentially nutritionally
significant, components?
Answering these questions accurately and unambiguously, particu-

larly for the fungal fermentation stage, is not easy for a number of
reasons:
1. A fermentation is a dynamic process with bean components

being degraded while mould compounds are simultaneously
synthesised and it is difficult to separately measure these two
processes.
2. The difficulty, or even impossibility, of accurately measuring
the amount of fungal biomass makes differentiating between
bean components utilised and fungal components synthesised
very problematic.
3. In the absence of special precautions, tempe normally has a

substantial bacterial population and, consequently, it is not
possible to unequivocally attribute observed chemical changes
to activities of the mould. Evidently, to draw unambiguous
conclusions, experiments should be conducted with tempe
free of bacteria.
1.1.7.1 Changes during Hydration of Beans Traditional processes gen-

erally hydrate beans at ambient temperatures but some modified
processes use higher temperatures (Nout and Rombouts 1990). The
imbibition of water by dry seeds is a physical process and occurs
equally in live seeds and in seeds that have been killed by heat or
other means (Mayer and Poljakoff-Mayber 1989). Imbibition is faster
at higher temperatures but the final seed volume is similar (Mayer
and Poljakoff-Mayber 1989). Irrespective of temperature, soya beans
imbibe water to approximately double the initial seed weight. At
25°C, but not at 60°C, 80°C or 100°C, the process occurred more rap-
idly both in beans pretreated by steaming for 15 min and in dehulled
beans than in raw whole beans (Steinkraus et al. 1965), presumably
because the testa was rendered more permeable or was absent.
Imbibition of water is, of course, the start of the germination pro-
cess and is accompanied by increases in respiration rate, albeit from
a very low level, and in the activities of certain enzymes (Bewley and
Black 1994). Suparmo and Markakis (1987) showed that, following
an initial soaking for 15 h at an unstated temperature (presumably
ambient, 20–25°C), a further 6 h incubation in air at 25°C produced
only a slight elongation of the radical. Rupture of the testa occurred
at 12 h and the radical emerged after 18 h, yielding 1 cm long rootlets
by 24 h. Similarly, Mulyowidarso et al. (1991a) noted the presence
of small ‘germination buds’ on the embryos of soya beans soaked for
24 h at 30°C. Thus, it may be concluded that changes in chemical
composition due to germination-related metabolic activities are likely
to be minimal within the soaking times and temperatures gener-
ally used. Nevertheless, some hydrolysis of isoflavone glucosides, due
to β-glucosidases in the beans, may occur during the soaking stage
(Matsuura and Obata 1993; Matsuura et al. 1969), though the extent
is quite small at temperatures below 30°C and no such hydrolysis was
observed by Wang and Murphy (1996) during the soaking of soya

beans for 10–12 h at 24°C.
In seeds generally, the imbibition of water is accompanied by
leakage of solutes, including sugars, organic acids, ions and amino
acids and, in soya beans, also of certain protease inhibitors and lec-
tins (Bewley and Black 1994). As might be expected, the higher the
temperature and the longer the duration of soaking, the greater the
amount of material leached from soya beans (Wang et al. 1979). The
major compounds leached included the oligosaccharides, sucrose,
raffinose, melibiose and stachyose (Table 1.10), organic acids (Table
1.11) and other unidentified compounds. Included in these unidenti-
fied compounds is the loss of about 10% of the isoflavones (Matsuura
et al. 1969; Wang and Murphy 1996). Only negligible amounts of
trypsin inhibitors and lectins were lost from the soya beans during
soaking (Egounlety and Aworh 2003; Wang et al. 1979). Leaching of
Table 1.10 Concentrations of Soluble Carbohydrates in Whole Soya Beans after Soaking in Tap
Water at 30°C for 24 h
CONCENTRATION AND CHANGE IN CONCENTRATION
AFTER SOAKING
IN THE PRESENCE OF IN THE ABSENCE OF
ANTIBIOTICSa ANTIBIOTICSb
BEFORE SOAKING
(G KG−1 DRY (G KG−1 DRY CHANGE (G KG−1 DRY CHANGE
CARBOHYDRATE MATTER) MATTER) (%) MATTER) (%)
Fructose 6 6.5c + 8 5 − 17
Glucose 2 8.6d + 220 3 + 50
Galactose 4 11 + 175 9 + 125
Sucrose 57 18 − 68 9 − 84
Melibiose 4 2.2 − 45 −e − 100
Stachyose 31 16 − 48 11 − 65
Raffinose 12 5.4 − 55 6 − 50
Total 116 68 − 41 34 − 71
Source: Adapted from Mulyowidarso, R.K., G.H. Fleet and K.A. Buckle. 1991a. International Journal
of Food Science and Technology 26:595–606.
a Microbial growth was inhibited.
b Natural fermentation occurred.
c Concentration was 22 at 18 h.
d Concentration was 21 at 18 h.
e Not detected.
Table 1.11 Concentrations of Organic Acids in Whole Soya Beans before and after Soaking in
Tap Water at 30°C for 24 h
CONCENTRATION (G [KG DRY MATTER]−1)
AFTER SOAKING AT 30°C FOR 24 H
IN THE PRESENCE OF IN THE ABSENCE OF
ACID BEFORE SOAKING ANTIBIOTICSa ANTIBIOTICSb
Formic 1.3 0.6 1.0
Acetic 1.2 0.8 1.1
Propionic 1.8 1.0 1.3
Butyric 0.7 0.5 0.5
Pentanoic 2.2 1.5 0.7
Lactic 0.8 0.7 7.3
Malic 0.8 0.4 3.7
Citricc 0.7 1.2 1.0
Oxalic 0.4 0.2 0.2
Tartaric 0.7 0.3 0.4
Source: Adapted from Mulyowidarso, R.K., G.H. Fleet and K.A. Buckle. 1991b. International Journal
of Food Science and Technology 26:607–614.
a Microbial growth was inhibited.
b Natural fermentation occurred.
c After 6 h concentrations were 1.3 in the presence of antibiotics and 3.0 in the absence of
antibiotics.
minerals from soya beans has not been reported on but Penaloza et al.
(1991) found that soaking lupin seeds in running tap water for ≥7 days
led to loss of ~98% of the potassium, which hampered subsequent
growth of the tempe mould. Phytic acid either increases or remains
unchanged during soaking (Table 1.12). The increase in concentration
is presumed to be due to enzymic activities in the beans as it is too
great to be a consequence simply of the loss of other materials.
1.1.7.2 Natural Lactic Acid Bacterial Fermentation The leaching of

materials from the beans during soaking creates a sugar and nutri-
ent-rich solution. In such habitats a well-characterised microbial suc-
cession occurs. First, aerobic bacteria grow but, as oxygen is quickly
depleted, these are soon replaced by fast-growing facultatively anaero-
bic bacteria, such as members of the Enterobacteriaceae, which grow
by fermentation and produce organic acids. The accumulation of acids
and consequent lowering of the pH value (depending on the degree
of pH buffering present) leads to the slowing and eventual inhibition
Table 1.12 Concentrations of Phytic Acid in Soya Beans during Tempe Production
CONCENTRATION (G [KG DRY MATTER]−1)
EGOUNLETY SUDARMADJI SUTARDI
AND AWORH AND MARKAKIS AND BUCKLE RIET ET AL.
MATERIAL (2003) (1977) (1985a) (1987)
Whole dry beans 13 14 11 16, 17
Soaked beans 17 14 17
Dehulled, cooked cotyledons 15 12 15 10, 12
Tempe 8.8 9.6 6.8–7.5 4, 2
of their growth and allows acid-tolerant lactic acid bacteria to become

the dominant microbes. A succession of lactic acid bacterial species
may occur having increasing levels of acid tolerance, as is the case in
many vegetable fermentations. Typically, the order of microbial suc-
cession in sauerkraut fermented at 15–20°C is Leuconostoc mesenteroi-
des, Lactobacillus plantarum and L. brevis. In addition, acid-tolerant
fungi may grow, with aerobic moulds and film yeasts at the surface
and fermentative yeasts in the depths.
Therefore, in traditional tempe manufacturing processes, a natural
fermentation occurs during the soaking stage that is ultimately domi-
nated by lactic acid bacteria (Table 1.13) and which results in acidi-
fication of the beans to pH 4.1–5.2. Maximum bacterial populations
of 109–1010 cfu mL−1 are achieved (Table 1.13). However, spontaneous
lactic acid bacterial fermentations do not occur reliably in temperate cli-
mates (Nout et al. 1987b; Steinkraus 1996). Nout et al. (1987b) showed
that reliable fermentations could be promoted by inoculating the soak
water with soak water from a previous batch, a procedure referred to
as ‘back slopping’. When this process was repeated a large number of
times they obtained a relatively stable microbial population that, at 19°C
and 25°C, was dominated by Lactobacillus plantarum and Saccharomyces
dairensis whereas at 37°C Pediococcus spp. dominated. In a more recent
study, using DNA-derived DGGE profiles, the prevalent bacteria after
10 cycles of incubation and back-slopping were Pediococcus pentosaceus,
Leuconostoc fallax and Wiesella cibaria (Yan et al. 2013).
With the exception of L. buchneri, Pediococcus spp. and Weisella
cibaria, the predominant lactic acid bacteria in Table 1.13 are homo
fermentative, producing mainly lactic acid from sugars. This might
suggest that the frothing that is normally observed during soaking is
Table 1.13 Microbial Flora of Soya Bean Soak Water with and without Inoculation with Previous Soak Water at Different Temperatures
48
MICROBIAL CONCENTRATION (LOG CFU ML−1)

SOAK CONDITIONS LACTIC ACID BACTERIA
STAPHYLO-
BEAN INOCUL- PREDOMINANT GENERA/ ENTERO- COCCUS BACILLUS
MATERIAL TEMP. (°C) TIME (H) ATION CONCN. SPECIESa BACTERIACEAE SPECIES SPECIES YEAST pH SOURCEb
Whole 20 24 None 9.1 Lactobacillus casei ~ 7 ~ 8.1 < 1 ~ 4.1 − c 1
Whole 30 24 None ~ 8.8 Streptococcus faecium, L. ~ 9.1 ~ 7.5 < 1 ~ 3.5 − 1
casei, S. dysgalactia
Whole 37 24 None 10 L. casei, S. faecium, S. ~ 7.8 ~ 9.2 < 1 ~ 3.8 − 1
dysgalactiae
Dehulled 30 24 None > 9.0 Streptococcus spp ~ 8 − c – ~ 6.5 4.5– 2
5.2
Dehulled 14 24 Yesd – – – – – 4–? 4.9 3
Dehulled 19 24 Yes 9.8–10.1 Lactobacillus plantarum, < 1–2.8 – – 4.1–7.3e 4.4– 3
L. coprophilus, L. casei 4.5
Dehulled 25 24 Yes 9.9–10.0 L. plantarum < 1–2.1 – – < 1 4.2 3
– 7.3e
Dehulled 30 24 Yes – – – – – – 4.1 3
Dehulled 37 24 Yes 9.4–9.6 Pediococcus sp., < 1 – – < 1 4.1– 3
Lactobacillus 4.2
fermentum, L. casei, L.
acidiphilus
Indigenous Fermented Foods of Southeast Asia
Dehulled 30 24 Yes – Pediococcus pentosaceus, – – – – 4.1 4
Leuconstoc fallax,
Weisella cibaria
Dehulled 45 24 Yes – – – – – – 4.3 3
a In order of decreasing prevalence.
b 1, Mulywidarso et al. (1989); 2, Ashenafi and Busse (1991a); 3, Nout et al. (1987b); 4, Yan et al. (2013).
c No data.
d Soak water included 2.5 or 25% previous soak water (back slopping), repeated until microflora was stable.
e Saccharomyces dairensis (= Nanmauvia dairenensis).
T em p e a n d Rel at ed P r o d u c t s
49
due to carbon dioxide production by yeasts, or to unidentified hetero-

fermentative lactic acid bacteria. Although the concentrations of yeast
cells are lower than the concentrations of lactic acid bacteria, their
larger size means that yeast populations of 106−7 cfu mL−1 are compa-
rable in biomass to bacterial populations of 109−10 cfu mL−1. In such
cases, the fermentation might more accurately be described as a lactic
acid bacterial and yeast fermentation.
There appear to be no reports in the international literature of the
microbial flora of soak waters from traditional Indonesian producers.
One might expect that producing tempe repeatedly in one location
would result in the selection of a relatively constant microflora and it
would certainly be interesting to have some data on what it is.
Mulyowidarso et al. (1991a, b) studied the release of organic acids
and sugars during the soaking of soya beans at 30°C for 24 h in the
absence and in the presence of antibiotics (to inhibit microbial growth)
(Tables 1.10 and 1.11). It is evident that the monosaccharides, fruc-
tose, glucose and galactose, and the oligosaccharides, sucrose, meli-
biose and stachyose, were utilised by the microbial flora and that the
quantities released from the beans were more than enough to support
the observed microbial populations. During hydration, substantial
amounts of organic acids are also released and some of these may also
serve as nutrient sources for microorganisms (Table 1.11). Certainly,
many lactic acid bacteria are known to be capable of utilising citrate
(Starrenburg and Hugenholtz 1991).
Major products of the microbial fermentation appear to be lactic
and malic acids (Table 1.11). The lactic acid is, evidently, the product
of lactic acid bacteria but the observed increase in malic acid concen-
trations, both in natural and pure culture fermentations by L. casei
and other bacteria, is curious. Many lactic acid bacteria are able to
utilise malate in the malo-lactic fermentation, decarboxylating it to
lactic acid (Poolman et al. 1991), but there are no other reports on
production of malic acid by lactic acid bacteria. There are, also, no
other reports on the metabolic activities of microorganisms in these
natural soak water fermentations.
The fermentations achieve a final pH of 4–5 and the cotyledons
are acidified to pH 4–5, compared with a pH of ~6.5 for cotyledons
hydrated in the absence of bacterial fermentation.
1.1.7.3 Acidification of Soya Beans by Soaking and/or Cooking in Acid

Solutions An alternative to acidification by natural fermentation is
to soak or cook the beans in a solution of lactic acid or acetic acid
(Steinkraus et al. 1965). Cotyledons soaked and cooked in lactic acid
(30 mL [85% solution] per 1 kg beans in 3 L water) had a pH of 4.8–
5.0 and mould growth on cotyledons acidified with lactic acid was
not visibly slowed until the pH was adjusted to below 3.5. With acetic
acid (7.5 mL glacial per 1 kg beans in 3 L water) the pH of the cotyle-
dons was 6.1–6.3 and mould growth was greatly slowed with 15 mL
glacial acetic acid while 3.75 mL was insufficient to prevent bacterial
spoilage of the tempe (Steinkraus et al. 1965). Other treatments used
include autoclaving at 121°C for 10 min in 0.11 mol L −1 lactic acid to
obtain cooked, sterile, acidified cotyledons (Ruiz-Terán and Owens
1996a). This procedure yielded kernels with a pH value of 4.1–4.6 and
containing 1.6–3.0 g lactic acid (kg dry matter)−1 (Ruiz-Terán 1995;
Sparringa and Owens 1999b).
1.1.7.4 Dehulling Hulls represent 6–10% of the dry matter of soya

beans and, thus, their removal, along with leaching losses during
soaking and cooking, constitute a substantial loss of dry matter
(Table 1.14).
Steinkraus et al. (1960) noted that the mould would not grow on
unskinned beans. However, Ko and Hesseltine (1979) pointed out
that it is not essential to remove the hulls and, in fact, some traditional
producers deliberately do not remove all the hulls in order to achieve a
higher yield and offer a lower price (Ko and Hesseltine 1979). Owens
and Hewitt (unpublished data) also made perfectly good tempe with
yellow Canadian number 2 soya beans, which were dehulled but the
detached hulls were not removed and were fermented along with the
kernels. Hence, the failure of the mould to grow on unskinned soya
beans is due to the inability of the hyphae to penetrate the testa, as
observed by Penaloza et al. (1992a) with quinoa seeds, rather than due
to an inhibition of growth by testa components.
Egounlety and Aworh (2003) noted that tannins were exclusively
associated with hull and were totally removed by dehulling but they
did not investigate whether or not the tannins were antagonistic to
growth of the mould.
Table 1.14 Soluble Carbohydrate, Total Water Soluble Losses and Total Dry Matter Losses
during Preparation of Sterile, Cooked Soya Bean Cotyledons
AMOUNT OF MATERIAL LEACHED INTO THE
WATER (G [KG INITIAL DRY SOYA BEAN]−1)b
YIELD OF DRY MATTER
SOLUBLE (G [KG INITIAL DRY
MATERIALa CARBOHYDRATESC TOTAL SOYA BEAN]−1)b
Hydration water 45 73
Hydrated beans 930
Wash water 14 42
Dehulled, washed 770
cotyledons
Cooking solution 24 54
Cooked cotyledons 730
Total 83 169
Source: Adapted from Ruiz-Terán, F. and J.D. Owens. 1999. Journal of the Science of Food and
Agriculture 79 (2):249–252.
a Soya beans were hydrated in purified water at 100°C for 30 min, dehulled and washed, and
cooked in 0.11 mol L−1 lactic acid solution at 121°C for 10 min.

b Amounts are expressed relative to the amounts present in the initial dry soya beans.
c Sucrose + raffinose + stachyose + fructose + glucose.
1.1.7.5 Cooking Cooking is essential to inactivate various antinutri-

tional components of soya beans, including trypsin inhibitors, lectins
and lipoxygenases. As mentioned earlier, moist heating at 100°C is
effective in inactivating all these factors (Belitz et al. 2009; Egounlety
and Aworh 2003; Liu 1997). In addition, some further leaching of
soluble material occurs, although not as extensive as that occurring
during the initial hydration (Table 1.14). Wang and Murphy (1996)
observed that about 40% of total isoflavones were lost into the cook-
ing water (boiling for 20 min).
Cooking at 100°C also kills vegetative microorganisms but, poten-
tially, allows bacterial endospores to survive. Cooking at higher tem-
peratures, such as 121°C for 10 min, will normally eliminate spores
and may be used to produce sterile cooked cotyledons (Ruiz-Terán
and Owens 1996a). Nout et al. (1985) noted that boiling in lactic
acid solution would, in practice, yield sterile cotyledons and Ruiz-
Terán and Owens (1996a) showed that boiling in the presence of
0.11 mol L −1 lactic acid at pH 3.0 was effective in killing spores of
Bacillus stearothermophilus and sterilising cotyledons.
Tuncel and Goktan (1990) measured concentrations of 1.2% (v/v)

lactic acid and 0.2% (v/v) acetic acid in soak water (inoculated with
previous soak water) incubated at 25°C for 24 h and having a final pH
of 4.4. Yan et al. (2013) recorded concentrations of titratable acidity
of 0.83% (as lactic acid) in soak water of dehulled soya beans. If this
acidity was entirely due to lactic acid, it would equate to a concen-
tration of 0.09 mol L −1. However, at the pH value of the soak water
(~4.1) only 37% of the lactic acid (pKa, 3.87) would be in the toxic
undissociated form. Thus, it is not clear whether or not procedures
that include boiling cotyledons in soak water, such as used by WJB
Home Industry in Bogor (Efriwati et al. 2013), would kill bacterial
spores. Certainly, it would be interesting to know whether the acid
concentrations, pH value and boiling times are sufficient to kill or
reduce the numbers of bacterial spores.
1.1.7.6 Overall Losses of Dry Matter Prior to the Mould Fermen

tation Growth of the mould inevitably involves the utilisation of
some material, due to its respiration to generate energy and support
mould growth, and there is little that can be done to affect the overall
loss at this stage. However, the processing steps prior to mould fer-
mentation are possibly amenable to manipulation to attempt to mini-
mise losses of solids. Steinkraus et al. (1960) recorded total solid losses
of 23.5% prior to inoculation in a laboratory tempe process using
overnight hydration followed by wet dehulling and cooking of the
cotyledons. Very similar losses of 22.2% were observed in a small fac-
tory process using dry dehulling of the beans followed by a short
hydration period and cooking (Steinkraus 1996). Ruiz-Terán and
Owens (1999) observed even larger losses (27%) in a laboratory pro-
cess for the production of sterile hydrated cotyledons. They investi-
gated the sources of the losses in detail and, in their process, losses of
soluble materials accounted for 63% of the losses (Table 1.14).
Hesseltine et al. (1963a) investigated ways to minimise these losses
by making tempe starting with full-fat soya bean grits and using the
minimum amount of water, including re-using the soak water to cook
the beans, but found that the mould then failed to grow well on the
grits. If the soak water was discarded and the grits cooked in fresh water
the grits then supported good growth of the mould. Similar results
were obtained with cotyledons (Smith et al. 1964). These observations
were attributed to the presence of a, as yet unidentified, water-solu-

ble inhibitor of mould growth in the soya bean grits which had to be
washed out to facilitate good mould growth (Hesseltine et al. 1963b).
1.1.7.7 Changes during Mould Fermentation

1.1.7.7.1 General Characteristics of the Fermentation Sudarmadji and
Markakis (1978) suggested that the fermentation may be divided into
three stages: a first stage of rapid fungal growth and chemical changes
in the substrate; a second mature tempe stage with little change in
fungal biomass or in most chemical components of the soya bean
cotyledons; and a third deterioration stage of mycelia senescence and
rapid chemical change.
1.1.7.7.2 Analysis of Fungal Growth and Chemical Changes during Tempe

Fermentation
1.1.7.7.2.1 Growth of the Fungus, Penetration of Beans, Assessment
of Growth Rate and Biomass Although Steinkraus et al. (1960)
described the mould growth as being superficial with little penetra-
tion between or into cells, it is now clear that the hyphae do, in fact,
penetrate deeply into the cotyledons (Jurus and Sundberg 1976; Ko
and Hesseltine 1979; Varzakis 1998). On an average there was one
hyphal invasion per 800–1000 µm2 with a large number of hyphae
penetrating 300–700 µm into the cotyledon, equivalent to 10–22%
of the cotyledon width. Some hyphae penetrated as far as 1500 µm
and the maximum observed was 2000 µm. Hyphae grew primarily
between cells but whether or not they penetrated the cell walls and
grew along the middle lamella or grew only in intracellular spaces
could not be determined (Jurus and Sundberg 1976; Varzakis 1998).
Varzakis (1998) claimed to see haustoria but Jurus and Sundberg
(1976) could not detect any. Jurus and Sundberg (1976) suggested that
the deep penetration, at least partially, accounted for the rapid physi-
cal and chemical changes that occur during tempe fermentation.
1.1.7.7.2.2 Assessment of Fungal Biomass in Tempe Assessment

of biomass and growth kinetics of filamentous moulds in solid state
fermentations, such as tempe, is inherently difficult because mycelium
is not uniform in composition but includes actively growing hyphal
tips, full of active protoplasm, and old hyphae that have been emptied
of their cytoplasm. Among the methods that have been applied to

assess biomass in tempe are the direct weighing of mycelium separated
from cotyledons by enzymic hydrolysis (Charles and Gavin 1977);
estimation of hyphal length by direct microscopy (Feng et al. 2005);
assay of ergesterol, a membrane component (Feng et al. 2005; Nout
et al. 1987c), though Nout et al. (1987c) opinioned that the method
should not be used with tempe because the substrate had such a large
effect on the ergesterol content of the fungus; assay of glucosamine,
a component of cell wall chitin/chitosan (Ruiz-Terán and Owens
1996b; Sparringa and Owens 1999c); real-time monitoring of con-
ductance (Penaloza et al. 1992b); and real-time quantitative pcr (Feng
et al. 2007). Estimates of the fungal biomass (g dry biomass [kg dry
tempe]−1) present in mature tempe range widely from 27–50 (Nout
et al. 1987c), 59–79 (Ruiz-Terán and Owens 1996b; Sparringa and
Owens 1999c) to 70–170 (Charles and Gavin 1996). Given a net dry
matter loss of ~100 g (kg initial dry matter)−1 during the fermentation
one might expect a biomass production of not greater than 100 g (kg
initial dry matter)−1, assuming a maximum conversion efficiency of
0.5 g dry biomass (g substrate utilised)−1, which would translate into
a concentration in tempe of 111 g dry biomass (kg dry tempe)−1. For
fresh tempe with 64% moisture content this would equate to about
70 g dry biomass (kg tempe)−1.
The widely ranging values for biomass concentrations in tempe tend
to confirm that none of the methods provide an accurate measurement
of biomass in tempe. Nevertheless, they can be used to monitor the
dynamics of fungal growth in tempe. It is evident (Figure 1.24) that
there is an initial phase of mycelia growth until the tempe is mature
and then no further increase in biomass occurs even though, after a
period, degradation of the substrate resumes.
1.1.8 Chemical Changes during Tempe Fermentation
As mentioned earlier, the unambiguous interpretation of data on

chemical changes occurring in tempe is difficult due to the presence
of bacteria and yeasts.
The other problem is the publishing of data only as proximate anal-
yses. A microbial fermentation is a dynamic process with materials
being degraded by the microbes and new biomass being synthesised.
100
Change (g/(kg initial dry matter) 50
–50
–100
–150
–200
–250
0 24 48 72 96 120 144 168 192
Time (h)
Figure 1.24 Changes in the amounts of biomass and total dry matter during tempe fermentation
with R. oligosporus NRRL 2710 at 30°C. Amounts are relative to the amount of material at the start
of the fermentation. Δ, biomass; ◻, total dry matter. (Adapted from Ruiz-Terán and Owens [1996b],
using a conversion factor of 12 g dry biomass per g glucosamine [Sparringa, R.A. and J.D. Owens.
1999c. Glucosamine content of tempe mould. Rhizopus oligosporus. International Journal of Food
Microbiology 47:153–157.])
Simple proximate analyses can fail entirely to indicate the scale of such
activities and it is far more informative to relate changes in chemi-
cal components to the amount present at the start of the fermenta-
tion (Berghofer and Werzer 1986; Ruiz-Terán and Owens 1996b).
The well-established and easiest way to do this is to relate measured
concentrations to the amount of ash present at the start of the fermen-
tation. The key assumption is that the total quantity of mineral ele-
ments is unaffected by microbial metabolism and can, therefore, serve
as a proxy for the initial amount of other components. Table 1.15
illustrates how simple proximate analyses may give an entirely mis-
leading impression of the metabolic activities occurring in a fermen-
tation. Even though there is a large reduction in the amount of lipid,
proximate analysis under-estimates the full scale of the reduction.
Similarly, with protein, where none has been utilised, an apparent
increase is recorded (Table 1.15). If ash concentrations are available, it
is a relatively easy matter to calculate the amounts of each material for
each sample time and to present the data as g analyte (kg initial dry
matter [i.e. amount of analyte present relative to that present at the
start of the fermentation])−1.
Table 1.15 Model Data Presented as Amount of Analyte and Concentration of Analyte during a
Microbial Fermentation
TIME (H) ASH LIPID PROTEIN DRY MATTER
AMOUNT OF ANALYTE (G)

0 50 300 400 1000
12 50 250 400 950
24 50 200 400 800
36 50 100 400 700
PROXIMATE ANALYSES (G DRY ANALYTE [KG DRY MATTER]−1)

0 50 300 400 1000
12 53 263 421 1000
24 62.5 250 500 1000
36 71.4 143 571 1000
These considerations are of little consequence in many micro-

bial fermentations, such as lactic acid bacterial vegetable fermenta-
tions, where only a small percentage of the substrate is utilised by the
microbes, but has become very important in fermentations where up
to 20% or more of the substrate is metabolised, such as with tempe.
It is, of course, true that simple analysis of lipid, protein and others
still conceals how much is derived from the substrate and how much
is from the new microbial biomass.
A third problem with data on chemical changes during tempe fer-
mentation is that different authors carry out fermentations at differ-
ent temperatures and for different lengths of time, since the point at
which tempe is considered ‘mature’ and constitutes kernels bound into
a solid block by fungal mycelium is not well defined. Hence, while
there is relative agreement regarding the qualitative changes during
the fermentation, there is considerable variation in the quantitative
changes reported.
1.1.8.1 Changes in Major Chemical Components during Tempe Fermentation

1.1.8.1.1 Proximate Analyses It is evident (Table 1.16) that, while
there are relatively small changes in the concentrations of crude pro-
tein (i.e. total N × constant) or of crude lipid (includes free fatty acids),
there are large increases in the concentrations of free fatty acids and
amino acids. Also evident is a considerable decrease in phytic acid con-
tent and a small decrease in the concentration of soluble carbohydrates.
Table 1.16 Composition of Cooked Soya Bean Kernels and Tempe Made from Thema
HYDRATED AND
MATERIAL COOKED KERNELS TEMPE MEAN CHANGE (RANGE)
CONCENTRATION (G [KG WET
WEIGHT]−1) (%)
Moisture 620–640 610–640
CONCENTRATION (g [kg DRY MATTER]−1)

Crude protein 400–480 390–500 +0.3 (−2.5– +2.8)
Free amino acids and peptides 9 75 +800
Crude lipid 240–290 230–260 −3.5 (−12– +4)
Free fatty acids 0.4–4 28–145 3300 (+900– +11,000)
Soluble carbohydratesb 19–26 15.5–25 −8.6 (−4– −18)
Phytic acid 10–12 2–4 −70 (−60– −80)
a Based on data selected from Riet et al. (1987), Ruiz-Terán (1995), Ruiz-Terán and Owens (1996b),
Sparringa (1999), Sparringa and Owens (1999b,d), Steinkraus (1996) and Wagenknecht et al.
(1961).
b Monosaccharides + sucrose + stachyose + raffinose.
As mentioned earlier, these analyses hide, to a large extent, the

true scale of the metabolic activities of the mould and the effects it
produces on the substrate.
1.1.8.1.2 Changes Relative to the Amounts of Compounds Present in the

Unfermented Soy Kernels Around 70% of the loss of dry matter in
mature tempe is due to loss of lipid (Figure 1.25), presumably due to
utilisation/oxidation of fatty acids by the mould, as most tempe moulds
are unable to utilise glycerol (Table 1.4). Loss of protein due to oxida-
tion of amino acids with release of ammonia accounts for only about
10% of the total dry matter loss. Together, the utilisation of fatty acids
and amino acids account for 83–94% of the total dry matter loss during
the fermentation. It is evident, therefore, that the major compounds
utilised as energy sources by the mould are fatty acids along with small
amounts of amino acids. Certainly, lipase and protease enzyme activity
is present from the earliest stages of the fermentation (Figure 1.26) and
activities of both also continue after the period of active fungal growth.
1.1.8.1.3 Lipids Of the major fatty acids in soya lipids (Table 1.8),

tempe moulds are able to utilise palmitic, oleic, linoleic and linole-
nic acids but not stearic acid, as sole sources of carbon and energy
–20
Change (g/kg initial dry matter)
–40
–60
–80 Mature
–100 tempe
–120
–140
–160
–180
0 12 24 36 48 60 72
Time (h)
Figure 1.25 Losses of dry matter, total lipid and protein during tempe fermentation with
R. oligosporus NRRL 2710 at 30°C. ◻, total dry matter; Δ, lipid; ○, protein. (Adapted from Sparringa,
R.A., and J.D. Owens. 1999d. Journal of Agricultural and Food Chemistry 47:4375–4378.)
(Table 1.4). This would suggest that, in the absence of any preferences,

one might expect to see an accumulation of stearic acid during the
fermentation. However, stearic acid accounts for only ~4% of the total
fatty acids and any increase may well not be detectable.
Lipases of Rhizopus species have a high selectivity and, in aque-
ous systems, they primarily catalyse the hydrolysis of ester bonds at
30
25
Enzyme activity (arbitrary units)
20
15
10
–5
–12 12 36 60 84 108 132
Time (h)
Figure 1.26 Lipase and protease activity during tempe fermentation with R. oligosporus NRRL
2710 at 30°C. Δ, lipase activity; ◻, protease activity. (Data adapted from Ruiz-Terán, F. and J.D.
Owens. 1996b. Journal of the Science of Food and Agriculture 71:523–530.)
260
240 Mature
Concentration (g/kg initial dry matter)

220 tempe
200
180
160
140
120
100
80
60
40
20
0
–20
–12 0 12 24 36 48 60
Time (h)
Figure 1.27 Concentration of lipid components during tempe fermentation with R. oligosporus
NRRL 2710 at 30°C. ◻, triacylglycerol; Δ, free fatty acids; ○, diacylglycerol; ◇, monoacylglycerol;
×, free glycerol. (Data from Lado and Owens, unpublished.)
positions 1- and 3- of tri-, di- and monoacylglycerols but not at the

2-position (Haas and Joerger 1995; Haas et al. 1999). Hence, hydro-
lysis of acylglycerols is a sequential process with progressive formation
of diacylglycerols and mono acylglycerols with relatively little release
of free glycerol (Figure 1.27). Where free glycerol is formed, this is
thought to be a consequence of migration of fatty acids from the 2
position to positions 1 and 3 of the acylglycerol prior to hydrolysis
(Haas and Joerger 1995).
1.1.8.1.4 Proteins There is a general agreement that some soya

protein is hydrolysed during the fermentation and that free amino
acids accumulate (Baumann and Bisping 1995; Steinkraus 1996) but,
apart from the data of Sparringa and Owens (1999d) (Figure 1.28),
there has been little attempt to present an overview of what happens.
The calculations on the protein data in Figure 1.28 involve various
assumptions and, hence, the curves cannot be viewed as an accurate
representation of the fate of soya protein during the fermentation but
are an attempt to indicate the relative scales of the main processes that
occur. These estimates suggest that after 46 h incubation, ~25% of the
initial soya protein had been utilised or hydrolysed. Of this, about
65% was present in the tempe as amino acids and peptides, about 25%
was assimilated into fungal protein, and only about 10% was oxidised
30
15
Amount (g/kg initial dry matter)

0
–15
–30
–45
–60
–75
–90
–105
–12 0 12 24 36 48 60 72
Time (h)
Figure 1.28 Fates of soya protein during tempe fermentation with R. oligosporus NRRL 2710
at 30°C. ○, estimated fungal protein; ◻, protein oxidised to ammonia (ammonia-N × 5.7), Δ, net
decrease in protein; ×, estimated total amount of protein degraded. (Data adapted from Sparringa,
R.A., and J.D. Owens. 1999d. Journal of Agricultural and Food Chemistry 47:4375–4378.)
with release of free ammonia. This suggests that amino acids serve as
only a very minor source of energy.
1.1.8.1.5 Alkalinisation during Tempe Fermentation During tempe

fermentation the pH value increases from an initial value of ~4.5 with
acidified kernels to ~6.5–7.0 in mature tempe and continues to rise
on further incubation and may reach values up to 8 (Ruiz-Terán and
Owens 1996b). The limited shelf life of tempe is associated with this
increase in pH and the development of an ammoniacal odour at pH
values of 7.5–8, when ammonia (pKa, 9.25) starts to volatise. This rise
in pH value is generally attributed to protein degradation (Steinkraus
1996) and is, presumably, due to the oxidation of amino acids when
they are used as energy sources:
RCH(NH+3 )COO− + nO2 → nCO2 + yH2O + NH4+ + OH−
Other possible alkali-generating metabolic reactions include the

uptake and oxidation of anions, such as lactate:
CH3CHOHCOO− + 3O2 → 3CO2 + 2H2O + OH−
Sparringa and Owens (1999b) investigated these possibilities and
concluded that ammonium production could account for 30–40% of
the observed pH increase at 28 and 46 h incubation but almost 100%

at 72 h. In spite of the fact that R. oligosporus NRRL 2710 is not able
to use lactate as sole carbon and energy source (Table 1.4), the lactate
concentration in the bacteria-free tempe was reduced to 20% of the
initial concentration. Nevertheless, this accounted for less than 3% of
the overall alkali production.
1.1.8.1.6 Oligosaccharides Although it is clear that the major loss of

the oligosaccharides, stachyose, raffinose and sucrose, occurs during
the hydration and cooking stages (see Figure 1.29), the fate of the
remaining oligosaccharides during the fermentation stage depends on
the properties of the mould (Rehms and Barz 1995) and, possibly,
on any bacteria present. R. oligosporus NRRL 2710 is not able to uti-
lise stachyose, raffinose or sucrose (Table 1.4) and, accordingly, the
amounts of these compounds are essentially unchanged during the
fermentation. With other mould strains that do utilise these com-
pounds it is possible to obtain a total removal of the oligosaccharides
if this is desired (Rehms and Barz 1995).
Autoclaving at 121°C, and possibly cooking at 100°C, causes some
hydrolysis of sucrose (Ruiz-Terán and Owens 1999) and the resultant
fructose and glucose are consumed by the mould.
60
50
Amount (g/kg initial dry matter)
40
30
20
10
–10
Beans, Beans, Kernels Kernels, Tempe
whole hydrated cooked
Figure 1.29 Fates of oligosaccharides during processing of soya beans to make tempe, using
R. oligosporus NRRL 2710 at 30°C. ◻, sucrose; Δ, stachyose; ○, raffinose; ×, glucose + fruc-
tose. (Data adapted from Ruiz-Terán, F. and J.D. Owens. 1999. Journal of the Science of Food and
Agriculture 79 (2):249–252.)
1.1.8.2 Changes in Minor Chemical Components in Bacteria-Free

Tempe Vitamin concentrations in tempe are generally higher than
in soya beans (Djurtoft and Nielsen 1983; Kao and Robinson 1978;
Murata et al. 1967; Roelofsen and Talens 1964), presumably because
dormant seed tissues are partially replaced by actively growing fun-
gal mycelium. Neither fungi nor plants synthesise vitamin B12 (cya-
nocobalamin) (Singleton and Sainsbury 2001) and so the vitamin
is absent from bacteria-free tempe (Liem et al. 1977; Suparmo and
Markakis 1989).
To date, no studies on phytate concentrations in bacteria-free tempe
have been reported. Some phytic acid is lost during the hydration and
cooking of the beans prior to fermentation (Table 1.12).
1.1.8.2.1 Isoflavones It is well documented that concentrations of

the aglycons, daidzein and genistein, increase during the fermenta-
tion and that there are decreases in the concentrations of the corre-
sponding glycons, daidzin and genistin (György et al. 1964; Nakajima
et al. 2005; Wang and Murphy 1996), but it is not clear if the hydroly-
sis is due to fungal or bacterial enzymes (it is presumed that soya bean
β-glucosidase is inactivated during cooking) as experiments with
assured bacteria-free tempe have not been reported. György et al.
(1964) claimed that ‘Enzymatic hydrolysis of the glucosides through
the action of Rhizopus oryzae has been proved by the use of aque-
ous extract of the fungus on boiled extracts of unfermented beans
(unpublished observations)’ but did not provide, and have not subse-
quently provided any supporting evidence (the R. oryzae referred to
is probably R. oligosporus NRRL2710). Ebata et al. (1972) purified a
β-glucosidase from R. oligosporus NRRL2710 that hydrolysed genis-
tin to genistein.
1.1.8.3 Changes in Tempe Containing Bacteria In the absence of pro-

cedures to exclude bacteria, most tempe made by traditional and
commercial methods include a diverse population of bacteria with
concentrations of 105–109 cfu g−1. Several investigations have been
conducted to determine the roles bacteria play and, in particular,
to determine whether or not their presence confers any benefits.
Certainly, in general terms, it is clear that bacteria are not required to
make good tempe and very early on Steinkraus et al. (1960) established
that tempe made with mould only and free of bacteria or other micro-
organisms was similar to, and possibly superior to, traditionally made
tempe.
1.1.8.3.1 Vitamins Vitamin concentrations in tempe are gener-

ally higher than in soya beans (Djurtoft and Nielsen 1983; Kao and
Robinson 1978; Murata et al. 1967; Roelofsen and Talens 1964), not
least because dormant seed tissues are partially replaced by actively
growing fungal mycelium. Particular attention has been given to lev-
els of vitamin B12 (cyanocobalamin) in tempe because it is an essen-
tial vitamin for humans but is synthesised in nature only by bacteria
(Singleton and Sainsbury 2001) and is not present in plant foods or
fungi. Tempe commonly, but not invariably, contains good amounts
of vitamin B12 (Djurtoft and Nielsen 1983; Liem et al. 1977) and it is
clear that contaminating bacteria are responsible for its production in
tempe (Liem et al. 1977; Okada 1989; Wiesel et al. 1997). Tempe is a
good meat substitute for vegans, who are at risk of B12 deficiency due
to not eating animal products. Since natural bacterial contamination
cannot reliably ensure adequate B12 levels, the deliberate addition of
bacteria to the fermentation has been investigated. The most effective
vitamin B12 producers include Klebsiella pneumonia (Keuth and Bisping
1993), Citrobacter freundii (Keuth and Bisping 1993; Wiesel et al. 1997),
Brevibacterium epidermidis (Wiesel et al. 1997), Bacillus megaterium and
Propionibacterium shermanii (Krusong et al. 1991). While it would seem
undesirable to inoculate tempe with a Gram-negative bacterium due to
the presence of endotoxins in the cell walls, use of some of the other
species would appear to be a perfectly acceptable means of ensuring the
consistent presence of vitamin B12 in tempe for vegans. The concentra-
tions achievable in fresh tempe are 50–100 ng (g dry matter)−1, making
it easy to obtain the adult recommended daily allowance of 1.5 (UK) or
2.4 (USA) μg by eating ~80 g or less of fresh tempe daily.
1.1.8.3.2 Phytate Some authors have reported reductions in phy-

tate concentrations during tempe fermentation, ranging from 20% to
83% of the initial concentrations (Riet et al. 1987; Sudarmadji and
Markakis 1977; Sutardi and Buckle 1985a,b) (Table 1.12). All the
tempes may be presumed to have included bacteria and so it is not
known whether the phytase activity was derived from the mould or
from contaminating bacteria. To date, no studies have been reported

on the phytase activity of bacterial contaminants of tempe but, if
it is eventually concluded that the moulds do not produce phytase,
this could be another area where the deliberate addition of phytase-
producing bacteria might be beneficial.
1.1.8.3.3 Isoflavones Sometimes, but not invariably, 6,7,4′-trihy-

droxyisoflavone, a potent antioxidant, is present in tempe (György
et al. 1964; Klus et al. 1993). It is formed from daidzein and glycetein
by certain bacteria. Rhizopus spp and yeasts are not able to effect the
transformation (Klus et al. 1993). Hence, if the presence of 6,7,4′-tri-
hydroxyisoflavone was deemed desirable this could be achieved by the
addition of an appropriate bacterium.
1.1.8.3.4 Minor Components Saponins are glycosylated alkaloids,

steroids or triterpenes and legumes are the main source in the human
diet (Belitz et al. 2009; Wang 2008). Concentrations are highest in
the seed coat and hypocotyl and consequently the concentration in
tempe is about half that in raw beans (Fenwick and Oakenfull 1983).
Some saponins complex cholesterol, ergesterol and 7-dehydrocholes-
terol and have been shown to reduce plasma cholesterol in animal
models by increasing excretion in faeces.
Nout et al. (1993) investigated the formation of biogenic amines
in tempe prepared with autoclaved kernels. During soaking at 4°C
some tyramine was produced (concentrations of 250 mg [kg dry mat-
ter]−1), presumably due to the activity of endogenous soya bean tyro-
sine decarboxylase, but following fermentation with R. oligosporus
NRRL5905 concentrations of tyramine in tempe reached 1575 mg
[kg dry matter]−1, levels that are high enough to be of some concern.
1.1.8.3.5 Bacterial Acidification Ruiz-Terán (1995) investigated the

effect of adding Lactobacillus plantarum to acidified and non-acidified
kernels. With acidified kernels (pH, 4.1–4.4; lactic acid concentra-
tion, 3.0 g [kg dry matter]−1) growth of the bacterium was, to some
extent, inhibited and had no discernible effect on the fermentation,
on the concentration of lactic acid in the tempe or on its final pH
(5.6–6.7 after 36 h incubation at 30°C). However, with non-acidified
kernels (pH, 6.6–6.8) the bacterium grew well, reaching populations
of ~5 × 109 cfu g−1, and generating lactic acid concentrations of 2 g

(kg dry matter)−1 in the tempe. For cooked cotyledons with a moisture
content of 64%, a lactic acid concentration of 2 g (kg dry matter)−1
equates to 0.035 mol kg−1 but, at the pH values existing in incubating
tempe made with non-acidified cotyledons, very little of the lactate
would be present as the toxic, undissociated acid (pKa, 3.87) (Ruiz-
Terán 1995). These observations suggest that using lactic acid bacteria
to enhance the safety of tempe prepared with non-acidified kernels is
unlikely to be effective.
1.1.9 Nutritional Value
Studies on the role of tempe in nutrition were first carried out by

Jansen in 1924 (Shurtleff and Aoyagi 1985). Jansen showed that the
thiamine content of tempe was less than that of unfermented soya
beans. Jansen and Donath (1924) demonstrated that tempe protein
was highly nutritious when fed to animals. In 1955, Autret and van
Veen published their findings on the importance of tempe protein
for children in developing countries, since when many more scientists
have paid attention to tempe.
In 1950, van Veen and Schaefer published their article on van
Veen’s experiences in prison camps; prisoners who suffered from dys-
entery could not digest soya beans but even malnourished prisoners
suffering from dysentry were able to digest and tolerate the beans in
the form of tempe. Soemantri and Sudigbia (1985) described pre-
school children with chronic diarrhoea treated with a tempe-based
formula (composed of [%]: tempe, 58; wheat flour, 23; sugar, 15; corn
oil, 2; salt, 1.5; and emulsifier, 0.5) who had a significantly shorter
recovery period than those who were fed a milk-based formula. Most
of the chronic cases showed an increase in weight after the second
week of treatment. This observation is very important as diarrhoea
is the second-ranking cause of death in Indonesia (Report on Basic
Health Research 2010). In Sardjto hospital in Yogyakarta, children
under five years old are treated with tempe-rice gruel when they suffer
from diarrhoea.
Because of the beneficial effects of tempe on nutrition and health,
the Indonesian government recommended the exploration of tempe,
not only as a source of protein but also as a source of iron in preventing
anaemia and as source of antioxidant in preventing aging processes

and degenerative diseases (Anonymous 1995).
In essence, the tempe process converts soya beans, that in their
unprocessed state contain a variety of anti-nutritional factors, into a
nutritious and highly acceptable meat substitute that may be consumed
in a variety of ways. There have been a number of nutritional studies on
tempe that have been reviewed by Nout and Kiers (2005). However,
the nutritional value of tempe ultimately derives from the soya beans.
Soya protein is low in sulphur-containing amino acids but contains
sufficient lysine, which is deficient in most cereal proteins and, hence,
combinations of soya and cereal proteins complement each other. There
is no unambivalent evidence that tempe is nutritionally superior to soya
beans processed by other methods, though there are some processing/
fermentation induced changes that may confer some benefits (van Veen
and Steinkraus 1970).
1.1.9.1 Protein Quality The evaluation of the nutritional quality of

proteins is not a simple matter and different methods are bound to
give different results (Liu 1997). It is not the intention to review these
here but, based on bioassays with human subjects, Young (1991) con-
cluded ‘…. that well-processed soya-protein isolates and soya-protein
concentrates can serve as the major, or even sole, source of protein
intake and that their nutritional value is essentially equivalent to that
of food proteins of animal origin. Furthermore, under conditions of
an anticipated normal usage of soya protein, methionine supplemen-
tation is not only unnecessary, but also may even be undesirable for
young children and adults.’
The quality of protein in tempe was studied by Hackler et al. (1964).
These authors observed that the protein efficiency ratio (PER) was not
significantly different between tempe and unfermented soya beans,
suggesting that fermentation did not significantly improve the quality
of the protein. Bai et al. (1975) evaluated protein quality in terms of
net protein utilisation (NPU) and also noted no significant differ-
ence between unfermented soya beans (NPU, 57%) and tempe (59%).
Similarly, Sudarmadji and Markakis (1978) observed no statistically
significant differences between PER values of fried unfermented soya
beans and fried tempe. Using tempe as a source of protein in baby food
formula, Astuti (1982) found that the growth rate of rats fed with this
formula was not different from that of rats fed with a commercial baby
food formula. It is clear that protein utilisation is unaffected by the
tempe fermentation. A study by Hermana (1985) demonstrated that
mixing soya beans and rice in the ratio 3:7 can improve the protein
nutrition of malnourished children less than five years old.
Concentrations of individual amino acids in tempe fermented for
48 h decreased by 3.6–30% of the initial concentrations (Astuti 1992;
Murata et al. 1967; Stillings and Hackler 1965) while the concen-
tration of soluble nitrogen increased from 3.5 to 5.0 g kg−1 in unfer-
mented soya beans to 8.7–20 g kg−1 tempe fermented for 48 h (Astuti
1992; Murata et al. 1967). In tempe made from soya beans variety
Wilis, glutamic was the amino acid present at highest concentra-
tion (110 g kg−1) and the one present at the lowest concentration was
methionine (0.5 g kg−1). Methionine is known to be the limiting
amino acid in soya beans and tempe.
1.1.9.2 Lipids The fat content of soya bean tempe is nearly 25% (w/w)
on a dry matter basis (Table 1.16) and this might be seen as a disad-
vantage where there are attempts to reduce calorie intake. However,
the fatty acids present are primarily mono- or poly-unsaturated,
which are viewed as less undesirable than saturated fatty acids. They
also include a high percentage of the essential fatty acid, linolenic acid
(Table 1.8). If the tempe is fried in oil before consumption then this,
obviously, further increases the oil and calorie content.
1.1.9.3 Carbohydrates Carbohydrate is predominantly present as oli-

gosaccharides, which have been seen as a problem due to them poten-
tially being a cause of flatulence. However, they may be desirable if
they function as prebiotics, though the evidence for the beneficial
effects of prebiotics on health in healthy adults is still ambivalent.
In any case, around 90% of the oligosaccharides in the original soya
beans are lost during the processing into tempe (Figure 1.29).
1.1.9.4 Minerals Few experiments on the effects of tempe fermenta-

tion on mineral nutrition have been done. Riet et al. (1987) reported
that the iron, copper and zinc contents were unaffected by fermen-
tation. Astuti (1998) reported that soya tempe powder added to the
diets of anaemic rats led to enhanced serum levels of haemoglobins,
iron and zinc but not of copper compared to the non-supplemented

diet. Generally, the mineral content does not change during tempe
fermentation but a small loss of calcium was observed, possibly due to
its release from the bridge of phytate–protein complexes and removal
with water released during the fermentation. The calcium content of
unfermented soya bean was 2.3 compared with 2.0 g kg−1 in tempe
fermented for 48 h (Astuti 1992).
Iron in soya beans is present as organic iron, bound to proteins and
other compounds, such as phytic acid. Iron–protein complexes were
found in protein fractions in the range 5000–70,000 Da. In unfer-
mented soya beans, the highest amount of iron was in the 70,000 Da
fraction, whereas in tempe fermented for 96 h the highest amount of
iron was in the 5000 Da fraction, suggesting that during fermentation
iron was released from protein complexes (Astuti 1992). Soluble iron lev-
els also increased with fermentation. Soluble iron in unfermented soya
beans was 24% of total iron compared to 28%, 41%, 69%, respectively,
in tempe fermented for 24, 48 and 72 h. The increase in soluble iron
was attributed to the hydrolysis of phytic acid during the fermentation.
In the diet, tempe is consumed along with staple foods, especially
rice. Astuti (1992) reported that while iron availability is similar in
rice and tempe (62% and 59%, respectively), in a 1:1 mixture of rice
and tempe the iron availability was increased by 10%.
1.1.9.5 Antioxidants in Tempe Tempe made from black soya beans has
a higher antioxidant content than that made from yellow soya beans.
Black soya beans had antioxidant activity of 14% free radical inhibi-
tion compared to yellow soya beans with only 11% (Xu and Chang
2007). However, Nurrahman et al. (2012) observed much higher lev-
els of 37% in black soya bean tempe and 25% in yellow soya bean
tempe. The antioxidants in soya beans were phenolic compounds, tan-
nins and isoflavones (Xu and Chang 2007).
During fermentation the antioxidant activity increased in yellow
and black soya bean tempe (Jha et al. 1997; Nurrahman et al. 2011).
The extent of the increase depended on time of incubation, with
tempe incubated for ~42 h having a higher antioxidant capacity than
tempe incubated for 36 or 24 h. Astadi et al. (2009) claimed that an
antioxidant extract of black soya bean seed coat could be used to pre-
vent low-density lipoprotein oxidation in human blood plasma.
1.1.9.6 Functional Food Attributes The U.S. Food and Drug

Administration, since 1999, has allowed food manufacturers to make
health claims for consumption of soya bean protein. Current (2013)
model claim statements are ‘(1) 25 grams of soya protein a day, as part
of a diet low in saturated fat and cholesterol, may reduce the risk of
heart disease US Food and Drug Administration (2013). A serving
of [name of food] supplies_grams of soya protein’ and ‘(2) Diets low
in saturated fat and cholesterol that include 25 grams of soya pro-
tein a day may reduce the risk of heart disease. One serving of [name
of food] provides_grams of soya protein.’ Soya bean tempe obviously
meets these requirements. The regulation specifically does not iden-
tify any component(s) of soya beans that may be responsible for the
putative health benefits. The efficacy of the health claims have been
questioned (Balk et al. 2005) and Hill et al. (2008) concluded that
the beneficial effects of soya proteins ‘appear to be limited to modest
reductions in LDC-C, particularly in individuals with elevated serum
levels’.
There is presently much interest in the possible roles of bifidogenic
oligosaccharides (prebiotics) in promoting health but unequivocal
evidence that supplementation of diets with such compounds confers
health benefits to healthy individuals consuming a balanced diet is not
yet available. Soya beans and tempe contain oligosaccharides which
have been shown to be bifidogenic (Crittenden and Playne 1996).
Even more interest attaches to isoflavones. Soya bean products,
including tempe, can be a major source of these compounds in the
diet. During the tempe fermentation the glycones may be converted
into aglycones but this is, apparently, of little consequence for their
biological effects (Faughnan et al. 2004; Xu et al. 2000). Nevertheless,
the benefits or otherwise of isoflavones in foods is an area of consider-
able controversy (Hill et al. 2008; Kelly 2002).
1.1.10 Safety Considerations
The rise in pH value that occurs during tempe fermentation has impli-
cations for safety as it converts a material that is inherently safe due
to its low pH and the presence of organic acids into a product that
allows the growth of a wide diversity of organisms, including patho-
gens. Possible microbiological safety concerns include
1. Growth of pathogenic bacteria and/or production of micro-

bial toxins, especially of heat-tolerant ones, in soak water
2. Growth of spore-forming bacteria and the formation of heat-
resistant spores in soak water
3. Growth of pathogens and/or production of microbial toxins,
especially of heat-tolerant ones, in tempe
4. Production of mycotoxins by tempe moulds
5. Production of mycotoxins by contaminant moulds
Each of these is considered below, along with any evidence show-

ing that it is or is not a problem in practice.
1.1.10.1 Growth of Pathogenic Bacteria in Soak Water A great diver-

sity of microbes is able to grow in soak water, including pathogens.
Nout et al. (1988) showed that Staphylococcus aureus survived in newly
started soak water at 25°C (final pH 4.7) but not in soak water inoc-
ulated with recycled soak water (final pH 4.0). Although there are
no reports on growth or otherwise of pathogens in traditional soak
waters at ~30°C, it may be suggested that when soya beans are acidi-
fied to pH ~4.0, whether this is accomplished in inoculated soak water
at 20–25°C or by natural soaks at ~30°C, pathogens are not likely to
be present.
There are no reports on toxin formation in soak water but, since
the water is discarded and the kernels washed, it is possible that little
would remain with the kernels.
1.1.10.2 Growth of Pathogens and/or Production of Microbial Toxins in

Tempe In non-acidified soya beans (pH ~6) many pathogens are able
to grow unimpeded, including Bacillus cereus (Ashenafi and Busse
1991a; Chen, L.Y. et al. 2009; Nout et al. 1987a; Tuncel and Goktan
1990), Clostridium botulinum (Tanaka et al. 1985), Salmonella spp
(Ashenafi and Busse 1991b; Tanaka et al. 1985), Staphylococcus aureus
(Nout et al. 1988; Tanaka et al. 1985; Tuncel and Goktan 1990) and
Yersinia enterocolitica (Tanaka et al. 1985), and those able to produce
toxins do so (Nout et al. 1988; Tanaka et al. 1985; Tuncel and Goktan
1990).
Acidification of soya beans, by natural lactic acid bacterial fer-
mentation or by soaking and/or boiling in lactic and/or acetic acids
inhibits or slows growth of pathogens in tempe, but there is some

disagreement about the extent. Nout et al. (1987a) observed inhibi-
tion of growth of B. cereus in beans acidified to pH 4.85 in natural
soak fermentations and in beans acidified to pH 4.4 by soaking in
lactic acid, but Tuncel and Goktan (1990) observed growth of B.
cereus to populations of >109 cfu g−1 in beans acidified to pH 4.4 in
natural soak fermentations and in beans acidified in a solution of
lactic acid, 0.6%, and acetic acid, 0.06%, adjusted to pH 4.0 with
HCl. These latter conditions were sufficient to totally inhibit growth
of S. aureus.
The inhibition of bacteria by short-chain organic acids at low pH
values depends primarily on the concentration of undissociated acid
(Stratford and Eklund 2003). Acidification of soya bean kernels by
autoclaving at 121°C for 10 min in 0.11 molL −1 lactic acid solutions
yielded kernels with pH values 4.1–4.6 and containing 1.6–3.0 g lac-
tic acid (kg dry matter)−1 (Ruiz-Terán 1995; Sparringa and Owens
1999b). There do not appear to be any other reports that give informa-
tion on both the pH values and the concentrations of organic acids in
the beans when they were inoculated with the mould. These concen-
trations equate to concentrations of 0.18–0.03 mol kg−1 (wet weight)
in the kernels and at the initial pH values it can be estimated that
37 − 16% would be in the undissociated form. Undissociated lac-
tic acid concentrations of 3–10 mmol kg−1 are certainly in the range
where antimicrobial effects against spoilage and pathogenic bacteria
may be expected (Westby 1989).
Tempe, commonly, contains large populations of lactic acid bac-
teria and several researchers have investigated the effects of their
presence or absence on the growth of pathogens. Some restriction of
the growth of pathogens has been observed on some occasions but
in other cases no inhibition has been observed (Ashenafi and Busse
1991a,b; Nout et al. 1987a, 1988; Tuncel and Goktan 1990).
Tempe is consumed fried or incorporated into stews and other
dishes that are cooked and any vegetative bacteria and heat-sensitive
toxins will normally be destroyed. Hence, the main potential concerns
would appear to be the heat-stable toxins of B. cereus and S. aureus. In
practice, tempe appears to be entirely safe and there are no reports
of poisoning associated with tempe, other than bongkrek poisoning
related to the consumption of tempe bongkrek.
Much commercial tempe in the United States and the Netherlands

is made without acidifying the beans, presumably in order to mini-
mise costs. Under these conditions, scrupulous attention to hygiene
and the avoidance of microbial contamination is essential if prob-
lems, especially over-growth by Bacillus species, are to be avoided
(Tibbott 2004).
1.1.10.3 Production of Mycotoxins by Tempe Moulds Although mucora-

ceous fungi are not noted for the production of secondary metabo-
lites, some Rhizopus species are capable of forming mycotoxins. The
mycotoxins fall into two classes, the rhizoxins, which are antimitotic
polyketide macrolides (Figure 1.30), and the rhizonins, which are
heptatoxic cyclic peptides (Figure 1.31) (Rohm et al. 2010). Jennessen
et al. (2005) examined four strains of R. microsporus, six R. oligosporus
strains and four R. chinensis strains, grown on a variety of substrates, for
the production of secondary metabolites. Rhizoxins were produced by
all four R. microsporus strains, and one also produced rhizonins while
the six R. oligosporus and four R. chinensis did not produce any of these
metabolites. None of the strains produced other mycotoxins (e.g. afla-
toxins, ochratoxins, trichothecenes). While this suggested that strains
O O
R N
H HN
O O NH O
HN
O N O
O
Figure 1.30 Rhizoxin, an antimitotic polyketide macrolide produced by Burkholderia rhizoxinica,

a bacterial endosymbiont in Rhizopus spp.
N O
HO
O
O
O
O O
O
O
Figure 1.31 Rhizonin, a heptatoxic cyclic peptide produced by Burkholderia endofungorum, a

bacterial endosymbiont in Rhizopus spp.
used in food fermentations did not produce mycotoxins, Rohm et al.

(2010) found rhizoxins production in a strain of R. microsporus (CBS
111563) from a starter culture for sufu. They further showed that the
strain would make perfectly good sufu or tempe and that in both cases
it produced significant amounts of rhizoxins in the product. Thus,
while it is reassuring that no R. oligosporus strain has, so far, been
found to produce these toxic metabolites there is the potential for
toxin-producing strains to be present in tempe manufacture. Rohm
et al. (2010) described a PCR procedure that would allow such strains
to be indentified and excluded.
These ‘mycotoxins’ are, in fact, not synthesised by the mould but
by bacterial endosymbionts located in the fungal cytosol (Partida-
Martinez and Hertweck 2005) and identified as new Burkholderia
species, B. rhizoxinica and B. endofungorum (Partida-Martinez et al.
2007). Moulds cured of the bacteria no longer produced the metabo-
lites and the bacteria were shown to produce the toxins in axenic cul-
ture. Lackner et al. (2009) were able to isolate the symbiont in each
case from eight strains of toxin-producing moulds. In nature, rhizoxin
is the cause of rice seedling blight and this represents the first example
of a fungus hosting a bacterium to produce a virulence factor.
1.1.10.4 Bongkek Poisoning Garcia et al. (1999) and Lynch and

Dennis (2010) provide good overviews of bongkrek poisoning and
much of what follows is from their articles. Tempe bongkrek is made
either from coconut press-cake, the by-product after the oil is pressed
H
H
H
H
O H O
OH OH
O
H
O
HO
Figure 1.32 Bongkrekic acid, a toxin produced by B. cocovenenans.
from the flesh, or from the residue from making coconut milk. Its
consumption has long been associated with human intoxication and
the cause of many deaths, despite its production having been banned
in Indonesia since 1962 (Kuswanto 1988). The poisoning is a conse-
quence of the growth of a bacterium, Burkholderia cocovenenans (for-
merly Pseudomonas cocovenenans; Zhao et al. 1995) and the production
of two toxins, bongkrekic acid and toxoflavin. Bongkrekic acid is a
28-C chain molecule (Figure 1.32) while toxoflavin is an azapteridine
(Figure 1.33). Toxoflavin is a key factor in bacterial plant wilt caused
by Burkholderia glumae (Jeong et al. 2003).
The onset of illness occurs within a few hours of consumption
of the contaminated food and death can occur in as little as 20 h.
Although both toxins are potent poisons, bongkrekic acid is typically
produced in greater quantities than toxoflavin and is presumed to be
mostly responsible for human poisonings (Garcia et al. 1999).
A key question is, why does B. cocovenenans occur in tempe bong-
krek, and also occasionally in fermented maize meal (Garcia et al.
N N
N O
N N
Figure 1.33 Toxoflavin, a toxin produced by B. cocovenenans.

1999), and not in other foods. Garcia et al. (1999) suggested that, in
the case of tempe bongkrek, important factors were a high oil con-
tent and especially the presence of lauric (12:0), myrystic (14:0) and
palmitic (16:0) acids, although in experiments with single fatty acids,
oleic (18:1) acid promoted the formation of the highest concentra-
tions of bongkrekic acid. Buckle and Kartadarma (1990) showed that
growth of B. cocovenenans could be prevented in partially defatted
coconut medium by acidifying it with acetic acid to an initial pH of
4.5 and adding 2% NaCl.
1.1.11 Future Prospects and Research Needs
The demand for tempe in Indonesia has increased in recent years

due to economic growth and increasing population. This has led to
increases in domestic tempe production, which is likely to continue.
Tempe also has considerable potential as a vegetarian meat substi-
tute in other countries but, while it is produced on a small scale in
many countries around the world, only in a few areas has it made
much impact. As the world population continues to increase and the
cost of producing meats escalates, tempe may become an increasingly
attractive realtively low-cost protein source. Certainly, the ubiquity of
soya beans and the relatively simple technology involved in its produc-
tion should allow it to compete with other meat substitutes, such as
Quorn® mycoprotein.
Although tempe has been the subject of scientific research since at
least the 1950s, there are still many, quite basic, aspects that are not
clearly understood. These include the following:
1. There is still a great diversity of traditional processes used
to make tempe in Indonesia (e.g. see Seumahu et al. 2013).
In some respects, it is surprising that the producers have not
arrived at more agreement on what is the optimum process
or processes. It is evident that much work could be done to
determine the advantages and disadvantages of the different
procedures in relation to the final quality of the product and
the economics of production.
2. The microflora, microbial succession and biochemistry of soak
waters in Indonesia.
3. The source of malic acid, reported by Mulyowidarso et al.

(1991b), in soak water.
4. The significance of yeasts in soak water, the alcohol concen-
trations produced and the effects on the properties of the
beans or kernels.
5. Are the acid concentrations, pH value and boiling times suf-
ficient to kill or reduce the numbers of bacterial spores?
6. The yield and acceptability of tempe made including the hulls.
7. Is it possible to achieve safe and reliable vitamin B12 supple-
mentation in tempe by inoculation of appropriate bacteria, if
this were deemed desirable?
8. Is it possible to achieve safe and reliable reduction in phytic
acid in tempe by inoculation of appropriate bacteria, if this
were deemed desirable?
1.2 Tempe from Other Pulses

Mary Astuti
In Indonesia, tempe is not only made from soya bean but also other
pulses. Commercial non-soya bean tempes currently found in the
market were made from velvet bean and Leucaena leucocephala (lam-
toro) and sword bean. Velvet bean tempe was found in Kulonprogo
district, Yogyakarta Province and lamtoro tempe was found in
Wonogiri district, Central Java Province. Sword bean tempe was
found in Yogyakarta and Central Java provinces. Tempe gembus (tofu
waste tempe) was present in Yogyakarta and Central Javaan mar-
kets. According to some people, tempe gude (pigeonpea tempe) was
found in Central Java. Wing bean tempe was not found in any of the
markets.
1.2.1 Lamtoro (Leucaena leucocephala) Tempe
L. leucocephala (lamtoro) is used as animal feed and the seeds are also
used as human food in Central America, Indonesia and Thailand.
The seeds are ovoid in shape and have brown hulls and yellow ker-
nels. The hull:kernel ratio is 1:1 by weight and the seeds are low in
oil but rich in protein, which is concentrated in the kernel. The seeds
contain a large amount of galactomanan mucilage located in the

endosperm. They also contain mimosine, a toxic non-protein amino
acid, which may contribute as much as 14% of the total nitrogen
content of lamtoro seeds. Lamtoro seeds may be eaten in processed
or un-processed form and in Java they are fermented into tempe or
cooked as a side dish.
1.2.1.1 Preparation of Lamtoro Tempe The lamtoro seeds are boiled in

water to which wood ash has been added for about 2 h. The cooked
seeds are drained overnight and then the hulls are removed. The ker-
nels are soaked overnight, washed, soaked again until foam appears
on the surface, and then boiled for about two and a half hours. After
boiling, the kernels are drained, cooled, inoculated with mould inoc-
ulum, wrapped in teak leaf and incubated for 36–48 h.
Removal of the hulls is very difficult, such that not all of them can
be removed and some remain with the kernels that are fermented and,
consequently, the taste of lamtoro tempe is not as attractive as that of
soya bean tempe. Lamtoro tempe lacks mimosine, probably because
it is lost by leaching during the combination of washing, soaking and
boiling. Lamtoro tempe contains (g kg−1) fat, 5; protein, 110 and car-
bohydrate, 210 (Depkes 2006).
1.2.2 Velvet Bean Tempe (Tempe Benguk)
Velvet bean (Mucuna pruriens) is a tropical legume. The plant is an

annual climbing vine that can grow to more than 15 m in a season.
The mature pods have a hair-like covering that may cause itching
to people sensitive to it. Velvet bean seeds are slightly bigger than
chick peas and come in a variety of colours, including white, black,
brown and mottled. In Indonesia, velvet beans are processed into
tempe and soups.
1.2.2.1 Preparation of Velvet Bean Tempe Velvet beans are boiled in

water containing wood ash. The function of the ash is to absorb the
bitter-tasting gum-like compounds and to prevent them from attach-
ing to the kernels. After boiling for about 1 h, the beans are drained
and the seed coat is removed. The kernels are soaked for 3 days,
changing the soak water each day, to remove toxic materials. They are
then thoroughly washed and cooked for about 30 min until the tex-
ture of the kernels becomes tender. The kernels are drained, cooled,
sliced into smaller sizes, inoculated with mould inoculum, wrapped in
banana or teak leaf and incubated for 48 h.
Velvet bean tempe contains (g kg−1): protein, 100; fat, 13; carbo-
hydrate, 230; (mg kg−1): daidzein, 59; genestein, 78; and glysitein, 36
(Ariani and Handayani 2009).
1.2.3 Sword Bean Tempe (Tempe Koro)
There are two kinds of sword bean tropical legume, upright plants
with white seeds (Canavalia gladiata) and vines with red seeds
(Canavalia ensiformis). In Indonesia, the young green pods and
immature seeds of sword bean are cooked as a vegetable. Despite
their nutritional potential in terms of protein content, sword beans
contain anti-nutritional factors, such as haemagglutinin, tannins,
phytate and canavanine (Siddhuraju and Becker 2001). Canavanine,
a structural analogue of arginine, is a non-protein amino acid nat-
urally occurring in legumes. The toxic potential of canavanine in
legume seeds may be reduced by soaking, boiling and decanting the
water. The content of canavanine was reduced by 50% after soaking,
and by 34% after boiling and decanting (Ekanayake et al. 2007).
There are no reports of people in Java who have been poisoned after
eating sword bean tempe.
1.2.3.1 Preparation of Sword Bean Tempe Sword beans are cleaned by

boiling in water for 30 min and then drained. The beans are then
soaked in fresh water for 3–5 h, the soak water is decanted and the
seeds drained. The seed coats are removed, the kernels sliced into
2–3 smaller pieces, thoroughly washed, steamed for about 30 min,
drained, cooled, inoculated with mould inoculum powder (2 g [kg
sword bean kernels]−1), wrapped in banana leaf or perforated plastic
bags and incubated at ambient temperature for about 48 h. The pro-
duction of sword bean tempe is less prevalent than soya bean tempe
but more prevalent than velvet bean tempe and lamtoro tempe. It is
produced not only in Central Java but also in East and West Java.
The nutrient composition of sword bean tempe is (g kg−1): protein
290; fat 7.7; carbohydrate, 137; ash 31 (Depkes 2006).
1.2.4 Pigeon Pea Tempe (Tempe Gude)
Pigeon pea (Cajanus cajan L.) is a short-lived perennial that is widely

grown in the tropics and subtropics as a source of human nutrition.
It is a minor crop in Indonesia. The seed contains 20–25% protein.
Tempe made from pigeon pea has an acceptable appearance and tex-
ture, although the shelf life of pigeon pea tempe is less than that of
soya bean tempe. Pigeon pea tempe has a shelf life of 27 h while that
of soya bean tempe is more than 48 h. Widowati and Damardjati
(1986) reported that tempe made from pigeon pea 67% and soya bean
33% was as tasty as pure soya bean tempe.
1.2.4.1 Preparation of Pigeon Pea Tempe Pigeon peas are dehulled,

soaked in water for 24 h (bean:water ratio:: 1:4), the water decanted,
the kernels boiled in fresh water for 25 min, drained, cooled, inoc-
ulated with mould inoculum, wrapped in banana leaf or perforated
plastic bags and incubated for 27 h. The pigeon pea tempe has the
white colour of mycelium of R. oligosporus and a protein content of
~120 g kg−1 (Herastuti et al. 1997).
1.2.5 Lablab Bean Tempe (Tempe Koro Wedus or Tempe Kacang Komak)
Lablab bean or hyacinth bean (Lablab purpureus) is a tropical legume

existing in many varieties. The beans contain tannin, and phytate
and trypsin inhibitors, with considerable variation in concentrations
between different varieties. Soaking or cooking reduces the activity of
these compounds. In Indonesia, the beans are used to prepare tempe,
especially in West Java.
1.2.5.1 Preparation of Lablab Bean Tempe The preparation of lablab

bean tempe is almost the same as for velvet bean tempe. First, water
is mixed with ash and boiled, and then the cleaned beans are added
and boiled for about 30 min. The water is decanted and the beans
washed and dehulled. The kernels are soaked for 2 days, changing
the soak water each day to remove toxic materials, washed, steamed
for about 1 h, drained, cooled, inoculated with mould inoculum,
wrapped in perforated plastic bags, and incubated at room tempera-
ture for 36–48 h.
1.2.6 Tofu Waste Tempe (Tempe Gembus)
Waste from the manufacture of tofu is not only used as pig feed but
is also used as human food. Tempe gembus is prepared from tofu
waste in the following manner. The waste is pressed to remove water.
The press cake is then broken up into small pieces, steamed for 1 h,
drained, cooled, inoculated with mould inoculum of R. oligosporus or
a mixed culture, wrapped in perforated plastic bags, and incubated
at room temperature for 24 h. A typical analysis of tempe gembus is
(g kg−1): protein 34; fat, 2; carbohydrate, 120; crude fibre, 39; calcium,
1.4; phosphorous, 0.5 (Departemen Kesehatan 2009). Tempe gembus
is produced in Central Java and Yogyakarta provinces.
1.2.7 Tofu and Peanut Waste Tempe (Tempe Menjes or Tempe Enjes)
Tempe menjes, also called tempe enjes, is made from tofu waste and
waste of defatted peanut and is produced especially in Malang, East
Java. The process is similar to that for tempe oncom hitam. First,
defatted peanuts are split and pressed to remove any residual oil.
The beans are then soaked for 3–4 h in water to which waste soya
bean soak water (acidic due to its content of organic acids) is added to
obtain a pH value of 4.6–4.7. The soak water is acidified to prevent
undesirable bacterial growth. The soak water and any oil are decanted
and the beans are washed. They are then mixed with pressed tofu
waste and the mixture is steamed for 1 h, drained, cooled, inoculated
with R. oligosporus mould or a mixed culture, wrapped in perforated
plastic bags, and incubated at room temperature for ~48 h. Tempe
menjes is always consumed cooked, mixed with rice flour, salt and
garlic and fried.
1.3 Indonesian Oncom (Fermented Food Processing By-Products)

Kapti Rahayu Kuswanto
Oncom (‘on-chom’) is a traditional food of West Java, Indonesia. It is

a cake-like product made from by-products of the production of other
foods. The by-products used include peanut press cake, solid waste
from soya bean curd production, solid waste from tapioca manufacture
Figure 1.34 Red oncom, made from a mixture of peanut press cake, tofu press cake and onggok
(residue of cassava starch production). (Courtesy of J.D. Owens.)
(onggok) and dried coconut press cake. There are two types of oncom,
black oncom (oncom hitam), in which the principal mould is R. oli-
gosporus, and red oncom (oncom merah), in which the principal mould
is Neurospora intermedia var. oncomensis. Black oncom is greyish black
in colour while red oncom is orange to reddish orange (Figures 1.34
and 1.35). Both are soft compact cakes with a pleasant aroma and
Figure 1.35 Black oncom, made from a mixture of coconut press cake and tofu press cake.
relatively bland taste. Oncom merah is the only human food produced
with Neurospora spp.
Oncom and tempe are not clearly differentiated. However, oncom
is generally made from waste products and never from soya beans.
Some oncom is prepared utilising exactly the same microorganism,
R. oligosporus, as used in the production of tempe but most oncom uti-
lises Neurospora spp (Winarno et al. 1984). Accordingly, some people
refer to foods utilising Rhizopus spp as varieties of tempe and foods
utilising Neurospora spp as types of oncom. Nevertheless, while spor-
ulation of the Rhizopus is undesirable in tempe, in black oncom the
mould is allowed to sporulate.
Because oncom production uses waste streams to make human
food, it increases the efficiency of utilisation of the crop concerned.
In a world with ever-increasing hunger and ever-dwindling crop-
land, this carries great promise for many who could most use help.
Peanut press cake has for a long time been used for animal feed,
as the relatively high fibre content and undigestable components
makes it relatively undesirable for human food. Oncom, made by
growing mould on peanut press cake, converts it to an acceptable
human food that is used in the daily diet of some 25 million people
in Indonesia.
1.3.2 Places of Production, How Consumed and Role in Diet
Oncom is mainly produced in West Java and is an important ingre-

dient in the daily menu of the West Javanese, particularly among
those in lower income groups. Red oncom is produced locally in
West Java and particularly in the Bogor area. Oncom is still consid-
ered to be less useful than tempe but has great potential as a supplier
of protein for the middle to lower income groups because it costs less
than tempe.
Oncom is consumed as a side dish in the form of slices deep fried in
oil or as small portions in soups or other dishes. Oncom is rarely eaten
raw but is commonly on sale as dry, fried slices, alongside fried tempe
slices (Winarno et al. 1984). Oncom may also be incorporated into
sambal (a mixture of chili sauce and other spices), laksa (a coconut
milk soup) and other soups and sauces, and typical Pasundan foods
such as oncom dijero (oncom put inside grated cassava and fried).
Similar to many other fermented foods, the fermentation of oncom

is carried out using traditional methods in homes and as small-scale
industries (Figure 1.36). The raw materials for production of both
oncom merah and oncom hitam are peanut press cake, the solid resi-
due from soya bean milk and soya bean curd production, and the solid
waste from tapioca (cassava starch) manufacture. Black oncom is usu-
ally made from peanut press cake, sometimes mixed with the residue
of tapioca production (onggok cassava) to obtain a better texture and
softer product. Some black oncom is made with a mixture of coco-
nut press cake (40–60% [wet wt/wet wt]) and solid residue from soya
bean curd production (60–40%). Red oncom is usually made from
a mixture of 90% solid residue from soya bean curd production and
10% of solid waste from tapioca production (Saono et al. 1974) but in
some areas may be prepared from peanut press cake alone.
Solid waste from tapioca Peanut press cake Residue from soybean curd
production production
Soak in water for 1 h Soak in water for 3–4 h Press in a gunny sack
Wash and drain Wash and drain
Mix thoroughly Mix thoroughly
Steam for 1 h Steam for 1 h
Cool and form into flat Cool and form into flat
cakes in a wooden mold cakes in a wooden mold
Inoculate with dried red oncom Inoculate with dried black oncom
(Neurospora sp.) (Rhizopus sp.)
Place on woven bamboo trays Place on woven bamboo trays

and cover with banana leaves and cover with banana leaves
Incubate, 25–30°C, 36–48 h Incubate, 25–30°C, 36–48 h

Turn over after 24 h Turn over after 24 h
Oncom merah (red oncom) Oncom hitam (black oncom)
Figure 1.36 Flow sheet of the preparation of red oncom and black oncom in West Java. (Adapted
from Saono, S., T. Basuki and D.D. Sastraatmadji. 1996. In Handbook of Indigenous Fermented Foods,
edited by K.H. Steinkaus, p. 81. New York: Marcel Dekker.)
The raw materials are soaked, washed and steamed prior to being
formed into cakes and inoculated. The inoculum commonly used is
dried oncom that has been incubated for additional time to maxi-
mise sporulation by the mould(s). The mould grows throughout the
substrate to bind it into a porous cake with a relatively soft spongy
texture and bright orange-red or grey appearance with a pleasant and
distinctive flavour. The cakes are inverted after about 24 h incubation,
when sporulation is evident on the upper surface, so as to promote
sporulation on the under surface.
1.3.4 Microbiology
The essential moulds in red oncom are Neurospora species. Early

authors identified the species as N. sitophila (Dwidjoseputro 1961;
Saono et al. 1974) or N. crassa (Dwidjoseputro 1961) but Ho (1986),
who examined 71 Neurospora cultures from oncom, concluded that
they were N. intermedia. However, the oncom cultures differed from
wild N. intermedia strains in having bright saffron-yellow growth
and large macroconidia in contrast to the pink-orange growth and
small macroconidia of the wild strains, and Ho (1986) suggested that
the mould in oncom was a distinct form of N. intermedia and pro-
posed that it be called N. intermedia var. oncomensis. Turner (1987)
showed that the rich golden-orange or saffron-yellow strains, which
she designated ‘yellow’ ecotype, are associated with nutritionally rich
by-products of human activities, such as maize cobs, sugar refinery
waste and oncom, while the pinkish-orange or salmon-orange strains,
which she designated ‘orange ecotype’, are associated with burned
vegetation. Turner considered the two ecotypes to be conspecific and
did not propose that they be designated subspecies or varieties. It also
follows that the oncom strains should not be viewed as domesticated
strains of the fungus. Priatni et al. (2010) used molecular methods to
confirm the identity of one oncom strain as N. intermedia.
The R. oligosporus strains in black oncom appear to be similar to
those used for tempe.
Moulds are the only organisms required to make typical oncom
but yeasts and bacteria are invariably present in commercial samples
(Saono et al. 1974). Sastraatmadja et al. (2002) showed that it was
possible to make good quality oncom using single pure cultures of
a variety of moulds, including three strains of Neurospora spp, four

strains of Rhizopus spp and four strains of Mucor spp. Lactic acid
bacteria are found during the soaking of the materials or may already
be present in the wastes but there is no evidence that they play any
important role in the fermentation.
1.3.5 Biochemical Changes
There have been only a few studies of the biochemical changes occur-
ring during oncom fermentation (Gunawan et al. 1985). Peanut press
cake, a waste product of the peanut oil industry, is a highly variable
material containing 10–16% moisture, 6–20% oil, 38–51% crude pro-
tein, 14–20% carbohydrate, 5–8% crude fibre and 4–6% ash. Fresh
red and black oncom (mostly coconut and peanut press cakes) con-
tained 70% moisture, 3–9% oil, 20–30% crude protein, about 4%
carbohydrate, 2% fibre and 1% ash (Gandjar and Slamet 1972). The
main result of fermentation was the disappearance of much of the
available carbohydrate and some hydrolysis of protein and lipids.
Total protein content remained relatively constant during the fer-
mentation while the total fat content decreased slightly. Total carbo-
hydrate content also decreased from an initial 10% to 8.4% (Gandjar
and Slamet 1972).
The carbohydrates in peanut press cake consist largely of cellulose
and the oligosaccharides, sucrose, raffinose and stachyose. During the
fermentation the oligosaccharide concentrations are reduced but no
studies have been made on the mechanism(s). Probably, similar con-
siderations apply as for tempe.
The residue (okara) from the preparation of soya bean milk and soya
bean curd (tofu) is produced in huge quantities worldwide (O’ Toole
1999). Like peanut press cake, its composition is very variable and
widely different values are given by different authors. Based on the
data reviewed by O’Toole (1999), the composition of typical okara is
as follows (% dry matter): protein, 24–27; oil, 9–22; insoluble fibre,
40–58; ash, 3–3.7; monosaccharides, 0.6–0.7; sucrose, 1.3–2.3; stach-
yose, 0.9–1.4; raffinose, 0.3–0.4; starch 0.6 −0.8. However, reports
of compositions substantially different from these values were also
noted. Red oncom, made from okara, had an increased protein con-
centration (27% of dry matter compared with 22% in the okara), a
lower fat content (9% versus 15%) and a reduced insoluble fibre con-
tent (Matsuo 1997).
N. sitophila and R. oligosporus are active lipase producers, hydro-
lysing triglycerides to yield free fatty acids that accumulate to vari-
ous levels, depending on the substrate and fermentation conditions
(Fardiaz 1987). Distribution of fatty acids within the free fatty acid
fraction of fermented peanut press cake is, however, substantially dif-
ferent from that in non-fermented peanuts. The free fatty acid frac-
tion of N. sitophila ferments contained significantly higher levels of
saturated fatty acids, particularly palmitic and stearic, and a lower
level of linolenic acid. Oleic acid was essentially unchanged. These
differences were attributed to the action of 1,3-lipases, since satu-
rated acids are located primarily in the 1 and 3 positions and lin-
oleic acid is in the 2 position of peanut triglycerides (Beuchat and
Worthington 1974).
As mentioned earlier, the major benefit of the oncom fermentation

is the conversion of low value waste materials into acceptable human
foods. Growth of the mould will result in some modification of the
chemical composition but, as with tempe, the overall effect will be
a reduction in the total amount of nutrients due to processing losses
and their utilisation by the mould. However, there appears to be no
published information on the scale of these losses.
Oncom, prepared from peanut press cake or from soya bean resi-
dues, contains relatively high (20 − 30% dry matter) concentrations of
protein, reduced levels of oil and high concentrations of dietary fibre
compared with the raw peanuts and soya beans. Hence, these oncoms
constitute good and low-cost protein sources (Siswono 2002).
Peanuts and soya beans contain various anti-nutritional factors
but most of these will be reduced, if not eliminated, during pro-
cessing and fermentation. The phytic acid content of peanut press
cake was 1.4% on a dry basis, and decreased rapidly during fermen-
tation. Oncom, prepared with R. oligosporus, contained the least
amount of phytic acid (0.5%) after 72 h fermentation, while oncom
prepared with Neurospora intermedia contained 0.7%. Penta-, tetra-,
tri-, di- and mono-phosphate inositol as well as inorganic phosphate
and inositol were found in the fermented press cake (Fardiaz and
Markakis 1981).
Thiamin, riboflavin and pantothenate contents were unchanged
after fermentation from their values in the raw materials (Gandjar and
Slamet 1972; Quinn et al. 1975; van Veen and Steinkraus 1970) but
increases were observed in niacin and carotene (Gandjar and Slamet
1972; Quinn et al. 1975). Calcium concentration increased from
2.0 g kg−1 in soya bean curd waste to 2.3 g kg−1 in oncom, presumably
as a consequence of loss of material during the fermentation.
Fardiaz and Markakis (1981) evaluated the protein efficiency ratio
(PER) and the digestion coefficient of peanut press cake at 15% pro-
tein level in the diet of Wistar male rats before and after fermenta-
tion. Neither the PER (2.17) nor the digestion coefficient (~83%) were
affected by fermentation.
Red oncom, produced by fermenting defatted soya bean with
Neurospora intermedia, reduced cholesterol levels in rats, suggesting
the possibility of similar effects in humans (Setyobudi et al. 2007).
Priatni et al. (2010) studied carotenogenesis by N. intermedia N-1
in liquid culture. The maximum total production of carotenoids
was 24 µg g−1 spores. Five carotenoid compounds were identified in
spores of N. intermedia, including lycopene, neurosporen, g-carotene,
β-carotene and phytoene. Penicillin influenced carotene synthesis by
N. intermedia and particularly stimulated β-carotene biosynthesis
(maximum, 17 µg g−1 spores). Tests on mice showed that encapsu-
lated carotenoids extract from N. intermedia N–1 was non-toxic in an
acute toxicity test.
An important consideration in oncom preparation is the sanitary and

the hygienic conditions under which it is produced. In traditional
processes, poor sanitation and hygienic problems are liable to lead
to contamination with toxigenic moulds and pathogenic bacteria.
Aflatoxin-producing moulds are often naturally present on peanut
press cake and coconut press cake and when these materials are used
as substrates for oncom production a possible danger is the presence
of aflatoxin in the raw materials. Setyobudi et al. (2007) found 0.23–
5.0 mg kg−1 (dry wt) of aflatoxin B1 in samples of raw peanut press
cake. Market samples of black oncom contained aflatoxin (mostly B1)

from trace amounts to concentrations as high as 1800 mg kg−1, while
red oncom samples did not contain detectable amounts of aflatoxin
(Muhilal et al. 1970). The presence of very high levels in some sam-
ples suggests growth of aflatoxin-producing moulds in the oncom and
emphasises the importance of good sanitation and hygiene to avoid
contamination with undesirable moulds.
There are a several reports on the levels of aflatoxin being reduced
during the fermentation of oncom (Setyobudi et al. 2007; van Veen
et al. 1968). Neurospora intermedia var. oncomensis and R. oligospo-
rus reduced the aflatoxin produced by Aspergillus flavus. Van Veen
et al. (1968) found that N. sitophila (possibly N. intermedia) and R. oli-
gosporus could decrease the aflatoxin content of peanut press cake
by 50–70% during the fermentation. The moulds, R. oligosporus,
R. oryzae, Aspergillus oryzae, some osmophilic yeasts and several bac-
teria, including Flavobacterium aurantiacum, Lactobacillus spp, and
Propionibacterium spp have been reported to be capable of degrad-
ing aflatoxins (Sarjono et al. 2004; Setyobudi et al. 2007). However,
since high concentrations of bacteria are invariably present in these
mould fermentations the mechanisms of any detoxification are
unclear. Certainly, detoxification by microbes could be very impor-
tant since aflatoxins are relatively resistant to common food process-
ing treatments.
1.3.8 Industrialisation
A major problem for industrialisation of oncom production is the

relatively short shelf life of the product. Kept at room temperature,
traditional oncom lasts only for 2–3 days, albeit this is slightly lon-
ger than the shelf life of tempe. After this period, it becomes over-
fermented and the texture softens and changes in flavour occur. The
product is also susceptible to spoilage by contaminating microor-
ganisms, especially other fungi and bacteria, and good hygiene and
sanitation must be observed during processing. Currently, the pro-
cess is not mechanised and, thus, it tends to be rather labour inten-
sive. Because of the nature of the product, rough handling has to be
avoided to prevent it from becoming mushy and this also limits the
scale of production.
1.3.9 Future Work and Prospects
Oncom has received very little detailed study and there is much oppor-
tunity to gather basic information on the processes and for studies to
optimise the fermentation. Relevant questions include the following:
1. What are the nutrient losses during processing and how

might these be minimised? The soaking and washing of tapi-
oca waste and peanut press cake is liable to result in the loss of
soluble nutrients.
2. What are the chemical changes effected by the mould?
3. What determines whether Rhizopus sp. or Neurospora sp. is
the dominant organism in the fermentation?
4. Questions similar to those raised above for the tempe
fermentation.
1.4 Indonesian Dage

Dage is a Central Javan term for oncom-type products made from

a variety of food processing wastes. Most dage is made from pea-
nut press cake mixed with coconut press cake or cassava press cake
and the major mould in the fermentation is Mucor hiemalis (Kozaki
et al. 1998). Hence, this dage may be distinguished from black or red
oncom. The appearance is a greyish-white compact cake about 2–3 cm
in thickness with a bland taste.
Other ‘dages’ are fermented by Rhizopus spp (Kozaki et al. 1998)
and are evidently identical to West Javan oncom hitam. In some parts
of Indonesia dage is used to refer to quite different products made by
the fermentation of dehulled rubber seeds or legume seeds or over-fer-
mented oncom (Gandjar and Hermana 1972; Muchtadi et al. 1991).
These ‘dages’ will not be described here. The relationships between
tempe and related products, including dage, are summarised in Figure
1.37. It is evident that there is some association between the substrate
material(s) and the predominant fermenting mould, suggesting that
the properties of the substrate is a factor in selecting the prevailing
mould under the non-aseptic conditions of the fermentations.
Soybean Coconut presscake
Soaked, dehulled Soaked, pressed
Cooked, drained Steamed
Rhizopus Mucor Rhizopus
Tempe kedele Dage Tempe bongkrek
Peanut presscake
Soaked, pressed
Steamed
Mucor Rhizopus Neurospora
Dage Oncom hitam Oncom merah

(black oncom) (red oncom)
Figure 1.37 Summary of relationships between tempe and related products based on the raw
materials and the major moulds involved.
1.4.2 Places of Production
The making and consumption of dage is restricted to the areas of

Brebes, Tegal and Purwokerto in Central Java. It is consumed as a
side dish, after deep-frying, or is cooked along with other ingredients.
The method for making dage is similar to that for oncom (made in
West Java) and for ordinary tempe. Peanut press cake and coconut
press cake are soaked separately in water overnight to remove excess
oil and then washed several times. The residues are pressed to remove
water or are simply drained. The solid residues are then mixed and
sometimes the pH of the mixture is adjusted to 4.5–5.0 with vinegar.
The exact proportion of each material depends on the producer but,
generally, peanut press cake contributes the greater proportion (>50%
of the mixture). The combined material is steamed for 1–2 h. After
Peanut and coconut press cake
Soak overnight
Wash on a sieve
Drain
Steam 1–2 h
Cool
Inoculate with ragi dage
Mold into blocks, about 10 × 6 × 2.5 cm
Place on bamboo trays and cover with banana leaves
Incubate at room temperature, ~28°C, 36–48 h

turnover after 24 h
Dage
Figure 1.38 Traditional production of dage at Brebes, Central Java.
being allowed to cool, it is inoculated with ragi dage or with dage

from a previous batch. The inoculated material is formed into oblong
cakes 2–3 cm thick. The cakes are placed on bamboo trays, covered
with banana leaves and incubated at room temperature for 36–48 h
(Figure 1.38). The mould penetrates the substrate and knits the par-
ticles into a firm cake, though ones that are less firm than those of
tempe kedele (soya bean tempe). The cakes have a shelf life of 1–2 days
at ambient temperatures, similar to that of other tempe-type products
(Kozaki et al. 1996).
1.4.4 Microbiology
Kozaki et al. (1998) stated that the main microorganisms in the
dage fermentation were Mucor hiemalis and Pediococcus pentosaceus.
It has been suggested that the distinctive taste of dage is derived
from the activities of Mucor hiemalis. Lactic acid bacteria, especially
Pediococcus pentosaceus, were found in the soak water (Kozaki et al.
1998; Luwihana 1986) and, presumably, contribute to acidification of
the substrate and possibly to the flavour of dage.
1.4.5 Characteristics of the Substrates and Changes

during Processing/Fermentation
Peanut press cake from village and cottage industries contains 10–16%
moisture, 6–20% oil, 38–51% crude protein, 14–20% carbohydrate,
5–8% crude fibre and 4–6% ash. The main result of fermentation is the
disappearance of most of the available carbohydrate and some hydro-
lysis of protein and lipid. Total protein remains constant during fer-
mentation while total fat content decreases slightly. Total carbohydrate
also decreases, and the thiamin and riboflavin content before and after
fermentation remains unchanged (Gandjar and Hermana 1972).
While the waste materials used in dage production are highly variable
in composition, peanut press cake, the residue after the oil has been
expressed, has a very high protein content and dage made from it
is a correspondingly nutritious product. Coconut press cake contains
lower amounts of protein and much soluble carbohydrate (Table 1.17)
which are lost in washing the residue.
Peanut press cake from village and cottage industry contains
10–16% moisture, 6–20% oil, 38–51% crude protein, 14–20% carbo-
hydrate, 4–6% ash and 5–8% crude fibre. The main result of fermen-
tation is the disappearance of most of the available carbohydrate and
some hydrolysis of protein and lipid. Total protein remains constant
Table 1.17 Proximate Analysis of Coconut Press Cake and Dage

COMPOSITION
% WET WEIGHT a % DRY MATTER
MATERIAL COCONUT PRESS CAKE DAGE COCONUT PRESS CAKE DAGE
Water 13 77
Total sugars 35 12 41 45
Protein 19 4.6 22 18
Fat 7.8 4.1 9.0 16
Fibre 19 4.2 21 16
Ash 5.9 1.3 6.8 5.0
a Adapted from Kozaki, M., H. Iino, K.R. Kuswanto and M. Ichinoe, 1998. Bulletin of Showa University
7(3):55–63.
during fermentation while the total fat content decreases slightly.

Total carbohydrate also decreased and the thiamin and riboflavin
contents were unchanged.
The safety of a fermented food depends on the composition of raw

materials and the microorganisms that are involved in the fermen-
tation. These should not contain compounds that could disturb the
health of consumers and the microbes, whether they are required for
the process or are present as contaminants, should not contain cellular
components or produce compounds which are injurious to the con-
sumer’s health. Safety considerations with dage are essentially similar
to those for tempe and oncom.
A concern with all tempe-type products made from coconut pro-
cessing wastes is the possibility of bongkrek poisoning. Bongkrek
poisoning is a consequence of contamination by and growth of
B. cocovenenans and the production of the toxins, bongkrek acid
and toxoflavin. Tempe bongkrek is made from home-made, freshly
prepared coconut residue (from the wet processing of coconut meat
to make coconut milk) and it is this product that was, historically,
associated with bongkrek poisoning. Oncom/dage made from dried
coconut press cake, usually from the coconut oil industry, or from
mixtures of peanut press cake and coconut press cake has not been
responsible for bongkrek poisonings.
Because of the limited number of customers, dage is produced only by

small-scale family producers or by families for their own consumption.
However, possibilities exist for the use of specific Mucor hiemalis strains
to improve dage production and to produce dage with specific properties
which are superior to those of dage produced by natural fermentation.
References
Abe, A., Y. Oda, K. Asano and T. Sone. 2007. Rhizopus delemar is the proper
name for Rhizopus oryzae fumaric-malic acid producers. Mycologia 99
(5):714–722.
Agranoff, J. ed. 2001. The Complete Handbook of Tempe. Second ed. Singapore:
American Soybean Association.
Anonymous. 1995. Widyakarya Nasional Makanan Tradisional. Jakarta:
Kementrian Negara Urusan Pangan. (In Indonesian).
Anonymous. 1996. Proceedings 2nd Soybean Processing and Utilization Conference.
Bangkok, 8–13 January 1996.
Anonymous. 2010. Report of the Department for Trade and Industry,
Yogyakarta Province, Indonesia.
Ariani, S.R.D. and S. Handayani. 2009. Pengembangan produk tempe generasi
ketiga berkhasiat antioksi dan berbahan baku koro benguk (Mucuna pru-
riens L.D.C. var. utilis). Report. Surakarta: Univeritas Sebelas Maret. (In
Indonesian).
Ashenafi, M. and M. Busse. 1991a. Growth of Bacillus cereus in fermenting
tempeh made from various beans and its inhibition by Lactobacillus plan-
tarum. Journal of Applied Bacteriology 70 (4):329–333.
Ashenafi, M. and M. Busse. 1991b. Growth potential of Salmonella infantis
and Escherichia coli in fermenting tempeh made from horsebean, pea and
chickpea and their inhibition by Lactobacillus plantarum. Journal of the
Science of Food and Agriculture 55 (4):607–615.
Astadi, I.R., M. Astuti, U. Santosa and P.S. Nugraheni. 2009. In vitro anti-
oxidant activity of anthocyanin of black soybean seed coat in human low
density lipoprotein. Food Chemistry 112:659–663.
Astuti, M. 1982. Tempe as protein source in baby food formula. Report, Gadjah
Mada University, Yogjakarta, Indonesia.
Astuti, M. 1998. Soy and heart disease—effects independent of cholesterol
reduction: The role of tempe on lipid profile and lipid peroxidation
(Abstract). American Journal of Clinical Nutrition 68(6S):1522S–1523.
Astuti, M. 1992. Iron bioavailability of traditional Indonesian soybean tempe.
Dissertation, Tokyo University, Tokyo, Japan.
Autret, M. and A.G. van Veen. 1955. Possible sources of protein for child feed-
ing in underdeveloped countries. American Journal of Clinical Nutrition
3:234–243.
Bai, R.G., T.N. Prabha, T.N. Ramachandra Rao, V.P. Sreedhara and N.
Sreedhara. 1975. Studies on tempeh. Part I. Processing and nutritional
evaluation of tempe from a mixture of soybean and groundnut. Journal of
Food Science and Technology 12:135–138.
Balk, E., M. Chung, P. Chew et al. 2005. Effects of Soy on health outcomes.
Evidence Report/Technology Assessment No. 126. Rockville, MD:
Agency for Healthcare Research and Quality.
Baumann, U. and B. Bisping. 1995. Proteolysis during tempe fermentation.
Food Microbiology 12:39–47.
Belitz, H.-D., W. Grosch and P. Schieberle. 2009. Food Chemistry. Fourth ed.
Berlin: Springer-Verlag.
Benny, G. 2009. Zygomycetes. www.zygomycetes.org (accessed 26 April 2010).
Berghofer, E. and A. Werzer. 1986. Herstellung von tempeh mit einheimischen
bohnen. Chemie, Mikrobiologie, Technologie de Lebensmittel 10:54–62. (In
German).
Beuchat, L.R. and R.E. Worthington. 1974. Changes in the lipid content of
fermented peanuts. Journal of Agricultural and Food Chemistry 22:509–512.
Bewley, J.D. and M. Black. 1994. Seeds Physiology of Development and
Germination. Second ed. New York: Plenum Press.
Buckle, K.A. and E. Kartadarma. 1990. Inhibition of bongkrek acid and toxo-
flavin production in tempe bongkrek containing Pseudomonas cocoven-
enans. Journal of Applied Bacteriology 68 (6):571–576.
Casey, A. and G. Walsh. 2004. Identification and characterization of a
phytase of potential commercial interest. Journal of Biotechnology 110
(3):313–322.
Centraalbureau voor Schimmelcultures. n.d. CBS-KNAW Fungal Biodiversity
Centre. www.cbs.knaw.nl/Collections/Biolomics.aspx?Table=CBS±strain±
database (accessed 28 January 2014).
Central Bureau of Statistics. 2012. Food Balance Sheet. Jakarta: Central Bureau
of Statistics.
Charles, M. and J.R. Gavin. 1996. Engineering studies of tempe fermenta-
tion. In Handbook of Indigenous Fermented Foods, 1996, edited by K. H.
Chen, C.C., Y.C. Cho, C.C. Lai and W.H. Hsu. 2009. Purification and char-
acterization of a new rhizopuspepsin from Rhizopus oryzae NBRC 4749.
Journal of Agricultural and Food Chemistry 57 (15):6742–6747.
Chen, L.Y., X.M. Zhong and L. Li. 2009. Inhibition effects of lactic acid on the
growth of Bacillus cereus. Modern Food Science and Technology 25 (7):756–
759, 854.
Crittenden, R.G. and M.J. Playne. 1996. Production, properties and applica-
tions of food-grade oligosaccharides. Trends in Food Science and Technology
7:353–361.
Delaney, R., R.N.S. Wong, G. Meng, N. Wu and J. Tang. 1987. Amino acid
sequence of rhizouspepsin pI 5. Journal of Biological Chemistry 262
(4):1461–1467.
Departemen Kesehatan. 2009. Daftar Komposis Bahan Makanan. Jakarta:
Departemen Kesehatan. (In Indonesian).
Depkes, R.I. 2006. Daftar komposisi bahan makanan. Penerbit Bthara. (In
Indonesian).
De Reu, J.C., A.M. Griffiths, F.M. Rombouts and M.J.R. Nout. 1995. Effect
of oxygen and carbon dioxide on germination and growth of Rhizopus
oligosporus on model media and soya beans. Applied Microbiology and
Biotechnology 43:908–913.
Djurtoft, R. and J.P. Nielsen. 1983. Increase in some B vitamins, including B12,
during fermentation of tempeh, produced from cowpeas or soy beans.
Journal of Plant Foods 5 (3):135–141.
Dwidjoseputro, D. 1961. Studies on Monilia sitophila from Indonesia. Bulletin
of the Torrey Botanical Club 88:404–411.
Ebata, J., Y. Fukuda, K. Hirai and K. Murata. 1972. β-glucosidase involved in
the antioxidant formation in tempeh, fermented soybeans. Journal of the
Agricultural Chemistry Society of Japan 46:323–329.
Efriwati, A. Suwanto, G. Rahayu and L. Nuraida. 2013. Population dynam-

ics of yeasts and lactic acid bacteria (LAB) during tempeh production.
HAYATI Journal of Biosciences 2:57–64.
Egounlety, M. and O.C. Aworh. 2003. Effect of soaking, dehulling, cooking
and fermentation with Rhizopus oligosporus on the oligosaccharides, tryp-
sin inhibitor, phytic acid and tannins of soybean (Glycine max Merr.),
cowpea (Vigna unguiculata L. Walp) and groundbean (Macrotyloma geo-
carpa Harms). Journal of Food Engineering 56 (2/3):249–254.
Ekanayake, S., K. Skog and N.-G. Asp. 2007. Canavanine content in sword
beans (Canavalia gladiata): Analysis and effect of processing. Food and
Chemical Toxicology 45:797–803.
Fardiaz, D. 1987. Oncom fermented peanut presscake, a unique Indonesian
traditional fermented food. In Traditional Fermented Foods and their
Processing in Asia, edited by F. Yanagida, Y. Takai, S. Homma, S. Kato and
Y. Ando. Tokyo: Nodai Research Institute, Japan.
Fardiaz, D. and P. Markakis. 1981. Oligosaccharides and protein efficiency
ratio of oncom (fermented peanut press cake). Journal of Food Science
46:1970–1971.
Farley, P.C. and L. Ikasari. 1992. Regulation of the secretion of Rhizopus oli-
gosporus extracellular carboxyl proteinase. Journal of General Microbiology
138:2539–2544.
Faughnan, M.S., A. Hawdon, E. Ah-Singh, J. Brown, D.J. Millward and A.
Cassidy. 2004. Urinary isoflavone kinetics: The effect of age, gender,
food matrix and chemical composition. British Journal of Nutrition 91
(4):567–574.
Feng, X.M., A.R.B. Eriksson and J. Schnürer. 2005. Growth of lactic acid
bacteria and Rhizopus oligosporus during barley tempeh fermentation.
International Journal of Food Microbiology 104 (3):249–256.
Feng, X.M., V. Passoth, C. Eklund-Jonsson, M.L. Alminger and J. Schnürer.
2007. Rhizopus oligosporus and yeast co-cultivation during barley tempeh
fermentation—Nutritional impact and real-time PCR quantification of
fungal growth dynamics. Food Microbiology 24 (4):393–402.
Fenwick, D.E. and D. Oakenfull. 1983. Saponin content of food plants and
some prepared foods. Journal of the Science of Food and Agriculture 34
(2):186–191.
Gandjar, I. and D.S. Slamet. 1972. Tempe gembus hasil fermentasi ampas tahu.
Penelitian Gizi dan Makanan (Bogor) 2:70–79. (In Indonesian).
Gandjar, I. and Hermana. 1972. Some Indonesian fermented foods from waste
products. In Waste Recovery by Microorganisms, edited by W.R. Stanton,
pp. 49–54. Kuala Lumpur: University of Malaya.
Garcia, R.A., J.H. Hotchkiss and K.H. Steinkraus. 1999. The effect of lip-
ids on bongkrekic (Bongkrek) acid toxin production by Burkholderia
cocovenenans in coconut media. Food Additives and Contaminants 16
(2):63–69.
Graffham, A.J., M.H. Gordon, A. Westby and J.D. Owens. 1995. Nutrition of
tempe moulds. Letters in Applied Microbiology 21:223–227.
Graham, D.C.W., K.H. Steinkraus and L.R. Hackler. 1976. Factors affecting
production of mold mycelium and protein in synthetic media. Applied and
Environmental Microbiology 32:381–387.
Gunawan, C., L. Sutedja and Roestamsyah. 1985. Biochemical changes dur-
ing oncom fermentation utilizing mixed substrates. ASEAN Food Journal
1:128–133.
György, P., K. Murata and H. Ikehata. 1964. Antioxidants isolated from fer-
mented soybeans (tempeh). Nature 230:870–871.
Haas, M.J., D.G. Bailey, W. Baker, T.R. et al. 1999. Biochemical and molecular
biological characterization of a lipase produced by the fungus Rhizopus
delemar. Fett/Lipid 101 (10):364–370.
Haas, M.J. and R.D. Joerger. 1995. Lipases of the genera Rhizopus and
Rhizomucor: versatile catalysts in nature and the laboratory. In
Food Biotechnology: Microorganisms, edited by Y.H. Hui and G.G.
Khachatourians. New York: VCH Publishers.
Hackler, L.R., K.H. Steinkraus, J.P. van Buren and D.B. Hand. 1964. Studies
on the utilization of tempe by weanling rats. Journal of Nutrition 82:
452–456.
Handayani, S., 1997. Bioactive compound of Mucuna tempe. Proceedings
International Tempe Symposium, Bali, Indonesia, July 13–15.
Herastuti, S.R., B. Subardjo and M. Astuti. 1997. Development of superoxide
dismutase in pigeon pea tempe fermentation. Proceedings International
Tempe Symposium, Bali, Indonesia, July 13–15.
Hermana. 1985. Pengaruh konsumsi bahan makanan campuran dengan keda-
lai atau tempe terhadap anak batita pendirita kurang kalori-protein.
Presented at Seminar Nasional Hasil Penelitian Perguruan, Bandung,
25–28 February. (In Indonesian).
Hesseltine, C.W., C.L. Featherson, G.L. Lombard and V.R. Dowell. 1985.
Anaerobic growth of molds isolated from fermentation starters used for
foods in Asian countries. Mycologia 77:390–400.
Hesseltine, C.W., M. Smith, B. Bradle and S.D. Ho. 1963a. Investigations
of tempeh, an Indonesian food. Developments in Industrial Microbiology
4:275–287.
Hesseltine, C.W., R. De Camargo and J.J. Rackis. 1963b. A mould inhibitor in
soybeans. Nature 200 (4912):1226–1227.
Hill, A.M., H.I. Katcher, B.D. Flickinger, and P.M. Kris-Etherton. 2008.
Human nutrition value of soybean oil and soy protein. In Soybeans
Chemistry, Processing and Utilization, edited by L.A. Johnson, P.J. White
and R. Galloway. Urbana, IL: AOCS Press.
Ho, C.C. 1986. Identity and characteristics of Neurospora intermedia respon-
sible for oncom fermentation in Indonesia. Food Microbiology 3(2):
115–132.
Hocking, A.D. and B.F. Miscamble. 1995. Water relations of some Zygomycetes
isolated from food. Mycological Research 99:1113–1118.
Horiuchi, H., K.T. Yanai, T. Okazaki and K. Yano. 1988. Isolation and sequenc-
ing of a genomic clone encoding aspartic proteinase of Rhizopus niveus.
Journal of Bacteriology 170:272–278.
Jansen, B.C.P. and W.F. Donath. 1924. Metabolic studies on rats and digest-
ibility of the proteins of some foodstuffs. Mededelingen Burgerlijken
Geneeskundigen Dienst Ned.-Indie 1924:24–25. Cited by Steinkraus 1996.
Jennessen, J., K.F. Nielsen, J. Houbraken et al. 2005. Secondary metabolite
and mycotoxin production by the Rhizopus microsporus group. Journal of
Agricultural and Food Chemistry 53 (5):1833–1840.
Jennessen, J., J. Schnurer, J. Olsson, R.A. Samson, and J. Dijksterhuis. 2008.
Morphological characteristics of sporangiospores of the tempe fungus
Rhizopus oligosporus differentiate it from other taxa of the R. microsporus
group. Mycological Research 112 (5):547–563.
Jeong, Y., J. Kim, S. Kim, Y. Kang, T. Nagamatsu and I. Hwang. 2003.
Toxoflavin produced by Burkholderia glumae causing rice grain rot is
responsible for inducing bacterial wilt in many field crops. Plant Disease
87 (8):890–895.
Jha, H.C., S. Kiriakidis, M. Hoppe and H. Egge. 1997. Antioxidative con-
stituents of tempe. Proceedings of the International Tempe Symposium, July
13–15, Bali, Indonesia.
Jurus, A.M. and W.J. Sundberg. 1976. Penetration of Rhizopus oligosporus
into soybeans in tempeh. Applied and Environmental Microbiology 32:
284–287.
Kamajaya. 1986. Centhini Tambangraras-Amongraga, Vols. 3 and 12. Yogyakarta:
Penerbityayasan Centhini. (In Indonesian).
Kao, C. and R.J. Robinson. 1978. Nutritional aspects of fermented foods from
chickpea, horsebean, and soybean. Cereal Chemistry 55 (4):512–517.
Kasmidjo. 1990. Sekilas tentang potensi dan permasalahan tempe sebagai pan-
gan sumber protein nabati. Agrita 4:13–16. (In Indonesian).
Kawamura, S. 1967. Quantitative paper chromatography of sugars of the
cotyledon, hull and hypocotyl of soybeans of selected varieties. Kagawa
Daigaku Nogakubu Gakuzyutu Hokoku (Technical Bulletin of the Faculty
of Agriculture, Kagawa University) 18:118–131. (In Japanese). Cited by
Wolf and Cowan (1971) and Liu (1997).
Kelly, C. 2002. Soya and Health. London: The British Nutrition Foundation.
Keuth, S. and B. Bisping. 1993. Formation of vitamins by pure cultures of
tempe moulds and bacteria during the tempe solid substrate fermenta-
tion. Journal of Applied Bacteriology 75:427–434.
Kito, H., A. Abe, I.N. Sujaya, Y. Oda, K. Asano and T. Sone. 2009. Molecular
characterization of the relationships among Amylomyces rouxii, Rhizopus
oryzae, and Rhizopus delemar. Bioscience, Biotechnology and Biochemistry 73
(4):861–864.
Klus, K., G. Borger-Papendorf and W. Barz. 1993. Formation of 6,7,4′-tri-
hydroxyisoflavone (factor 2) from soybean seed isoflavones by bacteria
isolated from tempe. Phytochemistry 34 (4):979–981.
Ko, S.D. and C.W. Hesseltine. 1979. Tempe and related foods. In Economic
Microbiology, edited by A.H. Rose. London: Academic Press.
Kozaki, M., H. Iino and K.R. Kuswanto, 1996. Indonesian fermented food,
dage and its making. Bulletin of the Doctoral Course of Science for Living
System. Showa Women’s University 5:63–68.
10 0 Indigenous Fermented Foods of Southeast Asia
Kozaki, M., H. Iino, K.R. Kuswanto and M. Ichinoe, 1998. Microorganisms

of indigenous fermented food ‘dage’ in Indonesia. Bulletin of Showa
University 7(3):55–63.
Krusong, W., B. Yongsmith and P.C. Sanchez. 1991. Note: increased vitamin
B12 level in a modified tempeh process. The Philippine Agriculturist 74
(1):89–94.
Kuswanto, K.R. 1988. Studies on the semayi Indonesian traditional fermented
food. Memoirs of the Tokyo University of Agriculture 30:189–287.
Lackner, G., N. Möbius, K. Scherlach et al. 2009. Global distribution and
evolution of a toxigenic Burkholderia–Rhizopus symbiosis. Applied and
Environmental Microbiology 75:2982–2986.
Liem, T.H.I., K.H. Steinkraus and T.C. Cronk. 1977. Production of vitamin
B12 in tempe, a fermented soybean food. Applied and Environmental
Liou, G.Y., S.R. Chen, Y.H. Wei et al. 2007. Polyphasic approach to the taxon-
omy of the Rhizopus stolonifer group. Mycological Research 111 (2):196–203.
Liu, K. 1997. Soybeans: Chemistry, Technology and Utilization. New York:
Chapman & Hall.
Liu, X.-y., H. Huang and R.-y. Zheng. 2007. Molecular phylogenetic relation-
ships within Rhizopus based on combined analyses of ITS rDNA and
pyrG gene sequences. Sydowia 59 (2):235–253.
Liu, X.-y., H. Huang and R.-y. Zheng. 2008. Delimitation of Rhizopus varieties
based on IGS rDNA sequences. Sydowia 60 (1):93–112.
Luwihana. 1986. Microflora of dage, an Indonesian traditional fermented food
from peanut and coconut press cake. Thesis, Department of Food Science,
Gadjah Mada University, Yogjakarta, Indonesia.
Lynch, K.H. and J.J. Dennis. 2010. Burholderia. In Molecular Detection of
Foodborne Pathogens, edited by Y.C. Dong, pp. 331–344. Boca Raton:
CRC Press.
Manachini, P.L., M.G. Fortina and C. Parini. 1987. Purification and prop-
erties of an endopolygalacturonase produced by Rhizopus stolonifer.
Biotechnology Letters 9:219–224.
Matsuo, M. 1997. Preparation and components of Okara-ontjom, a traditional
Indonesian fermented food. Nippon Shokuhin Kagaku Kogaku Kaishi 44
(9):632–639. Abstract.
Matsuura, M. and A. Obata. 1993. β-glucosidases from soybeans hydrolyse
daidzin and genistin. Journal of Food Science 58:144–147.
Matsuura, M., A. Obata and D. Fukushima. 1969. Objectionable flavor of soy
milk developed during the soaking of soybeans and its control. Journal of
Food Science 54 (3):602–605.
Mayer, A.M. and A. Poljakoff-Mayber. 1989. The Germination of Seeds. Fourth
ed. Oxford: Pergamon Press.
Mertens, J.A., R.C. Burdick and A P. Rooney. 2008. Identification, biochemical
characterization, and evolution of the Rhizopus oryzae 99–880 polyga-
lacturonase gene family. Fungal Genetics and Biology 45 (12):1616–1624.
Mitchell, D.B., K. Vogel, B.J. Weimann, L. Pasamontes and A.P.G.M. van
Loon. 1997. The phytase subfamily of histidine acid phosphatases:
Isolation of genes for two novel phytases from the fungi Aspergillus terreus
and Myceliophthora thermophila. Microbiology 143 (1):245–252.
Muchtadi, T.R., T.B. Noor, D. Muchtadi and N. Andarwulan. 1991. Studies
on ‘dage’, an Indonesian traditional food derived from fermented Hevea
(Hevea brasiliensis L.) seeds. Chemical and microbiological changes dur-
ing fermentation. Indonesian Journal of Tropical Agriculture 2:85–89. Cited
by K.H. Steinkraus 1996, p. 336.
Muhilal, D., Karjadi and D.D. Prawiranegara. 1970. A study on aflatoxin con-
tents of peanuts and peanut products. Gizi Indonesia 2:162.
Mulyowidarso, R.K., G.H. Fleet and K.A. Buckle. 1989. The microbial eclogy
of soybean soaking for tempe production. International Journal of Food
Mulyowidarso, R.K., G.H. Fleet and K.A. Buckle. 1990. Association of bac-
teria with the fungal fermentation of soybean tempe. Journal of Applied
Bacteriology 68 (1):43–47.
Mulyowidarso, R.K., G.H. Fleet and K.A. Buckle. 1991a. Changes in the
concentration of carbohydrates during the soaking of soybeans for
tempe production. International Journal of Food Science and Technology
26:595–606.
Mulyowidarso, R.K., G.H. Fleet and K.A. Buckle. 1991b. Changes in the
concentration of organic acids during the soaking of soybeans for
tempe production. International Journal of Food Science and Technology
26:607–614.
Murata, K., H. Ikehata and T. Miyamoto. 1967. Studies on the nutritional value
of tempeh. Journal of Food Science 32:580–586.
Murphy, P.A. 2008. Soybean proteins. In Soybeans: Chemistry, Production,
Processing, and Utilization., edited by L.A. Johnson, P.J. White and R.
Galloway. Urbana, IL: AOCS Press.
Nahas, E. 1988. Control of lipase production by Rhizopus oligosporus under
various growth conditions. Journal of General Microbiology 134:227–233.
Nakajima, N., N. Nozaki, K. Ishihara, A. Ishikawa and H. Tsuji. 2005. Analysis
of isoflavone content in tempeh, a fermented soybean, and preparation of
a new isoflavone-enriched tempeh. Journal of Bioscience and Bioengineering
100 (6):685–687.
National Population and Family Planning Board. 2013. Indonesia Demographic
And Health Survey 2012. Jakarta: Ministry of Health. (In Indonesian).
Nout, M.J.R. and F.M. Rombouts. 1990. Recent developments in tempe
research. Journal of Applied Bacteriology 69:609–633.
Nout, M.J.R., G. Beernik and T.M.G. Bonants-van-Laarhoven. 1987a.
Growth of Bacillus cereus in soyabean tempeh. International Journal of
Nout, M.J.R. and J.L. Kiers. 2005. Tempe fermentation, innovation and func-
tionality: update into the third millenium. Journal of Applied Microbiology
98:789–805.
Nout, M.J.R., M.A. de Dreu, A.M. Zuurbier and T.M.G. Bonants-van-
Laarhoven. 1987b. Ecology of controlled soybean acidification for tempe
manufacture. Food Microbiology 4:165–172.
Nout, M.J.R., M.M.W. Ruikes and H.M. Bouwmeester. 1993. Effect of pro-
cessing conditions on the formation of biogenic amines and ethyl carba-
mate in soybean tempe. Journal of Food Safety 13:293–303.
Nout, M.J.R., S. Notermans and F.M. Rombouts. 1988. Effect of environ-
mental conditions during soya-bean fermentation on the growth of
Staphylococcus aureus and production and thermal stability of enterotoxins
A and B. International Journal of Food Microbiology 7 (4):299–309.
Nout, M.J.R., T.M.G. Bonants-van Laarhoven, R. de Dreu and I.A.G.M.
Gerats. 1985. The influence of some process variables and storage con-
ditions on the quality and shelf-life of soybean tempeh. Antonie van
Leeuwenhoek 51:532–534.
Nout, M.J.R., T.M.G. Bonants-van Laarhoven, P. de Jongh and P.G. de Koster.
1987c. Ergosterol content of Rhizopus oligosporus NRRL 5905 grown
in liquid and solid substrates. Applied Microbiology and Biotechnology
26:456–461.
Nurrahman, M. Astuti, Suparmo and M.N.H.E. Soesatyo. 2011. The effect of
black soybeans tempe and its ethanol extract on lymphocyte proliferation
and Ig A secretion in Salmonella typimurium induced rat. African Journal
of Food Science 5(14):775–779.
Nurrahman, M. Astuti, Suparmo and M.N.H.E. Soesatyo. 2012. Role of black
soybean tempe to increase activity antioxidant enzyme and lymphocyte
resistance to hydrogen peroxide, in vivo. Seminar Nasional Hasil Penelitian
2012, Universitas Muhammadiyah Semarang, Indonesia. (In Indonesian).
Ogawa, Y., S. Tokumasu and K. Tubaki. 2004. An original habitat of tempeh
molds. Mycoscience 45:271–276.
Okada, N. 1989. Role of microorganisms in tempe manufacture; isolation of
Vitamin B12 producing bacteria. Japanese Agricultural Research Quarterly
22:310–316.
O’Toole, D.K. 1999. Characteristics and use of okara, the soybean residue from
soy milk production—A review. Journal of Agricultural and Food Chemistry
47 (2):363–371.
Partida-Martinez, L.P., C.F. de Looss, K. Ishida et al. 2007. Rhizonin, the
first mycotoxin isolated from the zygomycota, is not a fungal metabolite
but is produced by bacterial endosymbionts. Applied and Environmental
Microbiology 73 (3):793–797.
Partida-Martinez, L.P. and C. Hertweck. 2005. Pathogenic fungus harbours
endosymbiotic bacteria for toxin production. Nature 437:884–888.
Penaloza, W., C.L. Davey, J.N. Hedger and D.B. Kell. 1991. Stimulation by
potassium ions of the growth of Rhizopus oligosporus during liquid-
and solid-substrate fermentations. World Journal of Microbiology and
Penaloza, W., C.L. Davey, J.N. Hedger and D.B. Kell. 1992b. Physiological
studies on the solid-state quinoa tempe fermentation, using on-line mea-
surements of fungal biomass production. Journal of the Science of Food and
Agriculture 59 (2):227–235.
Penaloza, W., J.N. Hedger and B.J. Thomas. 1992a. Low-temperature scan-
ning electron microscopy of hyphal colonisation of Chenopodium quinoa
T em p e a n d Rel at ed P r o d u c t s 10 3
during fermentation with Rhizopus oligosporus. Annals of Applied Biology

121:559–563.
Pitt, J.I. and A.D. Hocking. 1997. Fungi and Food Spoilage. Second ed. London:
Blackie Academic.
Poolman, B., D. Molenaar, E.J. Smid et al. 1991. Malolactic fermentation: elec-
trogenic malate uptake and malate/lactate antiport generate metabolic
energy. Journal of Bacteriology 173:6030–6037.
Priatni, S., M. Singgih, L.B.S. Kardono and T. Gusdinar. 2010. Molecular iden-
tification of Neurospora sp N-1 isolated from Indonesian red fermented
cake. Journal of Pure and Applied Microbiology 4 (2):527–531.
Prijono. 1930. Sri Tanjung, een Oud-javaansch Verhaal. The Hague: V. de
Nederlansche Boekensteendrukkerij.
Quinn, M.R., L.R. Beuchat, J. Miller, C.T. Young and R.E. Worthington. 1975.
Fungal fermentation of peanut flour: Effect on chemical composition and
nutritive value. Journal of Food Science 40:470–474.
Raffles, T.S. 1817 and 1830. The History of Java. London: John Murray.
Rehms, H. and W. Barz. 1995. Degradation of stachyose, raffinose, melibi-
ose and sucrose by different tempe-producing Rhizopus fungi. Applied
Microbiology and Biotechnology 44 (1/2):47–52.
Report on Basic Health Research. 2010. Jakarta: Ministry of Health. (In Indonesian).
Richardson, T. and D.B. Hyslop. 1985. Enzymes. In Food Chemistry, edited by
O.R. Fennema. New York: Marcel Dekker.
Riet, W.B. van der, A.W. Wight, J.J.L. Cilliers and J.M. Datel. 1987. Food
chemical analysis of tempeh prepared from South African-grown soy-
beans. Food Chemistry 25 (3):197–206.
Roelofsen, P.A. and A. Talens. 1964. Changes in some B vitamins during mold-
ing of soybeans by Rhizopus oryzae in the production of tempeh kedelee.
Journal of Food Science 29:224–226.
Rohm, B., K. Scherlach, N. Mobius, L. P. Partida-Martinez and C. Hertweck.
2010. Toxin production by bacterial endosymbionts of a Rhizopus
microsporus strain used for tempe/sufu processing. International Journal of
Food Microbiology 136 (3):368–371.
Ruiz-Terán, F. 1995. Studies on chemical changes in tempe. PhD thesis,
Department of Food Science and Technology, University of Reading,
Reading, UK.
Ruiz-Terán, F. and J.D. Owens. 1996a. Sterilization of soybean cotyledons by
boiling in lactic acid solution. Letters in Applied Microbiology 22:30–33.
Ruiz-Terán, F. and J.D. Owens. 1996b. Chemical and enzymic changes during
the fermentation of bacteria-free soya bean tempe. Journal of the Science of
Food and Agriculture 71:523–530.
Ruiz-Terán, F. and J.D. Owens. 1999. Fate of oligosaccharides during produc-
tion of soya bean tempe. Journal of the Science of Food and Agriculture 79
(2):249–252.
Rumphius, G.E. 1747. Herbarium Amboinense. Vol. 5, pp. 388–389. Amsterdam:
Joannis Burmanni.
Ryan, F.J., G.W. Beadle and E.L. Tatum. 1943. The tube method of measuring
the growth rate of Neurospora. American Journal of Botany 30:784–799.
Samson, R.A., J.A. van Kooij and E. de Boer. 1987. Microbiological quality of
commercial tempeh in the Netherlands. Journal of Food Protection 50:92–94.
Saono, S., I. Gandjar and H. Karsono. 1974. Mycoflora of ‘ragi’ and some other
traditional fermented foods of Indonesia. Annales Bogorienses (Bogor)
V:187–204.
Saono, S., R.R. Hull and B. Dhamcharee. 1986. A Concise Handbook of Indigenous
Fermented Food in the Asia Countries. Jakarta, Indonesia: The Indonesian
Institute of Science.
Saono, S., T. Basuki and D.D. Sastraatmadji. 1996. In Handbook of Indigenous
Fermented Foods, edited by K.H. Steinkaus, p. 81. New York: Marcel
Dekker.
Sarjono, E.S. Rahayu, S. Raharjo and K.R. Kuswanto. 2004. Detoxication
of aflatoxin B1 by extracellular enzymes of Aspergillus oryzae KKB4.
Indonesian Food and Nutrition Progress 11(1):30–34.
Sarrette, M., M.J.R. Nout, P. Gervais and F.M. Rombouts. 1992. Effect of
water activity on production and activity of Rhizopus oligosporus polysac-
charidases. Applied Microbiology and Biotechnology 37:420–425.
Sastraatmadja, D.D., F. Tomita and T. Kasai. 2002. Production of high-qual-
ity oncom, a traditional Indonesian fermented food by the inoculating
with selected mould strains in the form of pure culture and solid inocu-
lums. Journal of the Graduate School of Hokkaido Agricultural University
70(2):111–127.
Schipper, M.A.A. 1984. A revision of the genus Rhizopus. I. The Rhizopus sto-
lonifera group and Rhizopus oryzae. Studies in Mycology 25:1–19.
Schipper, M.A.A. and J.A. Stalpers. 1984. A revision of the genus Rhizopus. II.
The Rhizopus microsporus group. Studies in Mycology 25:20–33.
Seaby, D.A., A.R. McCracken and J.P. Blakeman. 1988. Experimental deter-
mination of requirements for the growth of edible Rhizopus species for
use in solid substrate fermentation systems. Journal of the Science of Food
and Agriculture 44:289–299.
Seif, E., J. Leigh, Y. Liu, I. Roewer, L. Forget and B.F. Lang. 2005. Comparative
mitochondrial genomics in zygomycetes: bacteria-like RNase P RNAs,
mobile elements and a close source of the group I intron invasion in
angiosperms. Nucleic Acids Research 33:734–744.
Setyobudi, F.M.C., I. Sardjono and J. Bohm. 2007. Degradation of aflatoxin
B1 by Neurospora sp. isolated from oncom. 29th Mycotoxin Workshop, May
14–17, Stuttgart, Germany.
Seumahu, C.A., A. Suwanto, I. Rusmana and D.D. Solihin. 2013. Bacterial
communities in tempeh as revealed by amplified ribosomal intergenic
sequence analysis. HAYATI Journal of Biosciences 2:65–71.
Shurtleff. W and A. Aoyagi. 1985. The Book of Tempeh. Second ed. New York:
Harper & Row.
Siddhuraju, P. and K. Becker. 2001. Species/variety differences in biochemi-
cal composition and nutritional values of Indian tribal legumes of genus
Canavalia. Nahrung/Food 45:224–233.
Singleton, P. and D. Sainsbury. 2001. Dictionary of Microbiology and Molecular
Biology. Chicester: John Wiley.
Siswono. 2002. Oncom menutup kekurangan energy protein. Gizinet Jakarta,

4 pages. (In Indonesian).
Smith, A.K., J.J. Rackis, C.W. Hesseltine, M. Smith, D.J. Robbins and A.N.
Booth. 1964. Tempeh: Nutritive value in relation to processing. Cereal
Chemistry 41:173–181.
Soemantri, A.G. and I. Sudigbia. 1985. Management of chronic diarrhoea with
tempe-based formula. National Symposium of Tempe, Jakarta, April 15–16.
Sorenson, W.G. and C.W. Hesseltine. 1966. Carbon and nitrogen utilization
by Rhizopus oligosporus. Mycologia 58:681–689.
Sparringa, R.A. 1999. Growth and Protein Utilisation by Rhizopus oligosporus
during tempe fermentation. PhD thesis, Department of Food Science
and Technology, University of Reading, UK.
Sparringa, R.A. and J.D. Owens. 1999a. Inhibition of the tempe mould, Rhizopus
oligosporus, by ammonia. Letters in Applied Microbiology 29:93–96.
Sparringa, R.A. and J.D. Owens. 1999b. Causes of alkalinization in tempe solid
substrate fermentation. Enzyme and Microbial Technology 25:677–681.
Sparringa, R.A. and J.D. Owens. 1999c. Glucosamine content of tempe mould.
Rhizopus oligosporus. International Journal of Food Microbiology 47:153–157.
Sparringa, R.A., and J.D. Owens. 1999d. Protein utilization during soy-
bean tempe fermentation. Journal of Agricultural and Food Chemistry
47:4375–4378.
Sparringa, R.A., M. Kendall, A. Westby and J.D. Owens. 2002. Effects of tem-
perature, pH, water activity and CO2 concentration on growth of Rhizopus
oligosporus NRRL 2710. Journal of Applied Microbiology 92:329–337.
Starrenburg, M.J.C. and J.A.D. Hugenholtz. 1991. Citrate fermentation by
Lactococcus and Leuconostoc spp. Applied and Environmental Microbiology
57:3535–3540.
Steinkraus, K.H., ed. 1996. Handbook of Indigenous Fermented Foods. Second ed.
New York: Marcel Dekker.
Steinkraus, K.H., J.P. van Buren, L.R. Hackler and D.B. Hand. 1965. A pilot-
plant process for the production of dehydrated tempeh. Food Technology
19:63–68.
Steinkraus, K.H., Y.B. Hwa, J.P. van Buren, M.I. Provvidenti and D.B. Hand.
1960. Studies on tempeh. An Indonesian fermented soybean food. Food
Research 25:777–788.
Stillings, B.R. and L.R. Hackler. 1965. Amino acid studies on the effect of
fermentation time and heat-processing of tempe. Journal of Food Science
30:1043–1048.
Stratford, M. and T. Eklund. 2003. Organic acids and esters. In Food
Preservatives, edited by N.J. Russell and G.W. Gould. New York: Kluwer
Academic.
Subchan, M. and M.G.I. Rukmi. 2007. Effect of tempe gembus on cholesterol
profile of hyperlipidemic rats. Medical Journal of Indonesia 16:205–211.
Sudarmadji, S. and P. Markakis. 1977. The phytate and phytase of soybean tem-
peh. Journal of the Science of Food and Agriculture 28 (4):381–383.
Sudarmadji, S. and P. Markakis. 1978. Lipid and other changes occurring
during the fermentation and frying of tempeh. Food Chemistry 3:165–170.
Suparmo, S. and P. Markakis. 1987. Tempeh prepared from germinated soy-

beans. Journal of Food Science 52 (6):1736–1737.
Suparmo, S. and P. Markakis. 1989. High-vitamin B12 tempeh by mixing fer-
mentation. In Trends in Food Biotechnology, edited by H.G. Ang, B.H.
Nga and K.K. Lim. Singapore: Singapore Institute of Food Science and
Technology.
Suprapti, L. 2003. Pembuatan Tempe. Yogyakarta: Penerbit Kanisius. (In
Indonesian).
Sutardi and K.A. Buckle. 1985a. Phytic acid changes in soybeans fermented
by traditional inoculum and six strains of Rhizopus oligosporus. Journal of
Applied Bacteriology 58 (6):539–543.
Sutardi and K.A. Buckle. 1985b. Reduction in phytic acid levels in soybeans
during tempeh production, storage and frying. Journal of Food Science 50
(1):260–261, 263.
Sutardi and K.A. Buckle. 1988. Characterization of extra- and intracellular phy-
tases from Rhizopus oligosporus used in tempeh production. International
Journal of Food Microbiology 6:67–79.
Tanaka, N., S.K. Kovats, J.A. Guggisberg, L.M. Meske and M.P. Doyle. 1985.
Evaluation of the microbiological safety of tempeh made from unacidi-
fied soybeans. Journal of Food Protection 48 (5):438–441.
Tannenmaum, S.R., V.R. Young and M.C. Archer. 1985. Vitamins and min-
erals. In Food Chemistry, edited by O.R. Fennema. New York: Marcel
Dekker.
Tibbott, S. 2004. Tempeh: the ‘other’ white beancake. In Handbook of Food
and Beverage Fermentation Technology, edited by Y.H. Hui., L. Meunier-
Goddik, Å.S. Hansen, J. Josephsen, W.-K. Nip, P.S. Stanfield and
F. Toldrá, pp. 583–594. New York: Marcel Dekker.
Tuncel, G. and D. Goktan. 1990. Effect of different methods of soaking
soya beans on the growth of Bacillus cereus, Klebsiella pneumoniae and
Staphylococcus aureus in tempeh. Journal of the Science of Food and Agriculture
53 (3):287–296.
Turner, B.C. 1987. Two ecotypes of Neurospora intermedia. Mycologia 79
(3):425–432.
US Food and Drug Administration. 2013. Guidance for Industry: A Food
Labeling Guide (11. Appendix C: Health Claims). http://www.fda.gov/
food/guidanceregulation/guidancedocumentsregulatoryinformation/
labelingnutrition/ucm064919.htm (accessed 7 March 2014).
Varzakis, T. 1998. Rhizopus oligosporus mycelial penetration and enzyme diffu-
sion in soya bean tempe. Process Biochemistry 33:741–747.
van Veen, A.G., C.W. Graham and K.H. Steinkraus. 1968. Fermented peanut
press cake. Cereal Science Today 13:96–99.
van Veen, A.G. and G. Schaefer. 1950. The influence of the tempeh fungus
on the soya bean. Documenta Neerlandica et Indonesica de Morbis Tropicis
11:270–281. Cited by Steinkraus 1966.
van Veen, A.G. and K.H. Steinkraus. 1970. Nutritive value and wholesomeness
of fermented foods. Journal of Agricultural and Food Chemistry 18:576–578.
Wagenknecht, A.C., L.R. Mattick, L.M. Lewin, D.B. Hand and K.H.
Steinkraus. 1961. Changes in soybean lipids during tempeh fermenta-
tion. Journal of Food Science 26:373–376.
Walther, G., J. Pawlowska, A. Alastruey-Izquierdo et al. 2013. DNA barcoding
in Mucorales: an inventory of biodiversity. Persoonia 30:11–47.
Wang, H.J. and P.A. Murphy. 1996. Mass balance study of isoflavones dur-
ing soybean processing. Journal of Agricultural and Food Chemistry
44:2377–2383.
Wang, H.L. and C.W. Hesseltine. 1965. Studies on the extracellular proteo-
lytic enzymes of Rhizopus oligosporus. Canadian Journal of Microbiology
11:727–732.
Wang, H.L., E.W. Swain, C.W. Hesseltine and H.D. Heath. 1979. Hydration
of whole soybeans affects solids losses and cooking quality. Journal of Food
Science 44:1510–1513.
Wang, H.L., E.W. Swain and C.W. Hesseltine. 1980. Phytase of molds used in
oriental food fermentation. Journal of Food Science 45:1262–1265.
Wang, T. 2008. Minor constituents and phytochemicals of soybeans. In
Soybeans: Chemistry, Production, Processing and Utilization, edited by L.A.
Johnson, P.J. White and R. Galloway. Urbana, IL: AOCS.
Westby, A. 1989. Lactobacillus plantarum as a microbial antagonist. PhD thesis,
Department of Food Science and Technology, University of Reading, UK.
Widagdo. 1915. Carane andamel tempe. Majalah Guru Desa. (In Indonesian).
Widowati, S. and D.S. Damardjati. 1986. Evaluasi mutu tempe gude dan
kedalai dalam beberapa formula campuran. In Seminar Hasil Penelitian
Taman Pangan, Bogor Vol. 1 Palawija, pp. 126–129.
Wiesel, I., H.J. Rhem and B. Bisping. 1997. Improvement of tempe fermen-
tations by application of mixed cultures consisting of Rhizopus sp. and
bacterial strains. Applied Microbiology and Biotechnology 47:218–225.
Winarno, F.G., Y. Haryadi and S. Budiatman. 1984. Special traditional foods
of Indonesia. In Proceedings of the IPB-JICA International Symposium on
Agricultural Product Processing and Technology. Bogor: JICA.
Wolf, W.J. and Cowan, J.C. 1971. Soybeans as a Food Source. London:
Butterworths.
Xu, B.J. and S.K.S. Chang. 2007. A comparative study on phenolic profiles and
antioxidants of legumes as affected by extraction solvent. Journal of Food
Science 72 (2):159–166.
Xu, X., H.J. Wang, P.A. Murphy and S. Hendrich. 2000. Neither background
diet nor type of soy food affects short-term isoflavone bioavailability in
women. Journal of Nutrition 130 (4):798–801.
Yan, Y., J. Wolkers-Rooijackers, M.J.R. Nout and B. Han. 2013. Microbial
diversity and dynamics of microbial communities during back-slop soak-
ing of soybeans as determined by PCR-DGGE and molecular cloning.
World Journal of Microbiology and Biotechnology 29:1969–1674.
Yoshida, S., F. Suzuki, T. Tsukiboshi and H. Shinohara. 2004. Cloning and
characterization of a gene rpg1 encoding polygalacturonase of Rhizopus
oryzae. Mycological Research 108 (12):1407–1414.
Young, V.R. 1991. Soy protein in relation to human protein and amino acid
nutrition. Journal of the American Dietary Association 91:828–835.
Zhao, N.X., C.F. Qu, E T. Wang and W.X. Chen. 1995. Phylogenetic evidence
for the transfer of Pseudomonas cocovenenans (van Damme et al. 1960)
to the genus Burkholderia as Burkholderia cocovenenans (van Damme
et al. 1960) comb. nov. International Journal of Systematic Bacteriology 45
(3):600–603.
Zheng, R.Y., G.Q. Chen, H. Huang and X.Y. Liu. 2007. A monograph of
Rhizopus. Sydowia 59:273–372.
2
S tarter C ultures
L I L I S N U R A I DA A N D WA R AW U T K RU S O N G
Contents
2.1 Indonesian Ragi 110

2.2 Ragi Tape 110
2.2.1 Production of Ragi Tape 111
2.2.3 Microbiology of Ragi Tape 112
2.2.3.1 Amylomyces rouxii 114
2.2.3.2 Yeasts 116
2.2.3.3 Bacteria 117
2.2.3.4 Factors Affecting Growth of
Microorganisms during Ragi Preparation 117
2.2.4 Industrialisation and Future Prospects 118
2.3 Ragi Tempe 119
2.3.1 Description of the Product 119
2.3.2 Production of Laru 120
2.3.2.1 Effect of Substrate on Growth and
Sporulation 122
2.3.2.2 Microbiology of Laru 123
2.3.3 Preparation of Usar 123
2.3.3.1 Microbiology of Usar 124
2.4 Thai Loog-Paeng Lao Starter for Rice Wine 125
2.4.2 Traditional Production Methods for Loog-Paeng Lao 126
2.4.3 Preparation of Pure Culture Starter 128
2.4.4 Microorganisms in Loog-Paeng Lao 129
2.4.4.1 Selection of the Microflora 131
2.4.5 Characteristics of the Substrate and Changes
during Processing 131
References 132
10 9
2.1 Indonesian Ragi

Lilis Nuraida
Dried starter culture, called ragi, is an important part in fermenting

some Indonesian traditional fermented foods, such as tape (fermented
glutinuous rice or cassava), tempe (fermented soya bean) and oncom
(fermented peanut press cake). To indicate the intended use of the ragi,
the specific name of the product is appended and so, for example, ragi
tape is for tape fermentation, ragi tempe is ragi for soya bean fermenta-
tion and ragi oncom is used to produce oncom. Dried baker’s yeast is
also called ragi. This section will describe ragi tape and ragi tempe.
2.2 Ragi Tape
Ragi tape is a dry starter culture used to make tape, fermented glutinous
rice (known as tape ketan or peuyeum ketan) or fermented cassava (known
as tape ketela /tape singkong or peuyeum ketela). Ragi is off-white in colour
and shaped into small disks or flattened balls about 2–3 cm in diam-
eter (Figure 2.1). They are sold in markets as single pieces or in plastic
bags containing 10 pieces with or without brand name labelling. Some
identify the place where the ragi was made; such as ragi Cianjur, ragi
Cirebon and so on. However, much ragi is sold without any brand name.
There is no documentation to establish when ragi was first made in
Indonesia, but it has certainly been a part of traditional fermentation
processes for hundreds of years.
Figure 2.1 Ragi tape (tape starter culture).

S ta r t er C ult ure s 111
2.2.1 Production of Ragi Tape
Ragi is made of rice flour with the addition of some spices and coco-
nut water or sugar cane juice. The microorganisms that develop during
ragi processing originate from previously made ragi (back slopping).
Until now, preparation of ragi is carried out in the traditional way as
a home industry (Figure 2.2), sometimes under unhygienic condi-
tions. To prepare ragi, rice flour is mixed with powdered or ground
spices, such as chilli, cinnamon, pepper, garlic (Allium sativum) and
galangal (Alvina galangal Sw.). The proportion of spices and also the
types of spices used vary among different ragi producers. For exam-
ple, Hesseltine (1983) described a recipe that included garlic, ginger,
galangal, bitter orange (Citrus aurantium L.) and pepper. The mixture
is roasted (using dry heat in a pan over an open flame and mixed dur-
ing roasting), cooled and then made into a thick dough by addition
of water and coconut water or sugar cane juice. Powdered, previously
Rice flour
Powdered or ground spices

(pepper, chilli, cinnamon,
garlic, galangal)
Mix
Roast
Cool
Water
Coconut water or sugar cane juice
Mix to make a thick dough
Powdered ragi inoculum
Mold into disks or flattened balls
Place on bamboo tray lined with banana leaf, cover with banana leaf
Incubate, room temperature (25–30°C), 2–3 d
Sun or air dry
Ragi tape
Figure 2.2 Traditional method for preparation of ragi tape.

made ragi, corresponding to about 5% of the rice flour, is mixed

with the dough (back slopping system). The dough is then shaped
into small disks or flattened balls with a diameter of approximately
2–3 cm. A variant method of inoculation is to sprinkle powdered,
previously made ragi over the balls. The balls are placed on bamboo
trays lined and covered with banana leaves. The tray is kept at room
temperature (25–30°C) for 2–3 days to allow the desirable microor-
ganisms to grow. The cakes are then sun-dried or air dried at room
temperature (25–30°C) for 3–4 days.
Rice flour mainly consists of carbohydrate (~80%) with an amylose

content of ~35% (Wanyo et al. 2009). Coconut water contains fruc-
tose, glucose and sucrose, with their concentrations varying with the
origin and maturity of the nuts (Vigliar et al. 2006; Yong et al. 2009).
The total sugar content of coconut water is 9.2–15 g l−1 and it also
contains minerals, such as sodium, potassium, chloride, calcium and
magnesium, and vitamins and amino acids (Vigliar et al. 2006; Yong
et al. 2009). Coconut water used in the preparation of ragi is usually
from tender coconuts. The presence of sugars and other nutrients in
coconut water, presumably, promote the growth of microorganisms
during the preparation of ragi.
Various herbs and/or spices are usually included in recipes for ragi
and other similar dry starters but their role(s) in the preparation of the
starters has not been clearly established. They may serve to inhibit the
growth of unwanted organisms and/or may stimulate the growth of
useful microorganisms in ragi (Hesseltine 1983). However, no com-
prehensive research has been conducted on either the development of
the microbial population or the effects of spices during the prepara-
tion of ragi.
2.2.3 Microbiology of Ragi Tape
A diverse flora of microorganisms is present in ragi (Table 2.1).

Hesseltine et al. (1985) stated that starters used for fermenta-
tions based on rice or cassava, such as ragi, murcha, loog-paeng
and chiu, regularly contain certain species of Mucor, Rhizopus and
Table 2.1 Microorganisms That Have Been Found in Ragi

CATEGORY GENERA/SPECIESa REFERENCE
Moulds Amylomyces rouxii (Chlamydomucor oryzae) Ko (1972), Ardhana and Fleet
(1989)
Rhizopus sp. Ko (1972)
Mucor sp. Ko (1972)
Mucor circinelloides Saono et al. (1974)
M. rouxii Saono et al. (1974)
M. javanicus Saono et al. (1974)
R. oryzae Saono et al. (1974)
Fusarium sp. Saono et al. (1974)
Yeasts Hyphopichia burtonii (Pichia burtonii, Ko (1972)
Endomycopsis chodati)
Saccharomycopsis fibuligera (Endomycopsis Ko (1972), Saono et al. (1974)
fibuliger)
Asterotremella humicola (Candida humicola) Saono et al. (1974)
Candida sp. Saono et al. (1974)
C. guilliermondii, Saono et al. (1974)
C. japonica (Filobasidium capsuligenum) Saono et al. (1974)
C. intermedia Saono et al. (1974)
C. parapsilosis Saono et al. (1974)
C. solani Saono et al. (1974)
Wickerhamomyces anomalus (Candida Saono et al. (1974), Ardhana
pelliculosa, Pichia anomala) and Fleet (1989)
Bacteria Bacillus coagulans Ardhana and Fleet (1989)
B. brevis Ardhana and Fleet (1989)
B. stearothermophilus Ardhana and Fleet (1989)
Acetobacter sp. Ardhana and Fleet (1989)
Pediococcus pentosaceus Sujaya et al. (2002)
Weisella spp. Sujaya et al. (2002)
Lactobacillus spp. Sujaya et al. (2010)
Enterococcus spp. Sujaya et al. (2010)
Clostridium perfringens Sujaya et al. (2010)
a Names in brackets are some previous and alternative names (Centraalbureau voor
Schimmelcultures 2013; Index Fungorum 2013).
Amylomyces. The predominant species found in four different com-

mercial ragis, according to Ardhana and Fleet (1989), were the yeast,
Wickerhamomyces anomalus (Candida pelliculosa), and the mould,
Amylomyces rouxii, with concentrations from 106 to 108 cfu g−1.
Saccharomycopsis fibuligera, reported from ragi by other research-
ers (Ko 1972; Saono et al. 1974), was not detected by Ardhana and
Fleet.
In spite of the diversity of microbes found in ragi, it should be

remembered that Ko, in 1972, showed that it was possible to produce
tape using only two fungal strains, A. rouxii and S. fibuligera, and
without any bacteria. However, later efforts to standardise ragi qual-
ity and the production process by adding known starter cultures as
innoculum during ragi preparation did not result in product similar in
quality to that obtained using traditional ragi inocula (Ardhana and
Fleet 1989; Siebenhandl et al. 2001).
2.2.3.1 Amylomyces rouxii Amylomyces rouxii is an amylolytic filamen-

tous member of the Mucorales that is propagated primarily by chla-
mydospores. At most, it forms abortive sporangia that may or may not
form abortive sporangiospores. Rhizoids may or may not be formed
near the base of the sporangiophores. It does not grow at 45°C but
growth occurs at 37–40°C (Ellis et al. 1976).
Amylomces rouxii is able to grow anaerobically (Hesseltine et al.
1985; Graffham et al. 1995) and Hesseltine et al. (1985) suggested
that this property may be one of the factors that promotes its growth
in starter balls. Under aerobic and anaerobic conditions in the pres-
ence of 5–10% CO2 growth it is strictly filamentous (Hesseltine et al.
1985). However, anaerobic growth in the presence of 30–100% CO2
is purely yeast-like (Bartnicki-Garcia and Nickerson 1962). A. rouxii
forms large numbers of chlamydospores that are very resistant to
dessication and may survive for up to 5 years at room temperature
(Hesseltine et al. 1988).
Several workers have noted similarities between A. rouxii and
Rhizopus oryzae (=R. arrhizus; Zheng et al., 2007) (Hesseltine
1983, 1985; Ellis 1985; Liou et al. 2007) and Hesseltine et al. (1985)
suggested that A. rouxii isolated from starter cultures from rice and
cassava-based fermentations were derived from Rhizopus spp. Abe
et al. (2007) noted that R. oryzae included strains that primarily
produce lactic acid and strains that primarily produce fumaric and
malic acids and proposed that the fumaric and malic acid-producing
strains be re-classified as R. delemar (=R. arrhizus var. delemar; Zheng
et al. 2007). Based on the sequence of the lactate dehydrogenase B
(ldhB) gene, ribosomal RNA-encoding DNA (rDNA), internal tran-
scriber spacer (ITS) and amplified fragment length polymorphism
(AFLP) genome fingerprinting, Kito et al. (2009) confirmed that
strains of A. rouxii from starters for Asian fermented foods, including

ragi tape, could be grouped into two distinct types, lactic acid produc-
ers and fumaric and malic acid producers. They suggested that the
lactic acid producers were derived from R. oryzae and that the
fumaric and malic acid producers derived from R. delemar through
domestication in starters. Based on their findings, Kito et al. (2009)
suggested that A. rouxii be re-classified as R. oryzae or R. delemar,
based on organic acid production, but for practical reasons, especially
in relation to industrial fermentations, they suggested that the use
of the name, A. rouxii, should be retained for the sporangiospore-
less variants.
A. rouxii characteristically hardly produces mature sporangia or
sporangiospores but forms abundant chlamydospores. Zheng et al.
(2007) considered that the abortive sporangia and abundant chla-
mydospores ‘… are the result of adaptation to fermentative envi-
ronmental conditions through many generations that are retained
under natural conditions’. They assigned the strain NRRL 5866
(=CBS 438.76, ex-neotype of A. rouxii) to R. arrhizus var. arrhi-
zus and this attribution has been confirmed by genetic studies (Liu
2007, 2008; Walther et al. 2013). However, not all strains previ-
ously described as A. rouxii have been re-examined and it is not
entirely clear whether all would now be classified as R. arrhizus
var. arrhizus or whether some would be classified as R. arrhizus var.
delemar. Certainly, both R. arrhizus var. arrhizus and R. arrhizus
var. delemar have commonly been isolated from ragi and similar
starter cultures (Zheng et al. 2007).
2.2.3.1.1 Metabolism of Amylomyces rouxii Amylomyces strains utilise

glycerol, maltose, sucrose and soluble starch, and form lactic acid from
glucose (Hesseltine 1983). Although A. rouxii is able to grow anaero-
bically there do not appear to be any reports on its metabolism when
grown under strictly anaerobic conditions. Abe et al. (2007) and Kito
et al. (2009) grew cultures in a medium containing glucose, 50 g l−1,
Difco yeast nitrogen base without amino acids, 6.7 g l−1, Difco casa-
mino acids, 5.0 g l−1, and CaCO3, 25 g l−1 at 25°C for 7 days with
shaking, presumably in air, and then analysed the culture filtrates. It
is evident from the analyses of Kito et al. (2009) that, under these con-
ditions, R. oryzae-related A. rouxii strains produced primarily lactic
acid (9–69% of glucose supplied) and ethanol (7–11%) but also trace
amounts of malic (0.4–1.4%) and fumaric acids (0–0.6%). It is note-
worthy that authentic R. orzyae strains did not produce any fumarate
(Abe et al. 2007; Kito et al. 2009). R. delemar-related strains produced
no lactic acid and glucose was converted to ethanol (10–29% of glu-
cose supplied) and small amounts of malic (2.6–5.6%) and fumaric
acids (0.4–1.0%). Authentic R. delemar strains produced considerably
higher concentrations of malic acid, representing 6.4–23% of the sup-
plied glucose (Abe et al. 2007). Presumably, much glucose was also
oxidised to CO2 and/or assimilated.
It would obviously be of interest to know exactly the natures of
the sugar fermentations under strictly anaerobic conditions and how
this relates to conditions in the ragi fermentation and the fermenta-
tion products produced in ragi. Ellis et al. (1976) noted that A. rouxii
NRRL 5192 converted about 30% of supplied glucose to lactate dur-
ing the first 3 days incubation and that the lactate was then utilised
over the next 4 days.
2.2.3.2 Yeasts Some characteristics of moulds and yeasts present

in ragi tape and involved in tape fermentation were described by
Hesseltine (1983). Both Hyphopichia burtonii and Saccharomycopsis
fibuligera ferment glucose, maltose and sucrose and are able to assimi-
late glucose, maltose, sucrose, cellobiose and soluble starch (Hesseltine
1983; Centraalbureau voor Schimmelcultures 2013). Fructose should
also be readily utilised by these microorganisms.
Saccharomycopsis fibuligera is a yeast-like ascomycete but growth
is filamentous and it produces large numbers of yeast-like conidia
borne on small projections on the hyphae (Pitt and Hocking 1997).
S. fibuligera secretes large amounts of extracellular amylases and
β-glucosidase and is a very effective hydrolyser of starch. In addition,
it produces a glucosidase capable of digesting raw starch (Chi et al.
2009). S. fibuligera also accumulates high concentrations of trehalose.
Trehalose is an important protective molecule against dehydration,
high or low temperature and ethanol and in many organisms is accu-
mulated under stress conditions, although this does not appear to be
the case with S. fibuligera (Chi et al. 2009). Nevertheless, the presence
of high concentrations of trehalose in S. fibuligera cells might suggest
a possible role in its survival in dried starter cultures. Limtong et al.

(2002) state that, under anaerobic conditions, S. fibuligera ferments
sugars to ethanol but Chi et al. (2009) state that it is not able to fer-
ment glucose to ethanol.
2.2.3.3 Bacteria Bacteria are also present in ragi. Three Bacillus

species and an Acetobacter species were found at concentrations of
103–04 cfu g−1 in four samples of commercial ragi from Indonesia
examined by Ardhana and Fleet (1989) and the concentrations
reached 105 cfu g−1 during tape fermentation. However, these con-
centrations seem too low for the bacteria to have much effect on the
fermentation.
Sujaya et al. (2002, 2010) suggested that lactic acid bacteria may
play a role in tape fermentation. They showed, using colony hybridisa-
tion with specific oligonucleotide probes, that Pediococcus pentosaceus
and Weisella spp. were the predominant lactic acid bacteria in ragi
while Lactobacillus curvatus and Enterococcus faecium were associated
with certain types of ragi. In the later study, using polymerase chain
reaction-denaturing gradient gel electrophoresis (PCR-DGGE), they
confirmed that lactic acid bacteria were the predominant bacteria in
nine samples of ragi tape collected from different places in Indonesia
(Sumatera, Jawa and Bali). In addition to Pediococcus pentosaceus,
Weisella spp., Lactobacillus spp. and Enterococcus spp. were detected
in all ragi samples. Bacillus spp., Clostridium perfringens, uncultured
bacteria, possibly Eubacterium moniliforme, Clostridium sardiniensis or
Clostridium barratii, and some bacteria related to swine manure were
also detected. The presence of bacteria related to swine manure was
considered to be a consequence of unhygienic practices during ragi
production.
2.2.3.4 Factors Affecting Growth of Microorganisms during Ragi

Preparation Hesseltine (1983) suggested that the rice flour–spices
mixture acts as a selective medium for the growth of amylolytic, alco-
hol- and acid-forming microorganisms and Hesseltine et al. (1985)
later suggested that at least three factors are responsible as selective
agents in the development of fungal starters, namely the selective
action of spices, anaerobic conditions and high CO2 concentrations.
Numerous studies have shown that spices, such as garlic, cinnamon,

pepper, chilli and galangal, exhibit antimicrobial properties against
certain bacteria, yeasts and moulds. Garlic and cinnamon inhibited
the growth of moulds, including Aspergillus flavus, some Penicillium
species and S. fibuligera (Nielsen and Rios 2000), and Pozzatti et al.
(2008) showed that cinnamon and ginger essential oils had anti-
fungal activity against some species of Candida. Cinnamon has also
been shown to inhibit the growth of Saccharomyces cerevisiae isolated
from poultry products (Brr and Mahmoud 2005). Garlic, galangal
and ginger have been reported to exhibit antibacterial activity against
Gram positive and Gram negative pathogenic bacteria (Tajkarimi
et al. 2010). Seesuriyachan (2011) examined the effects of carda-
mom, cinnamon, clove, galangal, garlic, ginger, leadwort (Plumbago
indica), lemon grass, licorice, long pepper (Piper chaba), mace, nutmeg
and pepper on growth of Hyphopichia butonii DA69 and S. cerevi-
siae Biot88, strains isolated from Thai loog-paeng starter. Cardamom,
galangal and mace inhibited growth of H. butonii and cardamom and
ginger inhibited growth of S. cerevisiae while growth of S. cerevisiae
was stimulated by pepper.
Amylomyces rouxii grows well under anaerobic conditions, especially
in the presence of 5% CO2 (Hesseltine et al. 1985). Presumably, con-
ditions inside the starter cakes become anaerobic due to respiration by
the microorganisms and Hesseltine et al. (1985) suggested that CO2
produced by bacteria and yeasts in ragi could promote the growth of
A. rouxii. However, there is an absence of data on the actual condi-
tions in the incubating cakes.
2.2.4 Industrialisation and Future Prospects
Tape is not a staple food and is not produced on a large scale and,
hence, the demand for ragi tape is limited. Ragi tape is produced by
home industries often under poor standards of hygiene. This leads
to variablity in the microorganisms present in ragi tape and conse-
quently the quality of the tape produced is not consistent. Besides
applying more hygienic production methods, industrialisation of ragi
tape still needs further research support, especially regarding the
main microorganisms responsible for development of the taste and
aroma atributes of a good tape.
This includes
1. Identification of the minimum collection of organisms required
for the production of tape with desired characteristics. This
is essential before research on how to prepare suitable starter
cultures can be undertaken.
2. Identification of the factors that lead to the selection of the
desired microflora.
3. What is the role(s) of the coconut or cane sugars?
4. What is the role(s) of the herbs and spices?
5. What is the nature of the microbial succession and the associ-
ated changes in the substrate?
6. What are the differences between the surface and interior of
the starter cake with respect to environment and microbial
growth? What is the relative role of surface aerobic microbial
growth and interior anaerobic flora? How is this affected/con-
trolled by the size of the cake?
7. What are the metabolic products when Rhizopus oryzae and
R. delemar strains of Amylomyces rouxii are grown under
strictly anaerobic conditions and what are the products when
they are grown in ragi?
2.3 Ragi Tempe

2.3.1 Description of the Product
There are two types of starter culture used to make tempe, namely,
laru and usar (Figure 2.3). Laru, or ragi tempe, is made from cas-
sava solid waste (onggok) that has been inoculated with tempe or pure
starter culture of Rhizopus spp., incubated, dried and sold in the form
of pieces or blocks or sometimes powdered. Usar is dried hibiscus
(Hibiscus tiliaceus) leaf, locally called waru, where the hairy under-
side of the leaf is overgrown by the moulds. Laru is the most com-
monly used type of tempe starter culture. Use of usar is limited to
the Yogyakarta area and it is not found in the markets in other areas.
Both preparations contain the moulds necessary to produce tempe but
variation in the types of moulds and other microoganisms present
occurs when preparation of the ragi does not use a pure culture.
Nout and Rombouts (1990) classified tempe starters into four cate-
gories, namely, natural starters, pure culture starters, semi-pure culture
Figure 2.3 Commercially available ragi tempe starter cultures: (a) laru pieces; (b) laru blocks;
(c) powdered laru; (d) usar.
starters and mixed-strain cultures. Natural starters include dried and

pulverised tempe of a previous batch, laru and usar. Pure culture start-
ers are prepared by growing a pure culture Rhizopus strain on a sterile
substrate, followed by drying and grinding. Semi-pure culture starters
are prepared by growing a pure culture of Rhizopus sp. on cooked, but
not sterilised, rice or soya bean. Mixed-strain starter cultures are pre-
pared by growing Rhizopus sp. culture together with lactic acid bacte-
ria or Klebsiella pneumoniae (produces vitamin B12). Semi-pure culture
starters are available commercially in the market, although only one
producer is well known at present. Most tempe producers use semi-
pure cultures to make second-generation starter culture using tapioca
solid waste as substrate. Steinkraus (1983) suggested that pure culture
starters should be used for large industrial scale production of tempe,
but the preparation of tempe starter under aseptic conditions is not
done in commercial practice at present in Indonesia.
2.3.2 Production of Laru
Most laru producers are also tempe producers, and laru is usually
made by them as a home industry. The substrate is commonly cas-
sava solid waste, as this substrate is cheaper than rice or rice flour.
The cassava solid waste is a by-product of cassava starch production.

Cassava starch is produced by pulping or grating cassava, followed
by pressing to separate the starch and the solid residue. The solid
residue is dried and locally called onggok. To make laru, the ong-
gok is moistened and then steamed (Figure 2.4). After cooling, it
is inoculated by mixing with powdered tempe or laru from a previ-
ous batch and then placed on a banana leaf-covered bamboo tray to
a depth of 2–3 cm and 15–20 cm in width. Many tempe producers
use semi-pure starter culture, as described above, as inoculum. The
tray is incubated at room temperature (25–30°C) for 3 days until the
mycelium with spores covers the surface of the cake. The mouldy
cake is cut into small pieces or into blocks about 5–10 cm wide and
15–20 cm long. The pieces of cake or small blocks are then sun-dried.
Some manufacturers mill the dried cakes to produce a powder. The
laru dried cakes or blocks are usually sold in bulk while the powdered
laru is sold in plastic bags.
In the newest method, laru is made using rice flour or dehulled
and polished rice grains as substrate and a pure culture of Rhizopus
oligosporus. Steamed rice grains are inoculated with a pure culture,
Dried tapioca press cake (onggok)
Water
Moisten tapioca
Steam
Powdered laru or tempe
Mix
Place on banana-leaf covered bamboo tray
Incubate 30–35°C, 3 d
Cut into blocks or small pieces
Dry
Mill
Powdered laru
Figure 2.4 Traditional method for preparation of laru (ragi tempe).

incubated for 3–4 days at room temperature (25–30°C), and then

oven-dried followed by milling to a powder. This is a semi-pure
starter culture according to the classification of Nout and Rombouts
(1990) as the rice is simply steamed rather than sterilised. Other pro-
cesses use rice flour which is roasted (using dry heat in a pan over an
open flame with mixing during the roasting), cooled, moistened with
water, inoculated with dried starter culture of Rhizopus sp., incubated,
dried and milled. However, these methods have not yet been widely
adopted by home industry laru producers. At present, some tempe
producers propagate their own laru by inoculating onggok moistened
with warm water (50–60°C) rather than using roasted rice flour. The
product is liable to have a more diverse microbial population than
conventionally produced laru as no heating is applied to the substrate.
The shelf life of dried laru tempe starter cultures stored at room
temperature (25–30°C) has been reported to be 4–6 weeks by Jutono
(1985) and 4 months by Tanuwidjaja and Roestamsjah (1985) while
Shambuyi et al. (1992) stated that dried starter culture with aw of 0.48
could be stored for up to 30 weeks at 5, 25 and 37°C. The shelf life
can be extended to 6 months or more by refrigerated storage at 4–5°C
(Ko 1985; Jutono 1985; Shambuyi et al. 1992). The majority (~90%)
of R. oligosporus spores in cooked rice substrate after 6 days incuba-
tion at 30°C are in a dormant state, while 5–6% are metabolically
active (Thanh and Nout 2002, 2004). The age of the spores at harvest
influences their stability during storage. Spores harvested after 3 days
incubation suffered bigger losses during 2 and 3 months storage com-
pared to spores harvested after 4 or 5 days (Thanh and Nout 2002).
2.3.2.1 Effect of Substrate on Growth and Sporulation Preparation of

tempe starters should aim to obtain the maximum concentration of
viable sporangiospores that can be stored in dry conditions without
significant loss of viability (Thanh and Nout, 2002). Shambuyi et al.
(1992) compared cassava roots, cowpeas, partially defatted peanuts,
rice and soya beans as substrates to support growth and sporulation
of tempe mould. Cassava and rice supported the best growth and
the most luxuriant sporulation of the moulds. Growth and sporan-
gia formation was quicker at 37 than at 25 or 30°C. After incuba-
tion for 3–4 days at 37°C followed by drying at 40°C for 16–18 h
and pulverisation in a coffee mill, the dried material contained spore
S ta r t er C ult ure s 12 3
concentrations of 1.6 × 107 cfu g−1 on cassava and 3.4 × 106 cfu g−1 on

rice. Using cooked rice as substrate, Thanh and Nout (2002) obtained
maximum spore contents of 6.3 × 109 cfu g−1 after 6 days incubation
at 30°C. Mild oven-drying caused only a small reduction in viability
from 1.9 × 108 cfu (g dry weight)−1 when harvested to 1.4 × 108 cfu g−1
after drying but pulverisation of the dried material caused mechanical
damage and significant loss of viability, with a final concentration of
2.7 × 107 cfu g−1.
2.3.2.2 Microbiology of Laru Ko (1985) reported bacterial concen-

trations of 104 –105 cfu g−1 in dried tempe starter cultures. Shambuyi
et al. (1992) suspected that spores of Bacillus and Clostridium species
are liable to survive and to be activated by the cooking processes used
to prepare substrates for starter cultures. Certainly, the presence of
undesirable bacteria, such as Bacillus spp. should be kept to a mini-
mum during tempe starter production.
2.3.3 Preparation of Usar
Usar or dried hibiscus leaf overgrown by mould is made by sandwich-

ing inoculated, dehulled, cooked soya beans between the undersides
of two hibiscus leaves. The undersides of the leaves are used as they
are covered with hairs to which the mycelium and spores attach. The
cooked soya bean is inoculated with previously made usar or laru.
The leaves are incubated for 3–4 days at room temperature (25–30°C)
until the bean cotyledons are covered by sporulating mycelium. The
mouldy beans are removed, leaving the leaves covered with mycelium
and spores. The leaves are then sun-dried.
Besides H. tilaceus, Hibisccus similis, Tectona grandis, Bambusa spp.
and Musa paradisiaca are also used to make usar in Yogyakarta Region
(Jutono 1985). However, T. grandis and Bambusa spp. are not really
suitable for usar making due to their brittleness after drying and their
smooth surface. Leaves of Hibiscus timilis, H. tillaceus and Tectona
grandis exhibited decreased mycelia growth compared with that on
H. tilaceus leaves but increased numbers of sporangiophores and chla-
mydospores (Jutono 1985).
A requirement for the leaves used to make usar is that they should
retain their firmness during 48 h incubation under high humidity
and not become brittle after sun drying (Nout et al. 1992). During
production of filamentous fungal starters, good mycelial growth is
required with maximal sporulation and for these aims aerobic condi-
tions are essential (Nout and Rombouts 1990). Hibiscus leaves serve
as a convenient attachment surface with good moisture retaining
capacity while also providing adequate aeration for fungal develop-
ment (Nout et al. 1992).
2.3.3.1 Microbiology of Usar Nout et al. (1992) identified the predom-

inant fungi on hibiscus leaves incubated with cooked soya beans for
7 days at 30°C and 100% relative humidity. Rhizopus spp., mainly
R. oryzae and R. oligosporus, was predominantly found in all samples
with leaves from Indonesia while Cladosporium sp. dominated the
mycoflora of hibiscus leaves from Nigeria, Ghana and the Netherlands
(from a cultivated tree in a greenhouse). It was suggested that the
widespread use of Rhizopus spp. in the manufacture of tempe results
in its prevalence in the air spora and that hibiscus leaves constitute
one of its natural reservoirs in Indonesia. In the absence of inocula-
tion with previous starter material, usar may be made by incubating
boiled, dehulled and cooked soya beans sandwiched between hibiscus
leaves. Rhizopus spp. only grow if cooked soya beans are present as
substrate and then dominate quickly. A higher specific growth rate
than other fungi and production of growth inhibiting substances by
Rhizopus spp. were suggested to be factors contributing to the domi-
nance of Rhizopus spp. in usar (Nout et al. 1992).
Usar contains microorganisms other than the tempe mould, includ-
ing aerobic bacteria, bacterial spores, lactobacilli, Enterobacteriaceae
and yeasts (Nout and Rombouts 1990).
Tempe cannot be separated from the Indonesian diet. Although tempe

is mainly produced by small-scale home industries, the number of
producers amounts to about 115,000 in total, mainly concentrated in
Java island, each with the capacity to ferment 10–50 kg soya beans per
day (Sukhaeri 2012). To support the production of good quality and
safe tempe, good quality laru is necessary. The main microorganism
responsible for tempe fermentation is Rhizopus oligosporus but other
microorganisms, such as yeasts, lactic acid and other bacteria are also
present in tempe (Nout and Rombout 1990; Nout and Kiers 2005).
Potentially, lactic acid bacteria may be beneficial in tempe if they
inhibit the growth of pathogenic bacteria (see Chapter 1). However,
to take advantage of this benefit requires the development of defined
mixed culture tempe starters but, as yet, the growth and survival
of lactic acid bacteria during starter development and their interac-
tion with Rhizopus and other microorganism has not been studied.
Moreover, defined mixed-culture tempe starters are currently not
commercially available.
2.4 Thai Loog-Paeng Lao Starter for Rice Wine

Warawut Krusong
Loog-paeng, loog-pang or look-pang is a dry form of starter cake. It

is used as a fermentation starter for the production of traditional
fermented products from starchy raw materials. Different loog-
paengs are produced for different fermented products, with the spe-
cific product being indicated by a suffix. Thus, loog-paeng khao-mak
is for the production of khao-mak, a sweetened, slightly alcoholic
fermented rice, loog-paeng lao is for the production of satoh, a Thai
traditional rice wine (see Chapter 4), loog-paeng nam-som is for vine-
gar production (nam-som means vinegar in Thai) (Mahintratep et al.
1987). Similar dry starter cakes are used in other Asian countries
with different local names, such as bubud in the Philippines, banh
men in Vietnam, ragi in Indonesia, ragi tapai in Malaysia and chu
in China. Sometimes, these starters are referred to as ‘solid-state
starter cakes’ due to the methods used in their preparation, which
involve using only small amounts of water to adjust the moisture
content to 50–55%. This section is focused on Thai loog-paeng lao
used as starter for satoh, a traditional rice wine.
Loog-paeng lao is a starch-based starter, off-white in colour, in the
form of hard dry round balls or hemispherical balls or flattened cir-
cular cakes with diameters of 1–3 cm and 1–1.5 cm thick Figure 2.5).
Each region of Thailand has its own method of producing loog-paeng
lao depending on the available raw materials and preferences.
Figure 2.5 Loog-paeng lao, a dry starter cake for production of Thai rice wine: (a) from rural
central Thailand; (b) from rural north-eastern Thailand. (Courtesy of W. Krusong.)
2.4.2 Traditional Production Methods for Loog-Paeng Lao
Loog-paeng lao is made at home by rural cottage manufacturers

and the recipes are kept secret among the manufacturer’s families.
Different manufacturers may have different preparation methods,
depending on available ingredients and local custom in the region.
Most loog-paeng lao from these manufacturers is prepared under
non-aseptic conditions.
The process (Figure 2.6) starts with steeping polished (i.e. dehu-
lled) rice in tap water for 2–3 h. The rice is then washed in fresh
tap water and ground, either after draining or as wet rice which is
pressed afterwards to remove water. The ground rice is thoroughly
mixed with selected herbs (Table 2.2) and dry, powdered loog-paeng
Polished rice
Wash
Steep, 2–3 h
Drain Grind
Grind Press to remove water
Ground rice
Selected herbs,
grind and weigh
Loog-paeng lao
dry, powdered
Mix
Water
Knead to make paste
Shape into balls/hemispheres/flattened cakes
Place on bamboo tray
Loog-paeng lao,
dry, powdered
sprinkle over cakes
Cover with cheesecloth
Incubate, room temperature (32–35°C), 3–5 days
Air or sun dry
Loog-paeng lao
Figure 2.6 Traditional method for preparation of loog-paeng lao. (Adapted from Lotong, N. 1992.
Seed Inoculum and Their Production Technology. Bangkok: Funny Publishing. [In Thai.])
lao from previous batches. Clean water is then added to make a stiff
paste which is shaped into small balls or hemispherical or flattened
cakes. The cakes are placed on bamboo trays (Figure 2.7) and dry,
powdered loog-paeng lao is sprinkled over them. The bamboo trays
are covered with cheesecloth to exclude flies and incubated at room
temperature (32–35°C) for 3–5 days. The cakes are air or sun-dried
and can then be kept, usually in glass jars, for up to at least 6 months
until used. In some areas, a powder form of loog-paeng lao is prepared
Table 2.2 Herbs Used in Preparation of Loog-Paeng Lao

HERB CONCENTRATION
(G KG−1 GROUND RICE;
THAI NAME WET WT, WET WT)
Garlic (Allium sativum) kra-tiem 16
Ginger (Zingiber officinale) khing 16
Galanga (Alpinia galanga) Kha 8
Liquorice (Glycyrrhiza glabra) cha-em 16
Pepper (Piper nigrum) prik-thai 2.5
Thai Long pepper (Piper retrofractum) diphi 2.5
Onion (Allium cepa) hua-hom 8
Source: Adapted from Tammarate, P. 1990. Fermentation process control of rice and molasses.
Seminar: Fermentation Control and Analysis of Alcoholic Beverages. Excise Department,
Ministry of Finance, Thailand. (In Thai.)
and this, according to the producers, also remains useable for at least
6 months.
2.4.3 Preparation of Pure Culture Starter
Pure culture starter cultures of A. rouxii (see Indonesian ragi for dis-
cussion on the taxonomy of A. rouxii) may be prepared by growing the
mould on rice bran. Coarse and fine rice brans are mixed and moist-
ened to 45–50% water content. The paste is sterilised at 121°C for
Figure 2.7 Dough-like cakes of loog-paeng lao placed on a bamboo tray for incubation. (Courtesy
of W. Krusong.)
Coarse rice bran (50%) Fine rice bran (50%)
Mix
Add water, moisture content to 45–50% (w/w)
Autoclave 121°C × 1 h
Cool to room temperature
Autoclave 121°C × 1 h
Cool to room temperature
Inoculate Amylomyces rouxii DK
Incubate 35–37°C, 3–5 d
Dry at 45°C, 1 d
Dried Amylomyces rouxii DK bran starter
Figure 2.8 Preparation of dried A. rouxii DK mould bran starter for satoh production. (Adapted
from Krusong, W. 2008. Final report on process development of fermented vinegar from organic rice.
Industrial Technology Assistance Program, National Science and Technology Development Agency,
Thailand. [In Thai.])
1 h, cooled, autoclaved again at 121°C for 1 h, cooled, and inoculated
with A. rouxii suspension. Following incubation at 35–37°C for 3–5
days, the mycelium covered rice bran is dried at 45°C. After cooling,
it is packed in bottles and may be stored at room temperature for 4–6
months (Figures 2.8 and 2.9) without losing viability.
2.4.4 Microorganisms in Loog-Paeng Lao
Loog-paeng lao contains a mixed culture of moulds, yeasts and bac-

teria that convert starchy materials into fermentable sugars and sub-
sequently to alcohol and organic acids (Lotong 1992; Tamang and
Sarker 1995; Steinkraus 1996).
The major moulds in loog-paeng lao are A. rouxii and Rhizopus
spp. The remainder are Aspergillus, Mucor, Actinomucor, Monascus and
Penicillium species and include Asp. oryzae, Asp. niger and R. oryzae
Figure 2.9 A. rouxii DK: (a) culture on potato dextrose agar medium, 30–32°C, 2 days; (b) cul-
ture on sterilised, moistened rice bran medium, 30–32°C, 5–7 days; (c) dried moulded bran starter.
(Courtesy of W. Krusong.)
(Sukhumavasi et al. 1975; Pichayanglura and Kulaprecha 1977; Lotong

1992; Limtong et al. 2005; Samappito et al. 2007).
Saccharomycopsis (Endomycopsis) fibuligera is the predominant yeast
found in loog-paeng lao (Limtong et al. 2002). Other yeasts that
may be present include Saccharomyces cerevisiae, S. chevalieri, Kloeckera
apiculata, Pichia anomala, Pichia burtonii, Pichia fabianii, Pichia heimii,
Candida rhagii, C. glabrata, C. tropicalis, C. guilliermondii and Torulaspora
globosa (Tammarate 1990; Limtong et al. 2002; Puangwerakul
2002; Artjariyasriwong 2003; Samappito et al. 2007).
The bacteria present are predominantly lactic acid bacteria, includ-
ing Pediococcus spp., Pediococcus pentosaceus and Lactobacillus spp.,
with populations in the range of 103–104 cfu g−1. (Artjariyasriwong

2003; Rittiplang et al. 2007). Acetic acid bacteria, such as Acetobacter
spp. and Gluconobacter spp., may also contaminate loog-paeng lao
(Artjariyasriwong 2003).
2.4.4.1 Selection of the Microflora The water content of the rice paste
is 50–55% and, presumably, this low water content is a major fac-
tor in favouring the growth of the desired mould(s) and filamentous
and non-filamentous yeast(s) in the rice paste, while inhibiting the
growth of most bacteria. Additionally, loog-paeng lao cakes may be
punctured with small holes to aid aeration and promote the growth of
moulds and yeasts.
It is also possible that anti-microbial activity of the herbs and their
essential oils may play a role in the inhibition of contaminant microbes
in loog-paeng lao. However, the inhibition is likely to be limited due
to the low concentrations of herbs used (Lotong 1992; Dorman and
Deans 2000; Škrinjar and Nemet 2009). Lotong (1992) suggested the
use of propionic acid as a selective agent. Obviously, it would be useful
to have more information on the water activity, pH value and other
properties of the paste to aid in understanding the factors that control
the ecology of the microbial population that develops.
2.4.5 Characteristics of the Substrate and Changes during Processing
The substrate is almost entirely raw rice starch which, in its uncooked,
ungelatinised state is relatively resistant to the activity of most
amylolytic enzymes, but may be hydrolysed by glucosidase from
Saccharomycopsis fibuligera or other fungi (Chi et al. 2009). Starch from
non-sticky rice is more popular than starch from sticky rice (Sanchez
et al. 1988). Lotong (1992) recommended the use of home-prepared
starch in preference to ready-made or manufactured rice starch. It
is also possible that the small amount of herbs added provides some
nutrients for microbial growth and their presence was observed to
stimulate growth of A. rouxii and Saccharomyces cerevisiae (Dung et al.
2005). However, there do not appear to be any reports on the nutri-
ents utilised in loog-paeng lao or similar types of starter cultures.
Similarly, while there are numerous reports on the microbes pres-
ent and their metabolic capabilities, there are no reports on actual
metabolic activities occurring during the growth of microbes during

the preparation of loog-paeng lao. Certainly, the relatively low num-
bers of lactic bacteria observed are insufficient to produce much acidi-
fication of the rice paste.
Satoh manufacturers are aware that the choice of loog-paeng lao

influences the quality and yield of the product and this is, presumably,
at least partially a consequence of the particular mix of microbes pres-
ent in the starter. There are many aspects of the satoh fermentation
that are not well understood and a contributory factor is the lack of
understanding of the roles of the microbes in loog-paeng lao. Satoh
can be produced using pure cultures but the characteristics of satoh
produced by the traditional process and that made with pure cultures
are different. Hence, obtaining better control of the satoh fermenta-
tion is, to some extent, dependent on obtaining better control of the
microflora in loog-paeng lao.
References
Abe, A., Y. Oda, K. Asano and T. Sone. 2007. Rhizopus delemar is the proper
name for Rhizopus oryzae fumaric-malic acid producers. Mycologia 99
(5):714–722.
Ardhana, M.M and G.H. Fleet. 1989. The microbial ecology of tape ketan
fermentation. International Journal of Food Microbiology 9:157–165.
Artjariyasriwong, S. 2003. TISTR: Development of loog-paeng lao for good
quality satoh production. Thailand Institute of Scientific and Technological
Research Newsletter 6:6–7.
Bartnicki-Garcia, S. and W.J. Nickerson. 1962. Induction of yeast-like devel-
opment in Mucor by carbon dioxide. Journal of Bacteriology 84:829–840.
Brr, A.A.-H. and Y.A.-G. Mahmoud. 2005. Anti-yeast effect of some plant
extract on yeast contaminating processed poultry products in Egypt.
Czech Journal of Food Science 23:12–19.
Centraalbureau voor Schimmelcultures. 2013. CBS-KNAW Fungal Diversity
Centre: Hyphopichia burtonii; Sacchraomycopsis fibuligera. Online at www.
cbs.knaw.nl/Collections/BioloMICS.aspx (accessed 26 March 2013).
Chi, Z., Z. Chi, G. Liu, F. Wang, L. Ju and T. Zhang. 2009. Saccharomyces
fibuligera and its applications in biotechnology. Biotechnology Advances
27:423–431.
Dorman, H.J.D. and S.G. Deans. 2000. Antimicrobial agents from plants: anti-
bacterial activity of plant volatile oils. Journal of Applied Microbiology 88
(2):308–316.
Dung, N.T.P., F.M. Rombouts and M.J.R. Nout. 2005. Development of defined
mixed-culture fungal fermentation starter granulate for controlled pro-
duction of rice wine. Innovative Food Science and Emerging Technologies
6:429–441.
Ellis, J.J. 1985. Species and varieties in the Rhizopus arrhizus-Rhizopus
oryzae group as indicated by their DNA complementary. Mycologia
77:243–247
Ellis, J.J., L.J. Rhodes and C.W. Hesseltine. 1976. The genus Amylomyces.
Mycologia 68:131–143.
Graffham, A.J., M.H. Gordon, A. Westby and J.D. Owens. 1995. Nutrition of
Hesseltine, C.W. 1983. Microbiology of oriental fermented foods. Annual
Review of Microbiology 37:575–601.
Hesseltine, C.W., C.L. Featherson, G.L. Lombard and V.R. Dowell. 1985.
Anaerobic growth of molds isolated from fermentation starters used for
foods in Asian countries. Mycologia 77:390–400.
Hesseltine, C.W., R. Rogers and F.G. Winaro. 1988. Microbiological studies
on amylolytic oriental fermentation starters. Mycopathologia 101:141–
155. Cited by Steinkraus (1996).
Jutono. 1985. The microbiology of usar, a traditional tempe inoculum.
Proceedings, Asian Symposium on Non-Salted Soybean Fermentation, pp.
50–59. Tsukuba, Japan, July 1985, Tsukuba Science City: National Food
Research Institute.
Kito, H., A. Abe, I.N. Sujaya, Y. Oda, K. Asano and T. Sone. 2009. Molecular
characterizaton of the relationships among Amylomyces rouxii, Rhizopus
oryzae, and Rhizopus delemar. Bioscience, Biotechnology and Biochemistry 73
(4):861–864.
Ko, S.D. 1972. Tape fermentation. Applied Microbiology 23:976–978.
Ko, S.D. 1985. Some microbiological aspects of tempe starters. Proceedings,
Asian Symposium on Non-Salted Soybean Fermentation, pp. 101–109.
Tsukuba, Japan, July 1985. Tsukuba Science City: National Food Research
Institute.
Krusong, W. 2008. Final report on process development of fermented vinegar
from organic rice. Industrial Technology Assistance Program, National
Science and Technology Development Agency, Thailand. (In Thai).
Limtong, S., S. Sintara, P. Suwanarit and N. Lotong. 2002. Yeast diversity
in Thai traditional fermentation starter (loog-pang). Kasetsart Journal
(Natural Science) 36:149–158.
Limtong, S., S. Sintara, P. Suwanarit and N. Lotong. 2005. Species diversity
of molds in Thai traditional fermentation starters (loog-pang). Kasetsart
Journal (Natural Science) 39:511–518.
Liou, G.Y., S.R. Chen, Y.H. Wei et al. 2007. Polyphasic approach to the
taxonomy of the Rhizopus stolonifer group. Mycological Research
111:196–203.
Liu, X-y., H. Huang and R.-y. Zheng. 2007. Molecular phylogenetic relation-
ships within Rhizopus based on combined analysis of ITS rDNA and
pyrG gene sequences. Sydowia 59:235–253.
Liu, X-y., H. Huang and R.-y. Zheng. 2008. Delimitation of Rhizopus varieties
based on IGS rDNA sequences. Sydowia 60:93–112.
Lotong, N. 1992. Seed Inoculum and Their Production Technology. Bangkok:
Funny Publishing. (In Thai).
Mahintratep, B., R. Laopaiboon and N. Lotong. 1987. Production of dry
inoculum for vinegar fermentation. Kasetsart Journal (Natural Science)
21:278–288.
Nielsen, P.V. and R. Rios. 2000. Inhibition of fungal growth on bread by volatile
components from spices and herbs, and the possible application in active
packaging, with special emphasis on mustard essential oil. International
Nout, M.J.R. and F.M. Rombouts. 1990. Recent developments in tempe
research. Journal of Applied Bacteriology 69:609–633.
Nout, M.J.R. and J.L. Kiers. 2005. A Review. Tempe fermentation, innova-
tion and functionality: Update into the third millenium. Journal of Applied
Nout, M.J.R., T.D. Martoyuwono, P.C.J. Bonne and G.T. Odamtten. 1992.
Hibiscus leaves for the manufacture of Usar, a traditional inoculum for
tempe. Journal of the Science of Food and Agriculture 58:339–346.
Pichayanglura, S. and S. Kulaprecha. 1977. Survey of mycelial molds in
loogpang from various sources in Thailand. In Handbook of Indigenous
Fermented Foods, Second ed., 1996, edited by K.H. Steinkraus, p. 455.
Pitt, J.I. and A.D. Hocking. 1997. Fungi and Food Spoilage. Second ed., pp.
103–105. London: Blackie Academic.
Pozzatti, P., L.A. Scheid, T.B. Spader, M.L. Atayde, J.M. Santurio and S.H.
Alves. 2008. In vitro activity of essential oils extracted from plants used as
spices against fluconazole-resistant and fluconazole-susceptible Candida
spp. Canadian Journal of Microbiology 54:950–956.
Puangwerakul, Y. 2002. Microbiology of sato and sato production. National
Food Institute Journal 5:64–80. (In Thai).
Rittiplang, J., P. Laopaiboon, K. Vichitpan and P. Danvirutai. 2007. Lactic acid
bacteria from indigenous loog-pang samples of northeastern Thailand
and their lactic acid production ability. Thai Journal of Biotechnology
8:44–48.
Saono, S., I. Gandjar, T. Basuki and H. Karsono. 1974. Mycoflora of ‘Ragi’ and
some other traditional fermented foods of Indonesia. Annales Bogorienses
V:187–204.
Samappito, S., R. Gainsaku and L. Butkhup. 2007. Isolation of pure molds
and yeast from loog-pang for the production of Thai traditional alco-
holic beverage. Journal of Science and Technology Mahasarakham University
26:125–132. (In Thai).
Sanchez, P., B.O. Juliano, V.T. Laude and C.M. Perez. 1988. Nonwaxy rice for
tapuy (rice wine) production. Cereal Chemistry 65:240–243.
Seesuriyachan, P. 2011. Effect of spices and herbs on the growth of Pichia bur-
tonii DA 69, and Saccharomyces cerevisiae Biot 88 using the application
of Plackett-Burman screening design. Chiang Mai Journal of Science

38:151–155.
Shambuyi, M., L.R. Beuchat, Y-C. Hung and T. Nakayama. 1992. Evaluation
of substrates and storage conditions for preparing and maintaining
starter cultures for tempeh fermentation. International Journal of Food
Siebenhandl, S., L.N. Lestario, D.Trimmel and E. Berghofer. 2001. Studies on
tape ketan, an Indonesian fermented rice food. International Journal of
Food Science and Nutrition 52:347–357.
Škrinjar, M.M. and N.T. Nemet. 2009. Antimicrobial effects of spices and
herbs essential oils. ACTA Periodica Technologica 40:195–209.
Steinkraus, K.H., ed. 1983. Handbook of Indigenous Fermented Foods. New York:
Marcel Dekker.
Steinkraus, K.H., ed. 1996. Handbook of Indigenous Fermented Foods. Second ed.
Sujaya, I.N., K.A. Nocianitri and K. Asano. 2010. Diversity of bacterial flora
of Indonesian ragi tape and their dynamics during the tape fermenta-
tion as determined by PCR-DGGE. International Food Research Journal
17:239–245.
Sujaya, I.N., S. Amachi, K. Saito, A. Yokota, K. Asano and F. Tomita. 2002.
Specific enumeration of lactic acid bacteria in ragi tape by colony hybrid-
ization with specific oligonucleotide probes. World Journal of Microbiology
and Biotechnology 18:263–270.
Sukhaeri S.P. 2012. Profile of tempe industry in Indonesia. Paper presented
at Seminar on Tempe Toward International Standardization. National
Standardization Agency of Indonesia (Badan Standarisasi Nasional),
Jakarta, March 21, 2012. (In Indonesian).
Sukhumavasi, J., K. Kato and T. Harada. 1975. Glucoamylase of strain of
Endomycopsis fibuligera isolated from mold bran (loog-pang) of Thailand.
Journal of Fermentation Technology 53:559–559.
Tajkarimi, M.M., S.A. Ibrahim and D.O. Cliver. 2010. Antimicrobial herb and
spice compounds in food. Food Control 21:1199–1218.
Tamang, J.P. and P.K. Sarker. 1995. Microflora of murcha: an amylolytic fer-
mentation starter. Microbios 81:115–122.
Tammarate, P. 1990. Fermentation process control of rice and molasses.
Seminar: Fermentation Control and Analysis of Alcoholic Beverages. Excise
Department, Ministry of Finance, Thailand. (In Thai).
Tanuwidjaja, L. and Roestamsjah. 1985. Preparation and utilization of pow-
der form inoculum for tempe fermentation. ASEAN Food Journal 1(1):
22–24.
Thanh, N.V. and M.J.R. Nout. 2002. Rhizopus oligosporus biomass, sporangio-
spore yield and viability as influenced by harvesting age and processing
conditions. Food Microbiology 11:91–96.
Thanh, N.V. and M.J.R. Nout. 2004. Dormancy, activation and viability
of Rhizopus oligosporus sporangiospores. International Journal of Food
Vigliar, R., V.L. Sdepanian and U. Fagundes-Neto. 2006. Biochemical profile

of coconut water from coconut palms planted in an inland region. Jornal
de Pediatria 82:308–312.
Walther, G., J. Pawlowska, A. Alastruey-Izquierdo et al. 2013. DNA barcoding
in Mucorales: An inventory of biodiversity. Persoonia 30:11–47.
Wanyo, P., C. Chomnawang and S. Siriamornpun, S. 2009. Substitution of
wheat flour with rice flour and rice bran in flake products: Effects on
chemical, physical and antioxidant properties. World Applied Sciences
Journal 7:49–56.
Yong, J.W.H., L.Ge, Y.F. Ng and S.N. Tan. 2009. The chemical composition
and biological properties of coconut (Cocos nucifera L.) water. Molecules
14:5144–5164.
Zheng, R.-Y., G.-Q. Chen, H. Huang and X.-Y. Liu. 2007. A monograph of
Rhizopus. Sydowia 59:273–372.
3
S wee t, S our, A lcoholi c
S olid S ubstr ate
Fun g al Fermentati ons
L I L I S N U R A I DA A N D J . DAV I D O W E N S
Contents
3.1 Indonesian Tape Ketan 137

3.1.2 Production Method 138
3.1.3 Microbiology of Tape Ketan 140
3.1.3.1 Fungi 140
3.1.3.2 Evaluation of Fungal Growth in Tape 142
3.1.3.3 Bacteria 144
3.1.4.1 Rice 144
3.1.4.2 Source of Fermentable Sugars 145
3.1.4.3 Chemical Changes in the Substrate 147
3.1.4.4 Production of Flavour Compounds 149
3.2 Indonesian Tape Singkong (Cassava Tape) 152
References 154
3.1 Indonesian Tape Ketan

Lilis Nuraida and J. David Owens
Tape ketan is a traditional Indonesian fermented food that is pro-

duced across Indonesia. It is prepared from steamed glutinous rice
(Oryza sativa var. glutinosa) inoculated with ragi starter culture prior
to incubation. The rice grains are partially liquefied and acquire a
13 7
sweet, acidic and slightly alcoholic taste. Tape ketan is consumed

directly after fermentation without further processing as an appetiser
or as a dessert or it is used as an ingredient to make cakes. It is not
normally consumed every day and hence is not part of the daily diet.
There are two types of tape ketan: black (purple) tape ketan and
white tape ketan. The white tape ketan is often green in colour; a
result of the addition of leaf extract of Sauropus androgynus (L.)
Merr. (local name, katuk) or Dracaena angustifolia (Medik.) Roxb.
(suji) or Pandanus amarylifolius (pandan). In Malaysia, especially in
Sabah, tape is recognised as an alcoholic beverage with a sweet–
sour–bitter taste and sometimes a sparkling mouth feel (Chiang
et al. 2006). In some areas, such as West Java, tape ketan is referred
to as peuyeum ketan.
Traditionally, tape ketan was prepared by the household for spe-
cial occasions, such as after the Ramadan festival. However, nowa-
days tape ketan is produced by small or household industries and
sold in supermarkets. The products are packaged in glass, enam-
elled or plastic containers. In some places, tape ketan is wrapped
in banana leaves or water apple leaves (local name, jambu air,
Syzigium samarangens, syn. Eugenia javanica) or packed in plastic
bags (Figure 3.1).
3.1.2 Production Method
The production process for tape ketan is outlined in Figure 3.2. There
is a slight difference in the method for preparing tape ketan from
white or from black glutinous rice. Polished, white glutinous dehu-
lled rice is soaked in water for about 1 h, while the whole, black glu-
tinous rice grains are soaked overnight, as they still have their hulls,
and more water and cooking is required to soften them. When white
glutinous rice is used as substrate, leaf extract of katuk (Sauropus
androgynous (L.) Merr.) or suji (Dracaena angustifolia (Medik.) Roxb.;
Figure 3.3) is usually added and mixed with the rice. The leaf extract
gives the tape a green colour and a particular flavour. The rice is
steamed, cooled and inoculated by sprinkling powdered ragi (see
Chapter 2) over it and mixing. The inoculated rice is then placed in
a container, covered with banana leaves such that they are in contact
Sweet, Sour, Alcoholic Solid Substrate Fungal 13 9
Figure 3.1 Tape ketan: (a) white glutinous rice tape wrapped in water apple leaf or plastic bags;
(b) black/purple glutinous rice tape. (Courtesy of L. Nuraida.)
with the surface of the rice and the lid put on. The containers used
to ferment tape ketan include plastic, enamelled or glass ones. In
some areas, small portions of the inoculated rice is wrapped in water
apple leaves (Syzigium samarangens) and placed in plastic containers
or buckets. Wrapping with water apple leaves confers a specific fla-
vour to the tape ketan. In some places, banana leaves are also used to
wrap tape ketan. The packets of inoculated rice are incubated at room
temperature (25–30°C) for 2–4 days. The duration of fermentation
depends on how much sweetness and alcoholic flavour is desired.
After 1 day of incubation, the taste is usually only slightly sweet and
the alcoholic flavour is still weak. With longer fermentation times,
the texture becomes softer and juicier and a strong alcoholic flavour
develops.
White tape ketan Black tape ketan

Katuk or suji leaves
White glutinous rice Black glutinous rice
Trim, wash
Clean, wash Clean, wash
Add water
(water:leaves :: 1:3) Soak Soak
(ambient temp. 25–30°C, (ambient temp.
~1 h) overnight)
Grind
Filter
Extract Mix
Steam ~30 min
Cool
Powdered ragi
Mix
Put in container
Cover with banana leaf and container lid
Incubate
(ambient temp., 25–30°C, 2–4 days)
Tape ketan
Figure 3.2 Production of white or black Indonesian tape ketan.
3.1.3 Microbiology of Tape Ketan
3.1.3.1 Fungi The important microorganisms essential for tape

ketan fermentation are Amylomyces rouxii (Rhizopus arrhizus; see
Chapter 2) and at least one species of yeast (Djien 1972; Cronk
et al. 1977; Ardhana and Fleet 1989). The yeasts most commonly
found in tape are Hyphopichia burtonii (Endomycopsis burtonii;
Cronk et al. 1977), Saccharomycopsis fibuligera (Endomycopsis fibu-
ligera; Djien 1972) and Candida beverwijkiae (Candida pellicullosa;
Ardhana and Fleet 1989). Other yeasts that may also be present
include Saccharomyces cerevisiae and Wickerhamomyces anomalus
(Hansenula anomala; Ardhana and Fleet 1989). Ardhana and Fleet
(1989) described Candida beverwijkiae (Candida pellicullosa) as the
dominant yeast in tape but also observed Wickerhamomyces anomalus
(Hansenula anomala) in some fermentations. Currently (Mycobank),
Sweet, Sour, Alcoholic Solid Substrate Fungal 141
Figure 3.3 Leaves of (a) katuk (Sauropus androgynus Merr.) and (b) suji (Dracaena angustifolia
(Medik.) Robx.). (Courtesy of L. Nuraida.)
Candida beverwijkiae is classified as the anamorph (asexual stage) of

Wickerhamomyces anomalus.
Although referred to as ‘yeasts’ or ‘yeast-like’, Saccharomycopsis fib-
uligera (Endomyces fibuliger) and Hyphopichia burtonii, in fact, grow
as filamentous mycelia that produce conidia from small projections
on vegetative hyphae and also sometimes by budding, yeast-like cells
(Pitt and Hocking 1997). Additionally, Candida species commonly
produce pseudomycelia. Thus, all the predominant ‘yeasts’ in tape are
actually mycelial or pseudomycelial fungi.
Amylomyces rouxii and Candida beverwijkiae are strongly amylolytic
organisms (Ardhana and Fleet 1989). Amylolytic activity is necessary
to facilitate their growth in starchy materials, such as rice flour used
as a substrate in ragi and tape. Hyphopichia burtonii, Saccharomycopsis
fibuligera and Saccharomyces cerevisiae are not amylolytic and, hence,
their growth in starchy materials depends on other microbes generat-

ing sugars and other nutrients (Cronk et al. 1979; Ardhana and Fleet
1989). Saccharomyces cerevisiae is a strong sugar fermenter and, being
ethanol tolerant, produces high ethanol concentrations, in contrast to
Wickerhamomyces anomalus, which is relatively intolerant of high etha-
nol concentrations (Ardhana and Fleet 1989). Moulds identified as
Amylomyces rouxii include strains that produce lactic acid and ethanol,
and strains that produce ethanol but no lactic acid (see Chapter 2) and
it would obviously be useful in experimental work to characterise the
particular strain(s) used. Wickerhamomyces anomalus and Candida bev-
erwijkiae, it is suggested, may contribute to the development of estery
aromas in tape ketan (Cronk et al. 1979; Ardhana and Fleet 1989).
The diversity of the microorganisms in ragi has been shown to affect
the quality of tape. Siebenhandl et al. (2001) noted that there were dif-
ferences in taste between tape made with traditional ragi and tape made
with a mixture of pure cultures of Amylomyces rouxii, Saccharomyces cere-
visiae and two strains of Wickerhamomyces anomalus (Pichia anomale).
Ardhana and Fleet (1989), using a mixture of Amylomyces rouxii and
the yeasts, Candida beverwijkiae (Candida pelliculosa), Saccharomyces
cerevisiae and Wickerhamomyces anomalus (Hansenula anomala), pro-
duced tape with a sweet, alcoholic flavour and an odour comparable
to that produced with a traditional ragi, but it lacked the moist, juicy
appearance and soft texture of traditionally produced tape.
3.1.3.2 Evaluation of Fungal Growth in Tape It is notoriously diffi-

cult to accurately assess mould populations in foods (see Chapter 1).
Conventional colony counts reflect only the number of mould propa-
gules, be these spores and/or hyphal fragments of various and indeter-
minate lengths, and may bear little relationship to the amount of fungal
biomass present, which is the only true indicator of fungal metabolic
activity that is occurring or has occurred. There is only one report of
an attempt to directly monitor the development of fungal biomass in
tape fermentation (Cook et al. 1991a). These authors inoculated auto-
claved rice with commercial Indonesian ragi and used a microscopical
method to monitor the length of mould hyphae, the length of yeast
hyphae, the number of mould chlamydospores and the number of yeast
cells during fermentation at 30°C for 72 h (Figure 3.4). After 72 h,
the length of fungal hyphae present was (km g−1) as follows: mould
16
14
Biomass (g/kg dry matter)

12
10
8
6
4
2
0
–2
–6 6 18 30 42 54 66
Time (h)
Figure 3.4 Growth of fungi during tape fermentation of sterilised glutinous rice inoculated
with commercial Indonesian ragi starter culture. ◻, mould hyphae; ○, mould chlamydospores; Δ,
yeast hyphae; ◇, yeast cells. At 0 h, biomass concentrations were (g kg−1): mould hyphae, 0.2;
yeast hyphae, < 0.1; mould chlamydospores, <0.1; and yeast cells, <0.1. (Adapted from Cook, P.E.,
J.D. Owens and G. Campbell-Platt. 1991a. Letters in Applied Microbiology 13:123–125.)
(presumed to be Amylomyces rouxii), 0.68; and yeast (presumed to be

Saccharomycopsis fibuligera), 2.1. The numbers of unicells present were
(cells g−1) as follows: mould chlamydospores, 1.4 × 106; and yeasts,
5.1 × 107. However, taking into account the sizes of each, the biomass
concentrations were (g kg−1) as follows: mould, 15.3; yeast, 3.3; mould
chlamydospores, 6.0 and yeast cells, 0.3. Thus, the estimated total
fungal biomass of 25 g kg−1 dry weight was 62% mould hyphae, 24%
mould chlamydospores, 13% yeast hyphae and only 1% yeast cells. The
authors suggested that the commonly observed higher colony counts of
yeasts relative to moulds (Ardhana and Fleet 1989; Siebenhandl et al.
2001) overestimate their importance in the fermentation.
In principle, enumeration of yeasts is achievable with appropriate
culture media but, in practice, counts in tape are not readily inter-
preted because many of the ‘yeasts’ that occur in tape are mycelial
yeasts that grow as unicells under certain conditions and as filamen-
tous organisms under certain other conditions (Pitt and Hocking
1997). Siebenhandl et al. (2001) observed maximum yeast counts of
3.3 × 106 cfu g−1 in tape inoculated with commercial Indonesian ragi
after 60 h incubation at 30°C while Ardhana and Fleet (1989) obtained
counts of 105–107 cfu g−1 of Candida beverwijkiae (Candida pelliculosa)
in tape inoculated with commercial ragi. Numbers of Saccharomyces
cerevisiae increased sharply in the initial stages of the fermentation

but did not exceed 105 cfu g−1. Wickerhamomyces anomalus (Hansenula
anomala) was not consistently found in fermentations made with com-
mercial ragis but in some it achieved a maximum count of ~104 cfu g−1
(Ardhana and Fleet 1989).
3.1.3.3 Bacteria Bacteria, including a diversity of lactic acid bacteria,

are commonly present in tape (Ardhana and Fleet 1989; Cook et al.
1991b; Sujaya et al. 2002, 2010). Sujaya et al. (2002, 2010) noted the
successive growth of different lactic acid bacteria, with Weisella spp.
growing first, followed by Pediococcus pentosaceus. Enterococcus spp.
developed concurrently with Weisella and Pediococcus. Lactobacillus spp.
were detected after 24 h fermentation and their numbers then decreased
gradually as the fermentation proceeded. However, Ardhana and Fleet
(1989) did not find lactic acid bacteria in tape fermentations inoculated
with ragi from Indonesia, although they detected lactic acid bacteria
at levels of 106 –107 cfu g−1 in some tape samples purchased from mar-
kets in Indonesia. The market samples also contained 105–107 cfu g−1
of acetic acid bacteria. Acetobacter spp. metabolise ethanol to acetic acid
and are often associated with yeasts in alcoholic fermentations.
Bacillus spp. may also occur in tape, no doubt because their spores
are commonly present in cereals and are able to survive the steam-
ing treatment during processing. Ardhana and Fleet (1989) detected
Bacillus coagulans, Bacillus brevis and Bacillus stearothermophilus in
tape. The Bacillus population increased during the fermentation but
the maximum population did not exceeded 105 cfu g−1. Many Bacillus
species produce extracellular amylases and proteases but the concen-
trations generally reported in tape do not appear to be high enough to
contribute to significant starch or protein hydrolysis. However, Cook
et al. (1991b) demonstrated that fermented and unfermented steamed
glutinous rice could support the growth of added Bacillus cereus to
populations up to 107–108 cfu g−1.

3.1.4.1 Rice Polished white glutinous rice contains more than

80% carbohydrate, primarily starch, but the unpolished or partially
Table 3.1 Chemical Composition of White Polished Glutinous Rice, Black Unpolished
Glutinous Rice and Cassava
COMPOSITION (G KG−1 WET WT)
WHITE POLISHED BLACK UNPOLISHED CASSAVA,
COMPONENT GLUTINOUS RICE, RAWa GLUTINOUS RICE, RAWb RAWa
Water 105 90 600
Energy (kcal) 3700 3690 1600
Protein 68 96 13.6
Total lipid 5.5 34 2.8
Carbohydrate 820 730 380
Fibre, total dietary 28 37.5 18
Total sugars ndc 4 17
Ash 4.9 12.5 6.2
a USDA.
b Puwastien (2013).
c No data.
polished black glutinuous rice (the dark colour is mainly due to the
high content of anthocyanins) is considerable richer in nutrients as
it still has the outer layers of the pericarp, seed coat, aleurone layer
and the embryo (Table 3.1). The anthocyanins in black glutinous rice
include ones that are suggested to have health-enhancing properties,
such as antioxidant, anti-inflammatory, anticancer and hypoglyce-
mic activities (Abdel-Aal et al. 2006). Nevertheless, both contain
only small amounts of fermentable sugars and the tape fermentation
is dependent on amylases produced by moulds to make fermentable
sugars available to the yeasts (Djien 1972).
3.1.4.2 Source of Fermentable Sugars Djien (1972) suggested that the

mould initiated the fermentation by converting starch into sugars,
followed by fermentation of sugars to alcohol and flavour compo-
nents by the yeasts. Siebenhandl et al. (2001) confirmed that starch
was first hydrolysed to maltose and glucose by the enzymatic activi-
ties of the mould. Glucose content increased with fermentation time,
from 1.4% dry matter basis to 69% after 60 h fermentation, whereas
the total glucose (including that in starch) decreased from 83% to
78% (Table 3.2).
Amylomyces rouxii, the key mould of tape fermentation, has amylo-
lytic activity, utilises sucrose, maltose and glycerol, and metabolises
glucose to lactic acid (anaerobic fermentation) (Hesseltine 1983). In rice
14 6
Table 3.2 Changes in Chemical Composition during Tape Ketan Fermentation of Glutinous Rice Inoculated with Indonesian ‘Tebu’ Commercial
Ragi Tape
PROPERTY CONCENTRATION (G KG−1 WET MATERIAL) AFTER INCUBATION FOR (H)a
0 5 15 24 36 48 60
DRY MATTER 415 ± 0.55b 395 ± 4 395 ± 1 395 ± 1 390 ± 2 410 ± 8 395 ± 1
−1
Concentration (g kg dry matter)
Ash 2.2 ± 0.003 2.3 ± 0.03 ndc 2.2 ± 0.03 2.1 ± 0.08 2.1 ± 0.06 2.0 ± 0.1
Crude protein 100 ± 0.7 105 ± 0.6 105 ± 1 105 ± 0.7 110 ± 0.1 110 ± 1.5 110 ± 0.5
Soluble protein 6.2 ± 0.11 6.3 ± 0.17 9.9 ± 0.2 28 ± 0.5 30.5 ± 0.09 30 ± 0.7 32 ± 0.3
Free glucose nd 14 ± 0.08 125 ± 0.8 395 ± 6 595 ± 11 655 ± 14 695 ± 6
Total glucosed 835 ± 8 840 ± 89 820 ± 16 805 ± 0.3 815 ± 4 780 ± 18 775 ± 7
Ethanol nd 1.65 ± 0.045 3.9 ± 1.2 17.5 ± 1.6 27.5 ± 0.4 27 ± 1.1 34 ± 0.4
Lactic acid 0.015 ± 0.004 0.025 ± 0.003 1.6 ± 0.08 3.6 ± 0.08 3.6 ± 0.007 2.9 ± 0.2 2.65 ± 0.009
pH 6.3 ± 0.006 6.2 ± 0.006 4.7 ± 0.006 4.0 ± 0.01 4.1 ± 0.006 4.3 ± 0.007 4.7 ± 0.01
Source: Adapted from Siebenhandl, S., et al. 2001. International Journal of Food Sciences and Nutrition 52:347–357.
a All determinations were done in triplicate, except for the 48 h fermentation, which were analysed 12-fold.
b Mean ± standard deviation.
c Not determined.
d After hydrolysis of starch.
fermented with Amylomyces rouxii alone, starch content decreased only

slightly during the first 12 h and then decreased sharply to 12% after
48 h (Cronk et al. 1977). The decrease in starch content was accompa-
nied by liquefaction of the substrate. When rice was fermented with
Hyphopichia (Endomycopsis) burtonii alone, liquefaction was observed
only in localised parts. When rice was fermented with both the mould
and the yeast, the decrease in starch content was slightly more rapid
during the first 36 h (reaching 48%) than with Amylomyces rouxii alone.
In the mixed culture fermentation, the concentration of reducing sugar
increased during the first 36 h and then decreased, presumably due
to its use by the yeast (Cronk et al. 1977). Other strongly amylolytic
organisms commonly found in tape are Bacillus sp. and Candida bever-
wijkiae (Candida pelliculosa) (Ardhana and Fleet 1989).
3.1.4.3 Chemical Changes in the Substrate The major changes that

occur during the fermentation are the hydrolysis of a large propor-
tion of the starch, accompanied by a partial liquefaction of the rice,
and the production of free sugars, ethanol and lactic acid (Table 3.2).
Thus, bland cooked rice is converted to an attractive, sweet, slightly
sour and slightly alcoholic snack or desert.
Information on the biochemical and chemical changes that occur
during tape fermentation is presented in Steinkraus (1983), Cronk
et al. (1977, 1979) and Siebenhandl et al. (2001). The major chemi-
cal changes in tape fermented by Indonesian commercial ragi starters
were a very large decrease in starch and increase in glucose concentra-
tions, a substantial increase in ethanol and soluble protein, and a small
increase in lactic acid (Cronk et al. 1979; Siebenhandl et al. 2001;
Table 3.2). Washed and steamed glutinous rice contains very low con-
centrations of free sugars but in fully fermented tape ketan ~85% of
the initial starch was converted to glucose and ~4% was fermented
to ethanol. The amount of lactic acid produced was very small, with
a maximum concentration of 3.6 g kg−1 dry matter after 24 h incu-
bation, equating to 1.4 g kg−1 in the wet material (Siebenhandl et al.
2001). In spite of the small amount of lactic acid produced, it sufficed
to reduce the pH value to 4.0 after 24 h incubation. That the produc-
tion of such a small amount of lactic acid was sufficient to reduce
the pH value to 4.0 is evidence of the virtual absence of pH buffer-
ing compounds in rice. The subsequent increase in pH to 4.7 at 60 h
was, presumably, the result of oxidation of lactic acid by the fungi,

as evidenced by the decrease in lactic acid concentration between 36
and 60 h incubation. Cronk et al. (1979) observed the production of
similarly small concentrations of titratable acidity (0.5% as lactic acid
after 48 h). Although some strains of Amylomyces rouxii are able to
grow anaerobically by lactic acid fermentation, the data here suggests
that it primarily grew aerobically and/or anaerobically by alcoholic
fermentation (see Chapter 2, Section 2.2.3.1). It is also possible that
some lactic acid was produced by lactic acid bacteria.
The amount of ethanol produced during the first 24 h of incuba-
tion is quite small. Cronk et al. (1977) observed undetectable levels
(<0.08% [w/w wet wt]) in rice fermented with Amylomyces rouxii alone
or in combination with different yeasts while Siebenhandl et al. (2001)
obtained 0.7%. The alcohol content then increases progressively, and
reached 1.1–3.0% at 48 h and 2.6–5.9% after 96 h, depending on the
mould–yeast combination used (Cronk et al. 1977). Siebenhandl et al.
(2001) noted that the increase in ethanol concentration correlated
with the yeast count and the decrease in total glucose (including that
present in starch). After 60 h fermentation, the ethanol content of
tape ketan reached 3.4% (dry matter; ~1.45% wet wt). It is noteworthy
that the strain of Amylomyces rouxii used by Cronk et al. produced
concentrations of ethanol on its own that were only slightly less than
those produced when it was grown with a yeast.
Both Cronk et al. (1979) and Siebenhandl et al. (2001) observed
a small apparent increase in the concentration of crude protein
(Table 3.3) but this was most likely a consequence of the loss of dry
Table 3.3 Chemical Composition of Cassava and the Fermented Product, Tape Singkong
COMPOSITION (G KG−1 WET WT)
COMPONENT STEAMED CASSAVAa TAPE SINGKONGb
Water 615 573.5
Energy (kcal) 1530 1700
Protein 12 67.5
Fat 3 5
Carbohydrate 364 346
Ash 6 8
a Mahmud et al. (2009).
b Rukmini (2003).
matter due to oxidation by the fungi. Since crude protein is essentially

total nitrogen content, no change would be anticipated. In contrast,
Siebenhandl et al. (2001) observed a substantial increase in soluble
crude protein (i.e. determined as N) from an initial 0.6–2.8% at 24 h
incubation and 3.2% (w/w) at 60 h, while Cronk et al. (1979) recorded
an initial decrease followed by an increase after 24 h, reaching ~3%
after 48 h and continuing to increase with increasing incubation
time. It is difficult to interpret these results in the absence of data for
ammonia, amino acids and ‘true’ protein concentrations and it would
be useful to know what kind of protein metabolism occurs. Soluble
nitrogen may represent amino acids and peptides produced by protein
hydrolysis and/or ammonia produced as a consequence of the oxida-
tion of amino acids. Amylomyces rouxii and Candida beverwijkiae pro-
duce extracellular proteases that release amino acids required for their
growth (Ardhana and Fleet 1989).
Siebenhandl et al. (2001) observed no detectable change in ash
concentration, suggesting that the amount of material oxidised by the
fungi and lost from the system was quite small. Cook et al. (1991a)
estimated that the final fungal biomass represented 25 g dry matter
kg−1 dry tape. This would imply that at least 50 g carbohydrate was
oxidised, giving a net loss of 25 g kg−1 dry matter. Such a small loss
would not produce a measurable increase in ash content.
The use of pure cultures of Amylomyces rouxii, Saccharomyces cerevi-
siae and Pichia anomale did not result in tape ketan having similar char-
acteristics to that prepared with commercial ragi (Siebenhandl et al.
2001). The pH value of tape prepared with the pure cultures was about
3.4 while the normal range for tape ketan is from 4.0 to 4.2. The pH
then increased and reached 4.7 after 60 h fermentation (Siebenhandl
et al. 2001). The pH of tape fermented using a mixture of Amylomyces
rouxii and Endomycopsis burtonii also showed a sharp decrease in pH
value during the first 36 h, with a minimum of 3.9 after 48 h. Beyond
48 h, the pH fluctuated between 3.9 and 4.3 (Cronk et al. 1977).
3.1.4.4 Production of Flavour Compounds Apart from the production

of glucose, lactic acid and ethanol, which give tape its sweet–sour
slightly alcoholic flavour, the microbes involved in the fermentation
may also produce other metabolic products that contribute to the fla-
vour and aroma of tape. Cronk et al. (1979) showed that Amylomyces
rouxii, alone or in combination with different yeasts, produced a vari-

ety of higher alcohols, including isobutanol (2-methyl-1-propanol),
2-methyl-1-butanol and isoamyl alcohol (3-methyl-1-butanol). Each
of these alcohols ‘has distinct organoleptic properties’ (Cronk et al.
1979) but the authors did not state whether or not the concentrations
present affected the organoleptic properties of the tape. However,
they observed that Amylomyces rouxii with Wickerhamomyces anomalus
(Hansenula anomala) or Wickerhamomyces subpelliculosus (Hansenula
subpelliculosa), but not other yeasts, generated substantial concentra-
tions of ethyl acetate in tape, sufficient to allow its strong odour to
be easily recognised.
The nutritional value of tape is essentially similar to that of the gluti-

nous rice raw material, with the main changes being the conversion
of much of the starch to glucose, ethanol and lactic acid Table 3.2).
Possibly, some protein is hydrolysed but the overall changes in protein
and amino acid content are unknown and need to be clarified. Cronk
et al. (1977) noted an increase in thiamine content from 0.4 mg kg−1
in steamed rice to 1.2 mg kg−1 in tape.
Traditional fermentations, such as tape fermentation, are not car-

ried out under aseptic conditions and are exposed to a great diver-
sity of contaminating microbes. The excessive growth of undesirable
microbes is largely controlled by the use of ragi starter culture and the
conditions under which the tape is prepared. Thus, it is behoven on
producers to use ragi free, as far as is possible, of unwanted microbes
and to maintain sanitary and hygienic conditions throughout the pro-
duction process. Sujaya et al. (2010) detected the presence of Bacillus
cereus and Clostridium perfringens in commercial ragi, with the obvi-
ous risk that pathogenic bacteria carried over from ragi might prolif-
erate during tape fermentation. Cook et al. (1991b) demonstrated that
Bacillus cereus could grow and reach populations of 107–108 cfu g−1 if
present in the rice prior to the fermentation. Although the pH value
fell to 3.5, the numbers of Bacillus cereus remained high. However, so
far, there have been no reports of food poisoning related to consump-

tion of tape ketan but the possible presence of pathogenic bacteria in
ragi and their ability to grow during tape fermentation indicate that
preventive measures should be applied, starting with ragi prepara-
tion, raw materials preparation and the fermentation process.
The key factors to reproducibly obtain a desirable product by natu-

ral fermentation are the quality of the ragi starter culture and the
conditions of the manufacturing process. Tape is still produced at
home industry level with, generally, inadequate sanitary practices.
This leads to inconsistency of microorganisms present during the fer-
mentation and results in inconsistent quality of the product. Beside
using ragi that historically has been known to produce good-quality
tape, controlling sanitation is also important to produce a good-
quality tape. Over-fermentation leads to a more liquidified product
and higher alcohol content. As tape fermentation is not intended to
produce rice wine, terminating the fermentation at the optimum glu-
cose and alcohol concentrations and when other flavour compounds
are also optimal is important. Hence, to obtain a good product every
time, the fermentation process and materials handling needs to be
standardised.
Currently, tape is sold not only in traditional markets but in super-
markets too. The supermarkets store tape under refrigeration but fer-
mentation continues and, during storage, the taste and texture of tape
gradually changes.
For future development to improve quality and safety of tape, some
areas need to be further elucidated, including
1. The nutritional value of tape as compared to glutinuous rice.

As tape may be made from white or black glutinuous rice, it
is also of interest to evaluate any nutritional improvement as
well as the development of bioactive compounds, especially in
black tape.
2. The interactions between the dominant microorganisms in
tape fermentation to identify their roles, especially in relation
to the development of tape flavour.
3. The growth and survival of pathogenic microorganisms.

Previous studies were directed to Bacillus spp. associated with
rice grains, but the occurrence and survival of pathogenic
microorganisms commonly present in the environment or
associated with humans has not been studied.
4. Methods for handling tape to preserve quality without the
occurrence of over-fermentation.
5. Protein metabolism during tape fermentation.
3.2 Indonesian Tape Singkong (Cassava Tape)

Lilis Nuraida
Tape singkong (tape ketela, cassave tape) is a sweet, sour and alcoholic
fermented cassava product. It has characteristics similar to that of tape
ketan but is made from cassava (Manihot utilisima). Little research
has been directed specifically to tape singkong. The starter culture
used to make tape singkong is similar to that for tape ketan, that is
ragi tape, and, hence, the microorganisms involved in tape singkong
fermentation should be similar to those in tape ketan. Cassava has an
even higher carbohydrate content than polished white rice (Table 3.1)
and the microorganisms involved and the chemical changes occur-
ring during tape singkong fermentation are generally presumed to be
similar to those in tape ketan.
Making tape singkong starts with cleaning and peeling of the cas-
sava tubers. The peeled tubers are steamed, cooled and placed in a
bamboo basket lined with banana leaf. Powdered ragi is sprinkled over
each layer of cassava. The cassava is covered with banana leaves and
they are then incubated at room temperature (25–30°C) for 2–3 days
(Figure 3.5). The fermentation may be carried out in bulk in a big bam-
boo basket about 50 cm in diameter, especially when the cassava is fer-
mented as whole tubers or it may be done in small baskets (Figure 3.6)
containing only a few tubers. In some cases, the cassava tuber is cut into
smaller pieces before steaming and is fermented in plastic containers or
in bundles of 4–5 small pieces wrapped in banana leaf.
Tape singkong is made using ragi similar to that used for tape ketan
(glutinuous rice tape) and, hence, the microorganisms mainly respon-
sible for the fermentation of tape singkong are presumed to be similar
to those in tape ketan. In addition, Barus et al. (2013) isolated several
Cassava tubers
Peel
Wash
Steam
Cool
Put in bamboo container lined with banana leaf
Powdered ragi
(sprinkle over cassava)
Cover with banana leaf
Incubate, ambient temperature, 2–3 days
Tape singkong
Figure 3.5 Preparation of Indonesian tape singkong (cassava tape).
Figure 3.6 Tape singkong (cassava tape): (a) in plastic trays and traditional woven bamboo basket;
(b) wrapped in plastic film in a woven bamboo basket lined with banana leaf. (Courtesy of L. Nuraida.)
amylolytic Bacillus species, including Bacillus megaterium, Bacillus

subtilis, Bacillus amyloliquefaciens and Bacillus thuringiensis, from tape
singkong obtained from several different areas in Indonesia. The
occurrence of Bacillus spp. in cassava fermentations has been noted
previously (Amoa-Awua and Jakobsen 1995; Obilie et al. 2003).
Cassava is ~95% carbohydrate with only a small amount of pro-
tein (Table 3.1) and the fermentation of cassava to tape singkong
significantly increased the protein content and nutritive value
(Table 3.3).
References
Abdel-Aal, E.M., J.C. Young and I. Rabaski. 2006. Anthocyanin composition
in black, blue, pink, purple and red cereal grains. Journal of Agricultural
and Food Chemistry 54:4696–4704.
Amoa-Awua, W.K.A. and M. Jakobsen 1995. The role of Bacillus species in the
fermentation of cassava. Journal of Applied Bacteriology 79:250–256.
Ardhana, M.M. and G.H. Fleet. 1989. The microbial ecology of tape ketan
fermentation. International Journal of Food Microbiology 9:157–165.
Barus, T., A. Kristani and A. Yulandi. 2013. Diversity of amylase-producing
Bacillus spp. from ‘tape’ (fermented cassava). HAYATI Journal of Biosciences
20:94–98.
Chiang, Y.W., F.Y. Chye and M.A. Ismail. 2006. Microbial diversity and
proximate composition of tapai, a Sabah’s fermented beverage. Malaysian
Journal of Microbiology. 2:1–6.
Cook, P.E., J.D. Owens and G. Campbell-Platt. 1991a. Fungal growth during
rice tape fermentation. Letters in Applied Microbiology 13:123–125.
Cook, P.E., M.M-A.L. Themba and G. Campbell-Platt. 1991b. Growth of Bacillus
cereus during tape fermentation. Letters in Applied Microbiology 13:78–81.
Cronk, T.C., K.H. Steinkraus, L.R. Hackler and L.R. Mattick. 1977. Indonesian
tape ketan fermentation. Applied and Environmental Microbiology
33:1067–1073.
Cronk, T.C., L.R. Mattick, K.H. Steinkraus and L.R. Hackler. 1979. Production
of higher alcohols during Indonesian tape ketan fermentation. Applied
and Environmental Microbiology 37:892–896.
Djien, K.S. 1972. Tape fermentation. Applied and Environmental Microbiology
23:976–978.
Mahmud, M.K., Hermana, N.A. Zulfianto, R.R. Apriyantono, I. Ngadiarti,
B. Hartati, Bernadus and Tinexcelly. 2009. Table of Indonesian Food
Composition. Jakarta: Indonesian Nutritionist Association. (In Indonesian).
Mycobank. n.d. http://www.mycobank.org/(accessed 31 August 2013).
Obilie, E.M., K. Tano-Debrah and W.K.A. Amoa-Awua. 2003. Microbial

modification of the texture of grated cassava during fermentation into
akyeke. International Journal Food Microbiology 89:275–280.
Pitt, J.I. and A.D. Hocking. 1997. Fungi and Food Spoilage. 2nd ed. London:
Blackie Academic.
Rukmini, A. 2003. The composition of some of Yogyakarta’s traditional fer-
mented foods. Proceedings of National Seminar and Annual Meeting of
Indonesian Association of Food Technologists, Yogyakarta, July 22–23,
GM-01 (In Indonesian).
Siebenhandl, S., L.N. Lestario, D. Trimmel and E. Berghofer. 2001. Studies on
tape ketan—An Indonesian fermented rice food. International Journal of
Food Sciences and Nutrition 52:347–357.
Steinkraus, K.H. 1983. Handbook of Indigenous Fermented Foods. New York:
Marcel Dekker.
Sujaya, I.N., S. Amachi, K. Saito, A. Yokota, K. Asano and F. Tomita. 2002.
Specific enumeration of lactic acid bacteria in ragi tape by colony hybrid-
ization with specific oligonucleotide probes. World Journal of Microbiology
Sujaya, I.N., K.A. Nocianitri and K. Asano. 2010. Diversity of bacterial flora
of Indonesian ragi tape and their dynamics during the tape fermenta-
tion as determined by PCR-DGGE. International Food Research Journal
17:239–245.
USDA. n.d. National Nutrient Database for Standard Reference, Release 26. http://
ndb.nal.usda.gov/ndb/foods/show/6399 (accessed 9 September 2013).
4
A lcoholi c B e v er ag es
N G O T H I P H U O N G D U N G , WA R AW U T
K R U S O N G A N D K A P T I R A H AY U
K U S WA N T O
Contents
4.1 Vietnamese Rice-Based Alcoholic Beverages 158

4.1.2 Places of production 158
4.1.3.1 Raw Materials 162
4.1.3.2 Starter Cultures 162
4.1.3.3 Distillation 165
4.1.4 Characteristics of the Microorganisms Involved 166
4.1.5 Current Developments and Future Prospects 166
4.1.5.1 Research Needs and Opportunities 167
4.2 Satoh: Thai Rice Wine 168
Fermentation 173
4.2.4.1 Amylolysis 173
4.2.4.2 Ethanol Formation 174
4.2.4.3 Volatile Compounds 174
4.3 Indonesian Ciu (Fermented Molasses) 176
4.3.1 History and Description of Product 176
4.3.1.1 Early History 176
4.3.1.2 Recent History 177
4.3.2.1 Molasses Raw Material 178
4.3.2.2 Fermentation 178
15 7
15 8 In d i g en o us F erm en t ed F o o d s o f S o u t he a s t A sia
4.3.2.3 Distillation 180

4.3.4 Chemical Changes during Fermentation 181
4.3.5 Future Developments 181
References 181
4.1 Vietnamese Rice-Based Alcoholic Beverages

Ngo Thi Phuong Dung
Rice-based alcoholic beverages are highly popular in Vietnam, as are

similar traditional fermented alcoholic beverages in other Southeast
Asian countries (Basuki et al. 1996; Hesseltine 1983; Nout and Aidoo
2002). In Vietnam, the vast majority of alcoholic beverages produced
from rice are ferments fortified with added alcohol or are distillates.
They are mostly produced at home using solid-state starters in tab-
let form and under non-sterile and marginally controlled conditions.
Their production is an important source of income for many farmer
families in rural areas.
4.1.2 Places of production
Depending on the available regional ingredients and manufacturing

procedures followed, rice-based alcoholic beverages are known under
a variety of local names. In Vietnamese, ruou, means alcoholic drink,
including unfortified ferments, ferments fortified with additional
alcohol and distilled products. In the Mekong Delta in the south,
ruou nep than (fermented purple glutinous rice beverage) is fermented
from nep than (purple glutinous rice) and is not distilled. In the north
and the south of the country ruou de and ruou nep are distilled prod-
ucts from fermented rice or glutinous rice, respectively. In the moun-
tainous districts of the central highlands, such as Da Lat, Buon Me
Thuot and Dac Lac, an ethnic minority (Thuong people) produce ruou
can that is made from fermented rice, maize or cassava, with or with-
out subsequent distillation. Table 4.1 summarises the raw materials
and major functional fungi involved in the fermentation processes of
some traditional fermented alcoholic beverages from different Asian
countries.
A l c o h o li c Be v er ag e s 15 9
Table 4.1 Raw Materials and Functional Yeasts and Moulds in Traditional Fermented
Rice-Based Beverages in Asian Countriesa
COUNTRY PRODUCT RAW MATERIALS YEASTS AND MOULDS PRESENT
China Mie-chiu, Rice, wheat, barley Aspergillus oryzae, Rhizopus spp.,
Shaohing Saccharomyces cerevisiae
Indonesia Brem Bali, Arak, Rice, glutinous rice, Amylomyces spp., Mucor spp., Rhizopus
Tuak, Ciu sap of palm trees, spp., Candida spp., Saccharomyces spp.
cane-sugar
Korea Korea Rice, glutinous rice, Aspergillus oryzae, Aspergillus sojae,
barley, wheat, millet Mucor spp., Rhizopus spp.,
Saccharomyces cerevisiae, Hansenula
anomala, H. subpelliculosa, Torulopsis
sake, T. inconspicua, Pichia polymorpha
India Fenni, Sonti, Ruhi, Rice, cashew, apple Mucor, Rhizopus
Madhu, Jnard
Japan Mirin, Sake, Rice, maize, barley, Aspergillus oryzae, Aspergillus awamorii,
Shochu, Umeshu plum Saccharomyces sake, Hansenula
anomala
Malaysia Tapai, Samsu Rice, glutinous rice Mucor spp., Rhizopus spp., Candida spp.,
Saccharomyces spp., Amylomyces rouxii,
Rhizopus spp., Endomycopsis spp.
Philippines Bubod, Roselle, Rice, roselle fruit, Aspergillus spp., Endomycopsis spp.,
Lambanog, Tuba, palmyra juice Hansenula spp., Endomycopsis fibuliger,
Tapoi, Tapuy Rhodotorula glutinis, Debaromyces
hansenii, Candida parapsilosis,
Trichosporon fennicum, Saccharomyces
ellipsoideus
Thailand Sato, Ou, Rice, glutinous rice Mucor spp., Rhizopus spp., Candida spp.,
Nam-Khao Saccharomyces spp.
Vietnam Ruou De, Ruou Rice, (purple) Mucor spp., Rhizopus spp., Aspergillus
Nep, Ruou Nep glutinous rice, spp., Saccharomyces ellipsoideus,
Than, Ruou Can maize, cassava Saccharomyces cerevisiae,
Endomycopsis fibuliger, Hansenula
anomala, Torulopsis candida
a Based on data of Hesseltine (1983), Luong (1998), Haard et al. (1999) and Nout and Aidoo (2002).
Although traditional rice-based alcoholic beverages have different

compositions depending on the formulation and processes used, the
principle of their manufacture can be characterised as a biochemi-
cal modification of cereal starches brought about by microorganisms
and in which fungi (yeasts and moulds) play essential roles. Moulds
produce the amylases that degrade the starch to dextrins and sugars,
and yeasts convert these sugars to alcohol (Lim 1991; Motarjemi and
Nout 1996; Nout and Aidoo 2002). In this section, emphasis will be
laid on the fermentation processes for the preparation of rice-based
alcoholic beverages and the small-scale production processes followed
in Vietnam.
The manufacture of rice-based alcoholic drinks can be char-
acterised as a biological process whereby rice (Oryza sativa L.) is
converted into an alcoholic beverage by physical, microbiological
and biochemical operations, including steaming, inoculation with
starter, mashing and fermentation. Depending on the fermenta-
tion performance, the alcohol content may reach up to 15% (v/v). By
distillation, products with 50–60% (v/v) alcohol may be obtained.
The general outline of traditional production processes is shown in
Figure 4.1.
Rice-based alcoholic beverages are produced predominantly at
artisanal home level. Although each producer has his own proce-
dures, depending on his individual experience and regionally avail-
able raw materials, in principle, all producers follow the same process.
Powdered, starch-based starter is mixed with steamed or cooked
gelatinised rice, which is then incubated under ambient conditions.
In the Mekong Delta of the south, the leading production area for a
variety of alcoholic rice drinks, the typical ambient temperatures of
28–32°C are favourable for fermentation. After an initial period of
uncontrolled aerobic solid-state fermentation, the now mould-cov-
ered mass is mixed with water and allowed to undergo a submerged
alcoholic fermentation. Nowadays, producers tend to prefer poly-
ethylene vessels as fermentation containers instead of the previously
used large glazed terracotta jars, as the former are cheap and more
convenient to use.
The most common kind of rice beverage is made from white rice
or white glutinous rice and is distilled after alcoholic fermentation,
yielding a colourless liquor with a bland taste. Others, such as purple
glutinous rice alcoholic beverage, are fermented and not distilled and
may be sold as a crude, cloudy liquid containing sediment or as a clear
filtered product. Normally, the alcohol content in undistilled ferments
is 7−10% (v/v; about 6–8% w/v), which is insufficient to preserve the
wine from spoilage by undesired microorganisms. To overcome this
problem, producers usually increase the alcohol content, either by add-
ing refined or crude cane sugar prior to the alcoholic fermentation or
A l c o h o li c Be v er ag e s 161
Rice
Wash, soak
Steam/cook until starch gelatinized
Starter (men) Cool
Grind to powder
Inoculate
Aerobic solid-state fermentation
Add water to just submerge the mold-covered mass
Alcoholic fermentation
Alcohol
Distill Grind and mix
Mature
Filter
Distilled beverage Fortified turbid beverage Fortified clear beverage
Figure 4.1 Manufacture of traditional Vietnamese distilled and undistilled rice-based alcoholic
beverages.
by adding distilled, concentrated alcohol to the final product to obtain

final alcohol concentrations of 15–35% (v/v) depending on producers’
and consumers’ requirements for the level of alcohol and the required
storage life.
Local producers agree that the incubation period required for the
fermentation depends on the weather. The hotter the weather, the
shorter is the incubation time. In the Mekong Delta region of South
Vietnam the temperature is normally around 30–33°C and ranges
from 28°C to 40°C.
Some producers mix the initial mouldy rice mass with distilled
alcohol instead of water, followed by submerged alcoholic fermenta-
tion for a few months and filtration. To this clear, filtered liquid, cane
sugar can be added if a sweet beverage is required and the solution is
cooked and filtered again. This kind of beverage stays good for a year
without spoiling and retains its taste.
Purple glutinous rice beverage is particularly interesting, as its

brown–red colour and sherry-like flavour makes it a very attractive
and characteristic product compared to the colourless and neutrally
flavoured regular distilled wines. Purple glutinous rice alcoholic
drinks are twice as expensive as colourless beverages because of the
high cost of purple glutinous rice.
4.1.3.1 Raw Materials Depending on available ingredients and

regional preferences, different kinds of agricultural starchy materials
may be utilised. The most popular materials are dehulled rice (Oryza
sativa L.), including whole kernels or broken ones, glutinous rice and
purple glutinous rice. The rice is first soaked to hydrate and soften
the starch granules prior to subsequent gelatinisation by steaming or
cooking, which makes the starch more available to enzymic hydrolysis
(Snow and O’Dea 1981). Either steaming or cooking may be used but
most producers prefer cooking because it takes less time than steam-
ing to completely gelatinise the starch, although operators need to be
experienced to add the appropriate amount of water.
4.1.3.2 Starter Cultures It is recognised by producers that the choice

of starter tablets influences the yield and quality of the wine. Several
local processors claim that using a combination of two or three dif-
ferent starters yields beverage of better quality, with a stronger, sweet
alcoholic taste and more attractive flavour, than is obtained with a sin-
gle starter. Their experience is that each different starter has its own
advantages and disadvantages. These dried starters normally include
yeasts, moulds and bacteria (Hesseltine et al. 1988; Nout and Aidoo
2002; see Chapter 2). A variety of starter cultures is available in the
markets in most Asian countries (Table 4.2).
The raw ingredients for the preparation of starter tablets can be
either rice flour or cassava flour or combinations of both. However,
the mixed flours are preferred by local producers. The flour mixture is
ground and thoroughly mixed with spices and herbs that are believed
to play a role in preventing the growth of undesirable microorganisms.
The spices and herbs used include mixtures of garlic, pepper, onion,
rhizomes and roots of oriental herbs, and producers jealously guard
their own secret recipes. The ratio of ground rice to mixed spices is
about 14:1 by weight. Water is added to form a dough-like mass with
Table 4.2 Starters for Alcoholic Beverages Used in Asian Countriesa

COUNTRY STARTER YEASTS AND MOULDS PRESENT
China Chu Aspergillus oryzae, Rhizopus spp.
India Bakhar Mucor spp., Rhizopus spp., Amylomyces rouxii
Indonesia Ragi Mucor spp., Rhizopus spp., Amylomyces rouxii, Aspergillus
spp., Candida spp., Saccharomyces spp., Penicillium spp.
Korea Nurook Endomycopsis fibuliger (Saccharomycopsis fibuligera),
Aspergillus oryzae, A. niger, Rhizopus spp., Mucor spp.,
Hansenula anomala (Wickerhamomyces anomalus), Pichia
anomala (W. anomalus), Penicillium spp.
Malaysia Tapai Amylomyces rouxii, Rhizopus spp., Endomycopsis spp.
Nepal Murcha Hansenula spp., Torulopsis spp., Rhodotorula spp., Rhizopus
spp., Mucor spp., Saccharomyces cerevisiae
Philippines Bubud Mucor spp., Rhizopus spp., Amylomyces rouxii, Candida spp.,
Saccharomyces spp., Endomycopsis spp., Torulopsis spp.
Thailand Look-pang Mucor spp., Chlamydomucor oryzae (Amylomyces rouxii),
Candida spp.
Vietnam Men ruou, Mucor spp., Rhizopus spp., Aspergillus spp.,
Men com ruou Saccharomyces spp., Endomycopsis spp., Penicillium spp.
a Based on data of Tamang and Sarkar (1995), Basuki et al. (1996), Steinkraus (1997) and Lee and
Fujio (1999). Names in parentheses are current names according to Mycobank.
moisture content of 55–60% (w/w), which is inoculated with dry pow-

dered starter from previous batches, followed by thorough mixing.
The inoculated dough is shaped into small flattened or ball-shaped
cakes about 4 cm in diameter and 1 cm thick. The cakes are placed on
a bamboo tray and covered with a thin layer of rice husks. According
to producers, this reduces overheating and facilitates aeration. The tray
is covered with a cloth and incubated in a ventilated place at ambi-
ent temperature (28–32°C) for 2–5 days during which time the dough
rises slightly and becomes covered with fungal mycelium. The cakes
are air- or sun-dried and then have a shelf life of several months. The
traditional process of starter production is summarised in Figure 4.2.
There are three main kinds of Vietnamese traditional starters (men):
starters without oriental medicinal herbs; starters supplemented with
oriental medicinal herbs; and starters supplemented with leaves con-
taining aromatic essential oils. In the Mekong Delta region, start-
ers supplemented with herbs predominate. Although these herbs are
primarily added to starters for the fragrant flavour that they convey
to the drink, they may also have antibacterial properties and protect
the fermentation against failure. Phuc (1998) and Dung et al. (2005)
Rice and/or cassava
Soak
Starter Grind Herbs/spices
Grind into flour Grind
Water Mix to thick paste

(water content, 55–60%)
Form small flattened or ball–shaped cakes
Put on tray and cover with layer of rice husks
Incubate at room temperature (28–32°C) in a ventilated place, 2–5 days
Air or sun dry
Starter
Figure 4.2 Traditional Vietnamese process for preparing starter for rice-based alcoholic beverages.
examined various combinations of herbal extracts and single extracts

for their effects on the growth of the moulds and yeasts. It was sug-
gested that some kinds of herbs have a stimulatory effect on mould
biomass and also on yeast concentrations. The herbs tieu hoi (fennel,
Foeniculum vulgare Miller) and dinh huong (clove, Syzygium aromati-
cum L.) proved to be especially stimulatory to the growth of moulds
and yeasts. Both of these herbs are commonly used in traditional pro-
cesses for alcoholic beverages in Vietnam because of their fragrant
flavours and assumed anti-bacterial properties. It is unknown whether
or not any antibacterial effects are of relevance to the fermentation.
Dung et al. (2005, 2006, 2007) addressed the problem of the poor
and variable quality of traditional starter tablets by investigating the
possibility of preparing stable mixed cultures of selected compatible
strains of mycelial fungi and yeasts. Pure cultures of isolates from
Vietnamese rice beverage starters, Amylomyces rouxii CBS 111757 and
Saccharomyces cerevisiae LU 1250, were selected as powerful glucose-
producing and superior fermentative strains, respectively. They were
shown to be compatible in mixed culture, which is important for the
production of starters with consistent quality, and a laboratory-scale
process to formulate defined mixed-culture starter granules was estab-
lished. The beverage produced with the experimental, dehydrated,
defined mixed-culture starter was assessed as superior to commer-

cially available beverages, particularly with respect to its flavour and
overall acceptability. The defined starter also performed well in pilot-
scale fermentations using rice, glutinous rice or purple glutinous rice
(Dung and Phong 2011).
4.1.3.3 Distillation The distillation equipment used in the manufac-

ture of distilled rice wines is handmade, simple and includes three
main parts, a boiling pot, a condenser with cooling water and a recep-
tacle to receive the condensed alcohol distillate. Normally, the distil-
lation is performed in two stages. At the beginning of the distillation
process, once the material is boiling, distillation is rapid and steady,
yielding a distillate with a high alcohol concentration. Later, the dis-
tillation rate decreases and in the final period of distillation only low
alcohol concentrations are obtained in the distillate. Alcohol levels
can reach up to 65% (v/v) in the first stage and approximately 25–30%
(v/v) in the latter stage. Depending on the preferences of producers
and customers, as well as local commercial demands, these two distil-
lates may be sold separately or may be blended.
In practice, producers try to collect as much distillate as possi-
ble having a high alcohol concentration in order to maximise their
income. However, this can lead to problems due to the presence of
harmful by-products in the final product. The distilling method
affects the yield and quality of the final distilled beverage, particularly
the chemical composition, including the content of alcohol, acetal-
dehyde, esters, methanol and furfurol. Based on field work at local
manufacturers and some preliminary screening tests, it was realised
that the boiling temperature and the output rate of condensed liquid
strongly affected the yield and quality of the final product. Using an
improved still, Dung (2009) investigated the effects of pressure and
distillation rate on the properties of the distillate. For an initial vol-
ume of 25–40 l, distillation at a pressure of 0.4 or 0.5 kPa and a rate
of 100 l h−1 yielded distillate that met the required physicochemical
standards in high yield and with less time and labour than with a
traditional still. Distillates from rice fermented with defined mixed
starter or commercial starters contained significantly less undesir-
able by-products, and especially less acetaldehyde, than traditionally
distilled products.
4.1.4 Characteristics of the Microorganisms Involved
The two essential stages involved in rice beverage production are the
saccharification of starch in an aerobic solid-state fermentation and an
alcoholic fermentation. Starters for rice wine fermentation generally
include mycelial fungi, yeasts and bacteria but the mycelial fungi and
yeasts receive most attention as they are crucial for starch degradation
and alcoholic fermentation.
The major moulds in traditional starters are Amylomyces rouxii,
Rhizopus spp. and Mucor spp., and the commonly present yeasts
are Saccharomyces cerevisiae, Hansenula spp., Endomycopsis filbuligera
(Saccharomycopsis fibuligera) and Candida spp. (Table 4.2). The moulds
produce α-amylase and amyloglucosidase (also called glucoamylase)
that hydrolyse starch to dextrins and maltose but mainly to glucose
(Crabb 1999; Nout and Aidoo 2002). α-amylase cleaves starch ran-
domly at 1,4-α-glycosidic bonds, giving malto-oligosaccharides as
final products. Amyloglucosidase liberates single d-glucose monomers
in the β-form from the non-reducing end of starch and preferentially
hydrolyses 1,4-α-glucosidic linkages (Schindler et al. 1998).
Yeasts conduct the alcoholic fermentation but some may also cause
spoilage due to other metabolic activities, such as metabolism of car-
bohydrates to compounds other than ethanol, metabolism of nitrogen
compounds, production of organic acids, degradation of lipids, pro-
duction of polyols or have negative effects on quality through autolysis
(Fleet 1993). Representative spoilage yeasts in alcoholic fermentations
include Pichia spp., Zygosaccharomyces spp. and Kluyveromyces spp. At
certain concentrations, ethanol inhibits growth and affects viability of
yeasts. It has been reported to have a variety of inhibitory effects on
yeast cells (Casey and Ingledew 1986; D’Amore et al. 1990; Sharma
1997). One of the major target sites is the plasma membrane and,
at certain high concentrations, ethanol alters membrane organisation
and permeability and causes leakage of cell components.
4.1.5 Current Developments and Future Prospects
Currently, for commercialisation, traditional fermented rice-based

products, including beverages, are required to be professionally exam-
ined to evaluate their quality, using officially recognised and stan-
dardised techniques. While it is recognised that the choice of starter
strongly influences yield and quality, there is not much knowledge on

the relationship between the microbiological composition of starters
and their performance. This limited knowledge poses an obstacle to
industrial development and, thus, starters have attracted the attention
of researchers in relation to the selection of superior, storable and safe
starters for small-scale fermentation processes. Advantages of defined
mixed-starter cultures have been described (Holzapfel 1997; Ndip
et al. 2001; Siebenhandl et al. 2001) and potentially include enhanced
yield, improved hygiene, predictability of fermentation processes and
control of safety and quality.
4.1.5.1 Research Needs and Opportunities Although the essence of tra-

ditional fermentations is that they have solved the practical problems
of how to, on most occasions, ensure a successful fermentation, it is
evident that the scientific basis of many aspects of the fermentations
are poorly understood or not understood at all. Among the many
questions that remain unanswered regarding traditional Vietnamese
alcoholic beverages are
1. What is the role(s) of herbs and spices in the preparation
of starters—are they nutrient sources or/and anti-microbial
agents aiding the selection of the desired microbes?
2. Do the herbs and spices in the starter make any contribution
to the characteristics of the final drink?
3. Processors claim that using a combination of two or three
different starters yields beverage of better quality than is
obtained with a single starter and that each different starter
has its own advantages and disadvantages. Is there any micro-
biological basis for this belief?
4. What determines the maximum alcohol concentration in the
fermentation—nutrient deficiencies, sensitivity of yeasts to
alcohol?
5. What gives purple glutinous rice its sherry-like properties and
how might this property be enhanced and controlled?
6. What are the concentrations of lactic acid and other bacteria
and do they play any beneficial/detrimental role(s)?
In conclusion, the essential factors to be considered for stimulat-
ing innovation for controlled manufacture of fermented rice beverages
with consistent good quality and high yields, include starter formula-
tion, fermentation conditions and the distillation operation. While
home-scale production units are convenient for trialing new proce-
dures, expanded marketing would benefit the small-scale rice bever-
age producers.
4.2 Satoh: Thai Rice Wine

Warawut Krusong
Thai rice wine, satoh, is a traditional, non-distilled, fermented alco-

holic beverage that contains not more than 15% (v/v) alcohol (Thai
Industrial Standards Institute 2003). Both clear and cloudy satohs are
produced in rural areas across Thailand but most of them are of the
unclarified, cloudy type. Prior to the relaxation of the liquor produc-
tion laws in 1999, rice wine was manufactured illegally. After revising
the law, there was an enormous increase in satoh production by small-
scale cottage producers and satoh has gained popularity throughout
the country. Being a traditional product, there are some differences,
especially in taste, between satohs produced in different regions and by
different producers. Sweet satoh and slightly sour satoh are examples.
Rice wine is produced in many Asian countries, being called tapuy
in the Philippines, jiu in China, ruou nep inVietnam, yakju in Korea
and sake in Japan (Sanchez et al. 1988; Sirisantimethakorn et al.
2008). Rice wine has as its starting material rice starch and this neces-
sitates an amylolytic stage prior to the fermentation of the resultant
sugars to ethanol, in contrast to fruit-based wines.
Satoh production can be divided into two stages, a solid-state fermen-

tation and a submerged fermentation. The first, or saccharification
stage, occurs under aerobic conditions to promote growth of moulds
and production of amylolytic enzymes which hydrolyse the starch of
steamed glutinous rice to fermentable sugars. The second stage is ini-
tiated by adding clean water to the saccharified rice. During this stage
an alcoholic fermentation occurs and yeasts convert glucose to ethanol
under anaerobic conditions.
Polished glutinous rice
Wash
Steep in tap water 6–12 h or overnight
Steam 30–60 min
Wash, drain, cool, spray with small amount of clean water
Loog–paeng lao starter, 1–2% (w/w)
Mix
Ferment, room temperature 1–3 days

(saccharification of cooked rice starch)
Water
(total soluble solids
To 24–25°Brix)
Ferment, room temperature, 4–14 days
(alcoholic fermentation)
Filter
Bottle
Pasteurise, 65–70°C, 30 min
Satoh
Figure 4.3 Traditional process for production of satoh, Thai rice wine. (Adapted from
Chaownsungket, M. 1978. Selection of the yeast and mold strains for rice wine production. M.Sc.
Thesis, Graduate School, Kasetsart University, Bangkok, Thailand. 116 p. [In Thai.])
The traditional process for making satoh (Figure 4.3) starts with
polished glutinous rice, which is washed, soaked in tap water (three
to four times the volume of the rice) for 6–12 h or overnight and
then steamed. The steamed rice is washed with clean water to cool
it to room temperature and then drained. The cooled, steamed glu-
tinous rice is very sticky and is sprayed with a small amount of clean
water to modify the texture and to incorporate some air into the
rice. This is necessary to enable mould growth after inoculation. The
rice is inoculated with 1–2% (dry wt/wet wt) powdered loog-paeng
lao (either from ground cakes or bought as a powder) and then put
into an earthenware jar. Normally, the jar is only one quarter filled
and the inoculated rice is punctured with holes for aeration. In rural
areas, satoh is produced using traditional loog-paeng lao starter (see
Chapter 2). Usually, the dried cake form of loog-paeng lao is used,
after grinding to a powder, but some small cottage manufacturers
use ready-powdered starter (Figure 4.4). Saccharification of the rice
occurs within 1–3 days, resulting in increases in the concentrations
of glucose, maltose and dextrins. The saccharified rice is mixed with
clean water in the ratio 1:2 to 1:3 rice to water to obtain a total sol-
uble-solids content of 24–25°Brix and the mixture is allowed to fer-
ment for 4–14 days. The mash is then filtered, bottled and pasteurised
at 65–70°C for 30 min (Figure 4.5).
Satoh may also be produced using pure cultures (Figure 4.6).
Polished glutinous rice is washed, soaked in tap water for 6 h or over-
night and steamed. The steamed rice is sprayed with clean water to
modify the texture and aerate the rice. After cooling to room tem-
perature, it is spread in a shallow layer, approximately 0.5–1.0 cm
deep, and allowed to surface dry before being transferring to a clean
container and covered. The rice is inoculated by mixing with 0.2%
(dry wt/wet wt) Amylomyces rouxii moulded bran and the surface is
sprinkled with a small amount of moulded bran (Figure 4.7a). The
fermentation is complete within 3–4 days at room temperature (30–
32°C). The rice starch is liquified and converted to dextrins, maltose
Figure 4.4 Powder form of loog-paeng lao used for satoh production by some small-scale cot-
tage manufacturers in northeastern Thailand, showing presence of added rice husk. (Courtesy of
W. Krusong.)
Figure 4.5 Bottles of satoh from central Thailand. (Courtesy of W. Krusong.)
and glucose achieving a total soluble-solids of 35–40°Brix (Figure

4.7b). The wort is diluted with clean water to adjust the sugars con-
tent to 20–24°Brix. It is inoculated with 0.5% (wet wt/wet wt) of pure
Saccharomyces cerevisiae culture and fermented for 6–7 days. After fil-
tering, bottling and pasteurising at 65–70°C for 30 min, the satoh is
ready for consumption.
4.2.3 Microbiology
Microorganisms found in loog-paeng lao starter include moulds,

yeasts and bacteria, especially lactic acid bacteria (Rittiplang et al.
2007; Sirisantimethakorn et al. 2008; see Chapter2).
Yeasts present in loog-paeng lao include Saccharomycopsis fibuligera
and/or Saccharomyces cerevisiae (Tammarate et al. 1982; Limtong et al.
2002; Puangwerakul 2002; Artjariyasriwong 2003). Alternatively,
pure cultures of Saccharomyces cerevisiae may be used in satoh produc-
tion (Figure 4.6). Saccharomycopsis fibuligera is an amylolytic yeast and,
under anaerobic conditions, ferments sugars to ethanol according to
Limtong et al. (2002), although Chi et al. (2009) state that it is inca-
pable of fermenting glucose to ethanol.
Lactic acid bacteria are one of the major contaminants in satoh and
are the cause of the sour taste of satoh produced in some areas. It is
not clear whether they derive from the relatively low numbers of lactic
Polished glutinous rice
Wash
Steep in tap water ≥6 h
Steam 30 min
Spray with small amount of clean water
Spread in thin layer and allow to surface dry
Amylomyces oryzae DK
mold bran starter 0.2% (w/w)
Mix
Amylomyces oryzae DK
mold bran starter Sprinkle surface
Ferment, room temperature (30–32°C) 3–4 days

(saccharification of cooked rice starch)
Water
Adjust sugar content to 20–24°Brix
Sacchromyces cerevisiae
starter culture in saccharified
rice medium, 0.5% (v/v)
Ferment, room temperature, 6–7 days
(alcoholic fermentation)
Filter
Bottle
Pasteurise, 65–70°C, 30 min
Satoh
Figure 4.6 Pure cultures process for production of satoh, Thai rice wine. (Adapted from Krusong,
W. 2008. Final Report of Process Development of Fermented Vinegar from Organic Rice. Industrial
Technology Assistance Program, Technology Management Center, National Science and Technology
Development Agency, Bangkok, Thailand. [In Thai.])
Figure 4.7 Saccharification of rice starch during satoh production: (a) steamed glutinous rice
inoculated with mould bran starter, showing surface sprinkled with bran starter and aeration holes;
(b) after incubation at 30–32°C for 3–4 days, showing unhydrolysed starch grains and wort formed
by saccharification of rice starch. (Courtesy of W. Krusong.)
acid bacteria commonly present in loog-paeng lao or are acquired dur-

ing the processing of satoh.
4.2.4.1 Amylolysis The initial aerobic fermentation involves the growth

of a variety of amylolytic moulds and yeasts, including Amylomyces
rouxii and Saccharomycopsis fibuligera. These produce starch-degrading
enzymes, such as α-amylase and amyloglucosidase that hydrolyse the
174 In d i g en o us F erm en t ed F o o d s o f S o u t he a s t A sia
cooked and gelatinised rice starch to dextrins, maltose and, mainly,

glucose (see Chapter 2) and the rice grains break down and liquefy.
4.2.4.2 Ethanol Formation The second fermentation is an anaerobic

alcoholic fermentation where the sugars are fermented to ethanol and
carbon dioxide by yeasts, primarily by Saccharomyces cerevisiae and/
or Saccharomycopsis fibuligera (see Chapter 2). The fermentation ceases
when the sugars are exhausted or when the ethanol concentration
reaches inhibitory levels (~15% v/v). If residual sugars remain, the
satoh has a sweet taste but it may also acquire a sour taste from lactic
acid production by lactic acid bacteria.
Sometimes the alcoholic fermentation does not proceed well and
a satisfactory ethanol content is not obtained. To ensure that a good
alcoholic fermentation does occur it is recommended that the wort be
inoculated with Saccharomyces cerevisiae (Krusong 2008).
4.2.4.3 Volatile Compounds The aroma of satoh is affected by many

factors, including the variety of rice/glutinous rice, properties of the
water, the starter culture, incubation conditions and so on. Aroma
compounds that have been detected include 4 aldehydes (acetalde-
hyde, 1-decanal, furan-2-carbaldehyde, 5-methyl furfural), 1 acetal
(1,1-diethoxyethane), 1 ketone (3-hydroxy-2-butanone), 9 alcohols
(ethanol, glycerol, 1-propanol, 2-methyl-1-propanol, 3-methyl-1-buta-
nol, 1-hexanol, 1-octanol, 2,3-butanediol, 2-furan methanol), 3 acids
(acetic acid, hexanoic acid, 3-methylbutanoic acid) and 9 esters (ethyl
acetate, ethyl butyrate, ethyl hexanoate, ethyl octanoate, ethyl decano-
ate, ethyl lactate, 2-methylbutyl decanoate, octyl acetate, 2-phenethyl
acetate) (Numchaisiwattana and Chaiseri 2006; Chuenchomrat et al.
2008; Sirisantimethakorn et al. 2008).
Following concerns regarding the quality and safety of satoh, the Thai
Industrial Standards Institute, Minister of Industry, set up a product
certificate (No. 3/BE2546, Satoh; Table 4.3) to allow regulation of
the industry and has formulated standards for the physical and chemi-
cal properties of good quality satoh. To establish a satoh factory in
Thailand requires approval from the Excise Department, Ministry of
Table 4.3 Standard Quality of Satoh Required to Comply with Product Certificate: Satoh
(No. 3/BE2546)
MAXIMUM ALLOWED QUANTITY OR
CATEGORY PROPERTY RECOMMENDED CHARACTERISTICS
Chemical properties Ethanol Not more than 15% (v/v)
Methyl alcohol Not exceeding 420 mg L−1
Sulphur dioxide Not exceeding 300 mg L−1
Sorbic acid or sorbate Not exceeding 200 mg L−1
Benzoic acid or benzoate Not exceeding 250 mg L−1
Copper Not exceeding 5 mg L−1
Iron Not exceeding 15 mg L−1
Lead Not exceeding 0.2 mg L−1
Arsenic Not exceeding 0.1 mg L−1
Ferrocycanide Not detectable
Physical and other properties Clarity Clear or turbid
Colour Natural colour from raw materials
used
Smell/aroma Natural smell/aroma from raw
materials used
Taste Natural taste from raw materials
used
Foreign matter None
Stability No foam in product in container
Hygiene in production Manufacture to comply with GMP
regulations
Source: Adapted from Thai Industrial Standards Institute, 2003.
Industry, as the product is categorised as an alcoholic beverage, and

the satoh processing plant will be audited for Good Manufacturing
Practices standard by the Thai Food and Drug Administration
(Thai-FDA).
Since its manufacture was legalised, satoh has become a popular drink
in some areas. However, the unique taste of satoh is not appreciated
by all consumers and some, after an initial enthusiasm, then cease to
drink it. This has accentuated the need for research both on the basic
science of the process and on the technology of satoh production.
Many manufacturers are concerned with the problems of obtaining
satoh of consistent quality from batch to batch and much research is
focused on this. The use of pure culture techniques, using specially

selected strains of yeasts and moulds to impart desirable aromas and
taste, are still required. Although many strains are available from cul-
ture collections, there is still a need for suitable strains of amylolytic
moulds and fermentative yeasts to make good satoh on the large scale.
However, the taste of satoh produced by the traditional process is dif-
ferent from that of satoh made with pure cultures. This is, presumably,
a consequence of the different microorganisms present in the starter
used but why these differences occur is not understood as yet.
The development of equipment for the larger scale production of
moulded bran starter is required. The expense of large-scale Japanese
automated koji-making equipment would not be justified, but some
level of appropriate mechanisation is needed.
There are many aspects of the basic science of the process that are
unclear, such as the relative roles of Amylomyces rouxii, Saccharomycopsis
fibuligera, Saccharomyces cerevisiae, lactic acid bacteria and other
microbes in amylolysis, lactate formation and formation of flavour
and aroma compounds. Increased knowledge on these basic aspects
would facilitate rational approaches to obtaining better control of the
fermentations and the characteristics of the product.
4.3 Indonesian Ciu (Fermented Molasses)

4.3.1 History and Description of Product
Ciu is a traditional Indonesian distilled beverage made from molasses

and is similar to Jamaican rum.
4.3.1.1 Early History The history of ciu begins in the seventeenth cen-
tury, when the Batavian Middle Kingdom began to cultivate a variety
of crops, including sugar cane and rice. These two commodities, along
with palm sap, were utilised to make alcoholic drinks. Production of
Batavia arrack was started at Batavia Götheborg in 1743 (Godeliva
2008). Batavia Arrack van Oosten Batavi contained 50% alcohol and
became famous in the eighteenth century throughout Indonesia. In
Raffles’ time (1811–1815) sugar from cane was manufactured by the
Chinese alone and they were also responsible for the manufacture of
distilled alcoholic drinks from molasses, including arrack and ciu,

which Raffles describes as an inferior kind of arrack, made in a sim-
pler and more economical manner (Raffles 1830). Batavian arrack was
well known in Europe and was exported in considerable quantities to
continental Europe (Raffles 1830).
Bekonang village, Sukoharjo district, not far from Solo in central
Java, claims to be the first locality to produce ciu by fermenting sug-
arcane molasses. Bekonang lies on the banks of the Bengawan Solo
River, the longest river on Java island, which in ancient times was the
main economic artery connecting the Javanese hinterlands with the
bustling port of Gresik on the Java Sea in the East. In 1980, at the
entrance to the village a rusty, official-looking sign said that this is a
home-industry alcohol-producing village (Godeliva 2008) and old tra-
ditions claim that Bekonang has been distilling ciu for centuries. The
name ciu indicates a Chinese origin and the distillers in Bekonang say
that their ancestors were making ciu before the arrival of the Dutch.
One of the grievances of the Javanese against the Dutch was their
influence on the growth of a drinking culture among the Javanese.
The colonialists stipulated that every village had to set aside at least
a fifth of their agricultural land to plant European commodities,
including sugarcane (Godeliva 2008). At the same time, many sugar
factories were built in Central and East Java, followed by the pro-
duction of alcoholic fermented molasses in Bekonang. Bekonang was
then already well known for production of ciu.
4.3.1.2 Recent History Although, today, no one admits to making

ciu, illegally produced alcohol still finds its way to drinking com-
munities throughout Java. In 1981, the government acknowledged
this and legalised Bekonang as a centre of alcohol production. This
acknowledgement was the secret of the survival of the distillery busi-
ness in Bekonang. Although production of ciu for drinking is illegal,
the production of 90% alcohol for sale to the pharmaceutical industry
and hospitals is permitted.
To make industrial-strength alcohol using traditional methods one
must first make ciu, which is the result of the first distillation. After
second and third distillations, the ciu is transformed to industrial-
grade alcohol, which is sold legally with assistance from the local
village cooperative. However, not all the ciu made in Bekonang is
further distilled to make industrial alcohol and some of it slips out and
is sold as illicit spirit. Though Islam has been the main religion in Java
since the founding of the Islamic kingdom of Demak in the fifteenth
century, there are some places in Java that still retain a clandestine
alcohol production and drinking culture. This sub-culture comes into
view when the press publishes images of government officials destroy-
ing illegally produced and distributed alcohol. Because it is illegal,
drinkers of traditional alcoholic drinks usually do their drinking in
private and out of sight. As things stand now, people drink ciu that
would otherwise be used to make industrial alcohol.
4.3.2.1 Molasses Raw Material Molasses is the non-crystallisable

residue remaining after sucrose purification from sugar cane. It has
some advantages as a raw material for alcohol production, being rela-
tively inexpensive, readily available and, unlike starch, not requiring
hydrolysis before being fermented to alcohol (Najafpour et al. 2004;
Sanchez and Cardona 2008). The molasses obtained after sugar cane
processing contain 30–40% sucrose, 10–25% glucose + fructose,
very low amounts of raffinose and betaine and about 5% aconitic
acid. The non-sucrose substances include inorganic salts, organic
acids and nitrogen-containing compounds.
4.3.2.2 Fermentation Molasses is diluted to a suitable sugar concen-

tration (15–16% w/v) and small quantities of nitrogen source (e.g.
ammonium phosphate, urea or ammonium sulphate) and sulphuric
acid are added to adjust the pH of the medium to about 5.0. An
actively growing Saccharomyces cerevisiae culture is added after
which the fermentation starts and is allowed to proceed for about
24–40 h at ambient temperature of 25–30°C (Figure 4.8). The yield
of ethanol ranges up to about 50% of the fermentable sugar present
in the medium. It is a non-aseptic process and the presence of unde-
sirable microorganisms in the process cannot be completely avoided,
not least because the raw materials include contaminant microor-
ganisms. Fortunately, the contamination problem is mitigated by
inoculation with a large yeast population and the yeast populations
Black strap molasses (dark brown cane molasses)
Dilute to 15–16% sugar
Boil
Fermentation, 25–32ºC, 3–5 days

(natural fermentation or add yeast)
First distillation
Ciu
(alcohol content 40 to 50% v/v)
Second and third distillations
Industrial alcohol
(more than 90% alcohol)
Figure 4.8 Preparation of ciu and industrial alcohol as done at Bekonang, Java, Indonesia.
grow quite rapidly and overwhelm many of the potentially compet-

ing organisms. Provided reasonable care is exercised, decreases in the
yield of ethanol resulting from competing reactions can be held to a
minimum.
The configuration of the fermentation tank has very little influence
on system performance but the proportions of the tank should not be
extreme. Because, typically, the weight of non-fermentable solids in
molasses is equal to or somewhat greater than the weight of ferment-
able material the fermentation tanks need to be larger than would
be the case with a purer fermentation substrate. Commonly, tanks
are upright cylinders (steel oil drums or plastic drums) with a height
somewhat greater than the diameter. Usually, the tanks are covered
and fitted with a fermentation trap. Often, this takes the form of a
kitchen U-tube waste pipe trap, plumbed via reducer fittings to the
top of the tank. The trap is filled with water, and as the CO2 bubbles
through it any ethanol is retained and can be recovered. The trap also,
of course, serves to exclude oxygen, flies and microbial contaminants
from the fermenting mash.
The optimum fermentation conditions are a temperature of 25–30°C
and a pH of 5–6. A primary concern is keeping the tank warm and for
this it should be insulated. However, since the fermentation process
produces heat, cooling may be necessary to keep the temperature in

the optimum range.
The expected alcohol yield from a 15–25% solution of fermentable
sugars is 6.8–11% by weight. Contamination with undesirable micro-
organisms may decrease the yield of alcohol as they compete with the
yeast for the sugar. Traditionally, a good batch of cane molasses will
produce around half its volume of ciu in a little over a week.
4.3.2.3 Distillation After the fermentation process is complete, the

fermented wort is transferred to a distillation still. A traditional still
consists of a large clay stove with a steel drum of boiling fermented
molasses on it. The drum has a copper tube that extends out of its top
and which plunges into a barrel of cold water to condense the steam
and ends with ciu dripping into the mouth of a plastic jerrycan. In
earlier times an earthenware pot would have been used but nowadays
metal barrels are invariably used. Some large factories now have mod-
ern manufactured stills.
The biggest expenditure, after purchase of the molasses, is the fire-
wood and the village houses which are engaged in the production of
ciu have larger than normal stacks of firewood in the sheds beside
and behind them. If the distillers do what they are legally permit-
ted to do, they must further process the ciu into industrial alcohol,
which requires the addition of caustic soda and two more distillations,
reducing the volume by nearly half.
4.3.3 Microbiology
The most widely used microorganism in commercial alcoholic fer-

mentations is Saccharomyces cerevisiae but in the ciu fermentation
Schizosaccharomyces sp. is dominant in natural fermentations with-
out inoculation, with yeast concentrations up to 8 × 108 cfu ml−1.
Seven strains isolated from different ciu factories and tested by
physiological and morphological methods were all identified as S.
pombe by Kozaki et al. (1992). S. pombe used to occur in traditional
West Indian rum fermentations but, with improvements in tech-
nology and hygiene, was later replaced by Saccharomyces cerevisiae
(Fahrasmane et al. 1988, 1996).
4.3.4 Chemical Changes during Fermentation
The principal chemical changes are the production of ethanol and car-
bon dioxide from the sugar of molasses. Data are not available about
changes in nutritional value, but certainly the growth of yeasts in the
molasses might be expected to contribute B vitamins.
Yeasts, under anaerobic conditions, convert glucose to ethanol:
C6H12O6 → 2C2H5OH + 2CO2 − 265 kcal

Glucose Ethanol Carbon dioxide
Yeasts metabolise glucose to ethanol primarily by the Embden–

Meyerhof pathway. The overall net reaction involves the production of
two moles of ethanol per mole monosaccharide fermented, but the yield
attained in practical fermentations does not usually exceed 90–95% of the
theoretical value. This is partly due to the requirement for some nutrient
in the synthesis of new biomass and cell maintenance-related reactions.
4.3.5 Future Developments
As the world becomes a global village, new markets interested in

the diversity of cultures appear. Surely, there is a business opportu-
nity in bottling, branding and cleaning up the image of traditional
Indonesian beverages like Bekonang ciu.
References
Artjariyasriwong, S. 2003. Development of loog-paeng lao for good quality
satoh production. Thailand Institute of Scientific and Technological Research
Newsletter 6:6–7.
Basuki, T., D.S. Dahiya, Q. Gacutan, et al. 1996. Indigenous fermented foods
in which ethanol is a major product. In Handbook of Indigenous Fermented
Foods, Second ed., 1996, edited by K.H. Steinkraus. New York: Marcel
Dekker.
Casey, G.P., and W.M. Ingledew. 1986. Ethanol tolerance in yeasts. CRC
Critical Reviews in Microbiology 13:219–280.
Chaownsungket, M. 1978. Selection of the yeast and mold strains for rice
wine production. M.Sc. Thesis, Graduate School, Kasetsart University,
Bangkok, Thailand. 116 p. (in Thai).
Chi, Z., Z. Chi, G. Liu, F. Wang, L. Ju and T. Zhang. 2009. Saccharomyces fibuligera
and its applications in biotechnology. Biotechnology Advances 27:423–431.
Chuenchomrat, P., A. Assavanig and S. Lertsiri. 2008. Volatile flavor com-

pounds analysis of solid state fermented Thai rice wine (Ou). Science Asia
34:199–206.
Crabb, W.D. 1999. Commodity scale production of sugars from starches.
Current Opinions in Microbiology 2:252–256.
D’Amore, T., C.J. Panchal, I. Russell, and G.G. Stewart. 1990. A study of etha-
nol tolerance in yeast. CRC Critical Reviews in Microbiology 9:287–304.
Dung, N.T.P. 2009. Study of improved distilling method for production of rice
alcohol. Journal of Science of Cantho University, Vietnam 11:365–373.
Dung, N.T.P., F.M. Rombouts and M.J.R. Nout. 2005. Development of
defined mixed-culture fungal fermentation starter granule for controlled
production of rice wine. Innovative Food Science and Emerging Technologies
6:429–441.
Dung, N.T.P., F.M. Rombouts and M.J.R. Nout. 2006. Functionality of
selected strains of moulds and yeasts from Vietnamese rice wine starters.
Dung, N.T.P., F.M. Rombouts and M.J.R. Nout. 2007. Characteristics of
some traditional Vietnamese starch-based rice wine fermentation starters
(men). LWT-Food Science and Technology 40:130–135.
Dung, N.T.P. and H.X. Phong. 2011. Application prospects for the innovation
of defined fungal starter in rice wine fermentation. Journal of Life Sciences
5: 255–263.
Fahrasmane, L., B. Ganou-Parfait and A. Parfait. 1988. Yeast flora of Haitain
rum distilleries. MIRCEN Journal of Applied Microbiology and Biotechnology
4:239–241.
Fahrasmane, L., B. Ganou-Parfait, F. Bazile and P. Bourgeois. 1996. Rum
technology and typicality factors in the French West Indies. Cahiers
Agricultures 5:83–88.
Fleet, G.H., ed. 1993. Wine Microbiology and Biotechnology. New York: Taylor
& Francis.
Godeliva D.S. 2008. Drinking in a village tradition. Newspaper of East Java
Province, December 30.
Haard, N.F., S.A, Odunfa and C.H. Lee. 1999. Fermented cereals. A global
perspective. FAO Agricultural Services Bulletin 138:63–97.
Hesseltine, C.W., R. Rogers and F.G. Winarno. 1988. Microbiological studies
on amylolytic oriental fermentation starters. Mycopathologia 101:141–155.
Holzapfel, W.H. 1997. Use of starter cultures in fermentation on a household
scale. Food Control 8:241–258.
Kozaki, M., F.L. Lee, K.R. Kuswanto and H. Iino. 1992. On the yeasts isolated
from “Chiu” moromi mash in Java Island. Bulletin of the Doctoral Course of
Science for Living System, Showa Women’s University 2:77–82.
Krusong, W. 2008. Final Report of Process Development of Fermented
Vinegar from Organic Rice. Industrial Technology Assistance Program,
Technology Management Center, National Science and Technology
Development Agency, Bangkok, Thailand. (In Thai).
Lee, A.C. and Y. Fujio, 1999. Microflora of banh men, a fermentation starter
from Vietnam. World Journal of Microbiology and Biotechnology 15:51–55.
Lim, G. 1991. Indigenous fermented foods in South East Asia. ASEAN Food
Journal 6:83–101.
Limtong, S., S. Sintara, P. Suwanarit and N. Lotong. 2002. Yeast diversity
in Thai traditional fermentation starter (loog-pang). Kasetsart Journal
(Natural Science) 36:149–158.
Luong, N. D. 1998. Microbiological Technology, Vietnam. Ho Chi Minh City:
Technology University.
Motarjemi, Y. and M.J.R. Nout. 1996. Food fermentation: a safety and nutri-
tional assessment. Bulletin of the World Health Organization 74:553–559.
Najafpour, H., H. Younesi and K. Syahidah. 2004. Ethanol fermentation in
an immobilized cell reactor using Saccharomyces cerevisiae. Bioresource
Technology 92:251–260.
Ndip, R.N., J.-F.K.T. Akoachere, L.L. Dopgima and L.M. Ndip. 2001.
Characterization of yeast strains for wine production. Applied Biochemistry
Nout, M.J.R. and K.E. Aidoo. 2002. Asian Fungal Fermented Foods. In The
Mycota. Vol. X Industrial Applications, edited by H.D. Osiewacz. Berlin:
Springer-Verlag.
Numchaisiwattana, S. and S. Chaiseri. 2006. Aroma compounds in rice wine
fermented by Amylomyces sp. M2 and Saccharomyces cerevisiae. Proceedings
of the 44th Kasetsart University Academic Conference: Agro-industry
section. Kasetsart University,Thailand. pp. 421–429. (In Thai).
Phuc, N.H. 1998. The Fermentation Processes of Traditional Foods, Vietnam.
Ho Chi Minh City: Agriculture Publisher.
Puangwerakul, Y. 2002. Microbiology of sato and sato production. National
Food Institute Journal 5:64–80. (In Thai).
Raffles, T.S. 1830. The History of Java, Second ed. London: John Murray.
Rittiplang, J., P. Laopaiboon, K. Vichitpan and P. Danvirutai. 2007. Lactic acid
bacteria from indigeneous loog-pang samples of northeastern Thailand
and their lactic acid production ability. Thai Journal of Biotechnology
8:44–48.
Sanchez, O. J. and C.A. Cardona, 2008. Trends in biotechnological produc-
tion of fuel ethanol from different feedstocks. Bioresource Technology
99:5270–5295.
Sanchez, P., B.O. Juliano, V.T. Laude and C.M. Perez. 1988. Nonwaxy rice for
tapuy (rice wine) production. Cereal Chemistry. 65:240–243.
Schindler, R., B. Lendl and R. Kellner. 1998. Simultaneous determination of
α-amylase and amyloglucosidase activities using flow injection analysis
with Fourier transform infrared spectroscopic detection and partial least-
squarer data treatment. Analytica Chimica Acta 366:35–43.
Sharma, S.C. 1997. A possible role of trehalose in osmotolerance and ethanol
tolerance in Saccharomyces cerevisiae. FEMS Microbiology Letters 152:11–15.
Siebenhandl, S., L.D. Trimmel and E. Berghofer. 2001. Study on Tape Ketan –
an Indonesian fermented rice food. International Journal of Food Sciences
and Nutrition 52:347–357.
Sirisantimethakorn, L., L. Laopaiboon, P. Danvirutai and P. Laopaiboon. 2008.

Volatile compounds of a traditional Thai rice wine. Biotechnology 7:505–513.
Snow, P. and K. O’Dea. 1981. Factors affecting the rate of hydrolysis of starch
in food. The American Journal of Clinical Nutrition 34:2721–2727.
Steinkraus, K.H. 1997. Classification of fermented foods: worldwide review of
household fermentation techniques. Food Control 8:311–317.
Tamang, J.P. and Sarkar, P.K. 1995. Microflora of murcha: an amylolytic fer-
mentation starter. Microbios 81:115–122.
Tammarate, P., J. Kumnuanta, N. Takada and Y. Oshima. 1982. A study on
yeasts isolated from toddy and loogpang and the selected strains for alco-
hol production. Proceedings of the 20th Kasetsart University Academic
Conference: Agro-industry section, Kasetsart University, Thailand, pp.
75–87. (In Thai).
Thai Industrial Standards Institute. 2003. Product certificate: Satoh. No.3/
BE2546. Ministry of Industry, Bangkok, Thailand, 7 p. (In Thai).
5
L acti c Veg e table and
Fruit Fermentati ons
L I L I S N U R A I DA , J . DAV I D O W E N S ,
J U N A I DA H A B U B A K A R A N D
K A P T I R A H AY U K U S WA N T O
Contents
5.1 Introduction 186

5.2 Tempoyak (Fermented Durian Fruit) 186
5.2.2 Production Methods 187
5.2.3.1 Metabolism of Lactobacillus durianis and
Leuconostoc durionis 190
5.2.4 Chemical Composition and Changes during
Fermentation 192
5.2.8 Research Needs and Opportunities 196
5.3 Indonesian Growol (Fermented Cassava) 197
5.3.2 Places of Production, How Consumed and
Role in Diet 197
5.3.3 Traditional Production Methods 198
5.4 Bruneian Budu Pakis (Fermented Fern Fronds) 202
5.4.1 Description of the Product 202
18 5
5.4.2 Traditional Production Methods 203

5.4.3 Substrate Changes during Fermentation 204
5.4.4 Characteristics of the Microorganisms 205
References 206
5.1 Introduction
A wide variety of lactic acid bacteria fermented vegetables and fruits

are prepared in Southeast Asian countries (see Chapter 1; Steinkraus
1996). While the available evidence suggests that the microbiology
is, in general, similar to that of other fermented vegetables (Hutkins
2006), there has been little research on the Southeast Asian products.
The three products described in this chapter, tempoyak (fermented
durian fruit), growol (fermented cassava starch) and budu pakis (fer-
mented fern fronds), are among the more interesting and distinctive
fermented vegetable and fruit products of the region.
5.2 Tempoyak (Fermented Durian Fruit)

Lilis Nuraida and J. David Owens
Tempoyak is fermented durian (Durio zibethinus) fruit pulp and has

a strong durian smell and sour taste. Owing to the strong durian
flavour, consumption of tempoyak is limited to persons who usually
eat and like durian. Tempoyak is popular in Indonesia, especially in
southern Sumatra and also in Malaysia. In contrast to durian fruit,
which is consumed fresh, tempoyak is usually used as an ingredient to
make a condiment as an appetiser or is used in the traditional cooking
of fish and vegetable dishes.
There is no information available on when the tradition of mak-
ing tempoyak began. Sumatra island has long been known as a
durian-producing area. Fermentation of durian to tempoyak serves
to preserve surplus, over-ripe or poor-quality durian fruit during the
durian season. While durian fruit must be eaten within days, tem-
poyak usually has a shelf life of several months. Raw tempoyak is sold
L ac ti c V eg e ta b l e a n d F ruit F erm en tati o ns 18 7
in traditional markets in earthenware or aluminium containers. To

make tempoyak condiment, called sambal tempoyak, raw tempoyak is
mixed with spices, such as chilli, garlic, shallot, turmeric, sometimes
anchovy, a drop of cane sugar and then stir fried.
5.2.2 Production Methods
Fermentation of durian into tempoyak is done in the traditional way

as spontaneous fermentation. Durian fruit pulp is placed in jars and
mixed with salt. The amount of salt added to the pulp varies from
one production place to the other, ranging from 2% to 15% (Yuliana
and Garcia 2009). However, two samples analysed by Leisner et al.
(2001) contained no added salt and the authors concluded that it is
possible to make acceptable tempoyak without the addition of salt.
The container is tightly closed and kept for 7 days at ambient tem-
peratures to allow fermentation to occur. Traditionally, the container
used was earthenware, but nowadays, plastic, aluminium or glass jars
are used.
As with other spontaneous fermentations, salt concentration and
anaerobic conditions are the key factors for the success of tempoyak
fermentation. Efforts to improve fermentation have been very lim-
ited. Amiza et al. (2004) evaluated the effect of salt concentration
on the sensory characteristic of tempoyak. The salt concentration
significantly affected the sourness, sweetness, colour and aroma of
tempoyak, with the most preferred tempoyak being that with 2% salt.
Amiza et al. (2006) compared the fermentation of pasteurised durian
pulp inoculated with an unidentified lactic acid bacteria isolated from
tempoyak with that of pulp allowed to ferment naturally. Both fer-
ments achieved similar maximum lactic acid bacterial populations
of ~1.5 × 109 cfu g−1 after 4 day incubation at room temperature. The
naturally fermented tempoyak did exhibit a lower titratable acidity
(2.35% as lactic acid) than the inoculated tempoyak (2.75%) but this
observation is un-interpretable in the absence of information on the
specific acids present. There have been no observations made on the
sensory characteristics of the products. Yuliana and Garcia (2009) used
Pediococcus acidilactici as a starter culture for tempoyak fermentation.
Although no difference in sourness and aroma was noted between
inoculated tempoyak and spontaneously fermented tempoyak, there

was a significant difference for general acceptability, with the inocu-
lated tempoyak being more acceptable than the spontaneously fer-
mented product.
5.2.3 Microbiology
As one would expect in a sugar-rich, near-neutral pH, anaerobic

environment, the fermentation is dominated by lactic acid bacteria,
with populations reaching 108–1010 cfu g−1 and the pH value falling
to 4.0–4.2 (Leisner et al. 2001; Wirawati 2002; Amiza et al. 2004;
Yuliana and Garcia 2009). Wirawati (2002) observed that concentra-
tions of lactic acid bacteria decreased as the age of tempoyak increased
and that lactic acid bacteria were not detected in tempoyak kept for
more than 4 weeks. Moulds and yeasts were present in tempoyak
that were more than 4 weeks old. The pH of tempoyak kept for more
than 4 weeks fell to below 3, and Wirawati (2002) suggested that the
decrease in pH killed the lactic acid bacteria and allowed moulds and
yeasts to grow. Apart from their tolerance of low pH, the growth of
moulds and yeasts in tempoyak kept for more than 4 weeks may also
be associated with an ability to assimilate lactic acid.
Amiza et al. (2004) examined the effect of salt concentration on
the concentrations of lactic acid bacteria. Higher levels of lactic acid
bacteria were found in tempoyak with 1% and 2% salt than in tem-
poyak with 0% and 3% salt. The maximum count was observed after
4 days fermentation and it then gradually decreased.
A variety of lactic acid bacteria have been isolated from tempoyak
(Table 5.1), including some unusual strains that are unable to ferment
glucose on their own and which have been proposed as two new spe-
cies, Lactobacillus durianis (Leisner et al. 2002) and Leuconostoc durio-
nis (Leisner et al. 2005).
Lactobacillus durianis (Leisner et al. 2002) is an obligately heterofer-
mentative, rod-shaped bacterium producing d- and l- isomers of lactic
acid and acetic acid. Growth on glucose is poor and without evident gas
production in APT broth (Leisner et al. 2001) but gas is produced from
gluconic acid. Acid is produced from a relatively small number of sugars,
including maltose and the pentoses, l-arabinose, ribose and d-xylose.
On d-xylose, growth occurred from 10°C to 42°C for two strains and
Table 5.1 Lactic Acid Bacteria Associated with Tempoyak

SOURCE OF
TEMPOYAK LACTIC ACID BACTERIA REFERENCE
Indonesia Lactobacillus plantarum Wirawati (2002)
L. coryneformis Wirawati (2002)
L. casei Wirawati (2002)
Malaysia Lactobacillus plantarum Leisner et al. (2001)
L. brevis Leisner et al. (2001)
L. mali Leisner et al. (2001)
L. fermentum Leisner et al. (2001)
Lactobacillus sp. Leisner et al. (2001)
Leuconostoc mesenteroides Leisner et al. (2001)
Lactobacillus durianis Leisner et al. (2002)
Leuconostoc durionis Leisner et al. (2005)
Philippines Lactobacillus plantarum Yuliana and Dizon (2011)
Lactobacillus sp. Yuliana and Dizon (2011)
Pediococcus acidilactici Yuliana and Dizon (2011)
Weisella paramesenteroides Yuliana and Dizon (2011)
from 15°C to 42°C for one strain. Dellaglio et al. (2006) showed that
Lactobacillus durianis is a later heterotypic synonym of Lactobacillus vac-
cinostercus, a species originally isolated from dried cow dung.
Leuconostoc durionis (Leisner et al. 2005) is an obligately heterofer-
mentative rod-shaped bacterium producing d-lactic acid and acetic
acid but not ethanol. All three strains exhibited limited growth with
glucose (without gas production), fructose, sucrose, ribose or xylose
in MRS broth. No gas production was observed from glucose alone
but abundant gas was produced in MRS broth with glucose and fruc-
tose. Growth occurred at 5°C and 35°C but not at 42°C. Endo and
Okada (2008) proposed that Leuconostoc durionis be reclassified as
Fructobacillus durionis. Fructobacillus durionis, along with other species
in the genus, is osmotolerant and grows well in the presence of 40%
fructose but growth is poor with 50% fructose.
Leisner et al. (2001) pointed out that the poor growth of Lactobacillus
durianis and Leuconostoc durionis on glucose could lead to their num-
bers being under-estimated in samples if media containing glucose
as the only fermentable carbohydrate were used. Endo and Okada
(2008) suggested that Leuconostoc durionis and related species may have
adapted to environments rich in d-fructose, such as flowers, fruits and
related products, and recommended that isolation of lactic acid bacteria
from such environments should be done using glucose and fructose as
substrates. It is also evident that their heterofermentative metabolism

may be missed if the usual gas-detecting media with glucose are used.
Among the predominant lactic acid bacteria found in tempoyak,
Lactobacillus plantarum was identified by all authors (Table 5.1).
Lactobacillus plantarum is very commonly present in the later stages
of vegetable fermentations (Pederson 1996; Li 2004; Maki 2004) and
contributes the biggest proportion of lactic acid in the final product.
Partly, this is because it is more tolerant of low pH values and high
acid concentrations than most lactic acid bacteria. Often, vegetable
fermentations are initiated by Leuconostoc mesenteroides, which exhibits
rapid growth over a wide range of temperatures and salt concentra-
tions but, of all the authors listed in Table 5.1, it was only isolated
from tempoyak by Leisner et al. (2001).
Apart from lactic acid bacteria, Yuliana (2005) reported the pres-
ence of Bacillus megaterium in tempoyak, but whether or not it had
grown in the durian pulp or was merely an incidental contaminant
was not established.
5.2.3.1 Metabolism of Lactobacillus durianis and Leuconostoc durio-

nis Lactobacillus durianis and Leuconostoc durionis share a common
property in that the growth of both is poor with glucose alone and
both appear to require the presence of an external electron acceptor to
utilise it effectively (Endo and Okada 2008). With an accessory elec-
tron acceptor, they produce equimolar amounts of lactic acid, acetic
acid and CO2 without any ethanol. Most heterofermentative lactic
acid bacteria ferment hexoses to lactic acid, ethanol and CO2. Ethanol
is formed from acetyl phosphate due to the need to re-oxidise NADH/
NADPH (Figure 5.1). If an extraneous electron acceptor is available
for the re-oxidation of NDAH/NADPH, the acetyl phosphate can
be converted to acetic acid with the formation of an additional ATP
molecule, giving an overall yield of 2 ATP per mole hexose fermented
(Figure 5.2). Compounds that may serve as electron acceptors include
oxygen, fructose and pyruvate:
O2 + 2H+ + 2e− → H2O2
C6H12O6 + 2H+ + 2e− → C6H14O6
fructose → mannitol
L ac ti c V eg e ta b l e a n d F ruit F erm en tati o ns 191
Glucose
C6H12O6
ATP 2NAD+
ADP 2NADH
CO2
C5H10O5-P
TRIOSE-P ACETYL-P
2ADP 2NAD+ NADH
2ATP 2NADH NAD+
CH3COCOOH CH3CHO
2NADH NADH
2NAD+ NAD+
CH3CHOHCOOH CH3CH2OH
LACTIC ACID ETHANOL

Overall reaction:
C6H12O6 + ADP → CH3CHOHCOOH + CH3CH2OH + CO2 + ATP
Figure 5.1 Heterolactic fermentation of glucose.
CH3COCOO− + 2H+ + 2e− → CH3CHOHCOO−
pyruvate → lactate
If glucose is fermented to lactate and acetate with fructose as elec-

tron acceptor, the overall reaction becomes
C6H12O6 + 2C6H12O6 + 2OH− → CH3CHOHCOO −
+ CH3COO− + 2C6H14O6 + CO2 + H2O
+ 2 fructose + 2OH− → lactate + acetate +
glucose
2 mannitol + CO2 + H 2O
Leuconostoc durionis produces large amounts of mannitol when

grown on glucose + fructose (Leisner 2005; Endo and Okada 2008)
Glucose
C6H12O6
ATP 2NAD+ ALTERNATIVE
ADP 2NADH E ACCEPTOR
CO2
C5H10O5-P
TRIOSE-P ACETYL-P
2ADP 2NAD+ ADP
2ATP 2NADH ATP
CH3COCOOH
2NADH
2NAD+
CH3CHOHCOOH CH3COOH
LACTIC ACID ACETIC ACID
Overall reaction:
C6H12O6 + 2ADP → CH3CHOHCOOH + CH3COOH + CO2 + 2ATP
Figure 5.2 Heterolactic fermentation of glucose with an alternative electron acceptor.
and Endo and Okada (2008) noted that it grew better with glu-
cose + pyruvate than with glucose alone. It also grew better aerobi-
cally than anaerobically and Endo and Okada (2008) suggested that
these observations indicated that Leuconostoc durionis could use all
three compounds as external electron acceptors.
5.2.4 Chemical Composition and Changes during Fermentation
The edible portion of durian is rich in sugars, with 15–20% of total

sugars, mainly sucrose (Table 5.2). Unlike most fruits, durian flesh is
not acidic but has a pH value of ~6.2 and this allows the fermentation
of the sugars in tempoyak to be a lactic acid bacteria one rather than
an alcoholic yeast fermentation. Research on tempoyak prepared in
the laboratory (Amiza et al. 2004; Yuliana and Garcia 2009) showed
an increase in lactic acid content concomitant with a decrease in pH
Table 5.2 Chemical Composition of Fresh Durian Fruit Flesh and the Fermented Product, Tempoyak
COMPOSITION (G KG−1)
COMPONENT FRESH DURIAN TEMPOYAK REFERENCE
Water 650 626 Ministry of Health (1992),
700 Mahmud et al. (2009)
Energy 1340 kcal 1260 kcal Ministry of Health (1992),
1100 kcal Mahmud et al. (2009)
Protein 25 11 Ministry of Health (1992),
Fat 30 22 Ministry of Health (1992),
Carbohydrate 280 255 Ministry of Health (1992),
Salt <0.1 20–150 Yuliana and Garcia (2009)
<0.1, <0.1 Leisner et al. (2001)a
Glucose 10 0b , 1 Leisner et al. (2001)
Fructose 13 0, 7 Leisner et al. (2001)
Sucrose 174 7, 2 Leisner et al. (2001)
Fructan 0 0, 2 Leisner et al. (2001)
Acetic acid 0.21 8.0, 7.0 Leisner et al. (2001)
1.4 Yuliana and Garcia (2009)
d-Lactic acid 0.18 9.8, 9.0 Leisner et al. (2001)
l-Lactic acid 0.16 9.0, 8.8 Leisner et al. (2001)
Lactic acid 3.4 Yuliana and Garcia (2009)
Malic acid 14.6 Yuliana and Garcia (2009)
Total acidity (as lactic acid) 24 Amiza et al. (2006)
pH 6.2 4.0, 4.2 Leisner et al. (2001)
a Data for two samples.
b Undetectable.
value during fermentation, with a significant decrease in pH occur-

ring during the first 2 days. Lactic acid concentrations in the final
products ranged from 2.8% to 3.6% with pH 3.9–4.1 (Amiza et al.
2004).
The main changes during fermentation are the conversion of sug-
ars to lactic and acetic acids (Table 5.2). Yuliana and Garcia (2009)
also observed an increase in malic acid, both in naturally fermented
tempoyak and in tempoyak inoculated with Pediococcus acidilactici, but
it is not obvious by what mechanism this might occur. The presence
or absence of ethanol in tempoyak has not been reported. However,
Leisner et al. (2001) did observe the formation of a small amount
of fructan in the fermentation (Table 5.2). Fructans are made by the

addition of fructose residues to a sucrose residue.
n sucrose → sucrose-(fructofuranose residues)n−1 + (n−1) glucose
Inulin is a linear fructan consisting mainly of (2 → 1)-β-linked
fructofuranose residues while levan is a linear fructan consisting
mainly of (2 → 6)-β-linked fructofuranose residues (Singleton and
Sainsbury 2001). Lactic acid bacteria that synthesise one or both
types of fructans are known (Anwar et al. 2010).
Although the durian fruit and the tempoyak differed in their ori-
gins, the data of Leisner et al. (2001) suggest that the fermentation
products assayed (d- and l-lactate, acetate and fructan; 26.8 g kg−1 in
total) represented only 15% of the sugars utilised (178 and 175 g kg−1).
Possible other products that were not assayed include ethanol and
mannitol but it seems unlikely that these, on their own, could account
for the sugars unaccounted for. However, if each mole of acetate mol-
ecule produced was associated with the production of 2 mol of man-
nitol, the production of 7.5 g kg−1 of acetate would be accompanied by
the production of 45 g kg−1 of mannitol.
The flavour of tempoyak has been attributed to a combination of
various sugars, organic acids and volatile compounds (Yuliana and
Garcia 2009) but is dominated by sourness due to the acids present.
Durian fruit has a very distinctive and characteristic aroma. Weenen
et al. (1996) studied the volatile compounds of three varieties of
durian fruits and identified 3,5-dimethyl-1,2,4-trithiolane as one of
the strongest S-containing durian odorants and ethyl 2-methylbu-
tanoate as the compound with the highest odour impact among the
non-sulphurous odorants. Li et al. (2012) made an exhaustive study
of the volatile and odour-active compounds in durian and identi-
fied 13 of the most aroma-active compounds in the fruit pulp. These
included ethyl (2S)-2-methylbutanoate and ethyl cinnamate, both
having extremely high odour factors, six S-containing compounds,
two other esters and three furans. Of these 13 compounds, Yuliana
and Garcia (2009) detected ethyl (2S)-2-methylbutanoate and eth-
ane-1,1-dithiol in naturally fermented tempoyak. Neti et al. (2011)
compared the volatile compounds in unfermented durian pulp and in
pulp naturally fermented for 8 days at room temperature (30°C). They
did not observe any dramatic changes and most of the compounds in
durian pulp were also present in tempoyak, although there were some
compounds detected in the one but not the other. This is as expected,
as tempoyak is described to have a strong durian aroma. However,
one curious observation was the failure to detect 3,5-dimethyl-1,2,4-
trithiolane in durian pulp while it was present in fermented pulp in
high concentrations.
To date, no changes in aroma compounds during tempoyak fer-
mentation have been monitored systematically.
Yuliana and Garcia (2009) also studied flavour compounds in tem-
poyak inoculated with Pediococcus acidilactici but, as Pediococcus acidi-
lactici is a homolactic fermenter of sugars, it would not appear to be
the most appropriate starter organism to use and the observations lack
relevance to naturally fermented tempoyak.
The composition of tempoyak is essentially similar to that of durian

fruit, except for having higher concentrations of organic acids and
a lower concentration of protein (Table 5.1). Thus, it is primarily a
source of carbohydrate and energy. However, since tempoyak is usu-
ally eaten as a condiment, any contribution to energy intake is prob-
ably not significant.
Wirawati (2002) showed that several lactic acid bacteria isolated
from tempoyak were potential probiotic bacteria as shown by their
ability to resist low acid and bile salts, and to suppress enteropatho-
genic bacteria. However, to function as a probiotic, the bacteria need
to be alive and, as tempoyak is not usually eaten uncooked, this is not
a consideration in practice.
There are no reports of outbreaks of illness associated with the con-

sumption of tempoyak. However, since the fermentation is uncon-
trolled and carried out on a household scale, good hygienic and
sanitary practices should be followed while producing it. Certainly,
the production of organic acids and lowering of pH during fermenta-
tion will inhibit the growth of pathogenic bacteria in the final prod-
uct. As always, the ‘at risk’ period is during the initial stages of the
fermentation before the accumulation of organic acids and the estab-

lishment of acidic conditions occurs. Here, careful control of the envi-
ronment and salt concentration is required to avoid possible growth
and survival of pathogenic bacteria.
Tempoyak consumption is still limited to certain people who are used

to eating durian, especially in Sumatra. So far, no industrial produc-
tion of tempoyak or derivative products, such as sambal tempoyak,
has occurred. Sambal is usually eaten as an appetiser and, generally
speaking, all Indonesians like sambal. Industrial production of sam-
bal tempoyak could be one opportunity to provide support to durian
and tempoyak producers. In addition, by processing tempoyak into
bottled sambal, the product will have a longer shelf life. The other
area to be explored is the preservation of fresh tempoyak for use as an
ingredient in vegetable and fish dishes.
5.2.8 Research Needs and Opportunities
The role of microorganisms during tempoyak fermentation is still

unclear and requires further elaboration by monitoring the lactic acid
bacteria during fermentation and correlating microbial changes with
chemical changes. It would be particularly interesting to do this, using
modern molecular methods, and especially to look for Lactobacillus
durianis and Leuconostoc durionis.
The critical factors necessary for tempoyak fermentation also need
further study. For sauerkraut fermentation the suggested optimum
salt level is 2.25% but for tempoyak fermentation it has not been
standardised and a wide diversity of salt concentrations are used
by different producers, leading to much variability in the quality
of tempoyak. In particular, information on the role of salt in the
fermentation is needed and, especially, whether or not it may be
omitted entirely.
It is possible that Lactobacillus durianis and Leuconostoc durionis have
lost the ability to form ethanol and it would be interesting to confirm
or refute this with physiological and genetic studies.
Figure 5.3 Steamed, ready-to-eat growol (fermented cassava starch). (Courtesy of

K.R. Kuswanto.)
5.3 Indonesian Growol (Fermented Cassava)

Growol is a traditional Indonesian food made by submerged acid fer-

mentation of cassava (Manihot utillissima, M. esculenta) (Kuswanto
1994). Growol is a dough-like material with a sticky texture, a sour
taste and a strong and pungent flavour. It is white in colour after steam-
ing (Figure 5.3) and is primarily consumed as a substitute for rice.
Cassava is consumed as a major source of calories in certain parts

of Central Java. However, growol is produced and consumed only
in areas surrounding Wates and Kulonprogo in Central Java and in
the Yogyakarta area. Traditionally, growol was made in the villages
by women at home from cassava roots bought or grown locally. In
general, roots are not used after more than 48 h after harvesting due
to biodeterioration. Growol is primarily consumed in the form of
steamed product, along with other foods, for breakfast or lunch. It
is a staple item of diet for certain low-income people, who regularly
consume it once a day.
5.3.3 Traditional Production Methods
The processing of cassava into growol (Figure 5.4) is similar to that of

West Nigerian fu-fu (Westby and Twiddy 1992). The essential features
are (1) the soaking of peeled whole or large sections of cassava root in
water until they become soft; (2) disintegration of the tissues by forc-
ing them through a sieve and discarding the retained fibrous material;
(3) some washing of the starch in fresh water; (4) dewatering by plac-
ing the material in a sack and hanging up overnight or by pressing to
obtain a paste; and (5) cooking in boiling water to the desired consis-
tency. During the soaking stage, a variety of microorganisms grow and
acidify the material and confer other important flavour characteristics.
The production of growol is a local tradition and the equipment used
is still entirely traditional and unmodernised. Tubers from 1–2-year-
old sweet cassava are preferred. Sweet cassava is used because of the
lower content of cyanogenic glucosides than in non-sweet varieties.
Within 48 h after harvesting, the roots are processed by removing the
ends and chopping the remainder into pieces. The roots are peeled with
sharp kitchen knives and washed. Peeling losses are 25–30% on a wet
weight basis. Traditionally, the peeled whole roots are steeped in water
in a clay pot with excess well water and left to ferment naturally for 4–6
days with the water being changed everyday. It is important to change
the soak water everyday to prevent the development of a pungent fla-
vour. Owing to the content of cyanogenic glucosides, cassava tubers
Cassava root
Wash, peel and chop into pieces
Soak in water and change water every day
Ferment for 3 to 5 days (until the tuber is soft)
Wash and press under stone weights to remove water
Remove coarse fiber
Ferment starch
Steam for 30 to 60 min
Growol
Figure 5.4 Preparation of growol as practiced in Wates, Yogyakarta, Indonesia.

must be processed prior to consumption to lower the content of cya-

nide. During fermentation, the cyanide content of the roots is reduced
by 90–100% and the texture changes from hard and brittle to soft
and mushy, making it easier to extract the cassava starch. Using sieves
locally fabricated, usually from bamboo, the coarse fibres are removed
and discarded. The starch is then squeezed in the sieve to separate and
remove the inner cortical cellulosic materials and the starch is washed
4–5 times and drained, or the material is put in a bag and squeezed to
remove the liquid. After draining, the cassava pulp, composed mainly
of starch, is steamed for 1 h. The product has a distinctive flavour due
to the fermentation that occurs during soaking. The natural fermenta-
tion takes at least about 5 days to reach the required acidity (0.85% as
lactic acid) at ambient temperatures of about 30°C.
5.3.4 Microbiology
Rascana (1988) and Widowati (1994) studied the microorganisms

present during the soaking period of growol preparation. Oxygen
limitation in the submerged fermentation favours the growth of
lactic acid bacteria and, therefore, at the end of fermentation, the
pH is quite acidic with a pH of ~4.0. Owing to chemical changes
during fermentation, a succession of dominant microbes occurs.
The main microorganisms present during fermentation were lac-
tic acid bacteria and yeasts. The bacteria included Streptococcus,
Lactobacillus, Bacillus, Enterobacter, Acinetobacter, Moraxella and
Micrococcus species. On the first day of fermentation, the micro-
organisms involved were coryneform bacteria and Streptococcus,
Bacillus and Acinetobacter species. Amylase produced by Bacillus spp.
hydrolysed the starch to simple sugars, which were then fermented
with the production of acid and a lowering of the pH. During the
second day of fermentation, concentrations of Streptococcus and
Acinetobacter decreased while concentrations of yeasts, Lactobacillus
spp. and some other microorganisms increased. At the end of the
fermentation, after 5 days, the organisms that were present included
species of Lactobacillus, Streptococcus, Acinetobacter, Corynebacterium
and Enterobacter. After 6 days of fermentation, the microorganisms
identified were Streptococcus spp., coryneform bacteria, Moraxella
spp. and yeasts (Endang et al. 1996).
The lactobacilli found in growol and identified belonged to the

facultatively heterofermentative group of lactobacilli. They produced
dl-lactic acid, contained meso-diaminopimelic acid in their peptido-
glycan and were identified as belonging to the Lactobacillus plantarum/
Lactobacillus pentosus complex.
Westby and Twiddy (1992) and Brauman et al. (1996) studied the
microflora during the retting of cassava for the preparation of fu-fu.
The fermentation was dominated by lactic acid bacteria but, whereas
Westby and Twiddy noted that homofermentative organisms domi-
nated in the early stages and heterofermentative ones in the later
stages, Brauman et al. observed that the initial flora was first replaced
by Lactococcus lactis (homofermentative), then by Leuconostoc mesen-
teriodes (heterofermentative) and finally by Lactobacillus plantarum
(homofermentative on hexoses). Yeasts appeared after 48 h but were
not considered to play a significant role in the retting process.

The major biochemical activities that occur in the fermentation are

(1) softening of the cassava tissues; (2) acidification of the starch; (3)
production of flavour compounds; and (4) detoxification of any cyanide
present. Little work in these areas has been reported for growol and
most of the following information is derived from studies on fu-fu.
Freshly peeled cassava tuber has a water content of about 71.5%,
carbohydrate content of 27% and protein content of 0.74% (Akinrele
et al. 1965). It has a relatively high ascorbic acid content, ranging from
1.2 to 1.7 g kg−1 in fresh tubers, but ascorbic acid was undetectable in
growol. Amylase activity was found in a Lactobacillus strain isolated
from cassava soak water. This amylase was able to degrade the raw
starch in cassava (Kuswanto, unpublished), suggesting that, during
the soaking of cassava, some biodegradation of starch occurred. The
stickiness of growol after steaming is due to the dextrinisation and
gelatinisation of the starch. The pH of fresh cassava tuber is about 6.2
and this is reduced to 4.0 in 4 days of fermentation and remains at this
level till the end of the soaking period.
Softening of the plant tissue is a consequence of hydrolysis of pec-
tin of the middle lamella. Okafor et al. (1984) obtained softening of
cassava pieces only in the presence of a Bacillus sp. or a Corynebacterium

sp. Brauman et al. (1996) observed high levels of pectin methylester-
ase activity in cassava roots at the start and throughout the fermen-
tation, while much lower levels of polygalacturonate lyase appeared
only after 24 h fermentation. Brauman et al. did not identify the
sources of the enzymes and the exact mechanism(s) of tissue soften-
ing is not clear.
The fermentations were dominated by lactic acid bacteria and their
activities lowered the pH from an initial value of 6.1–6.5 to a value
of ~4.5 after 2–3 days fermentation and to ~4.0 after overnight dewa-
tering of the fermented material (Brauman et al. 1996; Westby and
Twiddy 1992). The principal fermentation products were lactic acid
(maximum concentration of 20 g [kg dry matter]−1), acetic acid (4.7)
and ethanol (1.8). Cassava contains significant amounts of sucrose
(20–40 g [kg dry matter]−1, glucose (13 g [kg dry matter]−1) and fruc-
tose (10 g [kg dry matter]−1) (Brauman et al. 1996) and these amounts
are sufficient to support substantial microbial growth without the
need for any hydrolysis of starch.
Brauman et al. (1996) noted the presence of low concentrations of
clostridia, similar to Clostridium butyricum, and ascribed the pres-
ence of butyric acid, reaching a concentration of 3.8 g [kg dry mat-
ter]−1 by the end of the fermentation, to their activities. However, the
maximum concentration detected was only 103.5 cfu [g dry matter]−1,
which appears to be too low to produce the concentration of butyric
acid observed.
To date, no studies have reported on the flavour compounds in gro-
wol or fu-fu but it may be presumed that lactate, acetate and ethanol
all make some contribution. Brauman et al. (1996) attributed the dis-
tinctive flavour of fu-fu to butyrate, which could also contribute to the
pungent aroma of growol. Aldehydes and esters also contribute to the
characteristic aroma of growol.
Growol is primarily a good source of calories and provides bulk in the

diet. Gelatinising and dextrinising of the starch may render it more
digestible. The protein content in cassava is only about 1% and it is
not clear how much of this is retained in growol. Growol is also low
in minerals and vitamins, particularly niacin, riboflavin (vitamin B2)

and ascorbic acid (vitamin C) (Endang 2000).
As with all cassava products, the cyanide content of the roots is an

important safety concern but there are no reports of cyanide levels
in growol. It is well established that the detoxification of cassava is a
consequence of the hydrolysis of the cyanogenic glycosides, liminarin
and lotaustralin, to the corresponding cyanohydrins by endogenous
linamarase, released by physical or enzymic tissue maceration, fol-
lowed by spontaneous hydrolysis of the cyanohydrins and evaporation
of the released HCN (Vasconcelos et al. 1990). Growol is tradition-
ally made from sweet varieties of cassava that contain only low levels
of cyanogenic glycosides and there is, therefore, little risk of cyanide
poisoning. Certainly, there are no recently documented cases of acute
poisoning due to growol consumption.
A problem with cassava is its limited shelf life once harvested, as after
48 h, biodegradation of the tubers begins. Growol is one example of
a product that preserves the cassava starch. However, many people
in growol-producing villages, and especially the younger generation,
do not like to eat growol and its production is, therefore, decreasing.
Hence, there would seem to be little prospect of its production being
expanded or industrialised.
5.4 Bruneian Budu Pakis (Fermented Fern Fronds)

Junaidah Abu Bakar
5.4.1 Description of the Product
Bruneians use the term budu to refer to foods with added salt and
water that are placed in containers and left at room temperature until
the desired changes occur. Pakis is Malay ‘fern’ and budu pakis is fer-
mented Diplazium esculentum, an edible terrestrial fern found grow-
ing wild on the banks of streams and on wet ground in open places in
Figure 5.5 Young pakis (Diplazium esculentum fronds). (Courtesy of J. Abu Bakar.)
the lowlands. The fern is a popular vegetable and is collected and sold
in local markets (Figure 5.5), including supermarkets in Brunei. The
curled tips and young, tender upper parts of the leaves or fronds are
stripped off the stalk and cooked as a leafy vegetable or used in salads.
To make budu pakis, fern tips are immersed in a salt–rice–water
mixture and allowed to ferment for at least 3 days before being con-
sumed. The fermented Diplazium esculentum (budu pakis) is crunchy,
sour and slightly salty. It is generally served uncooked, with meals
as a side dish. Not many people are now familiar with budu pakis,
although it is well known to the older generation, especially water
villagers and farmers. Many do not even know that pakis can be fer-
mented, as in earlier times, this fermented vegetable was only made at
home and was kept in a closed vessel.
5.4.2 Traditional Production Methods
Budu pakis is only made at the domestic level and is not commercially
available in local markets. Only the curled tips of the fern are used
and the rest of the frond is discarded. Fresh young pakis fronds are
trimmed and the tips are washed twice in water and drained. One
hundred gram of partially cooked rice plus 66 g of salt is boiled in 3 L
of water for 15 min. This provides a salt concentration of ~21 g L −1,
which is in the normal range for lactic vegetable fermentations. The
mixture is allowed to cool to room temperature (27–30°C). Clean
Pakis (Diplazium esculentum) fronds Water (3000 mL)
Trim pakis, keeping only young tips Rice, partially
Wash and drain tips

Salt, 66 g
Boil, 15 min
Cool to 27–30ºC
~200 mL quantities
Small clean bowls
Cover
Incubate, ambient temperature (27–30ºC), 3 days
Budu pakis
Figure 5.6 Traditional production of budu pakis (fermented fern frond tips).
containers, such as small plastic or ceramic bowls, are disinfected by

rinsing in boiling water and about 200 mL of the rice–salt mixture
is poured into each. The pakis tips are added to the container where
the tips sit loosely in the rice–salt–water mixture. The containers are
covered with aluminium foil, leaving a headspace of 5–7 cm between
the surface of the liquid and the aluminium foil, and stored at room
temperature (27–30°C). The fermentation takes about 3 days and the
product is then ready for consumption (Figure 5.6). Budu pakis is
normally consumed uncooked as a side dish and without further pro-
cessing. At room temperature, the period of acceptability only lasts for
1 week or less, after which the pakis turns yellowish and becomes very
sour (Abu Bakar and Md Noor 1992). If refrigerated, it will keep for
a month (Abu Bakar and Md Noor 1992).
5.4.3 Substrate Changes during Fermentation
The pH decreased from 6.9 at the beginning of the fermentation to 4.8

after 7 days of fermentation while the titratable acidity (expressed as lac-
tic acid) increased from 0.006 to 0.13 g kg−1. Reducing sugar (expressed
as d-glucose) increased from 0.14 g kg−1 to a maximum of 0.72 g kg−1

wet weight after about 96 h of fermentation and then declined to
0.27 g kg−1 at 7 days (Abu Bakar and Md Noor 1992). The final titrat-
able acidity was very low compared with the concentrations (5–20 g kg−1
as lactic acid) present in most vegetable fermentations, but could occur
if the buffering capacity of fern fronds is much lower than that of other
vegetables. The final pH (~4.8) was also much higher than is usually
observed in vegetable fermentations (typical final pH, 3.5–4.1).
There is little control exercised over the process and thus the salt
concentration and extent of the fermentation vary widely. This is evi-
dent when samples from home production are tasted; the budu are
commonly either too salty or too sour (Abu Bakar and Md Noor 1992).
5.4.4 Characteristics of the Microorganisms
Organisms isolated from budu pakis included Enterobacter cloacae,

Enterobacter aerogenes, Klebsiella pneumonia, Pseudomonas fluorescens,
Gram-positive cocci, yeasts and some moulds (Abu Bakar and Md
Noor 1992). A high count of up to 5.6 × 109 cfu g−1 of unidentified
lactic acid bacteria indicates that the fermentation is primarily due
to lactic acid bacteria (Ayres et al. 1980).
There are relatively few chemical analyses of the frond tips and those
that have been done differ considerably in the values reported (Table 5.3).
According to Maranon (1935), the leaves are a good source of protein,
vitamin B, iron, calcium and an excellent source of phosphorus.
Budu pakis will, presumably, have a composition similar to pakis
but with the addition of salt and rice starch.
As with other vegetable fermentations, the low pH will inhibit

the growth of spoilage and food poisoning bacteria. However, the
observed acid concentrations, if correct, are very low and one would
question if they are sufficient to inhibit all potential spoilage organ-
isms or pathogens.
Table 5.3 Composition of Fresh Pakis (Frond Tips of Diplaziun esculentum)

COMPOSITION (G [KG WET WEIGHT]−1)
SOURCE OF DATAa
COMPOUND Ab Bb C
Moisture n.d.d 907 876b
Dry matter n.d. 93 124b
Protein n.d. 9.3 18c
Crude fat n.d. 10 0.2c
Fibre n.d. 35 4.8c
Ash n.d. 55 15c
K 4.1 5.4 n.d.
P 0.84 n.d. n.d.
Mg 0.19 n.d. n.d.
Ca 0.14 1.1 n.d.
Na n.d. 0.15 n.d.
Fe 0.009 0.03 n.d.
Mn 0.00003 0.006 n.d.
Cu 0.00002 0.003 n.d.
Zn 0.000006 0.02 n.d.
a (A) Department of Agriculture, Sarawak (1992); (B) Abu Bakar and Md Noor (1992);
(C) Seal (2012).
b Original data.
c Values calculated from original data.
d No data.
References
Abu Bakar, J. and L.K. Md Noor. 1992. Studies on the nutrient, microflora and
chemical changes on the fermentation of budu pakis (Diplazium esculen-
tum). (Unpublished).
Akinrele, I.A., A.S. Cook and R.A. Holgate. 1965. The manufacture of gari
from cassava in Nigeria. In First International Congress of Food Science and
Technology Vol. IV, edited by J.M. Leith. New York: Gordon and Breach.
Amiza, M.A., J. Zakiah and L.K. Ng. 2004. Effect of salt on tempoyak fer-
mentation and sensory evaluation. Journal of Biological Sciences 4:650–653.
Amiza, M.A., J. Zakiah, L.K. Ng and K.W. Lai. 2006. Fermentation of tem-
poyak using isolated tempoyak cultures. Research Journal of Microbiology
1:243–254.
Anwar, M.A., S. Kralj, A.V. Piqué, H. Leemhuis, M.J.E.C. van der Maarel and
L. Dijkhuizen. 2010. Inulin and levan synthesis by probiotic Lactobacillus
gasseri strains: Characterization of three novel fructansucrase enzymes
and their fructan products. Microbiology 156:1264–1274.
Ayres, J.C., J.O. Mundt and W.E. Sandine. 1980. Microbiology of Foods. San
Francisco: Freeman.
Brauman, A., S. Kéléké, M. Malonga, E. Miambi and F. Ampe. 1996.

Microbiological and biochemical characterization of cassava retting, a
traditional lactic acid fermentation foo-foo (cassava flour) production.
Applied and Environmental Microbiology 62:2854–2858.
Dellaglio, F., M. Vancanney, A. Endo et al. 2006. Lactobacillus durianis
Leisner et al. 2002 is a later heterotype synonym of Lactobacillus vac-
cinostercus Kozaki and Okada 1983. International Journal of Systematic
and Evolutionary Microbiology 56:1721–1724.
Department of Agriculture, Sarawak. 1992. Wild Fruits and Vegetables in
Sarawak. Second ed. Kucing, Sarawak: Hua Daily News Bhd.
Endang S.R. 2000. Lactic acid bacteria in fermented foods of Indonesian ori-
gin. Agritech 23(2):75–84.
Endang S.R., T.F. Djafar and D. Wibowo. 1996. Lactic acid bacteria from
indigenous fermented foods and their antimicrobial activity. Indonesian
Food and Nutrition Progress 3(2):26–34.
Endo, A. and S. Okada. 2008. Reclassification of the genus Leuconostoc and pro-
posals of Fructobacillus fructosus gen. nov., comb. nov., Fructobacillus durionis
comb. nov., Fructobacillus ficulneus comb. nov. and Fructobacillus pseudo-
ficulneus comb. nov. International Journal of Systematic and Evolutionary
Hutkins, R.W. 2006. Microbiology and Technology of Fermented Foods. Ames,
Iowa: IFT Press/Blackwell.
Kuswanto, K.R. 1994. Food fermentation of cassava in Indonesia. In Proceedings
of International Workshop on Application and Control of Microorganisms
in Asia, edited by K. Komogata, K. Yoshida, T. Nakase and H. Osada.
Tokyo: Institute of Physical and Chemical Research, and Japan Institute
of Science and Technological Exchange Centre.
Leisner, J.J., M. Vancanneyt, G. Rusul, B. Pot, K. Lefebvre, A. Fresi and T.K.
Tee. 2001. Identification of lactic acid bacteria constituting the predomi-
nating microflora in an acid-fermented condiment (tempoyak) popular in
Malaysia. International Journal of Food Microbiology 63:149–157.
Leisner, J.J., M. Vancanneyt, K. Lefebvre, K. Vandemeulebroecke, B. Hoste, N.E.
Vilalta, G. Rusul and J. Swings. 2002. Lactobacillus durianis sp. nov., isolated
from an acid-fermented condiment (tempoyak) in Malaysia. International
Journal of Systematic and Evolutionary Microbiology 52:927–931.
Leisner, J.J., M. Vancanneyt, R. van der Meulen, K. Lefebvre, K. Engelbeen,
B. Hoste, B.G. Laursen et al. 2005. Leuconostoc durionis sp. nov., a hetero-
fermenter with no detectable gas production from glucose. International
Journal of Systematic and Evolutionary Microbiology 55:1267–1270.
Li, J.-X., P. Schieberle and M. Steinhaus. 2012. Characterization of the major
odor-active compounds in Thai durian (Durio zibethinus L. ‘Monthong’)
by aroma extract dilution analysis and headspace gas chromatography-
olfactometry. Journal of Agricultural and Food Chemistry 60:11253–11262.
Li, K.-Y. 2004. Fermentation: Principles and microorganisms. In Handbook
of Food and Beverage Fermentation Technology, edited by Y.H. Hui, L.
Meunier-Goddik, A.S. Hansen, J. Josephsen, W-K. Nip, P.S. Stanfield
and F. Toldrá. pp. 595–609. New York: Marcel Dekker.
Mahmud, M.K., N.A. Hermana, Zulfianto, R. Rozanna, Apriyantono, I.

Ngadiarti, B. Hartati, Bernadus and Tinexcelly. 2009. Table of Indonesian
Food Composition. Jakarta: Indonesian Nutritionist Association. (In
Indonesian).
Maki, M. 2004. Lactic acid bacteria in vegetables fermentation. In Lactic
Acid Bacteria, Microbiological and Functional Aspects. Third ed. Edited by
S. Salminen, A. von Wright and A. Ouwehand. New York: Marcel Dekker.
Maranon, J. 1935. Nutritive mineral value of Philippine food plants. Philippine
Journal of Science 58:317–358.
Ministry of Health. 1992. List of Food Composition. Jakarta: Ministry of Health,
Republic of Indonesia. (In Indonesian).
Neti, Y., I.D. Erlinda and V.G. Virgilio. 2011. The effect of spontaneous fer-
mentation on the volatile flavour constituents of durian. International
Food Research Journal 18:625:631.
Okafor, N., B. Ijioma and C. Oyolu. 1984. Studies on the microbiology of
cassava retting for foo-foo production. Journal of Applied Microbiology
56:1–13.
Pederson, C.S. 1996. Historical perspectives of the sauerkraut literature. In
Handbook of Indigenous Fermented Foods. Second ed., edited by K.H.
Rascana, P. 1988. Microflora of growol. Thesis, Faculty of Agricultural
Technology, Gadjah Mada University, Yogyakarta, Indonesia.
Seal, T. 2012. Evaluation of nutritional potential of wild edible plants, tradi-
tionally used by the tribal people of Meghalaya state in India. American
Journal of Plant Nutrition and Fertilization Technology 2(1):16–26.
Singleton, P. and D. Sainsbury. 2001. Dictionary of Microbiology and Molecular
Biology. Third ed., pp. 324, 403, 436. Chichester: John Wiley.
Steinkraus K.H., ed. 1996. Handbook of Indigenous Fermented Foods. Second ed.,
Vasconcelos, A.T., D.R. Twiddy, A. Westby and P.J.A. Reilly. 1990. Detoxification
of cassava during gari preparation. International Journal of Food Science and
Weenen, H., W.E. Koolhaas and A. Apriyantono. 1996. Sulfur-containing vol-
atiles of durian fruits (Durio zibethinus Murr.). Journal of Agricultural and
Food Chemistry 44:3291–3293.
Westby, A. and D.R. Twiddy. 1992. Characterization of gari and fu-fu prepara-
tion procedures in Nigeria. World Journal of Microbiology and Biotechnology
8:175–182.
Widowati, S. 1994. Microorganisms involved in growol (submerged fermented
cassava) production. Proceedings of the Cassava Biotechnology Network. Vol.
II pp. 702–715.
Wirawati, C.U. 2002. Potency of lactic acid bacteria isolated from tempoyak
as probiotic. Thesis, Bogor Agricultural University, Bogor, Indonesia.
(In Indonesian).
Yuliana, N. 2005. Identification of non-lactic acid bacteria associated with
tempoyak (fermented Durian). Journal Mikrobiology Indonesia 10:25–28.
(In Indonesian).
Yuliana, N. and E.I. Dizon. 2011. Phenotypic identification of lactic acid bac-
teria isolated from tempoyak (fermented durian) made in the Philippines.
International Journal of Biology 3:145–152.
Yuliana, N. and V.V. Garcia. 2009. Influence of Pediococcus acidilactici as a
starter on the flavour of tempoyak (fermented durian). Indian Journal of
6
L acti c Fermented
R i ce N o od les
R E N U P I N T H O N G A N D J . DAV I D O W E N S
Contents
6.1 Thai Fermented Rice Noodles (Kanomjeen) 212

6.1.2 History 214
6.1.3 Places of Production, Scale of Production, How It
Is Consumed and Role in Diet 215
6.1.3.1 Places of Production and Scale of
Production 215
6.1.3.2 How It Is Consumed and Role in Diet 215
6.1.4.1 Traditional Small-Scale Production of
Fermented Rice Noodles 216
6.1.4.2 Current Production of Fermented Rice
Noodles 217
6.1.4.3 Production of Unfermented Rice
Noodles 228
6.1.5.1 Characteristics of the Microorganisms 230
6.1.5.2 Selection of Starter Cultures 230
6.1.6.1 Rice 231
6.1.6.2 Sources of Fermentable Carbohydrates 237
Fermentation 239
6.1.7.1 Fermentation Products 240
6.1.7.2 Physico-Chemical Changes 241
6.1.7.3 Effects of Fermentation on Starch
Properties 243
6.1.7.4 Changes in Volatile Compounds 245
211
6.1.7.5 Shelf Life of Fermented Noodles 245

6.1.7.6 Yield 245
6.1.10.1 Unpolished Rice Noodles 248
6.1.10.2 Commercial Fermented Rice Noodle
Starch 249
6.1.10.3 Dried Noodles 249
References 251
6.1 Thai Fermented Rice Noodles (Kanomjeen)

Renu Pinthong and J. David Owens
Traditional Thai fermented rice noodles or kanomjeen are made from

broken, non-sticky rice (non-glutinous rice, Oryza sativa L. subsp.
indica), usually after being stored for 3–4 months. The traditional pro-
cess embraces three fermentation stages. The rice is washed, soaked,
drained and the moist grains are then stored in bags for 2–3 days,
when the first fermentation occurs. The softened rice grains are
washed, wet milled with added salt and the slurry is allowed to stand
overnight, when the second fermentation occurs. The starch sediment
is de-watered by pressing in bags for 2–3 days, allowing a third fer-
mentation to occur. The fermented rice starch is partially gelatinised,
kneaded and extruded as streams of 1.5–2 mm in diameter into boil-
ing water. The noodles are briefly cooked, washed, cooled and drained.
Most commercial fermented rice noodles are prepared by family
businesses using family labour. The fermented noodles are distributed
to retail traders and sold extensively in villages and towns and in open
markets across Thailand. Since it can be instantly prepared for serving
when ordered by consumers, kanomjeen is regarded as a fast-food alter-
native to cooked plain rice and can be eaten with a great variety of con-
diments and soups (Piyakul and Nuthong 1999; Figures 6.1 and 6.2).
There are two kinds of kanomjeen, kanomjeen pang muk and kanom-
jeen pang sod. Kanomjeen in Thai means fermented rice noodle, pang
means flour or starch, muk means fermented and sod means fresh or
L ac ti c F erm en t ed Ri c e N o o d l e s 213
Figure 6.1 Fermented rice noodles with nam-ngeow accompaniments, minced pork with thua-
nua and tomato in a spicy soup. (Courtesy of R. Pinthong.)
non-fermented. Thus, kanomjeen or kanomjeen pang muk are fermented

noodles and kanomjeen pang sod is unfermented noodles.
Some fermented noodles incorporate herbs, such as pandan
(Pandanus amaryllifolius), butterfly pea flowers (Clitoria ternatea) or
lemongrass (Cymbopogon citratus) leaf or stem extract, to provide a
variety of colours and tastes to consumers. These kinds of products
are attractive and popular.
In northern Thailand, fermented rice noodles are called kanomsen,
which probably derives from kao-nom-sen (sen meaning long-stranded
Figure 6.2 Fermented rice noodles with kangkeowan-kai accompaniments, chicken, blood, Thai
eggplant (Solanum melongena L.), Thai pea eggplant (Solanum torvum Sw.) and jelly ear mushroom
(Auricularia auricular-judae) in spicy green curry. (Courtesy of R. Pinthong.)
rice in Thai). They are called kao-pun in northeastern Thailand, nom-

jeen in most of Southern Thailand, lasa in some southern provinces
and kanomjeen in central Thailand (Piyakul and Nuthong 1999).
6.1.2 History
How and when fermented rice noodle production began in Thailand

is unknown. However, it has probably existed since at least the
Ayutthaya capital period (ad 1350–1767) as there are canals called
Klong Kanomjeen and Klong Nam-ya (nam-ya is a popular soup,
made of minced fish in spicy coconut milk, served with fermented rice
noodles) in Sena district, Ayutthaya province (Litisorn et al. 2005).
Since place names commonly derive from activities carried out at the
location, this does suggest that kanomjeen was being made in the
Ayutthaya period. Certainly, fermented rice noodle production was
well established in Bangkok by the time of the reign of King Rama
the first (reigned 1782–1809), as the occurrence of a big royal fer-
mented rice noodle feast to celebrate the completion of the construc-
tion of the Emerald Buddha Temple in ad 1809 at Ratanakosin is
recorded (Payakaranon 1999). Jeen in Thai means ‘Chinese’ but King
Rama the third (reigned 1824– 1851) suggested that the inclusion
of jeen in kanomjeen did not mean that kanomjeen originated from
China but rather was a traditional Thai product. S. Plynoi, a national
artist, researched many old legends and suggested that kanomjeen is
derived from kanomjin, a Maawn word (Mon or Maawn, an eth-
nic group from Myanmar). In Maawn language, kanom means being
combined into a bundle and jin means cooked. Thus, kanomjin, which
Thai people pronounce with a longer accent on the jin to become
kanomjeen, could indicate a cooked bundle-like food (Hongsawiwat
et al. 2009). Another possibility is that kanomjeen derives from the
Thai word kanom, meaning dessert, as kanomjeen used to be eaten as a
dessert with syrup in southern Thailand. Kanomjeen has nothing to do
with dessert nowadays because it is generally consumed accompanied
by various kinds of spicy or non-spicy dressings as the main dish.
While it seems certain that rice noodles originated in China
(Huang 2000) and spread to the neighbouring countries, Huang
makes no mention of fermented noodles. A type of fermented rice
noodles, sour mifen, is produced in southern China (Lu and Collado
2010) and it is certainly possible that the method was introduced to

Thailand by Chinese immigrants.
6.1.3 Places of Production, Scale of Production, How It Is

Consumed and Role in Diet
6.1.3.1 Places of Production and Scale of Production There were 53 fer-

mented rice noodle factories registered with the Ministry of Thai
Industry as of June 2013 (Department of Industrial Information
Center). The numbers of registered kanomjeen factories in differ-
ent provinces were southern provinces: Trang, 2; Pattani, 2; Yala, 1;
Narathiwat, 4; Chumphon, 2; Songkhla, 1; Phuket, 2 and Krabee,
1; central provinces: Bangkok, 3; Chanthaburi, 1; Singburi, 3;
Chachoengsao, 1; Kanchanaburi, 2; Suphanburi, 1 and Phetchaburi,
1; northern provinces: Nakhon Sawan, 1; Chiang Rai, 1 and
northeastern provinces: Kalasin, 5; Khon Khan, 5; Udonthani, 1;
Chaiyaphum, 5; Roi Et, 2; Buriram, 1; Ubolratchathani, 4; and Surin,
1. In addition, there are many homescale producers, making from 50
to 1000 kg per day, who are not registered with the Thai Department
of Industry. Apart from fermented rice noodle factories, there are also
factories that produce fermented rice starch. Fermented rice starch,
with a moisture content of 40–50%, is a product obtained by pressing
fermented wet-milled rice under heavy weights for 2–3 days to remove
water and is sold as a starting material for the production of fermented
rice noodles.
6.1.3.2 How It Is Consumed and Role in Diet Fermented rice noodles is a

popular dish in all regions of Thailand. Thai people enjoy eating them
with various spicy and non-spicy accompaniments. Non-spicy examples
are saow-nam, consisting of chopped ginger, pineapple, ground dried
shrimp and coconut milk, and nam-prig, comprising minced peanut and
various non-spicy herbs in coconut milk. There are many popular spicy
accompaniments, including nam-ya, a mixture of minced fish in spicy
coconut milk, nam-ngeow (Figure 6.1), minced pork, pork blood and
tomato flavouring with grilled and ground thua nao (Bacillus-fermented
soya beans; see Chapter 10) in a spicy soup supplemented with deep-
fried pork skin and pakchee-farung leaves (Eryngium foetidum L.), kang
keow-wan, chicken, pork or beef in spicy coconut milk with ground
green herbs (Figure 6.2), musamun soup, made from chicken or beef
and peanuts in spicy coconut milk, Thai papaya salad and so on. Young
children, who do not take spicy dishes, like eating fermented noodles
with boiled egg and fish sauce or soya sauce.
Fermented rice noodles have a slightly darker colour and tougher
and more elastic texture than non-fermented rice noodles, which is
attractive to some consumers. However, the consumption of non-
fermented noodles is much greater than that of fermented noodles,
possibly because many consumers find the white appearance and soft
texture of the non-fermented noodles more appealing and appetising
than the slightly darker and chewier fermented noodles.
The traditional production method includes three main processes, the

conversion of rice grains into a smooth starch slurry, the partial gela-
tinisation of the starch and the extrusion of the starch cream into
boiling water to produce noodles. During processing, there are at least
three stages when a bacterial fermentation occurs.
6.1.4.1 Traditional Small-Scale Production of Fermented Rice Noodles

6.1.4.1.1 Fermentation of Rice Broken polished rice, that has usually
been stored for 3–4 months, is washed with clean water to remove dust
and foreign materials and then soaked in water for 3 h. The grains are
transferred to a cloth bag, which is placed in a tray containing water
up to about 5% of the height of the bag for 3 days. The bag is inverted
once or twice daily to keep the broken rice grains thoroughly moist.
A natural fermentation takes place and there is a gradual increase
in temperature of the rice in the middle of the bag, the rice colour
changes from white to pale yellow and then dark yellow and there is a
development of some sliminess and a fermented odour. This treatment
allows rice grains to undergo the first natural fermentation, which
softens them and facilitates starch extraction.
6.1.4.1.2 Separation and Fermentation of Starch The fermented rice

grains are washed thoroughly to remove any dark-yellow colouration,
sliminess and fermented odour. They are then broken by rubbing the
starch through a linen cloth tied over the mouth of a jar while adding
brine (2–3% salt). The starch suspension is filtered through cheese

cloth and then transferred to a thick cloth bag that is tied and placed
under weights to remove excess water for one night. The resulting
dough is a firm lump of agglomerated starch with a moisture content
of 40–50%.
6.1.4.1.3 Partial Gelatinisation of the Fermented Rice Starch Dough A

lump of raw starch, supported in a bamboo mesh basket, is suspended
in boiling water for 5–10 min to gelatinise the starch on the surface
of the lump to a depth of ~0.5 cm. The mass is then pounded with
a wooden mortar and pestle to mix the surface-gelatinised starch
with the interior raw starch (Payakaranon 1999). This practice dis-
tributes the gelatinised amylose network around the uncooked starch,
with the gelatinised starch functioning rather like gluten in wheat
flour to bind all the starch granules together into a consistent and
smooth dough (Jane et al. 2010; Oupathumpanont et al. 2008).
6.1.4.1.4 Production of Noodle Strands Water equivalent to ~5% of

its weight is added to the partially gelatinised fermented starch dough
in a mortar and the mixture is pounded until it is smooth. The slurry
is filtered through a cotton cloth to remove any lumps and it is then
transferred to a cotton cloth extruder bag with a perforated metal
disc with about 30 holes of ~2 mm diameter sewn into the bottom of
it. Screwing the neck and squeezing the bag extrudes the rice slurry
through the holes as long noodle strands that fall into boiling water in
a bowl. The noodle strands are cooked in the boiling water for 45–75 s
until they float to the surface when cooked. The cooked noodles are
scooped out with a perforated bamboo basket and washed thoroughly
with cold water to remove residual starch and to cool them. They are
collected in bundles of 3–4 strands rolled around the fingers, some
excess water is squeezed out and the bundle coils are placed in woven
bamboo baskets lined with banana or papaya leaves. The finished fer-
mented noodles are ready for sale or for serving with various accom-
paniments to consumers (Payakaranon 1999).
6.1.4.2 Current Production of Fermented Rice Noodles The description

below is based on the processing of fermented rice noodles as car-
ried out by a factory located near Chiang Mai, northern Thailand, in
September 2012 (Figure 6.3). The factory makes both fermented rice
noodles, kanomjeen pang muk, and unfermented rice noodles, kanom-
jeen pang sod (Abngern 2012). The factory uses 540 kg of a mixture of
polished broken and whole rice grains as the raw material each day.
This yields about 540 kg of starch with 40–50% moisture content,
which is converted into approximately 1080 kg of unfermented rice
noodles. Fermented rice starch, about 360 kg, is prepared only once a
week and ~50 kg of this fermented raw starch is used to make ~100 kg
of fermented rice noodles daily.
Polished rice
(Broken: whole grains :: 50:50; three varieties of rice)
Wash once
Fermented noodles: Drain Unfermented noodles:
Soak 1 h Soak overnight

(first fermentation)
Polypropylene mesh bag, drain, two nights (first fermentation)
Add salt Wet mill Add salt

Stand overnight to sediment starch (second fermentation)
Transfer starch to finely woven cotton bags
De-water by pressing under stone weights, 3 nights (third fermentation)
Starch, 10 kg per polypropylene plastic bag
Heat in boiling water ~25 min (gelatinises outer layer of starch)

Knead (mix gelatinised and nongelatinised starch)
Hot or cold water to Second kneading (adjust cream temp. to ~40°C)
Filter through fine mesh bag attached to pipe outlet
Extrude into boiling water ~35 s
Cook noodles 45–75 s
Cool and wash noodles in cold water
Sort noodles in parallel bundles and put into perforated plastic baskets to drain
Fermented or unfermented noodles
Figure 6.3 Production of Thai fermented and unfermented rice noodles as practiced at S. Suthi
Kanomjeen Factory, 289/1 Moo 3, Tumbol Khilek, Mae-Rim District, Chiang Mai 50180, Thailand.
September 2012. Figures 6.4 through 6.14 and 6.16 through 6.20 were taken at this factory.
6.1.4.2.1 Preparation of Fermented Rice Starch A mixture of equal

parts broken polished rice and whole grain polished rice of three
different varieties are used (Figure 6.4). Broken rice is cheaper than
whole grain rice but, according to the manufacturer, a mixture of bro-
ken and unbroken selected varieties yields noodles with better texture
than using one variety of broken rice only. The rice is washed once and
then soaked for 1 h in a concrete tank (Figure 6.5). The rice is drained
and transferred to woven polypropylene bags. The bags are covered
with wet cloths and left to stand for 2–3 nights at ambient tempera-
ture (28–35°C). Water is poured over them twice a day to prevent the
surface drying out. The first fermentation occurs at this stage.
Figure 6.4 Broken and whole polished rice starting material. (Courtesy of J.D. Owens.)
Figure 6.5 Rice in the soaking tank. (Courtesy of J.D. Owens.)

Figure 6.6 Mill for wet grinding of rice. (Courtesy of J.D. Owens.)
The rice is then wet milled, using a mill with stone-grinding discs
(Figure 6.6). To prevent over-fermentation, salt (10 kg per 360 kg
rice) is added during the grinding to give a final concentration in the
suspension of 2.8%. The starch slurry is collected in a tank and left
to stand overnight to allow the starch to sediment, when a second
fermentation occurs (Figure 6.7). The supernatant liquid is siphoned
off and the starch sediment is pumped into cloth sacks. The open-
ings are tied tightly with rope and the sacks are pressed under heavy
weights for 2–3 days to express water, concomitantly with the third
Figure 6.7 Milled rice starch suspension soaking in the tank. (Courtesy of J.D. Owens.)
L ac ti c F erm en t ed Ri c e N o o d l e s 2 21
Figure 6.8 (a) and (b) De-watering starch flour under weights. (Courtesy of J.D. Owens.)
fermentation taking place (Figure 6.8a and b). The moist fermented
starch obtained has a light-cream colour and moisture content of
40–50%. The pressed, fermented rice starch is relatively stable and
may be kept for up to 5 days, after which moulds may grow on the
surface of the bags (K Pichit, personal communication).
6.1.4.2.2 Partial Gelatinisation of the Fermented Starch The fer-

mented starch dough is put into heat-resistant polypropylene plastic
bags (Figure 6.9), in 10 kg quantities, and the bags are immersed in
boiling water for ~25 min (Figure 6.10). This gelatinises the outer
layer of the starch and the surface 0.5–1.0 cm of the block becomes
light brown in colour (Figure 6.11). This discoloured zone, depending
on the thickness of the block, represents 15–40% of the total block
volume but how much of this is gelatinised starch is not known. To
Figure 6.9 Transferring de-watered starch to polypropylene bags, of 10 kg each. (Courtesy of
J.D. Owens.)
Figure 6.10 Cooking and partially gelatinising starch. (Courtesy of J.D. Owens.)
ensure that adequate gelatinisation has occurred, the extent of the

discoloured zone is monitored during cooking by cutting into a block.
The purpose of partially gelatinising the starch is to obtain some
starch gel to serve as a natural binding agent for the starch matrix,
allowing it to form firm, stable and elastic rice noodle strands. Rice
proteins lack the functionality of wheat gluten and are not able to
form a continuous viscoelastic dough. Too much or too little gelatini-
sation affects the firmness of the noodles. For example, excessive gela-
tinisation may result in undesirable short and broken strands while
Figure 6.11 Cooked starch showing outer discoloured region. (Courtesy of J.D. Owens.)
uniform, translucent white, straight strands with an absence of broken

strands are obtained from suitably partially gelatinised and kneaded
starch.
The partially gelatinised starch is then kneaded in a kneading
machine to mix the gelatinised and raw starch (Figure 6.12). The
dough is transferred to a second kneading machine and ~5% (w/w) of
hot water is added to adjust the temperature of the paste to ~40°C and
Figure 6.12 Mixing gelatinised and non-gelatinised starch in the first kneading machine.
Figure 6.13 Homogeneous starch paste in the second kneading machine. (Courtesy of J.D.
Owens.)
to obtain a homogeneous starch slurry (Figure 6.13). It is important

not to add too much water as the starch puree will not form strands
but may gel when extruded into boiling water. If too little water is
added, the puree is too concentrated and cannot pass through the
extruder holes or forms deformed noodle strands. The moisture con-
tent of the puree should be in the range of 55–60%.
The main aim of kneading is to mix the interior uncooked and
crumbly starch granules with the moist gelatinised outer starch gel.
When properly kneaded starch is parted, it exhibits no separate flour
crumbs and the starch is elastic and flexible without sticking to the
hands or the container, while insufficiently kneaded starch leads to
soft, deformed and broken noodle strands. Kneading is usually fol-
lowed by a resting stage to allow further hydration and redistribu-
tion of moisture in the starch. This maximises the formation of a
continuous-gel matrix with embedded starch granules and a firm,
stable and flexible final fermented rice noodle.
6.1.4.2.3 Preparation of Noodles The sticky and viscous starch cream

is filtered through cheese cloth to remove any residual lumps (Figure
6.14) and pumped into the extruder, comprising ~30 holes of 2–3
mm in diameter in an 8 cm steel plate. The cream/slurry is extruded
quickly with the extruder held not higher than 5–10 cm above the
surface of boiling water in a basin on a stove (Figure 6.15). Care is
taken to deposit the noodles into clear boiling water rather than on
top of the previously deposited noodles and to complete the extrusion
as quickly as possible. The desirable volume of boiling water is 10–20
Figure 6.14 Filtering starch cream into an extruder reservoir. (Courtesy of J.D. Owens.)
Figure 6.15 Extruding noodles into boiling water. (Courtesy of J.D. Owens.)
times the weight of uncooked noodle strands. If the volume of boiling

water is smaller, it takes longer to bring it back to boil and the noodle
strands stay close together without having enough relative movement,
resulting in a rough noodle surface and a lack of uniformity in cook-
ing. The noodle strands are left in the boiling water until they float up
to the surface, when they are scooped out with a perforated bamboo
basket (Figure 6.16) and then immediately cooled and washed free of
flour residue in 2–3 changes of cold water (Figure 6.17). The process
is a continuous cycle of extrude for 15–35 s, cook for 25–75 s, cool
Figure 6.16 Collecting noodles. (Courtesy of J.D. Owens.)
Figure 6.17 Cooling and washing noodles. (Courtesy of J.D. Owens.)
and wash the noodles. To prevent the noodle strands congregating in

large lumps, groups of 4–6 strands are collected, drained and aligned
into parallel bundles by rolling around 2 or 3 fingers (Figure 6.18).
Excess water is squeezed out and the noodle rolls are placed in a bowl
or a bamboo basket (Figure 6.19), layered with clean banana leaves.
Figure 6.18 Aligning noodles in bundles. (Courtesy of J.D. Owens.)
Figure 6.19 Final noodles in bundles. (Courtesy of J.D. Owens.)
For sale, 2–10 kg of noodles are loosely packed in plastic or bamboo

baskets and delivered to shops daily for retail sale.
According to the manufacturer, the fermented noodles are slightly
creamier in colour than unfermented noodles and have a different
odour and texture, being more elastic than unfermented noodles. The
fermented noodles are not as popular as the unfermented ones and
sold for 22 Bt kg−1 compared with 20 Bt kg−1 for unfermented noodles
(30.5 Thai Bt = 1 US$ at October 2012).
6.1.4.3 Production of Unfermented Rice Noodles The production of

unfermented rice noodles was similar to that for fermented noodles
except in the initial stages (Figure 6.3). The same mixture of broken
and whole rice grains was used. They were washed and drained but,
instead of being put in bags to ferment for 2 days, they were soaked
overnight, drained and then taken directly for wet milling. Thereafter,
the process was exactly the same as for fermented rice noodles.
The pH values and microbial flora (microscopical examination)
during the process were: supernatant liquid from grains soaked over-
night, pH 6, large and small bacterial cocci, rod-shaped bacteria and
a few budding yeast cells; supernatant liquid from milled rice after
standing overnight, pH 5, bacterial cocci and rods and large oval
yeast cells; uncooked, pressed starch, pH >4 –<5, a few bacterial rods
and no yeasts were seen; starch cream prior to extrusion, pH ~4,
rod-shaped bacteria and a few yeasts and final noodles, pH >5– < ~6
(Pinthong and Owens, unpublished observations). Although
described as unfermented noodles, it is evident that a (lactic acid)
bacterial fermentation occurred during the process and that the pH
of the starch cream prior to extrusion was, in fact, quite similar to
the pH of the starch cream for fermented noodles (pH 3.0–4.1).
No attempt was made to identify the yeasts observed but it is pos-
sible that some were aerobic film yeasts rather than fermentative and
a lcohol-producing species.
6.1.5 Microbiology
The microflora of Thai fermented noodles has been the subject of

some studies but very few of them have been published in scien-
tific journals. In contrast, all aspects of the production of Chinese
fermented noodles have received much attention, especially in col-
laboration with Japanese workers. Nevertheless, it is clear that the
fermentation is primarily a lactic acid bacterial one (Rosanaphaiboon
1987; Sribuathong et al. 2004; Suphichayangkoon et al. 2006) along
with yeasts (Sribuathong et al. 2004). Sribuathong et al. (2004)
observed lactic acid bacteria, Bacillus spp., yeasts and moulds at all
the fermentation stages. The main lactic acid bacteria found were spe-
cies of Lactobacillus, Leuconostoc, Enterococcus, Lactococcus, Pediococcus
and Streptococcus. Of these, the Enterococcus and Streptococcus isolates
could hydrolyse starch strongly. Lactic acid bacteria identified by

Suphichayangkoon et al. (2006) in fermented broken rice (i.e. the first
fermentation stage) included Lactobacillus plantarum, Lactobacillus cel-
lobiosus (L. fermentum; Dellaglio et al. 2004) and Leuconostoc lactis.
Yeasts are commonly present, including film yeasts that grow on
the surface of the liquid in the fermentation and sedimentation tanks
(Pinthong and Owens, unpublished observations).
Although the production of Chinese fermented rice noodles may
differ in some details from the traditional production of Thai fer-
mented noodles, it seems likely that the basic microbiological and
chemical changes are similar. Lu et al. (2008a) studied the micro-
flora from three factories in China of supernatants from polished
rice fermented at temperatures of 23.5–25.5°C over 3 days (samples
were collected at 0, 24, 48 and 72 h). Populations of lactic acid bac-
teria were 102.7 cfu g−1 on the polished rice and 102.6 cfu mL −1 in the
supernatant water at the beginning of fermentation. Within 24 h,
concentrations reached 107.9–108.6 cfu mL −1 and remained at levels of
108.1–10 8.7 cfu mL −1 to the end of fermentation.
Fermentations were dominated by Lactobacillus species, which
represented 83% of the 170 lactic acid bacterial isolates studied in
detail. The most frequently isolated species were L. plantarum, L. cel-
lobiosus, L. curvatus, L. fermentum, L. acidophilus, L. delbruckii subsp.
delbruckii and L. helveticus. Other lactic acid bacteria isolated included
Pediococcus pentosaceus, Enterococcus faecium, Enterococcus avium and
Leuconostoc species (Lu et al. 2008a). However, these population anal-
yses must be treated with some caution as no attempt was made to
monitor numbers of individual species during the fermentation.
Lu et al. (2008a) did not detect yeasts on the rice but yeasts were
present at concentrations of 101.9 –102.6 cfu mL −1 at time 0 h in super-
natants from rice fermentations in three factories. In two of the three
factories, concentrations of yeasts slightly increased during the fer-
mentation, reaching 102.7–103.5 cfu mL −1 after 3 days. In the third fac-
tory, the concentrations of yeasts initially increased but then decreased
and were at undetectable levels (presumably <102 cfu mL −1) after 48
and 72 h fermentation.
The yeast populations observed are quite small and increased dur-
ing fermentations only to 3.8, 7.2 and 15 times the initial populations.
Even allowing for the greater size of yeast cells, it would seem that the
biomass produced by the yeasts was insignificant compared to that of

the bacteria, suggesting that they played little role in the fermenta-
tion. Saccharomyces cerevisiae was by far the most frequently isolated
species.
Coliform bacteria were initially present at concentrations of 104.0 –
104.8 cfu mL −1 and then declined to undetectable levels at the end of
the fermentations (Lu et al. 2008a). This was, presumably, a conse-
quence of the low final pHs (4.0–4.1) and the presence of lactic and
acetic acids at concentrations of 5.9–6.4 and 2.1–3.9 g L −1, respec-
tively, at 72 h. Although lactic acid was present in the supernatants at
higher concentrations than acetic acid, it has a lower pKa (3.87) than
acetic acid (4.76) and it can be estimated that the concentrations of
the toxic undissociated lactic acid and acetic acid molecules were of a
similar order.
6.1.5.1 Characteristics of the Microorganisms Sribuathong et al. (2004)

noted that Enterococcus and Streptococcus isolates from fermenting
rice could hydrolyse starch strongly but neither provided any data
on the concentrations of bacteria nor on the concentrations of sugars
produced in the fermentation. Most of the microorganisms isolated
by Lu et al. (2008a) displayed α-glucosidase, β-glucosidase, lipase
and trypsin activities but they made no direct assessment of amylase
activity.
6.1.5.2 Selection of Starter Cultures Some effort has been directed to

looking for starter cultures that could affect reproducible fermenta-
tions of rice. Suphichayangkoon et al. (2006) compared fermenta-
tions singly inoculated with L. plantarum PD110, L. cellobiosus RE33
or L. lactis PD128 with a natural fermentation without any starter
(Figure 6.20). They demonstrated that the three lactic acid bacterial
strains could singly affect changes in pH value, titratable acidity, lac-
tic and acetic acid in fermenting broken rice in 24 and 48 h that were
similar to or faster than those occurring in natural fermentations.
Each of the three strains reduced the pH from an initial value of 6.7
to 3.8–4.5 within 24 h, whereas natural fermentation required more
than 48 h to produce similar pH values (Figure 6.20a). Since the
amount of acids produced in the natural fermentation was similar to
the levels with the pure culture ferments, it is not clear why the decline
in pH value should have occurred more slowly. The homofermentative

L. plantarum PD110 produced similar concentrations of lactic acid as
the natural fermentation but with considerably less acetic acid (Figure
6.20b and c). The heterofermentative strains, L. cellobiosus RE33 and
L. lactis PD128, produced substantially lower concentrations of lactic
acid than the natural or L. plantarum fermentations (Figure 6.20b).
With L. cellobiosus, the concentration of acetic acid formed was similar
to that in the natural fermentation but L. lactis produced distinctly
higher concentrations at 48 h fermentation (Figure 6.20c). The con-
centrations of acetic and lactic acids in the final noodles were very low,
evidently due to them having been leached out by the final washing
and draining steps. It is noticeable that the concentrations of acetic
acid in all the ferments were much lower than the concentrations of
lactic acid and this might suggest that some ethanol was also formed
in the heterolactic fermentations. The combination of a relatively high
concentration of lactic acid along with some acetic acid might also
suggest that the natural fermentation involved the activities of both
homofermentative and heterofermentative lactic acid bacteria.
The growth of undesirable organisms, including, yeasts, Bacillus
spp., staphylococci and coliform bacteria, was suppressed more
rapidly in inoculated fermentations than in natural fermentations
(Sribuathong et al. 2007), ensuring that the fermented noodles pro-
duced from inoculated fermentations conformed to Thai Community
Product microbiological standards (Thai Industrial Standards Institute
2004a). The pure culture-fermented starch also yielded lighter yel-
low and more elastic noodles with more acetic acid, ethanol and ethyl
acetate than the product obtained from the natural fermentation. The
authors suggested that the application of lactic acid bacterial start-
ers could shorten the fermentation times during the production of
fermented noodles while maintaining the same product qualities as
obtained by natural fermentation and further suggested that L. lactis
PD128 could potentially be used as a single culture starter for the
production of fermented rice noodles.
6.1.6.1 Rice The cost of broken rice is less than that of whole grains
and it is, therefore, preferred for rice noodle processing. It is also
(a) 7
5
pH value
3
Broken rice Fermented Fermented Sedimented Drained wet Fermented
24 h 48 h starch starch noodles
(b) 5.5
5
4.5
4
3.5
3
Lactic acid (g/L)
2.5
2
1.5
1
0.5
0
–0.5
Figure 6.20 Changes in (a) pH value; (b) lactic acid concentration; and (c) acetic acid con-
centration during processing of broken rice to fermented rice noodles by ◻, natural fermen-
tation; Δ, L. plantarum PD110; ⚪, L. cellobiosus RE33 and ×, L. lactis PD128. (Adapted from
Suphichayangkoon, N., W. Jirapakkul and O. Naivikul. 2006. Effect of lactic acid bacteria starter
culture on chemical properties in Thai fermented rice (kanomjeen) noodle process. In Proceedings of
44th Kasetsart University Annual Conference: Agro-Industry, pp. 356–362. Bangkok. http://kukr.lib.
ku.ac.th/Fulltext_kukr/KU0335074c.pdf (accessed 9 December 2013). [In Thai.])
considered that rice stored for more than 6 months makes better noo-
dles than newly harvested rice (Lu et al. 2003). Rice stored for 1–7
months exhibits decreases in moisture content (Kongkiattikajorn and
Photchanachai 2000), volatile 2-acetyl-1-pyrroline (Kongkiattikajorn
2008), phenolics and antioxidant capacity (Kongkiattikajorn et al.
(c) 5.5
5.0
4.5
4.0
3.5
Acetic acid (g/L)
3.0
2.5
2.0
1.5
1.0
0.5
0.0
Figure 6.20 (continued)
2011) and increases in hexanal, browning reactions (Kongkiattikajorn

2008) and gel consistency (Kongkiattikajorn and Photchanachai
2000). The rice should meet the requirements of the Thai FDA regu-
lations (Thai Food and Drug Administration 2004) regarding heavy
metal and pesticide residue contamination, microbiological specifica-
tions and ash content.
Traditionally, rice noodles are made from non-glutinous, indica-
type rice with a high amylose content (>22% amylose). Starch with
high amylase content, which yields a stable paste, behaves like a
chemically cross-linked material that exhibits restricted swelling and
solubilisation and gives brightly coloured, chewy and elastic noodles
with a low bulk density due to the low-swelling capacity. Starch gels
made from waxy japonica rice starch, which has fewer super-long
chains in the amylopectin than indica rice starch, retrograded more
slowly after gelatinisation than indica rice gels, even though both con-
tained similar amounts of amylose, and the paste was too sticky for
the production of rice noodles (Lu et al. 2009a). Thai rice varieties
with a high amylose content (25–33%) include Leuang Pratue123,
Leuang Oon, Bua-Yai and Pichit (Champangern 2004; Kongseree
2002; Makmoon et al. 2010).
6.1.6.1.1 Chemical Composition of Rice There are two main types of

rice (O. sativa):
1. Non-glutinous rice (O. sativa var. indica), called kaowjaow in

Thai, in which the starch is composed of 10–35% amylose and
65–90% amylopectin.
2. Glutinous rice (O. sativa var. glutinosa), called kaowneow in
Thai, where the starch is mainly amylopectin and only 5–7%
amylose is present.
The non-glutinous rice varieties grown in Thailand may be further

subdivided based on their amylose contents and their cooking proper-
ties (Amornsin 2003; Kongseree 2004):
1. Rice with a low amylose content (9–20%) and a tender

and smooth texture after cooking, such as varieties 105,
Pathumthani1, RD15 and RD21.
2. Rice with a moderate amount of amylose (20–25%) and a
moderately tender and smooth texture after cooking, such as
varieties RD7, RD23, Suphanburi2 and Suphanburi60.
3. Rice with a high amylose content (25–33%), which is crumbly
with a hard texture after cooking, such as varieties Chainat1,
Leuang Pratue123 and Suphanburi90 (Ninchan 2005).
6.1.6.1.2 Composition and Nutritional Value of Rice Rice seed or

paddy consists of four parts: the seed coat or husk which covers the
grain (16–28% of the grain weight); an inner kernel membrane com-
posing pericarp and testa (3–5%) and aleurone (~1%) layers; a germ
or embryo (1.5–2.5%) and the endosperm, the inner white material
which is the main carbohydrate component (69–73%) (Zhou et al.
2002). Collectively, the kernel membranes, embryo and endosperm
constitute the caryopsis. Unpolished or brown rice is milled to remove
most of the hull but most of the bran, comprising the pericarp, testa,
aleurone layer and embryo, remains attached to the rice kernels. In
polished rice, the bran is entirely removed (Zhou et al. 2002). The
nutritive values of Thai rice, fermented rice noodles and related prod-
ucts are shown in Table 6.1. The amounts of several nutrients are
considerably reduced in the processed products compared with the
amounts in the raw materials but, since rice is normally consumed
Table 6.1 Nutritive Values of Thai Rice, Fermented Rice Noodles and Related Products
FER
100% COOKED MENTED UN-FER
NUTRITION FACTS BROWN RICE WHITE BROKEN RICE RICE MENTED
(SERVING: 100 G) (UNPOLISHED) RICEa RICE (STEAMED) NOODLES NOODLES
Calories (kcal) 368 353 357 141 77 135
Moisture (g) 12.7 12.4 11.8 65.4 80.7 67.1
Protein (g) 6.6 6.4 6.0 2.8 0.9 2.5
Total fat (g) 2.3 0.9 1.4 0.5 0.1 0.8
Carbohydrates (g) 77.6 79.9 80.1 31.2 18.2 29.5
Dietary fibre (g) 1.7 (0.2)b (0.2) (0.1) (0.1) (0.1)
Ash (g) 0.8 0.4 0.7 0.1 0.1 0.1
Calcium (mg) —c 0 55 0 7 10
Phosphorus (mg) 66 130 90 11 14 19
Iron (mg) — 0.9 1.8 0.5 0.9 2.7
Vitamin B1 (mg) 0.34 0.26 0.13 0.01 Trace 0
Vitamin B2 (mg) 0.11 0.43 0.1 0 0.02 Trace
Niacin (mg) 1.4 1.6 0.6 1.5 0.4 0.3
Source: Nutrition Division. 2001. Nutritive Values of Thai Foods, edited by S. Bunwisut et al., pp.
9–10. Bangkok: Department of Health, Ministry of Public Health. http://nutrition.anamai.
moph.go.th/temp/files/Nutritive%20Values%20of%20Thai%20foods.pdf (accessed 12
November 2013).
a Polished rice containing ≥60% whole kernel with ≥40% longer than 7 mm and ≤4.5% broken rice
(Thai Agricultural Standard 2012).

b Values in brackets are crude fibre.
c Not determined.
along with other foods, these nutrient losses are not generally of any
consequence.
6.1.6.1.2.1 Carbohydrates Carbohydrates present include starch,

cellulose, hemi-cellulose, pentosan and sucrose. Most of the cellu-
lose, hemi-cellulose and pentosan are present in the husk and bran
and are removed when rice grains are dehulled and polished (FAO
and Juliano 1993). Polished rice is ~90% starch. Free sugars are found
abundantly in the embryo (8–12% dry matter) and in bran (5.5–6.9%)
(Champagne et al. 2004).
Starch is the major constituent of polished rice, at about 90% of the
dry matter, and starch granules fill most of the central space within the
endosperm cells. The main variation in composition is caused by the
relative proportions of amylose and amylopectin in the starch gran-
ules (Jane et al. 2010). Amylose content varies a great deal between
varieties, from a low of 0–2% in waxy rice (milled rice, dry mass basis)
to a high of >25% in non-waxy rice (Kongseree 2004).
6.1.6.1.2.2 Fibre Fibre, present as cellulose, hemi-cellulose and

pentosan comprises 35–46% of the rice husk and 7–11.5% of rice bran
(Champagne et al. 2004).
6.1.6.1.2.3 Protein Protein is the second major nutrient after car-

bohydrate, constituting 19–27% of the embryo and 6 − 8% of the
whole grain. Polished rice retains 6 − 7% protein. Rice protein is high
in the sulphur-containing amino acids, cysteine and methionine, but
low in lysine (Likitwattanasade and Hongsprabhas 2010; Shih 2004).
6.1.6.1.2.4 Fat Fat is found mainly in the embryo and in the

bran, which is the raw material for rice bran oil. The major fatty acids
present are oleic, linoleic and palmitic acids (Zhou et al. 2002).
6.1.6.1.2.5 Minerals and Vitamins Most minerals and vitamins

are present in the embryo and rice bran rather than in the endosperm,
and, therefore, polished rice is relatively deficient in minerals and
vitamins (Table 6.1).
6.1.6.1.3 Water Water varies in hardness, alkalinity and pH value,

which in turn affects flour hydration, noodle properties, starch gela-
tinisation and texture of the finished product. The ions in water have
a significant impact on the gelatinisation of starch during steaming
or boiling. Excessively hard water is undesirable because it retards the
hydration of flour particles (Kraidej et al. 2002). Very soft water is
objectionable since it lacks gel-strengthening minerals and tends to
yield a soft, sticky starch. Water of hardness 75–180 mg L −1 (Thai
Ministry of Health 2000) with pH 6.4–7 is preferred to maintain the
normal yellowish-white colour of fermented noodles.
6.1.6.1.4 Salt Either rock salt or sea salt can be added at the grind-
ing stage to stop the fermentation. Apart from slowing or stopping the
fermentation salt, 1–7% of the original rice weight, helps to eliminate
strong fermented odours, solubilise albumin and globulin proteins,
improve water absorption by the starch, decrease or mask flavour
off-notes (metallic taste, bitterness) and enhance mouth-feel (full-

ness, thickness), sweetness, balance and saltiness (Marianne 1985).
Salt also reduces growth of microorganisms, slows enzymic activi-
ties, oxidative discolouration and spoilage in warm and humid envi-
ronments, thereby extending the shelf life of fresh noodles (Abngern
2012; Kraidej et al. 2002; Likitwattanasade and Hongsprabhas 2010).
6.1.6.2 Sources of Fermentable Carbohydrates The possible sources of

fermentable sugars are sugars in the rice and/or sugars produced by the
hydrolysis of starch. Polished non-glutinous rice does contain some
sugars. Reported values include total sugars of 1.2 g kg−1 dry matter
(USDA) and reducing sugars (includes glucose, fructose and maltose
but not sucrose) of 3.5 g kg−1 dry matter (Lu et al. 2008a; Table 6.2).
These concentrations would appear to be insufficient to account for
the concentrations of lactic and acetic acids observed in the fermenta-
tions. Observed concentrations of titratable acid (as lactic acid) range
from 5.7 to 10.1 g L −1 (Niyomvit 1985; Oupathumpanont et al. 2009;
Table 6.4) and reported concentrations of organic acids are lactic acid,
~5 g kg−1 (Suphichayangkoon et al. 2006; Figure 6.20b), 6.3 g L −1
(Lu et al. 2008b); and acetic acid, ~3 g kg−1 (Suphichayangkoon et al.
2006; Figure 6.20c) and 0.4 g L −1 (Lu et al. 2008b).
Lu et al. (2008b) observed increases in the concentration of reduc-
ing sugars in the fermentation supernatant from initial values 0.9
Table 6.2 Chemical Components of Raw and Fermented Rice Grains and Their Supernatants
CONCENTRATION (G KG−1 DRY MATTER)
NATURALLY FERMENTED RICE GRAINS (72 H)
WHOLE RAW-POLISHED
COMPONENT RICE GRAINS (INDICA) FACTORY A FACTORY B FACTORY C
Total starch 890 ± 9 900 ± 11 910 ± 2 900 ± 7
Amylose 205 ± 7 215 ± 3 210 ± 6 220 ± 5
Reducing sugar 3.5 ± 0.1 31 ± 1 36.5 ± 2 38 ± 1
Protein 45 ± 3 36 ± 1 39 ± 2 32 ± 6
Lipid 9 ± 1 7 ± 1 6 ± 3 7 ± 2
CONCENTRATION IN SUPERNATANT LIQUID (G L−1)

Reducing sugar 0.12 ± 0.01 3.5 ± 0.03 3.5 ± 0.11 4.1 ± 0.07
Source: Lu, Z.H. et al. 2008a. Journal of Applied Microbiology 105:893–903.
a Fermented rice grains were from three factories in Chengde City, Hunan Province, China. Raw rice
material and tap water in all factories were from the same supplier.
Table 6.3 Types of Lactic Acid Bacteria Capable of Hydrolysing Rice Starch during Thai
Commercial Production of Fermented Rice Flour for Noodles Production
TYPES OF LACTIC ACID BACTERIA CAPABLE OF HYDROLYSING RICE
STARCHa
(% OF TOTAL STARCH HYDROLYSING ISOLATES)
HOMO- HETERO- HOMO- HETEROFER
FERMENTATIVE FERMENTATIVE FERMENTATIVE MENTATIVE
PROCESS STAGE RODS RODS COCCI COCCI
Soaked, fermented polished 74–76 12–15 0–2 6–14
rice grains
Rice slurry (after wet milling) 74–80 4–17 0 6–16
Drained, wet flour 77–78 5–12 0 10–19
Source: Modified from Oupathumpanont, O. et al. 2009. Kasetsart Journal (Natural Sciences)
43:557–565.
a Able to grow and produce acid in rice flour medium.
to 1.6 g L −1 during the first 24 h of a natural laboratory fermenta-

tion while Lu et al. (2008a) (Table 6.2) recorded increases in con-
centration from an initial 0.12 to 3.5–4.1 g L −1 in rice fermentation
supernatants from three factories after 72 h fermentation. These
authors concluded that some hydrolysis of starch occurred and Lu
et al. (2008b) showed that this was concomitant with an increase in
α-amylase activity.
While amylase activity is not common among lactic acid bacteria, it
has been described in strains isolated from a variety of starch-process-
ing systems (Petrova et al. 2013) and Oupathumpanont et al. (2009)
found that a high proportion of lactic acid bacteria isolated from rice
fermentations could hydrolyse rice starch and produce acid in fer-
mented rice flour preparations (Table 6.3). Among homofermentative
Table 6.4 Titratable Acidity of Unfermented and Fermented Rice Flour Slurries
TITRATABLE ACIDITY (AS LACTIC ACID; G L−1)a
FERMENTED RICE FLOUR SLURRIES
FERMENTED BY FERMENTED BY COMMERCIAL
UNFERMENTED NATURAL L. PLANTARUM L. PLANTARUM NATURAL
RICE FLOUR FERMENTATION P1 P39 FERMENTATION
SLURRY 24 H 24 H 24 H 48 H
0.8 ± 0.4c 5.7 ± 0.5b 10.3 ± 0.4a 11.0 ± 0.6a 11.2 ± 0.2a
Source: Modified from Oupathumpanont, O. et al. 2009. Kasetsart Journal (Natural Sciences)
43:557–565.
a Means of five replicate samples. Means with different letters were significantly different (P < 0.05).
rods, L. plantarum P1 and L. plantarum P39 produced the highest

titratable acidity, 10 − 11 g L −1 (as lactic acid), in rice slurry after
24 h (Table 6.4). Sribuathong et al. (2004) observed lactic acid bac-
teria, Bacillus spp., yeasts and moulds at all the fermentation stages.
The main lactic acid bacteria found were species of Lactobacillus,
Leuconostoc, Enterococcus, Lactococcus, Pediococcus and Streptococcus. Of
these, the Enterococcus and Streptococcus isolates could hydrolyse starch
strongly. Lu et al. (2008a) noted that most of the microorganisms
isolated from rice starch fermentations displayed α-glucosidase activ-
ity, suggesting that they could hydrolyse starch, but did not examine
isolates directly for the ability to hydrolyse starch.
Chavana et al. (1991) studied the proximate composition during the

processing of naturally fermented rice noodles (Table 6.5). The major
Table 6.5 Chemical Changes during Three Commercial Natural Fermentations of Polished
Broken Rice to Make Fermented Rice Noodles
PROCESSING STAGE
BROKEN RAW STARCH
BROKEN RICE AFTER WET PARTIALLY STARCH
NON- FERMENTED GRINDING GELATI SLURRY FERMENTED
GLUTINOUS 2 DAYS, AND NISED PRIOR TO RICE
PROPERTY RICEa 34–40°C DE-WATERING STARCH EXTRUSION NOODLES
Moisture (%) 11.6–14.2 30.2–31.6 43.7–47.8 44.4–44.6 50.5 69.3–73.7
pH n.d.b 3.3–3.4 3.0–3.8 3.0–3.3 3.0–3.3 4.5
PROXIMATE ANALYSIS (% DRY MATTER)

Proteinc 7.0–7.7 6.5–7.2 5.4–5.95 3.4–5.7 5.8 4.4–5.6
Fat 1.0–1.2 0.67–0.95 1.5 n.d. 1.5 0.6–1.2
Ash 0.57–0.77 0.29–0.41 0.39–1.7 0.20–0.36 1.7 0.21–0.70
Fibre 0.64–0.73 0.41–0.62 n.d. 0.47–0.62 n.d. 0.79–1.3
Starch 87.5–91 89–91 90–90 91–91 89 90–91
Amylose 27–29 29–30 28–32.5 32 28.5 30–32
Source: Modified from Chavana, S., P. Tungtrakul, O. Naivikul et al. 1991. Changes in the chemical
composition of kanomjeen during processing. In Proceedings of 29th Kasetsart University
Annual Conference, pp. 417–425. Bangkok. (In Thai.)
a Factories were located in Chacherngsao Province (rice variety, Chaghauy), Chaiyapoom Province
(Pleagdang) and Phuket Province (Esan.)

b No data.
c N × 5.95.
changes were a progressive increase in water content with each pro-

cessing step, a substantial decrease in protein content, some loss of
minerals and acidification during the initial fermentation. The pH
value was then maintained in the range 3.0–3.8 throughout the subse-
quent stages and only with the final washing of the noodles did it rise
to 4.5. The increase in ash content in the starch slurry prior to extru-
sion was, presumably, due to the addition of salt to the slurry, which
was subsequently removed when the noodles were washed.
Lu et al. (2003) examined the changes in whole polished rice grains
(5 kg in 10 L tap water) naturally fermented in the laboratory at 35°C
for 27 h, by which time, the pH had fallen to 4.0 and the titratable
acidity had reached ~11 g L −1 in the supernatant liquid. The starch
and amylose content of the grains did not change significantly during
the fermentation but the crude protein content of the grains decreased
from 77 to 52 g kg−1 dry matter and the lipid content decreased from
8 to 4.3 g kg−1 dry matter, along with a concomitant increase in free
fatty acids. The ash content was also reduced from 3.5 to 2.1 g kg−1
dry matter, presumably due to minerals diffusing from the grains into
the water phase. Lu et al. (2008a) observed similar reductions in the
concentrations of protein and lipid in commercially fermented rice
grains (Table 6.2). Presumably, the reduction in protein content was
due to loss of water-soluble proteins while the reduction in lipid con-
tent and increase in free fatty acids suggests the occurrence of lipolytic
activity.
6.1.7.1 Fermentation Products The fermentation is primarily a lactic

acid bacterial one and consequently, the main metabolic products are
lactic and acetic acids. The production of acetic acid implies that at
least some of the lactic acid bacteria are heterofermentative species
and this is supported by the isolation of heterofermentative species
by several authors (Oupathumpanont et al. 2009; Sribuathong et al.
2004; Suphichayangkoon et al. 2006). If extracellular electron accep-
tors are not available, heterofermentative lactic acid bacteria may also
produce ethanol instead of acetate (see Figure 6.1, Chapter 1). The
other possible source of ethanol is fermentative yeasts and, although
S. cerevisiae and other fermentative yeasts were isolated by Lu et al.
(2008a), the numbers present were low. There do not appear to be any
reports on ethanol levels in rice fermentations.
Suphichayangkoon et al. (2006) monitored pH and organic

acids during natural fermentations and in fermentations by single
lactic acid bacterial strains. The pH value was reduced to below 5
within 24 h and decreased further to 4 or below in the fermented
sedimented starch (Figure 6.20a). The final pH of the noodles was
slightly higher, presumably due to the leaching of acids from them.
Much lower pH values of 3.0–3.8 were observed by Chavana et al.
(1991) (Table 6.5).
As expected, the concentrations of lactic and acetic acids increased
during the fermentations but only reached maximum values of ~5
g kg−1 of lactic acid and ~3 g kg−1 of acetic acid (Figure 6.20b and
c). These are quite low concentrations compared to those present in
many fermented foods and the low pH values observed are probably a
consequence of the low pH-buffering capacity of the starch substrate.
With a low-buffering capacity, only a relatively small amount of acid
is required to drop the pH value to levels that inhibit further growth
and acid production. The concentrations of lactic and acetic acids in
the final fermented noodles were very low, at 0.5–0.8 g kg−1, as they
were, presumably, leached into the large volumes of boiling water used
to receive the extruded noodles and the cold water used to wash the
noodles.
6.1.7.2 Physico-Chemical Changes Fermented rice noodles, especially

those made from aged rice that has been stored for several months,
are generally described as having better texture than those made
from unfermented rice, being more elastic and chewy than unfer-
mented noodles (Lu et al. 2009a). Youngprasitiporn (1989) noted
that fermented rice noodles made from rice fermented for only 1 day
was a dull turbid-white colour, had a cooked rice-like odour and fla-
vour and had a rough and non-flexible texture. Noodles from rice
fermented for 2 days were creamy white and had a fermented odour
and flavour with a soft springy texture. Noodles from rice fermented
for 3 days were even shinier, creamy yellow in colour and with a more
distinct odour and flavour together with a soft springy and flexible
texture.
Lu et al. (2003, 2008b) measured some physical properties of
laboratory-made noodles. The fermented noodles were slightly

more elastic than unfermented noodles, with maximum extensions
of 11–17% and 8.2–8.4%, respectively, of the initial lengths of the

noodles. The observed maximum tensile stress values (maximum ten-
sile force required to break the noodle/initial cross-sectional area of
the noodle) ranged from 54 (2003) to 97 (2008b) kPa for fermented
noodles and from 120 (2008b) to 150 (2003) kPa for unfermented
noodles. However, while measurements made on one occasion were
quite reproducible (standard deviations of 5 or 6 replicate determina-
tions were 1.5–7.5% of the means), there are large differences between
the values recorded in 2003 and those found in 2008b. Sensory evalu-
ation of the 2003 samples described the fermented noodles as soft,
pliable, clear, white and chewy while the unfermented noodles were
described as firm, not pliable, opaque and crumbly.
Lu et al. (2008b) investigated the effects of a number of treatments
of rice grains prior to making fermented noodles. Treatment of rice
grains with α-amylase, pectinase or cellulase or the addition of glu-
cose or maltose to rice flour tended to have detrimental effects on the
tensile strength and textures of noodles made from them. In contrast,
rice grains treated with protease, lipase or lactic acid (pH 4) yielded
noodles having similar tensile strengths and organoleptic properties
to those made from naturally fermented rice. The authors concluded
that the purity of the starch is an important factor in determining
noodle texture.
Lu et al. (2005, 2007, 2009a) found that the amorphous amylopec-
tin portion within starch granules seemed to be the most susceptible
to hydrolysis by microorganisms and suggested that this was a key fac-
tor in inducing modification of the thermal and rheological properties
of fermented rice flour gels and noodles. Gels made from fermented
rice flour formed earlier and hardened much quicker than those made
from unfermented flour. During storage at 4°C for 1, 3 and 6 h, hard-
ness increased from 295 to 460 to 599 kPa in gels made from fer-
mented starch compared with from 313 to 482 to 816 kPa for those
made from unfermented starch. Upon storing for more than 1 h, the
hardness of both increased, indicating retrogradation of amylopectin
which hardened and impaired the desirable texture of rice noodles,
resulting in an undesirable, crumbly breakage of the gel structure.
Gels made from fermented starch were also less brittle than gels made
of unfermented starch. During storage, brittleness increased from 3
to 66 to 78 kPa in fermented starch gels compared with from 49 to
70 to 137 kPa for unfermented starch gels. After storage for 1 h, the
texture of fermented starch gel was more elastic but less firm than that
of unfermented starch gels.
In conclusion, there are evident differences in texture and appear-
ance between fermented and unfermented noodles. It is recognised
that the properties of the rice starch strongly influence the charac-
teristics of rice noodles, fermented or unfermented, but the specific
effects are not well defined and precise quality standards for rice for
noodles production had not been formulated by 2010 in China (Lu
and Collado 2010) or to date in Thailand. Additionally, the role played
by storing rice on the properties of fermented noodles is not clear.
6.1.7.3 Effects of Fermentation on Starch Properties Fermentation

reduces the gelatinisation temperature of rice flour (Lu et al. 2009b;
Oupathumpanont et al. 2008; Wong et al. 2000). Oupathumpanont
et al. observed significant reductions in gelatinisation onset, peak and
conclusion temperatures (Table 6.6) but the reductions observed by Lu
et al. were only significant for gelatinisation onset temperature (69.8°C
for fermented rice flour and 70.7°C for unfermented flour) and peak
temperature (73.6°C and 74.3°C, respectively). However, the standard
deviations reported by Oupathumpanont et al. are much larger than
those reported by Lu et al. and it is surprising that the observed tem-
perature differences were significant.
Oupathumpanont et al. (2008) measured many properties of rice
flour fermented by L. plantarum P1 (Table 6.6). The pasting proper-
ties, essentially reflecting properties of the starch, of the fermented
flour were substantially different from those of unfermented flour. The
fermented material had a lower pasting temperature (the temperature
at the onset of increase in viscosity, an indication of the minimum tem-
perature required to cook the test sample and an indicator of energy
cost), lower peak viscosity (occurs at the equilibrium point between
swelling of starch granules, causing viscosity to increase, followed by
rupture of starch granules and polymer leaching, causing viscosity to
decrease; a measure of the water-binding capacity of the starch or mix-
ture; water binding is lower with a decreased peak viscosity), higher
breakdown (the rate and extent of reduction from the peak viscosity to
a minimum value, known as the holding strength, hot paste or trough
viscosity; typically, at 95°C, a high breakdown means that the starch
Table 6.6 Effect of Fermentation by L. plantarum P1 at Room Temperature (35°C) for 24 h on
the Physico-Chemical Properties of Rice Flour
UNFERMENTED RICE FLOUR FERMENTED
PROPERTY RICE FLOURa WITH L. PLANTARUM P1a
pH 6.25 ± 0.06a 3.5 ± 0.08e
Titratable acidity (as lactic acid, g kg−1) 0.3 ± 0.4e 10.1 ± 0.3a
Amylose content (g kg−1) 300 ± 5b 335 ± 4a
Solubility at 95°C (%) 13e 28a
Swelling power at 95°C (g g−1) 16.5a 15e
PASTING PROPERTIES
Pasting temp (°C) 82.8 ± 0.05a 81.4 ± 0.34c
Peak viscosity (RVU) 355 ± 1a 285 ± 2e
Final viscosity at 50°C (RVU)b 490 ± 0.7a 375 ± 3.6e
Breakdown (RVU) 64 ± 1.4e 71 ± 1.6a
Setback (RVU) 210 ± 3a 155 ± 2e
THERMAL PROPERTIES
Gelatinisation onset temperature, To (°C) 71.7 ± 0.8a 71.2 ± 0.4d
Peaking temperature, Tp (°C) 75.4 ± 0.9a 74.1 ± 0.4d
Conclusion temp, Tc (°C) 79.5 ± 0.5a 78.2 ± 0.9d
Enthalpy (J g−1 dry matter) 19.1 ± 0.8a 17.8 ± 0.8c
SEM starch granule images No Pits With Pits
Source: Adapted from Oupathumpanont, O. et al. 2008. Effects of Lactobacillus plantarum P1 on
the physico-chemical properties of rice flour. In Proceedings of 46th Kasetsart University
Annual Conference: Agro-Industry, pp. 473–480. Bangkok: Kasetsart University.
a Means of five replications. Values with different letters were significantly different (p < 0.05).
b Relative viscosity unit; 1 RVU = ~10 cP.
granules are more susceptible to disintegration than those in a sample

with a lower breakdown), lower setback (the rate of retrogradation and
an indicator for the texture and consistency of a starch paste during
cooling) and lower final viscosity at 50°C (an indicator of the ability of
the material to form a viscous paste or gel after cooking and cooling)
than the unfermented starch (Delwiche et al. 1996; Rogers 2010).
Lu et al. (2009b) also showed that the fermented rice flour was
more resistant to breakdown during heating and retrograded more
slowly than unfermented flour.
During fermentation by L. plantarum P1, the starch grains acquire
pits (Oupathumpanont et al. 2008) where the starch has, presumably,
been hydrolysed. At the same time, the percentage of soluble mate-
rial increased (Table 6.6) and the relative increase in amylose content
suggests that it was amylopectin components that were preferentially
hydrolysed (Lu et al. 2007). The fermented rice flour exhibited less
swelling power than the unfermented flour (Table 6.6), indicating
that it absorbed less water than unfermented rice flour, possibly due
to its higher amylose content. In contrast, Lu et al. (2005, 2008b)
observed only slight surface erosion on the surface of fermented rice
starch granules and insignificant changes in total starch or amylose as
a result of fermentation.
6.1.7.4 Changes in Volatile Compounds The major volatiles in fer-

mented noodles included 2-methylpropanoic acid, 3-methylbutanoic
acid, diacetyl, ethyl acetate, ethyl valerate, ethyl hexanoate, ethyl
heptanoate, heptanal, octanal, nonanal, decanal and 3-methylbu-
tanal (Keatkrai and Jirapakkul 2010). Most of the volatiles appeared
after 24 h fermentation of broken rice but esters only appeared dur-
ing the fermentation of the starch sediment. Volatiles contributing
significantly to flavour and aroma of fermented noodles included
acetic acid, butyric acid and isopentanoic acid. Unfermented bro-
ken rice contained significant amounts of ethyl acetate, heptanal and
nonanal but no 2-methylpropanoic acid or 3-methylbutanoic acid.
6.1.7.5 Shelf Life of Fermented Noodles Freshly made noodles have the
best quality and mouth-feel but the texture then deteriorates with
time due to progressive retrogradation of amylopectin (Kraidej et al.
2002; Lu and Collado 2010; Table 6.7). Freshly prepared noodles
have a shelf life of 2–3 days at ambient temperature, but this can be
extended to 1 week if the noodles are packed in plastic bags to exclude
air (Abngern 2012). Evidently, spoilage is primarily due to the growth
of aerobic microbes.
Vacuum-packed fermented rice noodles pasteurised at 70°C for
30 min and kept under refrigeration for 4 weeks, were moderately
sensorily acceptable (Muangprasit 2002) while fast-frozen noodles
may be acceptable for up to a year (Lu and Collado 2010).
6.1.7.6 Yield The main material losses occur during the conversion

of rice into fermented starch. Kraidej et al. (2002) reported that for
15 factories, mainly in northeastern Thailand, producing between 6
and 1000 t of fermented rice noodle starch per year as the final prod-
uct from broken rice, the material losses ranged from 20% to 38% of
Table 6.7 Characteristics of Fermented Noodles Stored at Ambient Temperature (27–33°C) and
at 4°C
STORAGE CONDITIONS PROPERTIES OF STORED NOODLESa
TITRATABLE TOTAL VIABLE
TEMP STORAGE ACIDITY COUNT
ERATURE TIME (G KG−1 AS (LOG
(°C) (DAYS) PH LACTIC ACID) CFU G−1) SENSORY PROPERTIES
27–33 1 3.9–5.6 0.6–1.5 1.5–5.2 White, mild fermented odour,
bland taste, elastic texture
and acceptable
27–33 2 3.9–4.7 0.8–1.5 1.0–5.7 White, stronger acidic odour,
slight acidic taste, moist and
less acceptable
4 1 n.d. n.d. 1.9–3.5 White, mild fermented odour,
bland taste, elastic texture
and acceptable
4 2 4.2–4.3 0.8–1.0 n.d.b White, mild fermented odour,
bland taste, texture a little
hard and acceptable
4 3 4.1 0.9–1.8 2.2–4.6 White, mild fermented odour,
slightly sour taste, texture hard
and friable and unacceptable
Source: Adapted from Kraidej, L. et al. 2002. Quality standard and Thai identity of kanomjeen (fer-
mented rice noodle) produced in industrial scale. Final research report presented to
National Center of Genetic Engineering and Biotechnology, Office of National Science and
Technology, Bangkok. (In Thai, Personal communication.)
a Total of 11 samples from six fermented rice noodle factories.
b No data.
the starting material. For a factory producing 690 t fermented starch

(45% moisture) from 575 t broken rice (13.5% moisture), the dry mat-
ter losses were 24% (Kraidej et al. 2002). It is evident that produc-
tion costs could be greatly reduced if these losses could be minimised.
Saitong (2005) suggested that much of the waste derived from residual
rice flour could be readily recycled to ethanol or biogas production. It
is also likely that the soluble materials released during the fermenta-
tion are leached out and lost in the subsequent manufacturing steps.
Fermented noodles are ~90% starch, 3.2–6.0% protein, 0.3–1.2%

fat, 0.5–1.3% fibre, 0.2–0.7% ash and 69–74% moisture (Table 6.5;
Chavana et al. 1991). Rice noodles have a higher protein content than
most other starch noodles (Lu and Collado 2010) and are also suitable
for persons allergic to gluten as gluten is absent from rice.
However, most minerals and vitamins in rice are present in the
embryo, testa, pericarp and aleurone layers rather than in the endo-
sperm and are removed with the bran. Consequently, polished rice
and foods made from it are relatively deficient in minerals and vita-
mins, and need to be supplemented with other foods (FAO and
Juliano 1993).
As a lactic acid bacterially fermented product, fermented rice starch

has an acidic pH value below 4.0 and down to 3.0 (Tables 6.5 and
6.6). At these pH values, 85–98% of the acetic acid and 43–88% of
the lactic acid are in the toxic undissociated forms and the concen-
trations are likely to be sufficient to inhibit the growth of potential
spoilage or pathogenic bacteria. Hence, the fermented starch is quite
stable and susceptible only to the growth of acid-tolerant yeasts and
moulds. The fermented noodles have a less acidic pH value and the
concentrations of lactic and acetic acids are quite low (Figure 6.20b
and c) and, judging by the short shelf life of the noodles, the concen-
trations of undissociated acids are not sufficient to ensure microbio-
logical stability.
Community product standard (TCPS 500/2547) issued by the
Thai Industrial Standards Institute (2004a) relates to rice noodles
packed in containers. It requires that rice noodles must be made from
polished non-sticky rice grains or brown (unpolished) non-sticky rice
grains that are fermented or unfermented, wet milled and pressed to
drain with a heavy weight. They may be mixed with other components,
such as pandan (P. amaryllifolius; Thai name, bai toey) or butterfly pea
flowers (C. ternatea; dok anchan in Thai). Alternatively, noodles may
be made from starch for rice noodle processing. The specifications
include the following: colour should relate to the natures and colours of
the components used and should be consistent; odour without a musty
or spoiled smell; good smell and taste characteristic of the natural
flavour of noodles without other undesirable odours and flavours; tex-
ture must be soft elastic or flexible without sticky crumbs; free of for-
eign matters, such as hair, dirt, sand, gravel, wood chips or fragments
and animal waste; if bleach, preservatives or thickening additives are

used, they should conform to permitted types and amounts; pH value,
fermented flour noodles, 3.0–4.5, non-fermented flour, 4.0–6.0 and
microorganisms, total microbial count <1 × 106 cfu g−1, Staphylococcus
aureus <100 cfu g−1, Bacillus cereus <100 cfu g−1 and Escherichia coli by
most probable number method <3 g−1.
Ngokpilai et al. (2012) surveyed the kanomjeen produced by 77
Thai production units. Only three contained total microbial counts
exceeding 1 × 106 cfu g−1, only one exceeded 100 cfu g−1 S. aureus and
one had more than three E. coli cfu g−1. Six samples had pH values
higher than the recommended 3.0–4.5. Of the 77 factories surveyed,
32 were found to be adding benzoic acid at concentrations <1000 mg
kg−1 (maximum permitted level) and only one factory added more
than 1000 mg kg−1. Sorbic acid was not detected in any sample. This
survey would suggest that the majority of producers are making good-
quality fermented rice noodles.
There are a large number of noodle manufacturers in Thailand, most

of whom are relatively small producers using more or less traditional
methods. In China, the production of fermented noodles has been
substantially mechanised, with a process in which rice is fermented
for 3–6 days, then ground and the slurry poured onto a conveyer belt,
passed through a steam tunnel, extruded, cooled and washed (Lu
and Collado 2010). This technology has been adopted in Thailand for
manufacturing many kinds of industrialised noodles, especially dehy-
drated ones (Vongsawasdi et al. 2009), such as vermicelli which has a
completely different texture from fermented rice noodles.
6.1.10.1 Unpolished Rice Noodles There is an increase in the commer-

cial availability of unfermented and fermented noodles made from
unpolished rice noodles, which are nutritionally superior to noodles
made from polished rice. They have a softer texture and shorter shelf
life, of about a day, than polished rice noodles. Fresh fermented and
unfermented noodles are currently on sale in Thai supermarkets in
small (~200 g noodles wet weight) polyethylene terephthalate (PET)
containers with or without separate condiments.
6.1.10.2 Commercial Fermented Rice Noodle Starch Fermented rice

noodle flour or starch is increasingly being used by small-scale com-
mercial noodle factories (personal communication) as starting raw
material for the manufacture of fermented rice noodles due to the
savings of time, labour and space compared with using the traditional
process starting with broken rice grains. Thai Community Product
standard TCPS 499/2547 (Thai Industrial Standards Institute 2004b)
specifies that the starch may be made from polished or unpolished
non-sticky rice and should have a moisture content of 35–50%. It
should, generally, be packed in 10 kg quantities in double-layer bags,
with an inner layer made of a roughly woven polyethylene or cloth to
allow excess water to leak out into an outer bag made of high-density
polyethylene. It should be stored in a cool place in the shade. Under
warm (30–35°C) and humid conditions, the starch quickly ferments
and gains a sour odour due to growth and acid production by lactic
acid bacteria. If it becomes too acidic, the starch can be soaked in
water for 3–4 h to reduce the acidity prior to cooking. Under low-
humidity conditions, moulds may grow on the surface of the starch
bag. If this is a problem, mould growth may be prevented by storing
the bags of pressed fermented starch in 14% brine. Proper conditions
of storage can maintain the starch quality for 2 weeks.
6.1.10.3 Dried Noodles The shelf life of fermented rice noodles can be
extended by drying to <12% moisture content and small packages of
~200 g of dried fermented or unfermented noodles in PET containers
are on sale in supermarkets currently in Thailand (Figure 6.21).
Figure 6.21 Dehydrated white fermented noodles. (Courtesy of Monratanin Pinthong.)

Although the numbers of manufacturers of fermented noodles in

Thailand have decreased in recent years, there is still a large number
of small producers making noodles by essentially traditional methods.
The production of unfermented noodles is much greater than that of
fermented noodles and it appears that most consumers make little
distinction between them. Of course, this may be because the so-
called unfermented noodles prepared by traditional methods have, in
fact, undergone some fermentation during their manufacture.
Fermented rice noodles have received relatively little fundamen-
tal scientific attention and there are several questions that need to
be addressed if their production is to be placed on a more secure
footing:
1. The microbiology of the process is not well understood and
molecular methods would allow the population dynamics of
the fermentation to be followed and the identification of the
relative roles of homofermentative and heterofermentative
lactic acid bacteria and yeasts.
2. What is the source(s) of amylase required to hydrolyse starch
to fermentable sugars?
3. What are the role(s) of salt in the process? What are the levels
of salt in the final product and how may these be optimised
and/or minimised?
4. What are the physico-chemical changes occurring (studies
are needed with pure cultures using sterilised substrate and
aseptic fermentation conditions)?
5. How much gelatinisation of starch occurs and what are the
optimum proportions of gelatinised and ungelatinised starch
for good noodle texture? How may the extent of gelatinisa-
tion be precisely controlled?
6. What are the optimum levels of pH and organic acids in the
final fermented noodles? It is evident that much of the organic
acids are leached out during the final extrusion and washing
processes and it is possible that this should be better controlled.
7. What determines the shelf life of fermented noodles at ambi-
ent and refrigeration temperatures? Is it microbial growth or
changes in texture due to retrogradation of starch?
8. What characteristics should the rice have for the production

of noodles of high and consistent quality?
9. How may material losses and the production of waste-water be
minimised? For example, dry-milled rice flours could be used
in place of wet-milled flours to reduce the material loss conse-
quent on eliminating the excess water used with wet milling and
could also generate less waste-water (Rosell and Collar 2007).
However, dry milling can also introduce undesirable properties
(Yeh 2004) and may not yield a satisfactory noodle product.
10. Could the differences between fermented and unfermented
noodles be accentuated, such that consumers would more
readily discriminate between them?
11. Is it possible to use polymer-producing lactic acid bacteria to
aid adhesion of the starch and improve or modify the texture
of the noodles, with possible benefits in reduced processing?
12. How may the quality of dried noodles be maximised?
13. Could a fermented germinated brown-rice noodle, nutrition-
ally superior to fermented polished rice noodle, be made and
marketed? The utilisation of germinated brown rice, prepared
by soaking brown rice for ~20 h at 30–40°C, has been increas-
ing. It is used to make many rice products, including germinated
brown rice balls, soups, breads, doughnuts, cookies and rice
burgers (Nip 2007). Germinated brown rice could be evaluated
for making fermented rice noodles (Songsermpong et al. 2011).
References
Abngern, S. 2012. S. Sutti Kanomjeen Factory, Mae-Rim District, Chiang
Mai, Thailand. (Personal communications).
Amornsin, A. 2003. Properties of waxy rice starch and rice grain: Process of
an instant waxy rice product. Thesis, The University of Georgia, Athens,
Georgia, USA, pp. 21–38.
Champagne, E.T., D.F. Wood, B.O. Juliano and D.B. Bechtel. 2004. The
rice grain and its gross composition. In Rice: Chemistry and Technology,
third ed., edited by E.T. Champagne, pp. 77–107. St. Paul, Minnesota:
American Association of Cereal Chemists.
Champangern, S. 2004. Application of texture and viscosity analyses to predict
adulteration in Khao Dawk Mali 105, classify group of amylose and stor-
age period of rice. MS thesis, Food Science and Technology Department,
Kasetsart University, Bangkok. (In Thai).
Chavana, S., P. Tungtrakul, O. Naivikul, M. Suwanna-adth, L. Kraidej, P.

Sirirote and P. Patananurak. 1991. Changes in the chemical composi-
tion of kanomjeen during processing. In Proceedings of 29th Kasetsart
University Annual Conference, pp. 417–425. Bangkok. (In Thai).
Dellaglio, F., S. Torriani and G.E. Felis. 2004. Reclassification of Lactobacillus
cellobiosus Rogosa et al. 1953 as a later synonym of Lactobacillus fermen-
tum Beijerinck 1901. International Journal of Systematic and Evolutionary
Delwiche, S.R., K.S. McKenzie and B.D. Webb. 1996. Quality characteris-
tics in rice by near-infrared reflectance analysis of whole-grain milled
samples. Cereal Chemistry 73(2):257–263.
Department of Industrial Information Center, Thai Ministry of Industry. n.d.
Registered fermented rice-noodle factories (kanomjeen factories). http://
userdb.diw.go.th/results1.asp (accessed 15 June 2013).
FAO and B.O. Juliano. 1993. Rice in Human Nutrition. Rome: Food and
Agriculture Organisation of the United Nations in collaboration with
The International Rice Research Institute. www.fao.org/docrep/t0567e/
T0567E00.htm#Contents (accessed 21 August 2013).
Hongsawiwat, N., J. Tubnium, N. Hongsawiwat, P. Pongpanta, O. Emsabye and
S. Rodyan, eds. 2009. Kanomjeen, third ed., pp.1–21. Bangkok: Sangdad
Press Ltd. (In Thai).
Huang, H.T. 2000. Science and Civilisation in China. Volume 6, Biology and
Biological Technology. Part V, Fermentations and Food Science, p. 491.
Cambridge: Cambridge University Press.
Jane, J.-L., C.C. Maningat and R. Wongsagonsup. 2010. Starch characteriza-
tion, variety and application. In Industrial Crops and Uses, edited by B.P.
Singh, pp. 207–235. Cambridge, MA: CAB International.
Keatkrai, J. and W. Jirapakkul. 2010. Volatile profile of kanomjeen, Thai fer-
mented rice noodles and the changes during the fermentation process.
Science Asia 36:46–51.
Kongkiattikajorn, J. 2008. Effect of storage time and temperature on volatile
aroma compounds and physicochemical properties of rice. In Proceedings
of 46th Kasetsart University Annual Conference: Science, pp. 478–485.
Bangkok: Kasetsart University. http://kulc.lib.ku.ac.th/kucon/index.php/
meeting/107-m4605 (accessed 9 December 2013).
Kongkiattikajorn, J. and S. Photchanachai. 2000. Effect of temperature on
physico-chemical properties of milled rice cv. Khaw Dawk Mali 105 dur-
ing storage. In Proceedings of 38th Kasetsart University Annual Conference:
Fishery and Science, pp. 395–402. Bangkok: Kasetsart University. (In Thai).
Kongkiattikajorn, J., V. Tosingharach, S. Mongkol and P. Rattanachaisit. 2011.
Analysis of tocopherol content and antioxidant properties of rice bran
and correlation to cereal color during storage. In Proceedings of 49th
Kasetsart University Annual Conference: Science, pp. 250–257. Bangkok:
Kasetsart University.
Kongseree, N. 2002. Rice standard. In Rice Quality and Determination of Thai
Adulteration in Khao Dawk Mali 105, edited by N. Kongseree, pp. 31–45.
Bangkok: Chirawatana Express Ltd. for Office of Industrial Economics,

Ministry of Industry. (In Thai).
Kongseree, N. 2004. Cooked rice quality. In Rice Quality and Determination of
Thai Khao Hom Dawk Mali, edited by N. Kongseree, pp. 41–61. Bangkok:
Chirawatana Express Ltd. for Office of Industrial Economics, Ministry
of Industry. (In Thai).
Kraidej, L., C. Wongalueng, M. Bunratanakorakit, P. Chimanak, W. Prapaitong,
P. Tungtrakul, S. Sonsoawapak and P. Sirirote. 2002. Quality standard
and Thai identity of kanomjeen (fermented rice noodle) produced in
industrial scale. Final research report presented to National Center of
Genetic Engineering and Biotechnology, Office of National Science and
Technology, Bangkok. (In Thai, Personal communication).
Likitwattanasade, T. and P. Hongsprabhas. 2010. Effect of storage proteins
on pasting properties and microstructure of Thai rice. Food Research
International 43:1402–1409.
Litisorn, S., C. Punpeng, S. Tangtongkum, P. Ketrom, R. Chitmontree and N.
Chiyapun. 2005. Debt problem solving of Klong Kanomjeen Culture
Community in Sena District, Ayuthaya Province. Final research report,
Office of Research Fund, Chiang Mai University, Chiang Mai, Thailand,
pp. 1–28. (In Thai).
Lu, Z.H., H.H. Peng, W. Cao, E. Tatsumi and L.T. Li. 2008a. Isolation, char-
acterization and identification of lactic acid bacteria and yeasts from sour
Mifen, a traditional fermented rice noodle from China. Journal of Applied
Lu, Z.H., L.T. Li, W. Cao, Z.G. Li and E. Tatsumi. 2003. Influence of natu-
ral fermentation on physico-chemical characteristics of rice noodles.
International Journal of Food Science and Technology 38:505–510.
Lu, Z.-H., L.-T. Li, W.-H. Min, F. Wang and E. Tatsumi. 2005. The effects
of natural fermentation on the physical properties of rice flour and the
rheological characteristics of rice noodles. International Journal of Food
Science and Technology 40:985–992.
Lu, Z.H., M.L. Yuan, T. Sasaki, L.T. Li and K. Kohyama. 2007.
Rheological properties of fermented rice flour gel. Cereal Chemistry 84
(6):620–625.
Lu, Z.H., T. Sasaki, N. Kobayashi, L.T. Li and K. Kohyama. 2009a. Elucidation
of fermentation effect on rice noodles using combined dynamic viscoelas-
ticity and thermal analyses. Cereal Chemistry 86 (1):70–75.
Lu, Z.-H., T. Sasaki, Y.-Y. Li, T. Yoshihashi, L.-T. Li and K. Kohyama. 2009b.
Effect of amylose content and rice type on dynamic viscoelasticity of a
composite rice starch gel. Food Hydrocolloids 23 (7):1712–1719.
Lu, Z.H., W. Cao, H.H. Peng et al. 2008b. Effect of fermentation metabolites
on rheological and sensory properties of fermented rice noodles. Journal
of the Science of Food and Agriculture 88:2134–2141.
Lu, Z.-H. and L.S. Collado. 2010. Rice and starch-based noodles. In Asian
Noodles: Science, Technology, and Processing, edited by G.G. Hou, pp. 393–
432. New York: John Wiley.
Makmoon, S., S. Limpiti and P. Theanjumpol. 2010. Using near infrared spec-
troscopy to detect adulteration in milled rice cv. Khao Dawk Mali 105 with
cv. Chainat 1. Agricultural Science Journal 41 (Suppl.1):377–380. (In Thai).
Marianne, G. 1985. Flavor effects of sodium chloride. Food Technology.
39(6):47–52, 56.
Muangprasit, P. 2002. Development of Thai rice noodle production process
to improve the shelf-life and the safety quality. MSc thesis, Kasetsart
University, Thailand.
Ngokpilai, W., T. Parinyasiri, A. Jindaprasert, V. Areekul and A. Swetwiwathana.
2012. A study of quality and safety of kanom-jeen (fermented and fresh
rice noodle). SWU Science Journal 28 (1):121–134. http://202.29.22.173/
ArticleScan/000472.pdf (accessed 9 November 2013). (In Thai).
Ninchan, B. 2005. Physico-chemical properties of rice flour and starch of dif-
ferent Thai varieties. MSc thesis, Department of Biotechnology, Faculty
of Agro-Industry. Kasetsart University, Bangkok, Thailand.
Nip, W.-K. 2007. Asian (Oriental) noodles and their manufacture. In Handbook
of Food Products and Manufacturing: Principles, Bakery, Beverage, Cereals,
Cheese, Confectionery, Fats, Fruits, Functional Foods, edited by Y.H. Hui,
pp. 539–564. New Jersey: John Wiley.
Niyomvit, N. 1985. Khanom-jeen. Food Journal 15 (3):123–129. Cited by
Oupathumpanont et al. 2009.
Nutrition Division. 2001. Nutritive Values of Thai Foods, edited by S.
Boonwisoothi, S. Poosimuang, P. Wacharangkun, C. Srichaiya, N.
Yothinpatana, S. Silapaprom, N. Chongchaitet et al., pp. 9–10. Bangkok:
Department of Health, Ministry of Public Health. http://nutrition.
anamai.moph.go.th/temp/files/Nutritive%20Values%20of%20Thai%20
foods.pdf (accessed 12 November 2013).
Oupathumpanont, O., T. Suwonsichon, V. Haruthaithanasan, P. Chompreeda
and W. Chantarapanont. 2008. Effects of Lactobacillus plantarum P1
on the physico-chemical properties of rice flour. In Proceedings of 46th
Kasetsart University Annual Conference: Agro-Industry, pp. 473–480.
Bangkok: Kasetsart University.
Oupathumpanont,O.,W.Chantarapanont,T.Suwonsichon,V.Haruthaithanasan
and P. Chompreeda. 2009. Screening lactic acid bacteria for improving the
Kanom-jeen process. Kasetsart Journal (Natural Sciences) 43:557–565.
Payakaranon, J. 1999. Kanomjeen. In Thai Cultural Encyclopedia 2, Central
Thailand, pp. 513–519. Bangkok: Thai Cultural Encyclopedia Foundation,
Siam Press Management. (In Thai).
Petrova, P.K.P. and G. Stoyancheva. 2013. Starch-modifying enzymes of lac-
tic acid bacteria—Structures, properties and applications. Starch-Starke
65:34–47.
Piyakul, T. and U. Nuthong. 1999. Kanomjeen. In Thai Cultural Encyclopedia
2, Southern Thailand, pp. 610–613. Bangkok: Thai Cultural Encyclopedia
Foundation, Siam Press Management. (In Thai).
Rogers, J., ed. 2010. Appendix D—Interpreting test results. In RVA-SA—
Installation and Operation Manual, pp. 57–60. Warriewood, Australia:
Perten Instruments.
Rosanaphaiboon, T. 1987. Khanom-jeen. In Traditional Foods and Their

Processing in Asia, edited by F. Yanagida, Y. Takai, S. Homa, S. Kato and Y.
Ando, p. 48. Tokyo: NODAI Research Institute.
Rosell, C.M. and C. Collar. 2007. Rice-based products. In Handbook of Food
Products and Manufacturing: Principles, Bakery, Beverages, Cereals, Cheese,
Confectionery, Fats, Fruits, Functional Foods, edited by Y.H. Hui, pp. 523–
538. New Jersey: John Wiley.
Saitong, P. 2005. Biogas production from fermented rice noodles (kanomjeen);
processing waste by two-stage anaerobic digestion. MSc thesis, Kasetsart
University, Thailand. (In Thai).
Shih, F.F. 2004. Rice proteins. In Rice: Chemistry and Technology, third ed.,
edited by E.T. Champagne, pp. 143–1162. St. Paul, Minnesota: American
Association of Cereal Chemists.
Songsermpong, S., C. Charunuch and O. Naivikul. 2011. Twin-screw extrusion
of pre-germinated brown rice: Physicochemical properties and gamma-
aminobutyric acid content (GABA) of extruded snacks. International
Journal of Food Engineering 7 (4):1–15.
Sribuathong ,S., D. Bundidamorn and S. Trevanich. 2007. Use of lactic acid
bacteria as starter culture for production of safety and quality khanom-
jeen (Thai fermented rice noodle). In Proceedings of the 2nd International
Conference on Fermentation Technology for Valued Added Agricultural
Products, Khon Kaen, Thailand. Abstract P1–9, p. 74. http://fervaap.
kku.ac.th/conference/images/Conference2007/FerVAAP2007_Abs.pdf\
(accessed 12 November 2013).
Sribuathong, S., S. Trevanich, W. Jirapakkul and O. Naivikul. 2004. Changes
in microbial population during kanomjeen process and characterization
of lactic acid bacteria for kanomjeen starter. Proceedings of 42nd Kasetsart
University Annual Conference: Fisheries and Agro-Industry, pp. 307–314.
Bangkok: Kasetsart University. (In Thai).
Suphichayangkoon, N., W. Jirapakkul and O. Naivikul. 2006. Effect of lactic
acid bacteria starter culture on chemical properties in Thai fermented rice
(kanomjeen) noodle process. In Proceedings of 44th Kasetsart University
Annual Conference: Agro-Industry, pp. 356–362. Bangkok. http://kukr.lib.
ku.ac.th/Fulltext_kukr/KU0335074c.pdf (accessed 9 December 2013).
(In Thai).
Thai Agricultural Standard. 2012. Thai Agricultural Standard TAS 4004-2012-
Rice, edited by Kongseri, N., T. Wongleelaset, S. Rachruangrabin, P.
Urairong, W. Srisuriyathada, C. Methakulwat, E. Punyakich et al. Office
of Thai Agricultural Standard and Food. Ministry of Agriculture and
Cooperatives, Bangkok. (In Thai). www.acfs.go.th/standard/download/
RICE-1.pdf (accessed 9 December 2013).
Thai Food and Drug Administration. 2004. Thai FDA Regulations Regarding
Heavy Metal and Pesticide Residue Contamination. Bangkok: Ministry of
Health, Thailand.
Thai Industrial Standards Institute. 2004a. Kanomjeen (Thai Rice Noodle) TCPS
500/2547. Thai Community Product Standard, Ministry of Industry,
Thailand.
Thai Industrial Standards Institute. 2004b. Pang Kanomjeen (Thai Rice Noodle
Starch) TCPS 499/2547. Thai Community Product Standard, Ministry of
Industry, Thailand.
Thai Ministry of Health. 2000. Standard for Community Drinking Water.
USDA National Nutrient Database for Standard Reference http://ndb.nal.
usda.gov/ndb/nutrients/index (accessed 11 March 2014).
Vongsawasdi, P., M. Noppharat, N. Hiranyaprateep and N. Tirapong. 2009.
Relationships between rheological properties quality of rice flour and ver-
micelli. ASEAN Journal of Food-Agro-Industry 2(2):102–109.
Wong, H.-H., D.-W. Sun, Q. Zehg and Y. Lu. 2000. Effect of pH, corn starch
and phosphates on the pasting properties of rice flour. Journal of Food
Engineering 46:133–138. Cited by Oupathumpanont et al. 2008.
Yeh, A.-I. 2004. Preparation and applications of rice flour. In Rice: Chemistry
and Technology, third ed., edited by E.T. Champagne, pp. 495–539. St.
Paul, Minnesota: American Association of Cereal Chemists.
Youngprasitiporn, N. 1989. Application of pregelatinized non-glutineous rice
(Oryza sativa) flour in kanomjeen. MSc thesis, Kasetsart University,
Thailand.
Zhou, Z., K. Robards, S. Heliwell and C. Blanchard. 2002. Review: Composition
and functional properties of rice. International Journal of Food Science and
7
L acti c Fermentati ons
o f F ish and F ishery
P roducts
L E O N A R DA S . M E N D O Z A
Contents
7.1 Introduction 258

7.2 Philippine Microbially Fermented Fish and Fishery Products 261
7.2.1 Resources 263
7.2.2.1 Materials for Microbial Fish Fermentation 265
7.2.2.2 Fermented Rice and Fish (Burong Isda) 268
7.2.2.3 Fermented Mudfish (Burong Dalag) 269
7.2.2.4 Fermented Milkfish (Burong Bangus) 271
7.2.2.5 Fermented Tilapia (Burong Tilapia) 273
7.2.2.6 Fermented Shrimp (Balao-Balao or
Burong Hipon) 275
7.2.2.7 Fermented Dried Mudfish (Tinapayan) 277
7.2.3 Physico-Chemical Changes during Fermentation 285
7.2.3.1 Changes in Balao-Balao and Burong Isda 285
7.2.3.2 Tapai Inoculum 285
7.2.3.3 Tapai a Umay 286
7.2.3.4 Tinapayan Fermentation 286
7.2.4.1 Sources of Microorganisms 289
7.2.4.2 Microbial Changes during Fermentation 290
7.2.4.3 Sources of Fermentable Sugars in
Rice-Fish Fermentations 294
7.2.6 Safety Concerns 297
7.2.6.1 Microbial Food-Borne Hazards 298
7.2.6.2 Non-Microbial Biological Hazards 300
257
7.2.6.3 Parasites 301

7.2.6.4 Control of Parasitic Infections 303
7.2.6.5 Chemical Hazards 304
7.2.6.6 Physical Hazards 304
7.2.8 Future Prospects and Recommendations 306
References 306
7.1 Introduction
In Southeast Asia, there are many microbially fermented fishery

products with a common feature, namely the proliferation of lactic
acid bacteria and/or moulds and yeasts in rice–fish/shrimp mixtures
(Table 7.1). Salt is added so that the raw fish stays stable until acid is
evolved, and some processors add sour fruits or vinegar to reduce the
pH of the mixture prior to fermentation. A few of the products are
inoculated with lactic acid bacteria or an association of moulds, yeasts
and lactic acid bacteria. In some cases, a little fully fermented rice–
fish from a previous batch is added to ensure the presence of desirable
lactic acid bacteria. However, many are not inoculated and depend
on a microbial flora from the air and utensils to initiate the fermenta-
tion. Angkak (red rice with Monascus purpureus) is sometimes added to
provide colour to the mixture. Since the techniques vary from country
to country and from one household processor to another, low-quality
products are often produced, compromising both product safety and
acceptability and, thereby, resulting in low economic returns.
Many species of fish are now available for microbial fermentation,
especially with the success of freshwater aquaculture in inland towns.
Since fermentation technology is a cheap method of preserving fish in
rural areas and fermented products can be a major source of protein for
inland communities, the technology is reviewed regularly to improve
the organoleptic acceptability and safety of the products. In addition,
marketing and distribution rules have recently been made stringent,
especially regarding the assurance of quality and safety. Thus, there is
a need to improve the prevailing technologies, focusing on the strong
points of the traditional techniques and identifying gaps in knowl-
edge. Microbially fermented fish and fishery products in Southeast
Table 7.1 Lactic Acid Bacterially Fermented Fishery Products of Southeast Asia
LOCAL NAMES OF PRODUCTS IN
COMMON
CHARACTERISTICS PHILIPPINES THAILAND VIETNAM MALAYSIA INDONESIA BORNEO/LAOS CAMBODIA
A. Fermented fish Burong isda e
Plaa-som ; j
Mam-ca (similar Ikan masim k Makas- Borneo Mam-chao (Phaak)g,h
Low or high salt. (dalag, hito, etc.)a Plaa-raa e; to burong isda) sar m Peka-sam k
Carbohydrate Pla-ra-kao-kuo f Beka-sam n
source: cooked rice (kao-rice; kuo-roasted) Wadi k
(hard or sticky) or or pla-ra-ram (ram-rice Picungan k
rice bran, or roasted bran) f; Koong-
rice flour; or minced chom e;
garlic or picungan Plaa-paeng daeng (red)e;
seeds Plaa-chom e
B. Fermented fish with Burong bangos Plaa-mum e (papaya); Mam-chao or Peka-sam Mam-ca-sat g
acidic fruits, vinegar (+ vinegar)b Khem-bak-nad e Mam-chauk (tamarind)k,l (green papaya;
or sugar to bring (pineapple and juice); pineapple + ginger)
down pH of mixture Pla-mam g (pineapple) Mam-ca-loc g
or enhance softening (+ pineapple)
of cured fish
C. Minced fish stuffed (Not commercial) Som-fak (+ minced
in casings or garlic)e; Koong-some;
L ac ti c F erm en tati o ns o f Fish
wrapped with leaves Som-fug f

D. Fermented internal Factory waste and Trash-fishn Mam-rout g
organs, factory trash fisheso (entrails);
waste or trash Mam-seing g
marine fish (fish eggs)
(continued)
259
Table 7.1 (Continued) Lactic Acid Bacterially Fermented Fishery Products of Southeast Asia
260
LOCAL NAMES OF PRODUCTS IN

COMMON
CHARACTERISTICS PHILIPPINES THAILAND VIETNAM MALAYSIA INDONESIA BORNEO/LAOS CAMBODIA
E. Cured or dried fish Tinapayan c Plaa-chao e; pla-jao (with
with pre-fermented look-paeng or
rice kao-mark)i
F. Fermented shrimps Burong hipon Mam-rouc i (water Cinca-luk l Cinca-luk k Laos: Padek g
(Balao-balao)d shrimp)
a Guevarra et al. (1978).
b Uyenco and Ajon (1982).
c Guerra (1992).
d Arroyo et al. (1978).
e Phitakpol (1989).
f Tanasupawat and Komagata (1995).
g Westenberg (1951).
h van Veen (1953).
i Adams et al. (1985).
j Sanchez (2008).
k Ruddle and Ishige (2010).
l Karim (1989).
m Borgstrom (1961).
n Putro (1989).
o Lopez (1989).
L ac ti c F erm en tati o ns o f Fish 2 61
Asia have been reviewed by Cooke et al. (1987), Lee (1989), Owens
and Mendoza (1985), Steinkraus (1996) and Wood and Hodge (1985)
7.2 Philippine Microbially Fermented Fish and Fishery Products
Three products are discussed here: burong isda, balao-balao and tinapa-
yan. The first two are similar in that the major fermenting organisms
are lactic acid bacteria, but they differ in that burong isda uses fish
while balao-balao is made from small shrimps. Both are produced
in the island of Luzon. Tinapayan is produced in Mindanao and is
the result of fungal and lactic acid bacterial fermentation. The tech-
nologies are traditional methods of preserving freshwater fish in rural
areas. The preservation of burong isda and balao-balao is attributed to
the combined effects of interrelated processing steps, including clean-
ing the fish, salting and the generation of lactic acid. Preservation
of tinapayan includes drying as an additional step. Cleaning reduces
the natural microbial flora and autolytic enzymes while salting and
drying preserve the fish prior to the generation of lactic acid. During
fermentation, anaerobic conditions and generation of acid and carbon
dioxide contribute to the preservation of the product.
The majority of the studies conducted on Philippine low-salt,
microbially fermented seafood products are on the production pro-
cess (Arroyo et al. 1978; Guevarra et al. 1978; Orillo and Pederson
1968), on understanding the role of the microorganisms (Mendoza
1985; Mendoza and Owens 1986; Solidum 1979; Vatana and Del
Rosario 1983) and on the use of an inoculum to hasten acid produc-
tion (Mabeza 1983; Sanchez 2008).
Many of the current products are simply acidic and although the acid-
ity is, to some extent, neutralised during cooking, the product’s accept-
ability cannot compete with fish sauce (patis) or fish paste (bagoong).
The microbially fermented products are produced in only a few prov-
inces, indicating a localised and, therefore, limited consumption.
Microbially fermented fishery products may be grouped on the basis
of microorganisms involved, into those fermented primarily by lactic
acid bacteria and those fermented by an association of fungi and lac-
tic acid bacteria. The lactic acid bacteria fermented products can be
further sub-divided into low-salt and moderately high-salt products
(Table 7.2). Minced milkfish flesh seasoned with spices and sugar and
Table 7.2 Categorisation of Microbially Fermented Fishery Products of the Philippines Based on the Major Microbial Groups Involved, Product Form and Salt Content
262
CHARACTERISTICS
SALT pH STARTERS/COLOURANT ADDITIONAL
CATEGORY (WT/WET WT) (INITIAL—AT 48 H) CARBOHYDRATES SPICES ADDED TREATMENTS
I. FERMENTATION MAINLY BY LACTIC ACID BACTERIA

A. Low salt: Fully fermented Sautéed and/or cooked
-Coloured red 3.5–6.0% 6.5–3.9 Cooked rice product, angkaka into different variants,
-Plain white 5.6–5.0 None bottled and heat
processed
B. Moderately high salt: Soaked in 10% 5.7–4.7
-Coloured red brine; Cooked rice Angkak; Sautéed or cooked as
-Plain white Cured in 10–20% Fully fermented product main dish
for 2 days or longer; Ginger and garlic may be added
Salt added to rice -none
C. Minced fish, stuffed in Salt, sugar 5.8–4.8 Flour, sugar, spices No starter added, food Fried
sausage casing colour
II. FERMENTATION INVOLVING MOULDS, LACTIC ACID BACTERIA AND YEASTS

A. Tinapayan 2.5% of rice– Rice: 4.4–5.8 Pre-fermented rice (tapay Inoculum (tapai) Fermented 1–2 weeks;
fermentation spices–fish mixture Fish: 6.8–6.1 a umay); white ginger, recently prepared sautéed in garlic until
garlic, lemon grass dry, packed in plastic
cups
Source: Modified from Owens, J.D. and L.S. Mendoza. 1985. Journal Food Technology 20:273–293.
a Dried, red-coloured rice due to growth of Monascus purpureus.
L ac ti c F erm en tati o ns o f Fish 263
stuffed in casing is a more recently developed variant but is not yet

sold on a commercial scale. The addition of colouring material, ang-
kak (red rice), is optional and depends on the processor and the fish
used. Mudfish (Ophicephalus striatus), catfish (Clarias macrocephalus),
milkfish (Chanos chanos) and tilapia (Tilapia mossambica) are sometimes
coloured to enhance the appearance but the uncoloured products are
equally common. Fermented shrimps are not coloured with angkak
because the shrimp turns red-orange as soon as acid is generated. All
microbially fermented fishery products in the Philippines are cooked
before consumption and shelf life extension is either by bottling and
heat processing or by cooking in oil until the moisture evaporates.
A great variety of fish and shrimp are fermented in the Philippines
(Tables 7.3 and 7.4). In this section, the traditional methods of micro-
bial fermentation of fishery products in the Philippines are reviewed,
with the aim of describing the processes, the functions of added ingre-
dients, the roles of the dominant microorganisms, the quality and
safety of the products and the role of microbially fermented fishery
products in enhancing the overall nutrition of consumers in rural areas.
7.2.1 Resources
Philippines is rich in natural resources. It is surrounded by water, with

a shoreline of 17,460 km, and is also noted for its inland waters, with
a total area of 750,000 hectares. Inland waters are major sources of
fish, especially during the rainy season. The location of freshwater
bodies largely determines the place where microbially fermented fish
are produced. Traditional microbial fermentations of freshwater fish
are practiced today around Laguna de Bay, Taal Lake, Lake Bato and
Candaba Swamps in Central Luzon. Rice, freshwater fish, salt and
spices abound in these areas and traditional microbial fermentations
provide a practical way of preserving excess catch and at the same
time modifies the flavour to create new foods.
The second largest island, Mindanao, has six lakes and major riv-
ers. Freshwater species abound in this mountainous region. During
the rainy season, the rivers and lakes overflow and adjacent fields are
inundated, expanding the nutrient-rich area available for freshwater
species. Mudfish, which hibernate during the dry season, come out
and feed and breed during the rainy season. Other species, including
Table 7.3 Local, English and Scientific Names of Fermented Fish Species
FISH SPECIES USED
FERMENTED
FISHa LOCAL PHILIPPINE NAME(S) b ENGLISH NAME SCIENTIFIC NAME
Burong (B) Bangus, Bangos (T, B, I, Pm, Milkfish Chanos chanos (Forskal)
bangos V), Awa, Banglot (I), Bangros
(B, V), Banglos (T, B, KT),
Buetil (Pn), Banglis (V)
B. dalag Dalag (T, I), Bundaki Striated murrel, Ophicephalus striatus
Bundalang, Bulig, Bukali striped snake- (Bloch)
(T), Lawag (KT), Haruan, head, mudfish
Haluan (V), Obog (V)
B. hito Hito (T, B), Ito (B, Pm), Pantat Freshwater catfish Clarias macrocephalus
(B, Pn), Alabiyog, Kawatsi (Gunther)
(K T), Alimudan (V)
B. kanduli Kanduli, Kanduling, Arahan Green sea catfish Arius thalassinus
(T), Kanduli (I) (Ruppell)
B. gurami Gurami Gourami Trichogaster pectoralis
(Regan)
B. tilapya Tilapya Tilapia Tilapia nilotica, Tilapia
mossambica
B. biya Biya (T), Bunog (I), Bakuli Goby Amblygobius phaena
(Cuvier and
Valenciennes)
Glossogobius guiris
B. carpa Karpa (T), Babangan (MST) Common carp Cyprinus carpio
Silver carp (Linnaeus)
Hypophthalmichthys
molitrix (Cuvier and
Valenciennes)
B. ayungin Ayungin (T, I) Silver perch Leiopotherapon plumbeus
Source: Adapted from Conlu, P. 1986. Guide to Philippine Flora and Fauna: Fishes. Vol. 1X. Diliman,
Quezon City, Philippines: Ministry of Natural Resources and the University of the Philippines.
a Product name is indicated by adding the local name to the term burong (fermented rice), isda
(Philipino ‘fish’).
b Area where term is used: T, Tagalog; I, Ilocano; B, Bicol; KT, Kayanin and Tagbanua; Pn, Pangasinan;
Pm, Pampanga; V, Visayan; MST, Maranaw, Samal and Tausog; CV, Cagayan Valley.
catfish, perch, tilapia and freshwater shrimps also multiply. The sea-
son of abundance coincides with difficult conditions for drying and
salting fish and availability of ice becomes scarce due to low demand
during the cool rainy season. Therefore, from necessity, a preservation
method by microbial fermentation became inevitable.
In the cold mountainous region of Luzon, a rice wine, called tapuy, is
produced. A similar drink, pangasi, is produced in Bukidnon, Mindanao
Table 7.4 Local, English and Scientific Names of Fresh and Brackish Water Shrimp Species
Used in the Processing of Fermented Shrimps
SHRIMP SPECIES
FERMENTED LOCAL PHILIPPINE ENGLISH NAME SCIENTIFIC NAME
SHRIMP PRODUCT NAME(S)
Balao-balao Tagunton Freshwater small Palaemon sp.
shrimp Penaeus indicus,
Macrobrachium spp.
Burong hipona Suahe (T)b Greasy back Metapenaeus ensis
Puti-an (T) White shrimp or Penaeus merguiensis
Ulang (T) banana shrimp Penaeus indicus
Freshwater shrimp Macrobrachium idella
Aramang (CV) Pink shrimp Metapenaeus sp.

Source: Adapted from Conlu, P. 1986. Guide to Philippine Flora and Fauna: Fishes. Vol. 1X. Diliman,
Quezon City, Philippines: Ministry of Natural Resources and the University of the Philippines.
a Burong (fermented), hipon (shrimp).
b Area where term is used: T, Tagalog; CV, Cagayan Valley.
to ward off cold. The inoculum used in the fermentation of pangasi, with
modification over time, is used in the fermentation of fish and even-
tually led to a different method of fish fermentation in the area, the
tinapayan. This type of fermentation involves the action of a select
group of moulds, yeasts and lactic acid bacteria (Guerra 1992). No
similar product is produced in Luzon.
The fish species caught in inland water bodies are normally bland
in flavour but rich in nutrients (Table 7.5). Practically all freshwater
species can be fermented but, unfortunately, low-salt fermentation
remains a household industry.
7.2.2.1 Materials for Microbial Fish Fermentation The materials required

for traditional fermentations are readily available in rural areas. The
microorganisms for acid fermentation are easily obtained from a selec-
tion of existing fermented agricultural products such as palm juices,
sugar cane, vinegar, vegetables, fruits, legumes, milk, etc. (Sanchez
2008), many of which are fermented in rural areas. A household wish-
ing to initiate a low-salt fish fermentation can buy a fermented product
and use it as an inoculum for fish fermentation. In time, an adapted
population of microorganisms for the fish fermentation is selected.
266
Table 7.5 Chemical Composition of Species Used in Rice-Fish/Shrimp Fermentations

COMPOSITION PER 100 G EDIBLE PORTION
SHRIMP,
MILKFISH MUDFISH (DALAG) TILAPIA CARP FRESHWATER SHRIMP, SMALL
(BANGOS) OPHICEPHALUS TILAPIA CYPRINUS (TAGUNTON) (SUAHE)
COMPONENT CHANOS CHANOS STRIATUS MOSSAMBICA CARPIO PALAEMON SP. METAPENAEUS ENSIS
Water, g 74 78 77 69 74 80
Protein, g 18 21 18 18 20 18
Fat, g 6.4 1.6 3.8 12.5 1.9 0.6
Ash, g 1.1 1.2 1.2 0.9 3.5 1.3
Calcium, mg 44 65 74 36 2350 110
Phosphorus, mg 195 200 185 145 380 175
Iron, mg 1.2 1.5 0.8 1.0 15 1.8
Retinol, ug 135 45 65 45 90 290
β-Carotene, μg 20 30 10 20 NDa 280
Total vitamin A, μg 140 50 67 48 ND 340
Niacin, mg 7.8 3.8 4.6 3.9 2.7 2.5
Riboflavin, mg 0.10 0.12 0.20 0.17 0.02 0.20
Thiamine, mg 0.02 0.09 0.06 0.01 0.02 0.03
Source: Philippines Food Composition Tables. 1997. Manila: Food and Nutrition Research Institute, Department of Science and Technology, Taguig, Metro Manila.
a Not detected by the assay used.

7.2.2.1.1 Species Fermented The raw materials for microbial fermen-

tation are mainly freshwater fish species although brackish water spe-
cies, such as milkfish, tilapia and shrimp, are also utilised. Production
of milkfish and tilapia in freshwater environments enhances the
quantities of raw material available during summer months. Apart
from the utilisation of freshwater and brackish water fishes, a vari-
ety of invertebrates and seaweed are also fermented. Many of these
invertebrate and seaweed products are consumed only in speciality
restaurants or are prepared for special occasions but, unfortunately,
their production has yet to be documented. In addition, marine trash
fishes and factory wastes can be preserved by microbial fermentation
(Lopez 1989).
7.2.2.1.2 Salt Salt, the initial preservative, is produced in coastal

towns in the Philippines. Solar salt production is common along beach
fronts and also in ponds that are used to raise fish during the rainy
season but which function as salt ponds in the summer. In north-
ern Luzon, salt is produced by boiling, thus destroying microbes that
could contaminate fermentation. Major considerations in the selec-
tion of salt for fermentation are (a) sodium chloride content ≥97.5%;
(b) low chemical impurities, such as calcium ≤0.6%, magnesium
≤0.1%, insoluble residue ≤0.5% (Zaitsev et al. 1969); (c) low microbial
contaminants, mould spores, halophilic bacteria or food-borne patho-
gens; (d) white in colour and free from filth and physical hazards; and
(e) dry and packed in acceptable plastic bags.
7.2.2.1.3 Carbohydrate Source The common carbohydrate source is

non-glutinous rice. A ‘hard rice variety’ is preferred, like IR8 or Wag-
wag, due to their higher amylose content (27%) compared to soft rice
varieties (Juliano et al. 1965). It is boiled or steamed to a soft porridge-
like consistency. Boiling in water gelatinises the amylose and amylopec-
tin, making them more available to the action of amylases. Glutinous
rice is sometimes used for fermentations intended for special occasions.
Other sources of starch, such as ground roasted rice, ground sweet potato,
corn starch and cassava starch can also support successful fermentations.
7.2.2.1.4 Spices and Flavour Enhancers Philippine foods are not as

spicy as those of their Southeast Asian neighbours. This is reflected
in the spices used in the fermentation of fishery products. Common

spices used are ginger, garlic, chilli, turmeric, lemon grass and white
ginger (Curcuma zedoaria Berg., Rose). These spices not only enhance
flavour and reduce fishy odour but are also reported to have inhibi-
tory effects on spoilage microorganisms. Acidic fruits, such as tama-
rind and pineapple, are added to lower the pH and make conditions
favourable for lactic acid bacteria, while papaya fruit, containing the
protease papain, is added to hasten the softening of the fish flesh.
This softening facilitates the absorption of acids, sugars, amino acids
and fatty acids and promotes the preservation of the meat and flavour
development. Angkak (red rice) may be used as a colouring material.
7.2.2.2 Fermented Rice and Fish (Burong Isda) Fermented fish and

fishery products with added carbohydrate and salt are called buro. The
products are named according to the raw materials used. To distin-
guish the species used, the local name of the fish suffixes the term
burong. Hence, burong isda is fermented fish and burong dalag is fer-
mented mudfish. The common names, English names and scientific
names of species used are shown in Table 7.3.
Burong isda is a mixture of salted fish, cooked non-glutinous rice,
angkak and a selection of minced spices, mostly ginger and garlic. The
fully fermented product in the Tagalog region is red in colour due to
angkak but in Central Luzon, angkak is omitted and the product is
white (Barile 1983). After fermentation, a strong sour odour is emit-
ted from the mixture, the rice is partially liquefied and the fish flesh
is soft and tender. The shelf life of low-salt fermented fish at ambient
temperatures is quite short. It is fully fermented after 5–8 days and
then becomes over-ripe, with too sour a taste and flesh that is too soft
and pasty. Fish enzymes soften the flesh, making absorption of acid
more efficient. Further, the connective tissues, which bind the muscle
fibres, disintegrate in acid (Zaitsev et al. 1969) making the flesh pasty.
A longer shelf life of 20–25 days was obtained in products with higher
salt contents than normally used (Mabeza 1983). The higher NaCl
concentration extracted water, kept the flesh firm and slowed down
the over-ripening. The fermented fish is sold in the market in plastic
bags or bottles. The product is always cooked before eating and the
usual preparation for the table is by sautéing in oil with spices and
serving as a condiment.
Recently, owing to the demand for buro by busy housewives and

to make it available in convenient outlets, the fermented product is
packed in bottles with various seasonings and heat processed at 121°C
for 20 min. The heat processing stops microbial and enzymatic activ-
ity and extends the shelf life to a year or more. Two processing plants
in Luzon produce heat-processed buro as an addition to the more
popular high-salt shrimp and fish pastes. These factories produce
their own buro in order to ensure the safety and quality of the heat
processed fermented products (Vocalan 2010, personal communica-
tion). Another practice, used by exporters, is to pack freshly prepared
rice–fish mixture in bottles and distribute them refrigerated at 0–5°C
until they reach their destinations. This enables them to market their
product as soon as the rice–fish mixture is prepared and delegates the
burden of fermentation to the consumer. When needed, the bottles
are removed from the refrigerator, the cap is loosened and the bottles
are incubated at room temperature to ferment.
Traditionally, only freshwater species from the wild were microbi-
ally fermented but, owing to the decline of fish catch in rivers and
swamps during the dry season, aquacultured species such as carp,
catfish, tilapia and milkfish are now fermented (Guevarra et al. 1978;
Sanchez 2008).
7.2.2.3 Fermented Mudfish (Burong Dalag) The most popular species

for microbial fermentation is the lean and fleshy mudfish, Ophicephalus
striatus. The fermentation of mudfish (burong dalag) is still governed
by the traditional practices of individual households and there are many
variations in the method of cooking rice, porridge-like or boiled until
dry; concentration of salt added; method and length of salting; and
spices and fruits added. Traditional processors sometimes add a small
amount of fully fermented fish as inoculum or simply depend on fer-
mentative organisms in the air, utensils and jars to initiate fermentation
(Barile 1983). Current fermentation procedures are still similar to those
described in the 1950s by Orillo and Pederson (1968; Figure 7.1). The
preparation for fermentation starts by killing the fish with a blow on
the head and the slime is removed by rubbing the skin with ashes or
salt. This is followed by descaling, eviscerating and cutting the fish into
steaks. The steaks are washed, drained and then salted with coarse salt,
4% of the wet weight of the steaks, which is allowed to permeate for
Mudfish (dalag, Ophicephalus striatus)
Descale, fillet, wash and cut crosswise into steaks

(waste is fermented separately for animal feed,
head is cooked into a favourite dish)
Salt, 4.0% wt of steaks

Mix thoroughly and set aside for 24 h
Remove undissolved salt and drain off brine

Rice + water (1:2 v/v) (final NaCl content ~3.4%)
Boil until soft, cool
Add angkak,
(2% wt of cooked rice)
Mix thoroughly
Mix rice + angkak with salted fish
(rice:fish :: 2:1 w/w)
Pack mixture tightly in glass jars

Ferment 7–10 days at 27–32ºC
Red-fermented mudfish (burong dalag)
Sauté with onions and spices. Serve as condiment

or as main dish with vegetables
Figure 7.1 Fermentation of mudfish. (Modified from Orillo, C.A. and C.S. Pederson. 1968. Applied
Microbiology 16(11):1669–1671.)
24 h. The following day, the excess salt is removed and the brine that
formed is drained off. Meanwhile, rice is washed, drained and water
added (rice:water about 1:2 [v/v], depending on the preferences of the
processor) and the rice is boiled until the water is absorbed or evapo-
rated. The cooked rice is cooled by spreading it in a winnowing basket
and ground angkak, 2% (w/w) of the cooked rice, is sprinkled on it. The
salted fish steaks are coated with rice mixture, two parts rice to one part
fish, and then packed tightly in dry glass jars. They are left undisturbed
in a dry place at 28–30°C for 7–10 days. The preservation of the highly
perishable fish before the generation of acid is attributed to the added
salt. The time required to generate acid from the unsalted rice is around
48 h, long enough for the fish to spoil if not properly salted.
Burong dalag is always cooked before consumption. Cooking
includes adjustment of the acidity by adding sugar to mask the strong
sour taste. Saltiness is improved by adding more salt or fish sauce.

Cooking expels the CO2 in the mixture, leaving a cheese-like flavour.
Burong dalag is generally consumed along with vegetables and spices
in a family meal consisting mainly of rice.
Variations in the preparation of burong dalag include filleting the
fish and cutting it into strips to increase the surface area available
for the diffusion of salt, increasing the amount of salt to 15% (w/wet
weight of fish) and shortening the salting time to 3–6 h (Sanchez
2008). Higher salt concentrations extract water faster and, thereby,
reduce the rate of spoilage of the fish.
7.2.2.3.1 Utilisation of Dead Wet Market Fish A technique of fer-

mentation for enhancing the sale value of mudfish that died while
awaiting sale in the wet market is practiced by vendors. This pro-
cedure includes cleaning the fish, cutting it into steaks, followed by
mixing one part salt to three parts fish steak. The salt extracts mois-
ture from the fish to form brine. The steaks are soaked in the brine
for 24 h, drained and then coated with cooked rice and angkak. The
mixture is stacked in bowls and displayed for sale. The rice turns sour
and soft and the shelf life of the fish is extended to 3–7 days.
7.2.2.4 Fermented Milkfish (Burong Bangus) The success of aquacul-

ture in freshwater fishponds has enhanced the volume and variety
of raw material available for fermentation, especially during the dry
season. Most commonly used are milkfish, tilapia and big head carp.
Milkfish can be acclimatised and grown in a wide range of water
salinities and is cultured in fish pens, lakes, rivers and reservoirs.
Representative fermentation methods for aquaculture species are
described by Guevarra et al. (1978), Mabeza (1983) and Sanchez
(2008).
Milkfish is becoming increasingly popular as a raw material for
processing, including microbial fermentation (Figure 7.2). The fish
is descaled, filleted and washed. Salt, 18% of the wet weight of the
fillet, is added and allowed to permeate for 2 h. The brine that is
formed is drained off, resulting in a final salt content of about 3%.
This concentration is sufficient to inhibit growth of some spoilage
bacteria but does not totally prevent growth of other spoilage organ-
isms (ICMSF 1980). Since the diffusion of low salt concentrations
Milkfish (bangos, Chanos chanos)
Descale, split, remove internal organs, fins, gills and false kidney
Pin bones are sometimes removed
Wash well
(400 g cleaned fillets per kg fish)
Salt, 18% wt of fillets

Stand for 2 h
Rice + water (1:1.8 w/w)
Boil until cooked
Add angkak,
(1.4% wt of cooked rice)
Mix thoroughly
Coat salted fillets with rice-angkak mixture

Pack coated fillets into jar with additional rice-angkak mixture
(put alternate layers of fish and rice to fill jar)
(Overall rice:fish :: 6:4 w/w)
Cover tightly
Ferment
(7–10 days at room temperature, 27–30°C)
Fermented milkfish
Sauté with garlic, onions, coconut milk and slices of eggplant

Cook until sauce is thick and vegetable is cooked
Season with pepper and fish sauce
Serve as main dish
Figure 7.2 Fermentation of milkfish. (Modified from Guevarra, G. et al. 1978. Milkfish (Bangos)
as Food: Handling, Freezing and Processing of Milkfish (Chanos chanos, Forskal). Manila: National
Science Development Board; Recipe from Vocalan, A. 2010. Personal interview. Balaw-balaw foods,
Angono, Rizal, Philippines.)
into the flesh is slow (Zaitsev et al. 1969), spoilage can occur during
the salting process. Fish is a highly perishable commodity, making
preservation a race between the rate of salt permeation and the action
of microorganisms and enzymes (Borgstrom 1961). Enzymatic activ-
ity is enhanced by low pH, leading to a rapid softening of the flesh
if the salt concentration is not sufficient to retard autolytic activity in
the flesh. Normally, it takes 2.5–3 days to bring down the pH of the
rice–fish mixture to 4.5 and below (Mendoza 1985). The first 3 days
are, therefore, critical if the salt content is low. Spoilage due to enzy-
matic activities and bacterial growth can also result in the produc-
tion of histamine, the proliferation of food-borne pathogens, and can
cause a reduction of eating quality. A prevailing spoilage odour can
cause the outright rejection of the product.
Some processors have modified the process by marinating the
salted fish in vinegar overnight before adding rice (Uyenco and Ajon
1982). The marinating process further lowers the moisture content of
the fish and the lower pH creates conditions favourable for the growth
of lactic acid bacteria.
7.2.2.5 Fermented Tilapia (Burong Tilapia) Tilapia is prepared for fer-

mentation in a slightly different way from that of other fish (Sanchez
2008). Tilapia is a slimy fish and is cultured using animal or chicken
manure as a pond fertiliser. The fish sometimes eat mud and it is nec-
essary to purge the fish by impounding them in a net and withholding
food for several days. This is intended to remove the intestinal con-
tents prior to harvest. Although the fish are purged before harvesting,
they are still liable to pick up food-borne pathogens from the mud and
water during harvesting and handling and, therefore, careful cleaning
is important before fermentation. Intestinal parasites are sometimes
present in the fish.
The tilapia are descaled and eviscerated (Figure 7.3). Remaining
slime is removed by rubbing the fish with salt, washing and draining.
The cleaned fish is filleted and the flesh cut into ~20 g strips. The fil-
lets tend to be of uneven thickness but the thickest part of the fillet
should not be more than 2 cm. The strips are soaked in 10% brine
for 30 min to extract water and blood, making the fillets firm and
white. Again, the fillets are washed well and drained. The presence of
parasites in the flesh can now be easily detected in the white and firm
meat. The second stage is salting. Brined fish strips are mixed with
4.5% salt. The salt is allowed to permeate the fish strips for 6–12 h
at 28–30°C and then the brine, along with extracted materials, is
drained off. Meantime, rice is boiled, cooled, blended with 1% salt
and the mixture is mixed with the salted fish, taking care to ensure
that every fillet is well coated with the rice–salt mixture. The mixture
is packed into clean, dry glass jars by layering rice and fish alternately
Live or fresh tilapia (Tilapia nilotica, T. mosambica)
Kill fish
Wash to remove dirt, slime and remnants of animal waste
Descale, remove internal organs, head, fins, gills, false kidney

Wash, rub with fine salt to remove slime
Split fish and fillet
Wash, drain and cut into strips ~20 g per piece.
(~300 g fish strips per kg fish)
Wash fish strips with 10% NaCl
Soak for 30 min to leach out blood (brine:fish :: 1:1 v/w)
Discard used brine, wash fish with tap water, drain

Salt, 4.5% wt of steaks
Mix thoroughly
Set aside for 6–12 h for salt to diffuse inside
Rice + water (1:2 v/v)
Boil until cooked, cool Drain
Add salt
Mix salted rice with salted fish
(rice:fish :: 2:1 w/w)
Mix thoroughly
Pack rice-coated fish strips into jar with additional rice

(alternate layers of fish and rice to fill jar, press to exclude air
after each addition, finish with layer of rice, sprinkle surface with 1% salt, cover)
(Overall rice:fish : ~6:4 w/w)
Ferment, 7 d, 28–30°C
Fermented tilapia
Cook as a main dish with coconut milk or sauté with spices and use as a condiment
Figure 7.3 Fermentation of tilapia. (Modified from Sanchez, P.C. 2008. Philippine Fermented
Foods: Principles and Technology. Quezon City: The University of the Philippines Press.)
and pressing it down with a spoon to remove air. Finally, 1% salt is

sprinkled on the surface of the top layer. The jar is covered and incu-
bated at 28–30°C in a dry, well-ventilated place for 7 days. The final
salt concentration in the mixture is about 6.5%.
The fermented tilapia is sautéed with spices and used as a condi-
ment for vegetables and meat or is cooked with coconut milk and
served as a main dish. The possibility that tilapia might be contami-

nated with hazardous microorganisms and/or parasites requires strict
compliance with good aquacultural and manufacturing practices.
7.2.2.6 Fermented Shrimp (Balao-Balao or Burong Hipon) In general,

fermented shrimps are referred to as burong hipon irrespective of
the species used. However, in certain provinces, the processors give
special names to distinguish the kind of shrimp used and to give an
identity to the product, such as balao-balao. Balao-balao originates
from Central Luzon and is prepared from small freshwater shrimps
(tagunton; Macrobrachium spp.). Another commonly fermented small
shrimp is suahe (Metapenaeus ensis) but the fermented product is sim-
ply called burong hipon. At one time or another nearly all species of
freshwater shrimps are fermented by rural women in order to stretch
the availability of food.
Tagunton shrimp, used for balao-balao fermentation, is 3–4 cm
in length with a hard shell and is caught in Laguna de Bay and in
rivers with aquaculture fish pens. It is a bottom dweller living as a
scavenger, feeding on detritus and decaying feed. Owing to the very
small meat yield, it commands a low price in the fresh fish market and
sometimes it is just fed to ducks. The value of the shrimp is improved
by fermentation. The shells become soft due to the acid, making the
whole shrimp edible. Other shrimp species are also fermented but
since large shrimps are generally expensive, only the small species are
commonly fermented.
There is no standard method for balao-balao fermentation. The
processes described by Arroyo et al. (1978), Solidum (1979), Uyenco
and Ajon (1982) and Vatana (1982) are still practiced today depend-
ing on the preferences of the household. Kitchen processors use their
experience to apportion ingredients, determine the time for the fer-
mentation and the storage conditions. Sanitary practices are wanting
and, except for the heat-processed products, household fermentations
are, in general, not inspected by regulatory bodies.
Live shrimps are used for the production of good-quality balao-
balao (Figure 7.4). Extensive cleaning and washing of the raw mate-
rial are necessary due to the scavenging habits of shrimps. The shrimps
are washed well to remove debris, mud, grass, snails and fish and the
antennae, rostrum and tail are cut off with scissors. Salt (~20% the
Live or very fresh shrimp
Wash, remove antennae and tail

Wash with fresh water
Salt, 20% wt of shrimp

Mix thoroughly
Set aside for 2 h for salt to permeate
Rice + water Remove excess salt, drain liquid
Boil until cooked
Add salt
Mix thoroughly
Mix salted rice with salted shrimps
(rice:shrimp :: 4.8:1 w/w)
Pack tightly in jars, cover tightly
Ferment
(7–10 days, 27–30°C)
Balao-balao
Sauté with spices and serve as condiment

or pack in bottles and preserve by heat treatment
Figure 7.4 Fermentation of shrimps. (Adapted from Arroyo, P.T. et al. 1978. Studies on rice-
shrimp fermentation: Balao-balao. Philippine Journal of Food Science and Technology 2:106–125 and
Vocalan, A. 2010. Personal interview. Balaw-balaw foods, Angono, Rizal, Philippines.)
weight of shrimp) is added to the shrimps and allowed to diffuse for

2 h, after which the liquid and excess salt are drained off. Cooked
rice is sprinkled with salt (3% w/wet weight) and the salted shrimps
are coated with the rice–salt mixture. The rice–salt–shrimp mixture
is tightly packed into clean dry glass jars, using a spoon to press the
mixture down and eliminate air spaces. The pressing is repeated after
12 h incubation. The mixture is fermented at ambient temperature
(28–30°C) for 7–10 days. As soon as acid is produced and the pH
falls to 4.7 in about 48 h, the shrimps turn red-orange and the pig-
ment diffuses into the rice, turning the fully fermented mixture pink,
which becomes more intense after cooking. Balao-balao has a shelf
life of 20–25 days at ambient temperatures. Balao-balao is prepared
for the table by sautéing the mixture in oil and spices and is served as
a condiment to flavour bland dishes.
Mabeza (1983) suggested that salt is added to the rice to ensure
stabilisation of fermentation. Pre-salting the rice does ensure that the
entire mixture is more or less uniformly salted and inhibits potential
spoilage bacteria in the rice without destroying any lactic acid bacteria
present. Failure to suppress spoilage bacteria can result in shorter shelf
life of the fermented product.
The acid causes the colour of the shrimp, tagunton, to change from
whitish-grey to red-orange. The colour change in the shrimp is the
result of the breakdown of protein–carotenoid bonds, releasing the
red pigment. As the fermentation progresses, the acidic and ester
aroma of the mixture increases and the evolution of gas subsides. The
rice disintegrates, turning cream in colour and the shell of the shrimp
softens, making the whole shrimp edible, and the colour of the shrimp
diffuses into the rice. Sautéing homogenises the rice–shrimp mixture,
turning it into a red-orange slurry. Apart from the production of acids
and esters, flavour development is attributed to the hydrolysis of pro-
teins into peptides and amino acids (Vatana 1982; Vatana and Del
Rosario 1983; Sanchez 2008).
7.2.2.7 Fermented Dried Mudfish (Tinapayan) Guerra (1992) des

cribed the tinapayan fermentation as practiced in the mountain-
ous regions of Mindanao, particularly Maguindanao Province. This
type of fermentation is not practiced in Luzon where other lactic
acid bacteria fermented fish products are produced. For a long time,
many of the municipalities in Mindanao remained quite isolated
and this could be one reason why the technology is still confined to
the Mindanao area.
Mudfish, the raw material in this fermentation, is a carnivorous
species found in rivers, lakes, swamps, rice fields, ponds, fish ponds
and slightly brackish water. The fish hibernates when the ponds dry
up and emerges again during the rainy season. It feeds on small ani-
mals, such as frogs, snails, insects and even its own fry. It reaches full
maturity in 2 years and grows to a length of 30–90 cm (Conlu 1986),
making the fillets appropriate material for fermentation. The fish can
acquire a muddy odour when caught in highly polluted water but this
objectionable odour is removed by impounding the fish in clean free-

flowing water, normally for 2 weeks.
There are seven major lakes in mountainous Mindanao, supporting
an abundance of freshwater fish. The lakes are inundated during the
rainy season, extending the area of fish production to rice fields and
ponds and spreading the supply of mudfish. In addition, big rivers and
tributaries flow to lowland swampy areas, such as Liguasan Swamp in
Cotabato, carrying juvenile mudfish to marshes with abundant food.
Mudfish can grow in lakes 1000 m above sea level and in foul water by
virtue of their labyrinth, an accessory breathing organ in addition to
the gills (Conlu 1986). They can survive the dry season buried in moist
mud and then re-emerge during the rainy season and rapidly increase in
number. During the dry summer season, mudfish is cultured in fresh-
water fish ponds, cages or fish pens, ensuring a continuous supply of live
mudfish. The live fishes are placed in cans with water and transported
to the market. Approximately 10 medium-sized fishes are placed in a 5
gallon kerosene or oil can, taking care to avoid suffocation, scratching
or wounding of the fish. Keeping the fish in cans for longer than 3 days
is avoided because the fish may die or suffer loss in quality. Undesirable
microbial proliferation on the skin can also occur.
Tinapayan processing aims to preserve the excess mudfish catch
and impart a distinct meaty flavour to the fish. The price of the result-
ing product is increased to at least 3 times the live weight cost (Guerra
1992). Tinapayan is distinct from burong isda because the fish is
dried before subjecting it to a fermentation process brought about by
an association of lactic acid bacteria, yeasts and moulds. The dried
fish fillet is mixed with prefermented rice and an assortment of local
spices. The finished product is described by Guerra (1992) as having
a moderately cheesy, soya sauce like, spicy, sour and slightly sweet
aroma. The shelf life of tinapayan is at least 1 month, after which it is
liable to become rancid and/or mouldy.
Tinapayan is marketed after frying in oil with additional spices or,
in some cases, it is sold uncooked. The fried type is dry and is packed
in polyethylene bags or plastic cups with a lid. It is served at home,
brought to the fields to be consumed as viand, or carried as a read-
ily available food (baon) during travel. Cooking stops microbial and
enzymatic activities, thus making the fermented product stable dur-
ing distribution.
Current concerns with regard to product safety, necessitates the

need to understand the stages of tinapayan processing in order to
avoid or minimise potential hazards.
7.2.2.7.1 Drying of Mudfish As soon as the fish are caught, they

are immediately processed for drying because the flavour is rapidly
impaired once the fish die. Hence, they are either marketed alive or
are preserved by drying before being transported to the lowlands for
fermentation. The live mudfish is killed by inflicting a blow on the
head. Smaller fishes are held with bare hands and given a twist to
kill the fish. Mudfish is a very slimy species and the slime is removed
by rubbing the fish with salt or wood ash. The fish is descaled, split
into butterfly fillets, and internal organs, gills, fins and false kidneys
are removed. A thorough washing with clean water completes the
removal of blood and slime. The cleaned fish is soaked in 20% (w/w)
brine for 3 h. This preliminary brining leaches out blood, makes the
colour of the meat lighter and the flesh firm but, unfortunately, also
removes soluble proteins. Salt penetrates the flesh and inhibits growth
of potential spoilage organisms (ICMSF 1980; Table 7.6). The salted
fish is dried in the sun to a water content of 30–40% (w/w) (Guerra
1992), packed in boxes and stored until brought to the market. The
dried fish are sold to middlemen who trade the dried mudfish fillets
to various consumers in urban centres.
7.2.2.7.2 Preparation of New Inoculum The success of tinapayan fer-

mentation is determined by the quality of the inoculum. So, proces-
sors start by personally producing a new batch of inoculum, locally
called tapai. The processors can predict the quality of the tinapayan
by observing the microbial growth and odour of the inoculum. If
unusual microbial growth and/or odour are observed, the inoculum
is destroyed to avoid producing a low-quality tinapayan. Destroying
the contaminated inoculum also limits the spread of spores of con-
taminants. Great care is taken over cleanliness and the conditions of
production, including the selection of preparation room, equipment
and ingredients. It should be appreciated that a successful processor
does not have sophisticated instruments and relies on her experience
and senses. Normally, methods for the production of good inoculum
are guarded as a ‘top secret’ and the guidance of an ‘expert’ is sought
Table 7.6 Minimum Levels of Water Activity, NaCl and pH Permitting Growth of Selected
Microorganisms
NACL
ORGANISM WATER ACTIVITY (% W/W) pH
Pseudomonas fluorescens/aeruginosa 0.97 5.0 5.6
Clostridium botulinum Type E 0.97 5.0 5.0a
Clostridium botulinum Type A 0.95 8.0 4.8a
Clostridium perfringens 0.95 8.0 5.0a
Clostridium botulinum Type B 0.94 9.4 4.8a
Escherichia coli 0.95 8.0 4.4
Bacillus cereus 0.95 8.0 4.4–5.0
Salmonella spp. 0.95 8.0 4.1–5.5a
Lactobacillus viridens 0.95 8.0 3.8–4.4
Lactobacillus plantarum 0.94 9.0 3.8–4.4
Pediococcus cerevisiae 0.94 9.0
Enterobacter aerogenes 0.94 9.0
Vibrio parahaemolyticus 0.94 9.0 4.8a
Rhizopus nigricans 0.93 11.0
Mucor plumbeus 0.93 11.0
Saccharomyces cerevisiae 0.90 14.2 2.35
Staphylococcus aureus 0.86 18.2 4.3–4.7a
Paecilomyces variotii 0.84 20.0
Aspergillus fumigatus 0.82 21.6
Aspergillus flavus 0.78
Halobacterium halobium 0.75 27.0
Source: Adapted from ICMSF. 1980. Microbial Ecology of Foods, Vol. 1. Factors Affecting the Life and
Death of Microorganisms. London: Academic Press.
a Genigeorgis and Reimann (1979).
when a processor is new in the trade. The expert explains what hap-
pens during fermentation and how the success of the fermentation can
be predicted by evaluating the odour and colour of the inoculum. The
use of newly prepared tapai ensures that contamination of the fermen-
tation with undesired microbes is reduced to a minimum.
Inoculum production starts by selecting the best available tapai in
the market or by purchasing samples from a respected processor. The
materials for the production of new inoculum are non-glutinous rice
dough, newly harvested ripe red hot pepper (Capsicum frutescens) and
tapai (Figure 7.5). Non-glutinous rice is washed several times with
clean water until the water runs clear. The rice is drained and com-
bined with 2% (wet wt/wet wt) fully ripe red hot peppers without
Non-glutinous rice
Wash 3 times in water

Drain, weigh
Fresh, red chilli pepper,
ripened on the plant.
Remove stalks, wash, drain
2% wt of rice dough
Combine and pound
Sieve through fine-meshed cloth,

pound remaining big particles
Water
Mix to a dough consistency
Shape into balls, place in a winnowing basket

lined with banana leaves and flatten balls
Tapai starter, pulverize
0.01% wt of dough
Sprinkle over the flattened balls
Cover with muslin cloth or turong dome
Incubate, room temperature (28–30°C), 2 days
Remove cover, allow to dry overnight

Sun–dry 3–5 days
Pack in plastic bags or jars

or tie in bundles with string
Tapai
Figure 7.5 Fermentation of tapai for preparation of tinapayan. (Modified from Guerra, M.R.
1992. Studies on tinapayan—An indigenous fish ferment in central Mindanao. MS thesis, College of
Fisheries and Ocean Sciences, University of the Philippines in the Visayas, Philippines.)
stalks. The peppers must be ripened on the plant before picking and
used immediately to ensure that only the natural microbial flora is
present and that contaminants acquired during storage are avoided as
far as possible. The rice–pepper mixture is pounded until very fine.
Coarse particles are separated by passing the mixture through a sieve
and pounded further until the particles are fully disintegrated. The
fine powder is made into a dough by adding 50% water to obtain a
paste thick enough to roll into ~20 g balls. The balls are arranged
in a clean, dry winnowing basket lined with clean banana leaves.
They are flattened to around 1 cm thick, appearing like small rice
cakes. Around 0.01% (dry wt/wet wt) pulverised commercial tapai is
sprinkled over the flattened cakes and the cakes are covered with a tra-
ditional wicker dome, called a turong, which fits the diameter of the
winnowing basket. The turong is covered with dried palm (anahaw)
leaves. This enclosure maintains a high humidity environment that is
favourable to the growth of moulds. The cakes are allowed to ferment
for 2–3 d at 37–42°C until they are covered with thick white myce-
lia. A good tapai, incubated for 48 h, smells sweet, slightly alcoholic
and fruity. At this stage, any tapai that shows big patches of black
or green are removed and destroyed, as these are contaminated. The
turong cover is then removed and the new tapai cakes are air dried for
10–12 h in the incubation room and then further dried under the sun
for 3–5 days, with occasional turning to obtain a uniform moisture
content. The dried inocula are small white cakes with a slightly convex
top, possibly caused by the leavening property of yeasts. The microbial
flora of tapai is primarily yeasts and moulds, with only small numbers
of bacteria (Table 7.7). The storage life of dried tapai is 1 month or
more provided they are kept dry.
7.2.2.7.3 Pre-Fermentation of Rice Pre-fermented rice, locally called

tapai a umay, is prepared using the newly produced inoculum. Fermen
tation of the rice releases sugars, lowers the pH and releases inhibi-
tory substances that slow down the growth of pathogenic and spoilage
Table 7.7 Microbial Populations of Market and Laboratory-Prepared Tapai, and of Red Hot Chilli
Peppers
MICROBIAL CONCENTRATION (LOG CFU [G WET WT]−1)
DRY MARKET LABORATORY RIPE, RED HOT CHILLI
MICROBIAL GROUP TAPAIa PREPARED TAPAIb PEPPERS
INITIALC FINALD
Moulds 9.7 8.6 9.5 6.6
Yeasts 8.6 6.5 8.5 2.5
Lactic acid bacteria 3.4 3.3 3.4 6.9
Aerobic bacteria 1.6 6.5 1.1 6.8
Source: Adapted from Guerra, M.R. 1992. Studies on tinapayan—An indigenous fish ferment in
central Mindanao. MS thesis, College of Fisheries and Ocean Sciences, University of the
Philippines in the Visayas, Philippines.
a Moisture content ~10%.
b Inoculated with market tapai and ripe fresh fruit of red hot chilli peppers.
c Wet mixture at 0 h.
d Dry tapai after 168 h incubation (moisture ~10%).
Non-glutinous rice
Wash 3 times in water, drain

Water
(water:rice :: 1.5:1)
Cook until soft and water is fully absorbed
Spread on winnowing basket to cool to ~40°C

Tapai starter, pulverize
Sprinkle over rice, mix thoroughly
Wrap 100–200 g portions in fresh alam leaves

Place packets in a turong lined with cotton cloth and alam leaves
Cover with alam leaves and cotton cloth
Incubate 14–18 h, 28–30°C
Fermented rice (tapai a umay)
Figure 7.6 Fermentation of tapai a umay for preparation of tinapayan. (Modified from Guerra,
M.R. 1992. Studies on tinapayan—An indigenous fish ferment in central Mindanao. MS thesis,
College of Fisheries and Ocean Sciences, University of the Philippines in the Visayas, Philippines.)
organisms (Figure 7.6). Rice is washed several times until the water
runs clear and is the boiled (rice to water, 1:1.5 wt/wt) until the rice
is soft and cooked. The cooked rice is spread on a clean dry win-
nowing basket to cool to 35–40 °C. The newly prepared inoculum,
1.5% the weight of the rice, is sprinkled over and mixed with the
cooked rice. Approximately 100–200 g quantities of rice–inoculum
mix are wrapped in leaves of alam (Melanolepsis multiglandulosa)
and incubated in a dry, warm place for 14–18 h to ferment. Guerra
(1992) monitored the internal temperature with a thermocouple
probe and noted that the packets wrapped in alam leaves maintained
a relatively uniform temperature of 47–48°C. Packets wrapped in
takip-asin (Macaranga grandifolia) or baguiang (taro) leaves, which
are also used, had lower internal temperatures of 44–45°C. It was
suggested that a more acceptable odour and flavour were obtained
at the higher incubation temperature. Whether or not the elevated
temperature selects a group of fermentative organisms was not stud-
ied and requires elucidation. During fermentation, the rice becomes
soft and watery and a sweet fruity and alcoholic odour is emitted.
7.2.2.7.4 Tinapayan Fermentation The dried fish is cut into strips,

washed, drained and weighed (Figure 7.7). An equivalent weight of
pre-fermented rice is removed from the leaf packets and combined
Spices Fermented rice (tapai a umay) Dried, salted mudfish

(% of fermented rice)
White ginger, thinly sliced, 18% Break lumps by hand
Garlic, peeled and crushed, 0.3%
Lemon grass, washed and crushed, 1.5% Wash fillets
Mix Soak in water till soft
NaCl
(2.5% wt of rice–fish mix)
Water:
(1.5% wt of rice–fish mix)
(rice–spices:fish :: 1:1)
Mix fish, rice-spices and salt
Put portion of rice-spice mixture in bottom of container (~10 cm deep)

Put alternate layers of fish and rice to fill jar
Cover surface with rice-spice mixture
(overall rice-spices:fish :: 1:1 w/w)
Ferment
(1–2 weeks, 28–30°C)
Fermented dried mudfish (tinapayan)
Flake fillets, sauté with garlic until nearly dry
Pack fried tinapayan in plastic cups
Figure 7.7 Fermentation of dried mudfish tinapayan. (Modified from Guerra, M.R. 1992. Studies
on tinapayan—An indigenous fish ferment in central Mindanao. MS thesis, College of Fisheries and
Ocean Sciences, University of the Philippines in the Visayas, Philippines.)
with finely diced spices (% in relation to rice): langkuas (white ginger,

Curcuma zedoaria), 18; garlic (Allium sativum), 0.3; and salay (lemon
grass, Cymbopogon citratus) cut 10–12 cm from the roots and crushed
to release the odour, 1.5. The spices-fermented rice mix is blended
thoroughly. Salt, 2.5% (wt/wet wt), and water, 1.5% of the total rice–
fish mixture, are added. At this stage, the mixture is quite lumpy. The
mixture is blended thoroughly to the consistency of a smooth batter,
such that individual rice grains are no longer visible. Batter is poured
into 2-L wide-mouthed jars to a depth of approximately 10 cm. Fillets
are dipped individually into the remaining batter and arranged inside
the glass jars. Fillets and batter are packed alternately into the jars
until the jars are full to the neck, leaving an air space above the mix-
ture and the closure. The jars are covered loosely, to allow the release
of gas that forms during fermentation, and incubated for 1–2 weeks at
room temperature (28–30°C). The dry fillets absorb the liquid of the
fermented rice and subsequently become soft and flaky. The odour of
the fermenting mixture is slightly acidic, sweetish and fruity.
Tinapayan is sold in the market in at least three forms. The first
is as a mixture of fish and batter in plastic bags. The second is the
fermented batter alone, after separating the fish flakes, which is used
as an additive to soups and as a flavouring ingredient during cooking.
The third is as sautéed dry flakes. The fermented fillets are flaked, by
breaking them into pieces by hand and sautéed with garlic until red-
dish brown in colour and almost all the water is evaporated, leaving
only the oil (Guerra 1992). The last form is the most in demand, not
only due to its flavour, but also because it has a longer shelf life than
the other forms. The cooking stops microbial and enzymatic activities
and the shelf life is determined by the oxidation of fat and growth of
moulds.
7.2.3 Physico-Chemical Changes during Fermentation
7.2.3.1 Changes in Balao-Balao and Burong Isda In balao-balao, the

initial physical changes are the formation of liquid in the mixture
accompanied by gas (Solidum 1979). The liquid facilitates the diffu-
sion of substances within the mixture and the gas displaces oxygen in
the mixture, making it anaerobic. The pH dropped to 3.4 and 3.8 in
mixtures with 3% and 6% NaCl, respectively, and total acidity (as lac-
tic acid) reached 2.0% after 7–9 day incubation at 28–30°C (Arroyo
et al. 1978).
The changes during burong isda fermentation are more or less sim-
ilar to those in balao-balao. Sensory evaluation of burong isda after
sautéing ranged from a pleasant, sour and cheese-like flavour to an
objectionable, putrid odour (Barile 1983). The strong acidic taste is
not favoured by some consumers.
7.2.3.2 Tapai Inoculum The major changes observed during tapai fer-
mentation were a substantial increase in the concentration of reducing
sugars, presumably due to hydrolysis of rice starch, and in the ethanol
content (Table 7.8). The high ethanol concentration after 48 h incuba-
tion may reflect the very high yeast concentrations present in tapai
Table 7.8 Chemical Changes in Fermenting Tapai Incubated at 28–30°C

CHANGES DURING FERMENTATION
FERMENTATION TITRATABLE ACIDITY ETHANOL REDUCING SUGARS MOISTURE
TIME (H) pH (%, AS LACTIC ACID) (G KG−1 WET WT) (G KG−1 WET WT) (%)
0 6.7 0.09 n.d.a 2.6 54
12 4.4 0.23 n.d. 5.1 52
24 4.1 0.62 0.6 14 42
48 4.1 0.39 12.9 14 39
96 5.2 0.52 0.2 14 33
168 (dried) 5.4 0.46 2.7 10.5 9.5
a Not determined. Microbial populations in final product were (log cfu g−1): mould propagules,
9.5; yeasts, 8.5; lactic acid bacteria, 3.4.
(Table 7.7) and, possibly, the lower concentrations later in the fermen-
tation were due to oxidation of ethanol and/or evaporative losses. The
early stages of the fermentation were accompanied by a rapid drop in
pH value and a relatively small increase in titratable acidity. The acid
content then decreased as the cakes were dried, presumably due to
utilisation by fungi, accompanied by an increase in pH value. During
incubation, a distinct increase in the thickness of the cakes, due to
release of gas, and a slight alcoholic-ripened-fruit odour was noted.
The odour became more intense from the second day onwards. It is
also evident that considerable loss of water occurs during fermenta-
tion, before the final drying stage, which renders the cakes stable at
room temperature.
7.2.3.3 Tapai a Umay Pre-fermentation of the rice (tapai a umay) lasts

for 24 h, and is characterised by the production of reducing sugars and
ethanol. The pH fell to 4.4 and there was a small increase in titratable
acidity of 0.3% (w/w) (Table 7.9). Fruity and alcoholic odours were
detectable by 12 h incubation and became more intense towards the
end of fermentation. The rice liquefied and there was an increase in
moisture content from an initial 64% to 67–68% after 24 h.
7.2.3.4 Tinapayan Fermentation A rapid fermentation occurs, mani-

fested by bubbling of the mixture and pH changes in both the rice
and fish portions (Table 7.10). Initially, during the first 6 days of
Table 7.9 Chemical Changes in Fermenting Rice (Tapai a Umay) Incubated at 28–30°Ca
CHEMICAL PARAMETERb
FERMENTATION TITRATABLE ACIDITY ETHANOL REDUCING SUGARS MOISTURE
TIME (H) pH (%, AS LACTIC ACID) (G KG−1 WET WT) (G KG−1 WET WT) (%)
0 5.8 0.1 n.d.c 6.3 64
6 4.5 0.2 n.d. 17 62
12 3.9 0.3 11.8 19 66
18 4.1 0.3 4.2 18 68
24 4.4 0.3 5.1 18 67
a The initial mixture was inoculated with laboratory-made tapai and wrapped in takip asin leaves.
The internal temperature reached 40–45°C.

b Means of triplicate fermentations.
c Not determined.
incubation, the pH of the fermented rice rose and that of the fish fell,
presumably due to diffusion of acid from the rice to the fish. The pH
and acid contents of the rice and fish then remained relatively constant
to the end of the fermentation. At the end, the rice was fully liquefied
and the fish flesh softened. The aroma of the uncooked fermented fish
was described by a set of trained panellists as soya sauce-like, cheesy,
sour, burnt sugar, spicy, sweet and with a discernible rancid odour.
Cooked samples, sautéed in oil until dry, were described as salty, sour,
sweet and spicy with traces of burnt sugar and a rancid taste. An
Table 7.10 Chemical Changes in Tinapayan during Fermentation for 15 Days at 28–30°C
RICE PORTION FISH PORTION
FERMENTATION TITRATABLE ACIDITY TITRATABLE ACIDITY
TIME (D) pH (%, AS LACTIC ACID) pH (%, AS LACTIC ACID)
0 4.4 0.25 6.8 0.5
3 5.8 0.6 6.1 0.5
6 6.2 0.4 5.9 0.6
9 5.2 1.1 5.0 0.9
12 5.5 1.0 5.2 0.8
15 5.1 1.1 4.6 1.0
aftertaste, characterised as a pricking sensation on the tongue was

detected but was not a reason for rejecting the product.
7.2.4 Microbiology
The primary objective of lactic acid fermentation of fishery products is

to preserve the raw material from microbial spoilage, while inciden-
tally modifying the eating characteristics of the food. Commonly, the
preservation is attributed entirely to the action of lactic acid bacteria,
which is not entirely true. Processing starts with cleaning and gutting
the fish. These two steps remove digestive enzymes and much initial
microbial flora. Dolendo et al. (1978) observed that the microbial flora
on the skin and intestines in milkfish was reduced to about 17% of
the initial levels by cleaning. Likewise, salting and/or drying have
large effects on enzymatic and microbial activities. Apart from the
direct effects of NaCl, salt lowers the aw and enhances the sensitivity
of microorganisms to acid (ICMSF 1980). Inadequate salting causes
spoilage of the fish and no amount of acid can then cover up the prod-
ucts of spoilage. Many products in the market are rejected due to their
putrid odour (Barile 1983).
Lactic acid, produced as a consequence of the proliferation of
lactic acid bacteria, is the major preservative in rice–fish fermenta-
tions. In traditional fermentations, like balao-balao with ~6% NaCl
and an incubation temperature of 28–30°C, a pH of 4.5 is reached
after 2–3 days (Arroyo et al. 1978). This suggests that preservation of
highly perishable fish before sufficient acid is produced depends on
the salt absorbed into the flesh. In the absence of acid, 3.5% NaCl,
as recommended for milkfish fermentation, is insufficient to inhibit
putrefactive organisms (Guevarra et al. 1978), whereas, in the pres-
ence of acid, proteolytic and spoilage organisms associated with natu-
rally fermented products may not tolerate salt concentrations of 2.5%
(Potter and Hotchkiss 1995). In cases where the generation of acid is
delayed, spoilage takes place.
The preservation of fish, therefore, involves several interrelated
steps during processing. Any stage that allows the spoilage of any of
the ingredients corrupts the quality of the product and may result in
eventual rejection. Spoilage of each of the ingredients needs further
study, alongside studies on the proliferation of acid-producing and
flavour-enhancing microorganisms. Apart from studies on the gen-

eration of acid, there are very few studies on other aspects of fish fer-
mentations, such as pre-fermentation spoilage of the fish, inoculum
development, shelf life with different packaging materials, factors
affecting acceptability of fermented products, and process safety. Since
the primary quality that consumers evaluate before buying a product is
its odour, it is important to minimise fish spoilage prior to and during
the fermentation. However, this area has received little study to date.
Ultimately, it is the acceptability of the finished product that deter-
mines the success of the technology. Currently, microbially fermented
fishery products are neither as stable nor as acceptable as high-salt fish
sauces or pastes and only a limited percentage of the population, even
in rural areas, eat bacterially fermented fishery products.
Some processors in Central Luzon marinate salted milkfish in vin-
egar to ensure preservation of the fish before combining it with rice
(Uyenco and Ajon 1982). The initial pH in marinated milkfish was
4.2 and this dropped to pH 4.0 after 48 h. However, the presence of
acetic acid at low pH may be unfavourable to the growth of lactic acid
bacteria.
7.2.4.1 Sources of Microorganisms The microbial flora of fish and

shrimp is influenced by the environment, the food available to them
and the feeding habits of the species. Freshwater cultivated species,
such as carp and freshwater shrimp, were found to harbour lactic acid
bacteria that could grow well in broth with pH 9, 6.5% NaCl (w/v)
and 40% (w/v) bile (Cai et al. 1999). Some of these bacteria were
identified as Lactococcus, Pediococcus and Enterococcus spp. and it was
suggested that they could be remnants of probiotics that are added
to the feeds, the pond or the environment of the larvae. Shrimp from
coastal brackish water, such as greasy backs (Metapenaeus ensis), pos-
sess a distinct microbial flora due to their scavenging nature. They
feed on decaying organic matter resulting in a higher microbial count
than fishes that are pelagic.
The majority of traditional fish fermentations are natural fermen-
tations without inoculation with starter cultures and the organisms
responsible for the fermentation derive from the raw materials, air,
utensils or surfaces of equipment used. Lactic acid bacteria can be
true inhabitants of the digestive system, gills, slime and skin of fish,
as shown by their isolation from carp and shrimp (Cai et al. 1999), rice
(Juliano 1993), spices (Guerra 1992; Sanchez 2008) and other ingre-
dients. Fish in wet fish markets may be contaminated with organisms
from containers, handlers and so on, including potential spoilage bac-
teria and food-borne pathogens (Seema et al. 1997).
Some traditional processors add a little fully fermented product to
a new fish/shrimp–rice mixture to ensure the presence of lactic acid
bacteria. At the same time, the fully fermented material may bring
amylases and acid to the new mixture. In many cases, the lactic acid
bacteria present in fully fermented products are acid tolerant (Uyenco
and Ajon 1982), which is advantageous for rapid acid formation.
7.2.4.2 Microbial Changes during Fermentation The microbial changes

during acid generation in uninoculated mixtures, whether of fish or
shrimps, show a succession of microbial populations (Arroyo et al.
1978; Mabeza 1983; Orillo and Pederson 1968; Solidum 1979;
Uyenco and Ajon 1982).
7.2.4.2.1 Burong Isda Before incubation, non-acid producers, lac-

tic acid bacteria and sometimes moulds were isolated but within 1
day Leuconostoc mesenteroides and Streptococcus spp. appeared and pre-
dominated until the fourth day of incubation (Orillo and Pederson
1968). Liquid and gas, presumably CO2, were formed. On the third
day, Pediococcus cerevisiae and Lactobacillus plantarum were isolated and
were present until the end of the fermentation.
Sakai et al. (1983a, b) isolated an assortment of non-acid-forming
microbes, including Bacillus spp., moulds, such as Aspergillus spp. and
Monascus spp., and yeasts, such as Candida tropicalis, Torulopsis mogii
and Saccharomyces cerevisiae. Sakai et al. suggested that these non-
acid formers could have contributed to the hydrolysis of starch and
generated reducing sugars. In addition to the above-named organ-
isms, at one time or another, microorganisms in burong isda included
Bacillus subtilis, Bacillus cereus, Streptococcus faecalis, Lactobacillus brevis,
Leuconostoc mesenteroides, Pediococcus cerevisiae, Lactobacillus fermentum
and Lactobacillus plantarum. The end of the fermentation was domi-
nated by acid-tolerant lactic acid bacteria.
Lactic acid bacteria in commercial fermentations by traditional
household processors in Central Luzon were examined by Uyenco and
Ajon (1982). The fish were marinated with vinegar overnight, mak-
ing the mixture acidic, pH 4.2, and probably reducing the number of
spoilage organisms and favouring lactic acid bacteria. The rice–fish
mixtures were stored in the laboratory and incubated for 7 days and
lactic acid bacteria were enumerated daily (Table 7.11). A microbial
count of 108 cfu g−1 and pH 4.2 was observed after 24 h but the count
declined to 106 cfu g−1 by the third day and the pH reached 3.5. The
decline in bacterial numbers could be due to the effect of low pH and
organic acids, attributed to the growth of Lactobacillus plantarum and
Lactobacillus confusus.
7.2.4.2.2 Balao-Balao Microorganisms in balao-balao have been

described by Solidum (1979), Uyenco and Ajon (1982) and Vatana
(1982). A sequence of microorganisms proliferated in the rice–shrimp
mixture and fermented it successfully without the addition of inocu-
lum, sugar or amylase. Arroyo et al. (1978) reported a final pH 3.4
and 3.9 in balao-balao with 3% and 6% NaCl content, respectively.
Table 7.11 Succession of Lactic Acid Bacteria and pH Changes in Commercial Burong Isda and
Balao-Balao Fermentations
BURONG ISDA (FERMENTED MILKFISH) BALAO-BALAO (FERMENTED SHRIMP)
TIME
(DAY) LOG CFU G−1 pH ISOLATESa LOG CFU G−1 pH ISOLATESa
1 8.1 4.2 Leuconostoc >8.0 5.5 Streptococcus
mesenteroides faecalis
2 7.8 4.0 Lactobacillus >8.0 5.0 Pediococcus
plantarum pentosaceus
3 7.6 4.0 Lactobacillus 6.3 4.5 n.d.b
confusus
4 7.4 3.9 Lactobacillus 5.7 3.9 Lactobacillus
plantarum plantarum
5 7.1 3.7 n.d. 4.6 3.7 Lactobacillus
plantarum
6 6.9 3.5 Lactobacillus 4.0 3.7 Lactobacillus
plantarum plantarum
Lactobacillus casei
7 7.1 3.5 Lactobacillus <2.0 3.6 n.d.
plantarum
Source: Adapted from Uyenco, F.R. and E.C. Ajon. 1982. Isolation and identification of lactic acid
bacteria from some fermented foods. Natural and Applied Science Bulletin 34:15–24.
a Predominant lactic acid bacteria isolated on glucose–yeast extract–peptone–agar with CaCO .
3
b No data.
Six percent NaCl allowed faster acid generation and yielded a more
acceptable product than fermentations with higher salt concentrations.
Solidum (1979) studied a balao-balao fermentation with a high
salt content. Salt added to the shrimp was 20% and to the rice 3%.
Streptococcus faecalis initiated the fermentation, followed by Leuconostoc
mesenteriodes on the second day, Pediococcus cerevisiae on the third day,
Lactobacillus brevis on the fourth day and Lactobacillus plantarum on
the fifth day. The higher salt concentration ensured the stability of the
mixture up to the third day, before the pH was brought down to 4.1.
7.2.4.2.3 Tapai The data in Table 7.7 suggest that the major micro-
organisms in tapai are yeasts and it is clear that their concentration
increased 100-fold during the period of incubation. Although Guerra
(1992) provided mould counts, these are likely to reflect the numbers
of mould spores present and give no information on actual mould
biomass. The concentrations of lactic acid bacteria are quite low, at
103.4 cfu g−1, and, given an inoculation rate of 0.2% of the cooked rice,
the numbers present in the inoculated rice fermentation will have been
even lower.
Red hot chilli, considered to be extremely spicy by panellists, in
tapai comprises 2% of the rice mixture. The compound capsaicin,
responsible for the red hot sensation when chilli is eaten, exhibits
antimicrobial activities against bacteria (Cichewicz and Thorpe 1996)
and fungi (De Lucca et al. 2002) and may play a role in selecting the
microbial flora of tapai. The pungent taste of chilli was not detected
in newly made tapai, suggesting that capsaicin had decomposed and/
or had been utilised (Guerra 1992).
Tapai is very similar to Thai look-paeng and Indonesian ragi
and it seems very likely that the microbial flora is also similar (see
Chapter 2).
7.2.4.2.4 Fermented Rice Guerra (1992) investigated the microbial

populations in fermenting rice (tapai a umay; Table 7.12). It is evident
that the inoculation of the rice with tapai provided a very high initial
yeast population, such that it only increased fivefold by the end of the
fermentation. However, lactic acid bacterial numbers increased rap-
idly during the first 6 h of incubation while numbers of other bacteria
declined throughout the fermentation.
Table 7.12 Microbial Populations of Fermenting Rice (Tapai a Umay) Incubated at 28–30°Ca
MICROBIAL POPULATION (LOG CFU G−1)b
FERMENTATION TIME
(H) YEASTS LACTIC ACID BACTERIA AEROBIC PLATE COUNTc
0 6.1 3.0 6.9
6 6.9 5.0 4.5
12 6.8 5.0 3.0
18 6.8 4.7 1.6
24 6.8 5.0 1.2
a The initial mixture was inoculated with laboratory-made tapai and wrapped in takip asin leaves.
The internal temperature reached 40–45°C.

b Means of triplicate fermentations.
c Nutrient agar with 0.05% starch medium.
Tapai a umay, a fermented rice is very similar to Thai khao mak and
Indonesian tapé and it may be presumed that the microbiology and
biochemical processes are also similar (see Chapter 3).
7.2.4.2.5 Tinapayan The microbial population was dominated by

yeasts for the first 6 days of incubation while lactic acid bacteria mul-
tiplied from an initial concentration of 102.9 cfu g−1 to 107.0 in 3 days
(Table 7.13). A concentration of 107.8 yeast cfu g−1 implies a substan-
tially larger yeast biomass than lactic acid bacterial biomass.
Table 7.13 Microbial Population of Tinapayan Cooked Rice–Fish Mixture during Fermentation at
28–30°C
CONCENTRATION IN RICE PORTION CONCENTRATION IN FISH PORTION
(LOG CFU G−1) (LOG CFU G−1)
FERMENTATION LACTIC ACID LACTIC ACID
TIME (DAYS) YEASTS BACTERIA YEASTS BACTERIA
0 6.2 2.9 1.0 1.0
3 7.8 6.8 6.9 6.5
6 6.8 6.2 6.3 5.5
9 7.0 7.1 6.5 6.7
12 6.5 7.1 5.0 6.2
15 5.8 7.1 4.9 6.3
In tinapayan processing, an association of moulds, yeasts and lactic

acid bacteria is responsible for liquefying the rice starch, for lactic acid
formation and for the development of the meaty and cheesy aroma
and flavour.
7.2.4.3 Sources of Fermentable Sugars in Rice-Fish Fermentations In

tinapayan fermentation, the tapai a umay fermented rice is the source
of fermentable sugars and of amylases that continue to function dur-
ing fermentation. However, there remains the question as to how
fermentable materials are formed in white burong isda and in balao-
balao where no inocula, amylases or sugars are added and yet the
mixtures successfully undergo lactic acid bacterial fermentations.
Mendoza (1985) investigated the fermentation of a minced
Decapterus macrosoma Bleeker (a carnivorous marine species)–cooked
rice–3% NaCl mixture inoculated with fully fermented balao-balao.
Rapid fermentation was observed, with the pH falling from an initial
value of 6.5 to a value of 3.5 in 48 h, while acid-producing bacteria,
presumed to be lactic acid bacteria, increased from 104.9 to 109.1 cfu g−1
(Figure 7.8). Lactic acid bacteria isolated from the fermentations were
shown to be capable of hydrolysing starch, completely or partially,
10 7
9
pH/titratable acidity (% as lactic acid)
6
8
Microbial concn. (log cfu/g)
5
7
6 4
5 3
4
2
3
1
2
1 0
–1 1 3 5 7 9 11 13
Time (days)
Figure 7.8 Development of natural microbial populations in fermenting minced fish (Decapterus
macrosoma)–cooked rice mixture. Values are means of duplicate fermentations. □, acid producing
bacteria (on tryptone–glucose–yeast extract agar with CaCO3); Δ, non-acid producing bacteria;
○, yeasts; ×, pH value; *, titratable acidity. (Adapted from Mendoza, L.S. 1985. The microbiol-
ogy of cooked rice and fish fermentation. PhD thesis, Department of Food Science and Technology,
University of Reading, United Kingdom.)
and growing and producing acid from the resultant sugars (Mendoza
1985; Mendoza and Owens 1986). Starch-hydrolysing lactic acid bac-
teria identified by Mendoza (1985) were of two types: Lactobacillus
casei subsp. pseudoplantarum group producing clearing only beneath
the colonies on APT-starch agar, suggesting that the amylases were
cell-bound and/or non-diffusible; the other type, Lactobacillus plan-
tarum and Lactobacillus brevis groups, produced clearings around the
colonies and could hydrolyse starch fully both in a complex medium
such as APT-starch broth and in a basal medium containing only
NaCl, phosphate and starch. It appears that the last group can initi-
ate fermentation as long as starch is present in the solution.
Starch hydrolysis by lactic acid bacteria from fermented milkfish
(burong bangos) was also demonstrated by Olympia et al. (1995). One
isolate was identified as Lactobacillus plantarum. Other starch-hydro-
lysing lactic acid bacteria that have been described include Leuconostoc
sp. from fish silage (Lindgren and Refai 1984), Lactobacillus amy-
lophilus from swine waste-corn and cattle waste-corn fermentations
(Nakamura and Crowell 1979; Nakamura 1981), strains from the
contents of chicken crops (Champ et al. 1983) and others (Petrova
et al. 2013).
The nutritional value of the product is derived from the composition

of the raw materials and ingredients used (Tables 7.5 and 7.14). Fish
protein contains all the essential amino acids and has high levels of
lysine and methionine, making it of comparable quality to eggs, meat
and chicken. The fish muscle is joined together by connective tissue
that gelatinises when the fish is cooked, making it easier to digest
than animal meat (Staub 1982). Fish are high in protein, minerals and
vitamins and are good sources of long-chain polyunsaturated fatty
acids, including docosahexaenoic acid (DHA) (Uauy et al. 2000).
Montano et al. (2001) examined preserved fishery products that
are available to women in the rural areas, for their DHA content,
including salted small shrimps, Ascetes spp., and fermented juvenile
marine fish, Siganus sp. Fully grown Ascetes spp. are small shrimps
of approximately 2–3 cm in length and very similar to the freshwa-
ter shrimp, tagunton, and the brackish water shrimp, suahe, that are
Table 7.14 Chemical Composition of Rice and Spices Used in Rice–Fish/Shrimp Fermentations
COMPOSITION PER 100 G EDIBLE PORTION (WET WT)
RICE, WELL RICE, GARLIC BULB GINGER ROOT CHILLI FRUIT
MILLED, GLUTINOUS (ALLIUM (ZINGIBER (CAPSICUM
COMPONENT BOILED SATIVUM) OFFICINALE) FRUTESCENS)
Water, g 67.5 11.5 66.5 89 72
Protein, g 2.1 6.9 7.0 1.1 4.8
Fat, g 0.2 0.8 0.3 0.8 2.2
Carbohydrate, g 30 80 24.5 8.5 9.0
Ash, g 0.4 0.5 1.6 0.8 12
Calcium, mg 11 26 28 32 65
Phosphorus, mg 36 95 120 30 89
Iron, mg 0.6 1.1 1.2 3.0 2.3
Ascorbic acid, mg NDa ND 7.0 4.1 9.0
Niacin, mg 0.5 2.0 0.4 0.6 1.8
Riboflavin, mg 0.02 0.03 0.08 0.04 0.25
Thiamine, mg 0.02 0.14 0.23 0.04 0.31
Source: Philippines Food Composition Tables. 1997. Manila: Food and Nutrition Research Institute,
Department of Science and Technology, Taguig, Metro Manila.
a Not detected by the assay used.
used in balao-balao production. The product from Acetes spp. showed

a high DHA content. Furthermore, the DHA content of the salted
products were not very different from that of fresh animals, indicating
that oxidation of the fatty acid is controlled by the anaerobic nature
of the product.
The high percentage of rice in fermented fish–rice mixtures also
contributes to the nutritive value of the product. Rice is not only the
main source of carbohydrates for acid generation but is also a source
of reasonable amounts of thiamine, riboflavin, niacin and vitamin E.
In Asian diets, rice is the main source of energy but contributes little
protein, iron or calcium, etc (Juliano 1993).
However, some nutritional value is lost during processing and fer-
mentation. The softening of the fish flesh as a result of spoilage allows
loss of water-soluble proteins during washing. Contamination with
moulds and yeasts during storage further reduces the nutritive value
of the finished product. However, there is little data on the scale of
these losses.
The contribution of fermented fishery products to the nutrition of
people depends on the daily consumption, which in turn is determined
by its acceptability, availability and affordability. It is the first two fac-

tors that limit the consumption of bacterially fermented fishery prod-
ucts. The odour of the products and their high acid content make
them unacceptable to some consumers. This is compounded by the
unavailability of the product in many parts of the country due to its
short shelf life. Two livelihood projects in the Philippines have started
to bottle low-salt, microbially fermented fishery products (Vocalan
2010) in order to widen its distribution. It is re-processed into two
variants and presented in enticing and functional packages but the
growth of the industry is not evident. In fact, data concerning the
development of this industry are just lumped with other fermented
fish data, an indication that growth is insignificant.
7.2.6 Safety Concerns
Apart from risks of physical and chemical contamination, the major

safety concerns with bacterially fermented fish/shrimps relate to food-
borne bacterial pathogens or their toxins, fish parasites that may also
infect humans, and the formation of bioactive amines. These prod-
ucts offer nutrient-rich anaerobic environments incubated at tempera-
tures conducive to rapid bacterial growth and of particular concern
is the possibility of contamination, growth and toxin formation by
Clostridium botulinum.
Cases of food poisoning associated with fermented fish have
been reported in Japan, the United States, the United Kingdom and
Australia (USFDA 2011). In Japan and Southeast Asia, poisoning
was associated with the consumption of raw fish (isushi, Tanikawa
1971; ICMSF 1980) and bacterially fermented fish (pla-som and som-
fak) from Thailand (Adams et al. 1985) while in other countries it
was attributed to pickled fish. The higher incidence of reported food
poisoning in developed countries can probably be explained by their
reporting systems being more effective than those in developing
countries, where illnesses are not recorded or rural folks attribute the
illness to other causes (USFDA 2011).
The Codex Alimentarius Commission (2001) defines hazards as
‘biological, chemical or physical agents in or condition of food with
a potential to cause an adverse health effect’. Hazards can be picked
up from the natural environment of the fish, equipment and contact
surfaces, from workers, developed during processing and storage, and

through cross-contamination from inputs.
7.2.6.1 Microbial Food-Borne Hazards

7.2.6.1.1 Clostridium botulinum Clostridium botulinum is a natural
inhabitant of soil and aquatic sediments (Hobbs 1976). It has been iso-
lated from freshwater fish ponds (Cann et al. 1975; Huss et al. 1974),
the marine environment (Tanikawa 1971) and from fish, crabs and
shellfish (USFDA 2011). Outbreaks of Type E botulism in northern
and temperate regions have mostly been associated with fermented
fish (Huss and Petersen 1980) and three quarters of the botulism cases
registered in Alaska were attributed to the consumption of fermented
seafood preparations (Gram 2003). Previously, isushi, a fermented
rice–fish mixture prepared in Japan and consumed without cooking,
was implicated in botulism poisoning (Tanikawa 1971). There are no
reports from Southeast Asian countries of cases of botulism related
to the consumption of fermented fish, possibly because the product is
cooked before consumption, thus destroying the toxin.
The optimum conditions for growth of Clostridium botulinum are
25–37°C, pH ~7.0 and 0.5–1.0% NaCl (Combase Predictor), indicat-
ing the suitability of fish as a medium for growth of the organism,
but once the fermentation has dropped the pH to 4.7–5.0, the growth
of Clostridium botulinum is inhibited (Combase Predictor). Hence,
the crucial period is the time between when anaerobic conditions are
first established and when sufficient acid has been produced to pre-
vent growth. A temperature–time combination of 79°C for 20 min
or 85°C for 5 min can inactivate the toxin, making ordinary cooking
sufficient to destroy it (Hauschild 1989). The toxin is heat sensitive at
pH 7 but is extremely stable in salty and acidic substrates. Preformed
toxin in a raw material may be found again in heavily salted, mari-
nated or fermented final products (Huss and Petersen 1980).
Clostridium botulinum may be found in the gastrointestinal tract
and on the skin of fish, so cleaning and eviscerating followed by
thorough washing helps to reduce the numbers of the microorgan-
ism to a minimum. Fish pond operators in Denmark subject the
fish to ‘self-cleaning’ before harvest by pond liming using 1–6 kg
quicklime m−2 (Huss et al. 1974). In the Philippines, purging in
clean water is practiced to remove the stomach contents of tilapia

before harvesting.
Philippine fermented fishery products have not generally been
implicated as causes of food-borne illnesses, as they are cooked before
consumption and only on rare occasions do Philippine consumers eat
uncooked products.
7.2.6.1.2 Staphylococcus aureus The natural source of these organ-

isms is man and animals and they are found in the nose, on skin and
in the intestinal tract, leading to their presence in air, dust, clothes
and foods. They can grow and produce toxin under both aerobic and
anaerobic conditions and most strains associated with food grow well
in the presence of 10% NaCl (Huss and Petersen 1980). Under anaer-
obic conditions, populations of ~106 cfu g−1 are required to produce
toxic levels of enterotoxin (Jay 1978). In rice–fish fermentations, the
initial preservative is salt at concentrations of 3.3–3.5% (Guevarra
et al. 1978; Orillo and Pederson 1968) and such levels are sufficiently
low to allow growth and toxin production by Staphylococcus aureus.
Toxin pre-formed before the rice–fish fermentation is, therefore, a
significant safety concern because of the stability of the toxin to heat.
However, there are no reports of staphylococcal poisoning associated
with fermented fish products.
7.2.6.1.3 Mycotoxins Potential vehicles for mycotoxins in fer-

mented rice–fish products are angkak and rice. Monascus strains used
in angkak are slow growing and contamination by other moulds can
readily take place (Wang and Lin 2007). In addition, Monascus pur-
pureus produces a variety of secondary metabolites, including some,
such as citrinin, that may be toxic. However, the level of citrinin is
undetectable or very low in daily food products containing Monascus
additives. No reports are available concerning mycotoxins in angkak
coloured fermented fish.
Rice is the major carbohydrate source in lactic acid fermentations
of fish–rice and shrimp–rice mixtures. Aflatoxin contamination can
occur in rice harvested during rainy weather (Vasanthi and Bhat
1990; Shotwell et al. 1966). Processing reduces the levels of toxin,
with dehulling removing 50–70% of the toxin and milling further
reducing the toxin content to 20–35% of the initial amount (Ilag and
Juliano 1982). The extent of destruction of aflatoxin B-1 in rice by
cooking depends on the temperature (Park and Zoe 2006; Rehana
et al. 1979). These authors explained that normal cooking at 100°C
for 30 min reduced the mycotoxin by 49% while pressure cooking for
5 min at 121°C reduced the toxin by 73%.
There are no reports on the aflatoxin content of rice–fish fermented
products but it is evident that only rice free of aflatoxin should be used
in the fermentation.
7.2.6.2 Non-Microbial Biological Hazards Scombrotoxin, histamine, is

produced in spoiled scombroid fishes, such as tuna and mackerel, in
non-scombroid fishes such as sardines, herring and mahi-mahi, and
in flying fishes and half-beaks of the Scomberesocidae family (Taylor
et al. 1989). Besides histamine, other biogenic amines, such as putres-
cence and cadaverine, may also be produced (Arena and Manca de
Nadra 2001) during spoilage. As the fish undergoes autolysis, the free
amino acids become available to spoilage bacteria and the formation
of bioamines, including histamine, becomes possible. The symptoms
of biogenic amine intoxication in general are migraine, brain haemor-
rhage, heart failure, urticaria, headache, ﬂushing, abdominal cramps
and hypertension (Kim et al. 2009).
Red tide toxins, saxitoxins, are produced by dinoflagellate blooms
(Furio et al. 2012; Trevino, 1998) and may contaminate small shrimps
such as suahe and Ascetes spp.
7.2.6.2.1 Histamine Formation The amino acid histidine is con-

verted into histamine by bacterial histidine decarboxylase (Taylor
et al. 1989). Histamine causes illness in humans, especially the elderly
and children, resulting in the formation of rashes on the upper body,
burning in or around the mouth, drop in blood pressure, headache,
nausea and heart palpitation (Kim et al. 2009; USFDA 2011). Fish
implicated in histamine poisoning are not limited to the Scombroidae
and Scomberesocidae families (Taylor et al. 1989). Other species from
freshwater aquaculture, like milkfish (Shiau et al. 2000) and salmon
(Hungerford 2010) showed histamine accumulation upon spoilage,
implying a possible risk of histamine formation if spoilage is allowed
to happen at any stage of the processing.
Histidine decarboxylase occurs widely in bacteria, including in

Enterobacteriaceae, clostridia, lactobacilli, Vibrio spp., Pseudomonas
spp. and Photobacterium spp. (Ababouch et al. 1991; Taylor 1986). The
last three are indigenous in the marine environment while Entero
bacteriaceae species are the most important biogenic amine-forming
bacteria in fish. These include Morganella morganii, Klebsiella pneu-
moniae, Proteus vulgaris and Hafnia alvei (Frank 1985). Inactivation
of microorganisms involved in decarboxylation of histidine is not an
assurance that formation of histamine is stopped, as the decarboxylase,
once formed, remains in the mixture (Hungerford 2010; Richard et al.
2008; ). Heat can inactivate the microorganisms and enzymes but once
histamine is formed it cannot be eliminated by cooking or even retort-
ing (USFDA 2011). Some lactic acid bacteria also produce histidine
decarboxylase, including Pediococcus cerevisiae, Lactobacillus spp. (Arena
and Manca de Nadra 2001; Rodwell 1953) and the halophilic lactic
acid bacteria, Tetragenococcus muriaticus and Tetragenococcus halophi-
lus (Kimura et al. 2001). The halophilic strains form histamine when
present in fermenting fish mixtures. Besas and Dizon (2012) observed
histamine levels exceeding the USFDA limit of 50 mg kg−1 formed by
halotolerant lactic acid bacteria in a substrate containing 10% NaCl.
Thus, the potential exists for the formation of toxic levels of histamine
in fermented fishery products but, so far, no cases have been reported.
7.2.6.3 Parasites The freshwater fish for lactic acid fermentations

are exposed to parasitic infections (Adams et al. 1997; Arthur and
Laman-Mayo 1997; Cross and Basaca-Sevilla 1991; Huss et al. 2003;
Petersen et al. 1993). There are many different kinds of parasites of
fish but only those that may infect humans through the consumption
of fermented fish products are discussed. These include nematodes
(round worms), cestodes (tapeworms) and trematodes (flukes). Their
life cycles involve a variety of hosts in fresh and marine water environ-
ments, including carnivorous fishes, crustaceans, copepods and mol-
luscs and, in some cases, land animals such as snails, crabs, dogs and
cats (Huss et al. 2003; Murrell 2002; WHO 1974).
Human infections have been attributed to the consumption of raw
fish, sashimi, pickled raw fish such as ceviche (fresh raw fish mari-
nated in citrus juices), marinated sardines, improperly cooked fish,
salted herring, cold smoked products, fermented fish, Japanese isushi
and inadequately cooked fish, crabs, shellfish and squids (Murrell

2002). There are no studies that report on the survival of parasites in
lactic-fermented fish or shrimps.
7.2.6.3.1 Trematodes (Flukes) The eggs of the parasite are excreted

by infected hosts and are ingested by snails, where they multiply asex-
ually and hatch into free-swimming ciliated larvae that are released
into the water. The larvae swim in the water until they lodge under
the scales of fishes, and encyst in the muscles and subcutaneous tis-
sues. Crustaceans are also intermediate hosts. Consumption of the
infected fish flesh by humans and carnivorous animals enables the
cysts to reach the stomach where the larvae, called metacercariae, are
freed by digestion of the tissues. The metacercariae move from the
stomach to the small intestine and enter the common bile duct (usually
4–7 h after ingestion) and mature in about 4 weeks (Murrell 2002).
Different species can parasitise the liver, the intestines or the lungs
and certain species may move to organs outside the digestive tract,
including the brain and the heart. Most human infections, seven mil-
lion were reported worldwide in 1994 (WHO 1995), are traced to the
consumption of infected aquacultured cyprinids (carp). The increase
in aquacultural operations in Southeast Asia, Japan and China appear
to be important factors in the spread of these parasites.
7.2.6.3.2 Nematodes (Round Worms) Like other parasites, round worms

infect humans upon consumption of infected fish, terrestrial or aquatic
molluscs or crustaceans that are raw, slightly salted, pickled, smoked
or partially cooked (Huss et al. 2003). In the life cycle of the round
worm, humans are abnormal hosts. The life cycle of the worm can
be completed without humans but infections happen due to the eat-
ing and food preparation habits of humans. Adult round worms live
primarily in the intestines and kidneys of fish-eating mammals, birds
and fish. The larvae encyst in the abdominal cavity of their interme-
diate mammal, avian or fish hosts without affecting them. However,
when the infected flesh is eaten by humans, they can cause intrac-
table diarrhoea. Capillaria philippinensis causes intestinal capillariasis
in humans. An unusual feature of the infection is the presence of all
stages of the parasite in the infected person (WHO 1974). Thailand
and the Philippines showed the largest number of cases of intestinal

capillariasis (Cross 2001).
7.2.6.3.3 Cestodes (Tapeworms) Adult tapeworms are the largest par-

asites of humans and reside in the small intestine. The adult parasite
releases its eggs into the intestine, from where they are excreted with
the faeces. Dogs, cats and pigs may also serve as hosts. The eggs develop
into embryos in water, which are eaten by copepods and develop into
procercoids. When the copepods are eaten by fish, the procercoids
enter the blood stream and reach muscles and other organs where
they develop into plerocercoids. The larvae can be transferred several
times along a chain of small fishes until they are consumed by bigger
carnivorous fishes. The plerocercoids are transferred to man through
the consumption of raw or under-cooked infected fish and develop
into adult tapeworms. The parasite competes with the host for vitamin
B12, which may lead to anaemia (Murrell 2002).
7.2.6.4 Control of Parasitic Infections The Code of Conduct for Res

ponsible Fisheries—Aquaculture (Bagarinao 2005) recommends
that good aquacultural practices be followed and that fish be cooked
properly before consumption. Since part of the life cycle of the par-
asites takes place in freshwater environments, spread of parasites
can be controlled by following the guidelines. Likewise, the guide-
lines on ‘Responsible Practices in Fisheries Post-Harvest’ (FAO
1998) include additional measures that can help in reducing parasite
infections. Huss et al. (2003) noted that release of live larvae as a
result of cutting the internal organs and belly wall can occur, lead-
ing to the transfer of larvae to other parts of the fish. Cleaning the
fish may require trimming the belly wall and some of the flesh to
remove infected tissues, depending on the extent of cyst infestation.
Candling and picking parasites from the flesh is acceptable if the fish
is to be heat processed.
Metacercariae larvae of the fluke, Opisthorchis viverrini, are killed
by temperature–time combinations of 70°C for 30 min or 80°C for
5 min (WHO 1995). The metacercariae are also inactivated by 13.6%
NaCl for 24 h or 20% NaCl for 5 h (Huss et al. 2003) but dry salting
does not reliably kill cysts embedded in the tissues (Francis-Floyd
2009). Anisakid larvae (see Huss et al. 2003), however, are relatively
heat sensitive and are killed by 60°C for 1 min (Bier 1976).
There is little information on the effects of lactic acid and low pH
or on the combined effects of acid, pH, salt and temperature on the
survival of parasites. Each of the these factors may contribute, to a
certain degree, to the inactivation of parasites, including salt concen-
tration, temperature of raw material storage, anaerobiosis during lac-
tic fermentation, and total acidity of the system. In view of the above,
it would be helpful to evaluate the totality of these effects on the sur-
vival of different stages of the parasites.
7.2.6.5 Chemical Hazards Fish are harvested from natural water

environments or from aquaculture ponds and these environments may
be contaminated with organic compounds, such as pesticides, antibi-
otics, dioxins, industrial wastes, cleaning chemicals etc., and/or with
inorganic chemicals, such as lead, mercury, copper, arsenic etc.
7.2.6.6 Physical Hazards These include any potentially harmful extra-

neous matter not normally found in food. The commonly reported physi-
cal hazards are filth (non-food safety hazard), glass, metal, wood, bones,
stones, plastic and so on. Solar salt, produced locally, is often the source
of physical hazards, such as glass fragments, wood, shells or stones.
Lactic acid fermentation methods for fish and fishery products are
practical and simple preservation techniques suitable for use in rural
areas. While the technologies have been in existence for a long time,
they have received relatively little attention and so knowledge of the
technology of fermentation is not widely known. In spite of moves to
re-process fermented fish and shrimp products so as to increase the
shelf life and allow wider distribution, the industry is slowly disap-
pearing from the rural landscape of the Philippines. The two types of
lactic-fermented fishery products in the Philippines, the buro/balao-
balao and the tinapayan types, meet with different constraints. The
demand for the wet buro type is less than for the dry, fried tinapayan
type. As a fish sauce, buro-type products compete with the high-salt,
enzymatically hydrolysed fish sauces and pastes. The high-salt products
are more in demand due to their stability and better appearance, tex-
ture and flavour while buro sauce is only acceptable to a limited set
of consumers in Central Luzon and in scattered places in Pangasinan
and Southern Luzon. The strong dominant putrid odour of the bac-
terially fermented product is perceived to be the result of fish spoilage
before acid is generated. This results in low product acceptability. The
short shelf life also hinders progress of the industry because of dis-
tribution problems. Bacterially fermented rice–fish/shrimp is still a
kitchen industry and, for the product to gain better acceptability, the
product quality needs to be improved. In spite of the merits of low-
salt fermentation as a tool to preserve the nutritional value of fishery
products, the acceptability of buro, plus its highly acidic nature, are
deterrent factors in the development of the industry.
In contrast, a wide group of consumers in Mindanao favour the dry,
meat-like flavour of tinapayan and the absence of pork in the formula-
tion of tinapayan gives it a special niche among Muslim consumers in
Mindanao. Hence, the prospects for this fermented product are more
promising, mostly due to its stability, meaty flavour and not too acidic
taste. Although the amount of tinapayan produced is low, it could expand
as long as the raw material is available or is supplemented by other lean
marine species. Mudfish is sold as a table fish in the wet market, limiting
the volume for fermentation. Owing to the seasonality of mudfish, other
lean marine fish could possibly be utilised as raw material.
It has always been assumed that fermentation is successful as long
as the acid is generated on time. However, for the safety-conscious
consumer, products with any objectionable property become objects
of scrutiny. Inspection of establishments producing these products,
no matter how small, is recommended to promote good Sanitation
Standard Operating Procedures (SSOP) and Good Manufacturing
Practices (GMP).
Before modifications are made to the production technology, it is
desirable to conduct training programmes to review the existing tech-
nology, to enhance the appreciation of the nutritional value of the
product, to increase understanding of the need to monitor product
quality and safety and to discuss the relevance of rules and regula-
tions concerning the marketing of such products. Toxin formation
before and during processing are not evaluated in this traditional
product and the cumulative effects of the interrelated factors in the
fermentation systems are not clear. The presence of biological hazards

and parasites in the environment of the raw material may necessitate
modifications in the technology.
7.2.8 Future Prospects and Recommendations
The buro-type products require a thorough study of spoilage of the

fish before acid is generated and during fermentation in order to
address the problem of objectionable odour and histamine formation.
Sanchez (2008) reported that some lactic acid bacteria are salt toler-
ant, growing in 10–18% NaCl. The significance of this is the pos-
sibility of adjusting the salt added to fish so that spoilage before acid
generation is avoided. The use of halotolerant lactic acid bacteria in a
mixture would enable the use of higher salt concentrations while still
allowing acid formation. To pursue this approach, the salt concentra-
tions that can inhibit spoilage organisms need to be elucidated.
Some salt-tolerant lactic acid bacteria were reported to possess
histamine-degrading capabilities (Sanchez 2008; Zaman et al. 2010).
This information could be pursued further in relation to the pos-
sible detoxication of histamine and other amines. The raw material
for tinapayan, being a carnivorous fish, could be the medium for the
spread of parasitic infections. In view of this, the handling and dry-
ing of the raw materials needs a thorough evaluation to determine the
effects of these processes on parasites.
The tinapayan fermentation involves both fungi and bacteria and
studies are needed to identify the predominant species present and
the natures and scales of their activities. The biochemical activities of
the fungi and bacteria involved in the fermentation have received little
study and information on them would help greatly in disentangling
the relative contributions of the fungi and the bacteria. The fruity
odours observed are, presumably, due to ester formation but the con-
ditions affecting their formation, so far, have received no attention.
References
Ababouch, L., M.E. Afilal, H. Benabdeljelil and F.F. Busta. 1991. Identification
of histamine-producing bacteria isolated from sardine (Sardina pilchar-
dus) stored in ice and at ambient temperature (25°C). Food Microbiology
8(2):127–136.
Adams, A.M., K.D. Murrel and J.H. Cross. 1997. Parasites of fish and risks
to public health. Revue Scientifique et Technique (Office International Des
Epizooties) 16:652–660.
Adams, M.R., R.D. Cooke and P. Rattagool. 1985. Fermented fish products of
South East Asia. Tropical Science 25:61–73.
Arena, M.E. and M.C. Manca de Nadra. 2001. Biogenic amine production by
Lactobacillus. Journal of Applied Microbiology 90:158–162.
Arroyo, P.T., L. Ludovico-Pelayo, H.T. Solidum, Y.N. Chiu, M. Lero and
E. Alcantara. 1978. Studies on rice-shrimp fermentation: Balao-balao.
Philippine Journal of Food Science and Technology 2:106–125.
Arthur, R. and S. Laman-Mayo. 1997. Checklist of the Parasites of Fishes of the
Philippines, FAO Fisheries Technical Paper 389. Rome: FAO.
Bagarinao, T.U., Ed. 2005. Regional Guidelines for Responsible Fisheries in Southeast
Asia—Aquaculture. Tigbauan Iloilo, Philippines: SEAFDEC, Aquaculture
Department.
Barile, E. 1983. Survey of lactic fermented fishery products in central Luzon
and their characterization. Thesis, University of the Philippines, Diliman,
Phillipines.
Besas, J. and E. Dizon. 2012. Influence of salt concentration on histamine
formation in fermented tuna viscera (dayok). Food and Nutrition Sciences
3(2):201–206.
Bier, J.W. 1976. Experimental Anisakis: Cultivation and temperature tolerance
determination. Journal of Milk Food Technology 39:132. Cited by USFDA
2011.
Borgstrom, G. 1961. Fish as Food. Vol. 1. Production, Chemistry and Microbiology.
New York: Academic Press.
Cai, Y., P. Suyanaindana, P. Saman and Y. Benno. 1999. Classification and char-
acterization of lactic acid bacteria isolated from the intestines of com-
mon carp and freshwater prawns. Journal General Applied Microbiology
45:177–184.
Cann, D.C., L.Y. Taylor and G. Hobbs. 1975. The incidence of Clostridium
botulinum in farmed trout raised in Great Britain. Journal of Applied
Bacteriology 39:331–336.
Champ, M., O. Szylit, P. Raibaud and N. Ait-Abdelkader. 1983. Amylase pro-
duction by three Lactobacillus strains isolated from chicken crop. Journal
of Applied Bacteriology 55:487–493.
Cichewicz, R.H. and P.A. Thorpe. 1996. The anti-microbial properties of chili
peppers (Capsicum species) and their uses in Mayan medicine. Journal of
Ethnopharmacology 52(2):61–70.
Codex Alimentarius Commission. 2001. Food Hygiene Basic Texts. Second ed.
Rome: Food and Agriculture Organization/World Health Organization.
Combase Predictor. n.d. Clostridium botulinum (non-prot) and Clostridium
botulinum (prot). http://modelling.combase.cc/ComBase_Predictor.aspx
(accessed 5 July 2013).
Conlu, P. 1986. Guide to Philippine Flora and Fauna: Fishes. Vol. 1X. Diliman,
Quezon City, Philippines: Ministry of Natural Resources and the
University of the Philippines.
Cooke, R.D., D.R. Twiddy and P.J.A. Reilley. 1987. Lactic acid fermentation
as a low-cost means of food preservation in tropical countries. FEMS
Microbiology Reviews 46:369–379.
Cross, J.H. and V. Basaca-Sevilla. 1991. Capillariasis philippinensis: A fish-
borne parasitic zoonosis. Southeast Asian Journal of Tropical Medicine and
Public Health. 22:153–157.
Cross, J.H. 2001. Fish and invertebrate-borne helminths. In Foodborne Disease
Handbook. Vol. 2. Second ed., edited by Y.H. Hui, S.A. Sattar, K.D. Murell,
W.K. Nip and P.S. Stanfield, pp. 249–288. New York: Marcel Dekker.
De Lucca, A.J., J.M. Bland, C.B. Vigo, M. Cushion, C.P. Selitrennikoff, J. Peter
and T.J. Walsh. 2002. Cay-1, a fungicidal saponin from Capsicum sp. fruit.
Medical Mycology 40:131–137.
Dolendo, A.R., E.P. Tongco, R.A. Roncal and M.C. Alameda. 1978.
Standardization of handling, icing and freezing milkfish. In Milkfish
(Bangos) as Food. Manila: National Science Development Board.
FAO. 1998. Code of conduct for responsible fish utilization. In FAO Technical
Guidelines for Responsible Fisheries. No. 7. Rome: FAO.
Francis-Floyd, R. 2009. The Use of Salt in Aquaculture. Cooperative Extension
Service, Institute of Food and Agriculture, University of Florida
Publication No. VF86.
Frank, H.A. 1985. Histamine-forming bacteria in tuna and other marine fish.
In Histamine in Marine Products: Production by Bacteria, Measurement and
Prediction of Formation, FAO Fisheries Technical Paper 252, edited by
B.S. Pan and D. James, pp. 2–3. Rome: FAO.
Furio, E.F., C.L. Gonzales and Y. Fukuyo. 2012. Widespread outbreak of toxic
red tide in the Philippines. National Fisheries Research and Development
Institute Report, BFAR, Metro Manila, Philippines.
Genigeorgis, C. and H. Reimann. 1979. Food processing and hygiene. In Food
Borne Infections and Intoxications. Second ed., edited by H. Reiman, pp.
613–713. New York: Academic Press.
Gram, L. 2003. Identification of hazards in seafoods. In Assessment and
Management of Seafood Safety and Quality, FAO Fisheries Technical Paper
No. 444, edited by H.H. Huss, L. Ababouch and L. Gram, Rome: FAO.
Guerra, M.R. 1992. Studies on tinapayan—An indigenous fish ferment in
central Mindanao. MS thesis, College of Fisheries and Ocean Sciences,
University of the Philippines in the Visayas, Philippines.
Guevarra, G., E. Marfori, V. Matias, P. de la Pena, N. Sanchez and M. de Guzman.
1978. Milkfish (Bangos) as Food: Handling, Freezing and Processing of Milkfish
(Chanos chanos, Forskal). Manila: National Science Development Board.
Hauschild, A.H.W. 1989. Clostridium botulinum. In Foodborne Pathogens,
edited by M.P. Doyle, pp. 111–189. New York: Marcel Dekker.
Hobbs, G. 1976. Clostridium botulinum and its importance in fishery products.
Advances in Food Research 22:136–185.
Hsieh, Y.T. and T.M. Pan. 2002. The analytic methods and assays for secondary
metabolites of Monascus products. Journal of the Biomass Energy Society of
China 21(1–2):63–71. (In Chinese). Cited by Wang and Lin (2007).
Hungerford, J.M. 2010. Scombroid poisoning: A review. Toxicon 56(2):231–243.
Huss, H.H., A. Pedersen and D.C. Cann. 1974. The incidence of Clostridium
botulinum in Danish trout farms. II. Measures to reduce contamination of
the fish. Journal of Food Technology 9:451–458.
Huss, H.H. and E.R. Petersen. 1980. The stability of C. botulinum type E toxin
in salty and/or acid environments. Journal of Food Technology 15:619–627.
Huss, H.H., L. Ababouch and L. Gram. 2003. Assessment and Management of
Seafood Safety and Quality, FAO Fisheries Technical Paper No. 144. Rome:
FAO.
ICMSF. 1980. Microbial Ecology of Foods, Vol. 1. Factors Affecting the Life and
Death of Microorganisms. London: Academic Press.
Ilag, L.L. and B.O. Juliano. 1982. Colonization and aflatoxin formation by
Aspergillus spp. on brown rice differing in endosperm properties. Journal
of the Science of Food and Agriculture 33:97–102.
Jay, J.M. 1978. Modern Food Microbiology, pp. 254–255. New York: Van Nostrand.
Juliano, B.O. 1993. Rice in Human Nutrition. Rome: International Rice
Research Institute and FAO.
Juliano, B.O., L.U. Onate and A.M. Del Mundo. 1965. Relation of starch com-
position, protein content and gelatinization temperature to cooking and
eating qualities of milled rice. Food Technology 6:1006–1011.
Karim, M.I.A. 1989. Fermented fish products in Malaysia In Fish Fermentation
Technology, edited by C.H. Lee, K.H. Steinkraus and P.J.A. Reilly. Seoul:
Yurim Publishing.
Kim, M.K., J.H. Moh and Huang, H.J. 2009. Biogenic amine formation
and bacterial contribution in fish, squid and shellfish. Food Chemistry
116:87–95.
Kimura, B., Y. Konagaya and T. Fujii. 2001. Histamine formation by Tetra
genococcus muriaticus, a halophilic lactic acid bacterium isolated from fish
sauce. International Journal of Food Microbiology 70:71–77.
Lee, C.L. 1989. Fermentation technology—A review. In Post-Harvest Tech
nology, Preservation and Quality of Fish in Southeast Asia, edited by
P.J.A. Reilly, R.W.H. Parry and L.E. Barile, Stockholm: International
Foundation of Science.
Lindgren, S. and O. Refai. 1984. Amylolytic lactic acid bacteria in fish silage.
Journal of Applied Bacteriology 57:221–228.
Lopez, C. 1989. Microbial ensilage of waste trash fish for animal feeds. In
Proceedings of the Workshop on Post-Harvest Technology, Preservation and
Quality of Fish in Southeast Asia, edited by P.J.A., Reilly, R.W.H. Parry
and L.E. Barile. Stockholm: International Foundation of Science.
Mabeza, A.M. 1983. Effects of pure culture inoculation on the quality of fer-
mented rice-fish (Burong isda) mixture. MS thesis, University of the
Philippines, Los Banos, Philippines.
Mendoza, L.S. 1985. The microbiology of cooked rice and fish fermentation.
PhD thesis, Department of Food Science and Technology, University of
Reading, United Kingdom.
Mendoza, L.S. and J.D. Owens. 1986. Microbiology of cooked rice-minced fish
fermentation. In The First Asian Fisheries Forum, edited by J.L. Maclean,
L.B. Dizon and L.V. Hosillos, pp. 443–446. Manila: Asian Fisheries Society.
Montano, N., G. Gavino and V.C. Gavino. 2001. Polyunsaturated fatty acid
contents of some traditional fish and shrimp paste condiments of the
Philippines. Food Chemistry 75:155–158.
Murrell, K.D. 2002. Fishborne zoonotic parasites: Epidemiology, detection
and elimination. In Safety and Quality Issues in Fish Processing. edited by
H.A. Bremner. Cambridge: Woodhead Publishing.
Nakamura, L.K. 1981. Lactobacillus amylovorous, a new starch hydrolyzing
species from cattle waste-corn fermentations. International Journal of
Systematic Bacteriology 31:56–63.
Nakamura, L.K. and C.D. Crowell. 1979. Lactobacillus amylophilus, a new starch-
hydrolyzing species from swine waste-corn fermentation. Developments
in Industrial Microbiology 20:531–540.
Olympia, M., H. Fukuda, H. Ono, Y. Kaneko and M. Takano. 1995.
Characterization of starch-hydrolyzing lactic acid bacteria isolated from
a fermented fish and rice food, ‘burong isda,’ and its amylolytic enzyme.
Journal of Fermentation and Bioengineering 80(2):124–130.
Orillo, C.A. and C.S. Pederson. 1968. Lactic acid bacterial fermentation of
burong dalag. Applied Microbiology 16(11):1669–1671.
Owens, J.D. and L.S. Mendoza. 1985. Enzymatically hydrolyzed and bacteri-
ally fermented fishery products. Journal Food Technology 20:273–293.
Park, J.W.Y. and K. Zoe. 2006. Effect of pressure cooking on aflatoxin in rice.
Journal of Agricultural and Food Chemistry. 54(6):2431–2435.
Petersen, F., H. Palm, H. Moller and M.A. Cuzi. 1993. Flesh parasites of fish
from central Philippines waters. Diseases of Aquatic Organisms 15:81–86.
Petrova, P., K. Petrov and G. Stoyancheva. 2013. Starch-modifying enzymes
of lactic acid bacteria—Structures, properties, and applications. Starch/
Stärke 65:34–47.
Philippines Food Composition Tables. 1997. Manila: Food and Nutrition
Research Institute, Department of Science and Technology, Taguig, Metro
Manila.
Phitakpol, B. 1989. Fish fermentation technology in Thailand. In Fish Fermen
tation Technology, edited by C.H. Lee, K.H. Steinkraus and P.J.A. Reilley.
Seoul: Yurim Publishing.
Potter, N.N. and J. Hotchkiss. 1995. Food safety, risks and hazards. In Food
Science, Fifth ed., edited by N.N. Potter and J. Hotchkiss, pp. 532–558.
New York: Chapman & Hall.
Putro, S. 1989. Fish fermentation technology in Indonesia. In Fish Fermentation
Technology, edited by C.H. Lee, K.H. Steinkraus and P.J.A. Reilley. Seoul:
Yurim Publishing.
Rehana, F., S.C. Basappa and V.S. Murthy. 1979. Destruction of alflatoxin
by different cooking methods. Journal of Food Science and Technology
16:111–112.
Richard, L., L. Curtis and J. Davis. 2008. The Food Safety Hazard Guidebook.
Cambridge, UK: Royal Society for Chemistry.
Rodwell. A.W. 1953. The histidine decarboxylase of a species of Lactobacillus.
Apparent dispensability of pyridoxae phosphates as coenzymes. Journal of
General Microbiology 8:233–237.
Ruddle, K. and Ishige, N. 2010. On the origins, diffusions and cultural context
of fermented fish products in Southeast Asia. In Globalization, Food and
Social Identities in the Asia Pacific Region, edited by J. Ferror. Tokyo: Sophia
University of Comparative Culture.
Sakai, H., G.A. Caldo and M. Kozaki. 1983a. The fermented fish food, burong
isda in the Philippines. Journal of Agriculture Science (Tokyo University of
Agriculture) 28:138–144. Cited by Sanchez (2008).
Sakai, H., G.A. Caldo and M. Kozaki. 1983b. Yeast flora in red burong isda,
a fermented fish food in the Philippines. Journal of Agriculture Science
(Tokyo University of Agriculture) 28:181–85. Cited by Sanchez (2008).
Sanchez, P.C. 2008. Philippine Fermented Foods: Principles and Technology.
Quezon City: The University of the Philippines Press.
Seema N.P., P.K. Surendaran and K. Gopakumar. 1997. Occurrence and dis-
tribution of lactic acid bacteria in fish and fishery environments. Fishery
Shiau, C.Y., Y.J. Pong, T.K. Chiou and Y.Y. Tin. 2000. Effect of starvation
on free histidine and amino acids in white muscles of milkfish, Chanos
chanos. Comparative Biochemistry and Physiology—Part B. 128:501–506.
Shotwell, O., C.W. Hesseltine, R. Stubblefield and W.G. Sorensen. 1966.
Production of aflatoxin in rice. Applied and Environmental Microbiology
4(3):405–428.
Solidum, M.H. 1979. Chemical and microbiological changes during fermentation
of balao-balao. Philippine Journal Food Science and Technology 3(10):47–48.
Staub, P.A. 1982. The Science of Philippine Foods. Quezon City: Abaniko
Enterprises.
Steinkraus, K.H., ed. 1996. Handbook of Indigenous Fermented Foods, New York:
Marcel Dekker.
Tanasupawat, S. and K. Komagata. 1995. Lactic acid bacteria in fermented foods
in Thailand. World Journal of Microbiology and Biotechnology 11:253–256.
Tanikawa, E. 1971. Marine Products of Japan. Tokyo: Koseisha-Koseidaku
Company.
Taylor, S.L. 1986. Histamine food poisoning: toxicity and clinical aspects. CRC
Critical Reviews in Toxicology 17(2):91–117.
Taylor, S.L., J.E. Stratton and J.A. Nordlee. 1989. Histamine poisoning (scom-
broid) food poisoning: an allergy like intoxication. Journal of Toxicology.
Clinical Toxicology. 27(45):225–240.
Trevino, S. 1998. Fish and shellfish poisoning. Clinical and Laboratory Science
11(5):309–314.
Uauy, R., P. Mena and C. Rojas. 2000. Essential fatty acids in early life, struc-
tural and functional role. Proceedings of the Nutrition Society 59:3–15.
Cited by Montano (2001).
USFDA. 2011. Fish and Fishery Products Hazards and Control Guidance.
Scombrotoxin (Histamine) Formation. Silver Spring, USA: US Food and
Drug Administration.
Uyenco, F.R. and E.C. Ajon. 1982. Isolation and identification of lactic acid
bacteria from some fermented foods. Natural and Applied Science Bulletin
34:15–24.
van Veen, A.G. 1953. Fish preservation in Southeast Asia. Advances of Food
Research 4:209–231.
Vasanthi, S. and R.V. Bhat. 1990. Aflatoxin in stored rice. International Rice
Research Newsletter 15(1):39–40. Cited by Juliano (1993).
Vatana, P. 1982. Biochemical and nutritional changes in fermented rice-shrimp
(Macrobrachium idella) mixture (balao-balao). Dissertation, University of
the Philippines Los Banos, Philippines. Cited by Sanchez (2008).
Vatana, P. and R. Del Rosario. 1983. Biochemical changes in fermented rice-
shrimp Macrobrachium edella mixture. Changes in protein fraction. Food
Vocalan, A. 2010. Personal interview. Balaw-balaw foods, Angono, Rizal,
Philippines.
Wang, T.H. and T.F., Lin. 2007. Monascus rice products. Advances in Food and
Nutrition Research 53:123–159.
Westenberg, J. 1951. Fishery products of Indochina, a compilation of literature
up to the Japanese invasion. Indo Pacific Fisheries Council Proceedings. 2nd
Meeting, FAO Technical Paper No 12:125–150. Bangkok: FAO.
WHO. 1974. Fish and shelfish hygiene. World Health Organization Technical
Report Series 550. Geneva: WHO in cooperation with FAO.
WHO. 1995. Control of fish-borne trematode infections. WHO Technical
Report Series 849. Geneva: WHO.
Wood, B.J.B. and M.M. Hodge. 1985. Yeast-lactic acid bacteria interactions and
their contribution to fermented foodstuffs. In Microbiology of Fermented
Foods Vol. 1, edited by B.J.B. Wood, pp. 263–293. New York: Elsevier.
Zaitsev, V., I. Kizevetter, I. Laganov, T. Makarova, L. Minder and V. Podsevalov.
1969. Fish Curing and Processing. Translated by A. Merindol. Moscow:
MIR.
Zaman, M.Z., F. Abu Bakar, J. Selamat and J. Bakar. 2010. Occurrence of bio-
genic amines and amines degrading bacteria in fish sauce. Czech Journal
of Food Science. 28:440–449.
8
L acti c M e at Fermentati on
W O N N O P V I S E S S A N G U A N , V E T H AC H A I
P L E N G V I D H YA , N I PA C H O K E S A J J AWAT E E
A N D J U N A I DA H A B U B A K A R
Contents
8.1 Nham, a Thai Fermented Pork Sausage 314

8.1.2 Microbiology of Nham Fermentation 315
8.1.3 Characteristics and Roles of the Substrates 318
8.1.3.1 Minced Pork and Cooked Pork Rind 318
8.1.3.2 Cooked Rice 319
8.1.3.3 Garlic 319
8.1.3.4 Sodium Chloride 320
8.1.3.5 Nitrite and Nitrate Compounds 321
8.1.3.6 Erythrobate 322
8.1.3.7 Phosphate Compounds 322
8.1.3.8 Other Ingredients 323
8.1.4 Physico-Chemical Changes during Fermentation 323
8.1.4.1 Sugars Utilisation 323
8.1.4.2 Acidification of Nham 324
8.1.4.3 Colour Development 326
8.1.4.4 Drip Formation 327
8.1.4.5 Proteolysis 328
8.1.4.6 Lipolysis and Lipid Oxidation 329
8.1.4.7 Volatile Compounds in Nham 330
8.1.4.8 Biogenic Amine Formation 332
8.1.5 Safety Considerations and Recommendations 334
8.1.5.1 Physical Hazards 334
8.1.5.2 Chemical Hazards 335
8.1.5.3 Biological Hazards 336
8.1.6 Development of Nham Starter Culture Technology 342
8.1.6.1 Safety Recommendations 343
8.1.7 Future Developments 345
313
8.2 Bruneian Belutak (Fermented Meat Sausage) 346

8.2.2 History, Places of Production, How Consumed
and Role in Diet 346
8.2.4 Characteristics of the Fermentation 348
8.2.7 Future 349
References 349
8.1 Nham, a Thai Fermented Pork Sausage

Wonnop Visessanguan, Vethachai Plengvidhya and Nipa Chokesajjawatee
Nham (also naem) is an indigenous fermented pork sausage of Thailand

that has gained popularity among the Thais and is widely consumed
owing to its unique texture, flavour, colour and taste. The basic ingre-
dients of nham are minced pork (52% w/w) and cooked pork rind
(35% w/w), which are mixed with fresh garlic (4.3% w/w), cooked rice
(4.3% w/w), salt (1.9% w/w), sugar (0.3% w/w), whole bird chilli (~2%
w/w), sodium tripolyphosphate (0.2% w/w), monosodium glutamate
(0.2% w/w), erythrobate (0.2% w/w) and potassium nitrite (0.01%
w/w). The mixture is thoroughly mixed and then tightly packaged
in banana leaves or plastic casing. The product is allowed to ferment
at ambient temperature over a 3–4-day period, during which time it
attains a final pH of 4.6 (Phithakpol et al. 1995; Valyasevi and Rolle
2002; Figures 8.1 and 8.2).
Mechanical processes for the production of nham are well devel-
oped. The traditional fermentation process is spontaneous but,
increasingly, starter cultures are now being used. Acid production
contributes to both intrinsic quality and safety in the production
of nham in that it imparts the sour taste, typical fermented aroma,
firmness of texture and prevents the growth of acid-sensitive patho-
gens. Nham is considered a local authentic in that its flavour and
texture varies with the region in which it is produced. It is con-
sumed in the raw state as a condiment, or is served cooked on its
L ac ti c M e at F erm en tati o n 315
Cooked pork skin Minced lean pork Cooked rice Garlic

(35) (52) (4.3) (4.3)
Whole bird chilli, ~2;

NaCl, 1.9; sucrose, 0.3;
Na3PO4, 0.2; MSG, 0.2;
Erythrobate, 0.2; KNO2, 0.01
Mix
Package in banana leaf or plastic tubing
Incubate, ambient temperature, 3–4 days
Nham
Figure 8.1 Traditional process for production of Thai nham (fermented pork sausage). Figures
are % (w/w) of the total mix. MSG, monosodium glutamate.
own, or as part of a main meal. In Thailand, nham is produced on

a commercial scale primarily by small-scale and cottage-level pro-
cessors, creating an estimated production value of USD20 million
annually.
8.1.2 Microbiology of Nham Fermentation
Fermentation of nham involves the successive growth of different micro-

organisms, dominated by lactic acid bacteria naturally present in nham
ingredients (Valyasevi et al. 2001). Studies have identified lactobacilli
(Lactobacillus plantarum, Lactobacillus pentosus and Lactobacillus sakei)
and pediococci (Pediococcus acidilactici and Pediococcus pentosaceus) as the
predominant microorganisms in nham fermentation (Tanasupawat and
Komagata 1995). Other microorganisms identified at the early phases
of the fermentation include Micrococcus varians, yeasts and moulds.
Khieokhachee et al. (1997) reported a total non-lactic acid bacterial
count of 106 cfu g−1 and a yeast count of 103 cfu g−1 during the initial
16 h of nham fermentation. Wiriyacharee et al. (1990) suggested that
Lb. plantarum and P. cerevisiae are important for acid production while
M. varians produces nitrate and nitrite reductases that are important
in the conversion of nitrate to nitrosylhemochrome, imparting a pink
colour to the product. Krairojananan et al. (1997) showed that most of
316
Figure 8.2 Preparation of nham. (a) main ingredients (from left to right): cooked pork skin in small strips, lean pork meat, rice, garlic, and chilli; (b) grinding pork meat
with seasonings through 2 mm holes: (c) grinding cooked rice with garlic through 2 mm holes; (d, e, f, g) adding to pork meat and mixing in order, seasonings, cooked rice and
garlic mix, shredded cooked pork skin, and chillies; (h) stuffing mixture into air-impermeable plastic tubing casing; (i) sealing end of sausage; (j) finished sausages ready for
incubation at room temperature (25–30°C) for 2–4 days. (Courtesy of W. Visessanguan.)
the lactic acid bacteria and other aerobic bacteria isolated from nham
were hydrogen peroxide producers, which caused discolouration and a
rancid off odour due to the oxidising properties of hydrogen peroxide.
Catalase production is, therefore, a desirable trait for bacterial strains
used in nham fermentations.
The use of molecular typing has greatly increased the ability to
discriminate between closely related bacterial strains (Valyasevi et al.
1997; Urbach et al. 1998). Based on this methodology, about 100
bacterial isolates obtained at 12 h intervals from nham fermented for
84 h, were differentiated and grouped on the basis of their genetic
similarities (Valyasevi et al. 1999). At 12 h of fermentation, a total
of eight different genetic groups were identified. With increasing
fermentation time, the number of genetic groups decreased, while
new genetic groups emerged. It is interesting to note that lactobacilli
strains increased to more than 50% of the population at 36 h prior to
declining in number. The highest rate of decline in pH occurred dur-
ing the initial 36 h (from an initial pH 6.2–4.9) and it then declined
slowly, reaching pH 4.6 at 84 h.
Different commercial brands of nham contained a variety of dif-
ferent lactobacilli, including Lb. acidophilus, Lb. cellobiosus, Lb. planta-
rum, Lb. pentosus, Lb. curvatus, Lb. sakei, Lb. delbrueckii, Lb. paracasei
and Lb. brevis (Valyasevi et al. 1999). Kunawasen (2000) used both
phenotypic and randomly amplified polymorphism DNA or RAPD
to identify lactic acid bacterial strains present during commercial
nham fermentations and showed that the dominant genetic groups
were lactobacilli, including Lb. acidophilus, Lb. cellobiosus, Lb. grami-
nis, Lb. plantarum, Lb. pentosus, Lb. curvatus, Lb. sakei, Lb. delbruck-
eii, Lb. paracasei and Lb. brevis, while Leuconostoc mesenteroides and
Pediococcus pentoaceus were present in much lower proportions.
In order to investigate the role of Lb. plantarum BCC 9546 dur-
ing nham fermentation, Laxananil et al. (2009) developed a recombi-
nant strain resistant to erythromycin and emitting green fluorescence.
When this strain was used as a starter culture for nham fermentation,
the numbers increased 10-fold during the first 12 h of fermentation,
reaching 107–108 cfu g−1 after 24 h, and then declining after 60 h and
reaching 105 cfu g−1 at 168 h.
Plengvidhya et al. (2008) used the combination of two powerful
PCR-based detection systems, intergenic transcribe spacer (ITS)-PCR
and repetitive extragenic palindrome (REP)-PCR, to study the

dynamics of the species responsible for the fermentation. They showed
that nham fermentation is more complex than previously reported
and isolated several additional lactic acid bacteria, such as Weissella
sp., Weissella pseudomesenteroides, Lactococcus lactis, Lactococcus garvieae,
Leuconostoc fallax and Leuconostoc citreum. They suggested that nham
fermentation could be divided into three stages of microbial succes-
sion: (1) an initial stage during 0–12 h incubation which involved
mainly the spoilage microorganisms of various species and Lactococcus
garvieae; (2) a second stage from 12–24 h of fermentation dominated
by Lactococcus lactis and; 3) a third stage from 24–72 h when Lb. plan-
tarum and P. pentosaceus were the predominant organisms and a fully
fermented and stable product with a final pH ~4.6 was obtained.
8.1.3 Characteristics and Roles of the Substrates
8.1.3.1 Minced Pork and Cooked Pork Rind Minced pork and cooked
pork rind are the major ingredients and comprise over 85% of the raw
mix. The ratio of minced pork to rind varies depending on the formu-
lation. Nham formulated with a high proportion of minced pork had
higher sensory scores for overall preference and texture likings than
formulations with lower proportions, but no significant differences
were observed in flavour and sourness likings (Visessanguan et al.
2005). Based on consumer preferences, the optimal minced pork to
rind ratio was 6:4. Lowering the amount of minced pork, particularly
to a ratio of 4:6, adversely affected the sensory texture of nham. The
quality of the minced pork used is also important for the quality of
nham. To minimise excess water affecting drip loss in nham, dark,
firm and dry meat is recommended. Lean pork meat trimmed of all
visible fat and connective tissues is recommended.
As important sources of protein (about 60% of nham dry weight),
the minced pork and cooked pork rind are mainly responsible for the
unique characteristics of nham and particularly for its texture and
colour. The most important component of meat is muscle in which
the myofibrillar proteins form about 60% of the total muscle protein
(Xiong 1997). Raw meat and cooked pork rind proteins exhibit a wide
range of functional properties. They are able to form networks and
structures, interact with other ingredients and thus play an important
role in the textural, sensory and nutritional quality of nham. Collagen

is the main component of the skin. Collagen and other connective tis-
sue proteins play a negligible role in gel formation in sausage batters.
However, they are typically included in formulations to improve water
binding, processing yield, juiciness and palatability, and to reduce
ingredient costs in processed meats (Osburn et al. 1997).
Owing to their different functional properties, varying the propor-
tions of minced pork and pork rind affects the chemical, physical and
sensory properties of nham as it plays an important role in quality
variation. As product costs are largely based on the quantity of meat
protein in formulations, increasing the proportion of rind in the for-
mulation can reduce the production cost. An increase in the propor-
tion of cooked pork rind resulted in higher moisture content, lipid and
initial pH value of nham but no significant effects on the character-
istics of the fermentation were observed (Visessanguan et al. 2005).
At the end of fermentation, nham with a high proportion of meat
exhibited higher force, hardness and cohesiveness than nham with
lower amounts of meat. However, incorporation of higher proportions
of cooked pork rind improved water-binding properties, leading to
decreased weight loss and less released water. Up to 43% pork rind
could be used in the formulation with no adverse effect on the textural
and sensory qualities of nham.
8.1.3.2 Cooked Rice Either glutinous or non-glutinous cooked rice

may be used in nham. Wiriyacharee et al. (1993a) noted that a nham
formulation with 3% cooked non-glutinous rice and 1% cooked gluti-
nous rice fermented at 30°C for 48 h underwent a rapid pH reduction
to 4.3 and increase in lactic acid to 1.1% (w/w). Wiriyacharee et al.
(1993a) claimed that cooked rice is a suitable source of carbohydrate
for nham production when using starter cultures, Lb. plantarum, P.
cerevisiae and Micrococcus varians. However, most lactic acid bacteria
are unable to hydrolyse starch and, although it is commonly assumed
to be the main carbohydrate source (Lee 1997), it has not been clearly
established that rice starch serves as a fermentable carbohydrate source
in nham fermentations.
8.1.3.3 Garlic Garlic is usually included in nham but the amount dif-

fers in different formulations. Garlic contains (g [kg fresh weight edible
portion]-1) inulintype fructans, 225-300; fructo-oligosaccharides,

1.0-2.0 (Judprasong et al. 2011); sucrose, 9.6–11; fructose, 1.2–2.4;
and glucose, 0.8–1.1 (Losso and Nakai 1997; Judprasong et al. 2011).
Baumgartner et al. (2000) characterised a high molecular weight fructan
from garlic belonging to the neoketose family. It has a (2 → 1)-linked
β-d-fructose backbone with (2 → 6)-linked β-d-fructose side chains.
Garlic enhanced both the growth of lactic acid bacteria and lactic acid
production in nham and formulations containing 5% garlic had higher
lactic acid contents than formulations without garlic (Swetwiwathana
et al. 1999), suggesting that garlic may serve as a source of fermentable
carbohydrate in nham fermentation. This would be in accordance with
the observation that garlic fructans are the major source of fermentable
carbohydrate utilised in some fish–rice–garlic fermentations (Palludan-
Müller et al. 2002).
In addition to its possible role as a source of fermentable carbohy-
drate, garlic possesses antimicrobial activities that could play a role as
a natural preservative to restrict the growth of spoilage microbes and
of food-borne pathogens, such as Staphylococcus aureus, Escherichia
coli, salmonellae and Listeria monocytogenes. Of these, Kumar and
Berwal (1998) found that Escherichia coli was the most sensitive and
L. monocytogenes was the least sensitive to garlic extracts. However,
to fully eliminate Salmonella sp. in the product, a combination of
8% garlic and Lb. plantarum 509 as a starter culture was necessary
(Bernbom et al. 2009). Corzo-Martínez et al. (2007) suggested that
various sulphur compounds are the main active antimicrobial agents
in garlic.
Several components of garlic and garlic extracts, including allin,
diallyl sulphide, allyl sulphide and propyl sulphide, have antioxidant
activity and are responsible for the antioxidant properties of garlic
in a dose-dependent manner (Yang et al. 1993). Sallam et al. (2004)
suggested that fresh garlic and garlic powder, through their combined
antioxidant and antimicrobial effects, are potentially useful in pre-
serving meat products.
8.1.3.4 Sodium Chloride Salt is usually added to the nham mix at ~2%
(w/w). At this level, NaCl exerts a partial bacteriostatic action, con-
tributes to an initial reduction in water activity to 0.96, improves pro-
tein solubilisation and water-holding capacity and imparts a typical
L ac ti c M e at F erm en tati o n 3 21
salty taste (Paterson et al. 1988). If the reduction in water activity was
entirely due to NaCl, a water activity of 0.96 would correspond to a
NaCl concentration of 7% (w/w) in the water phase. In fermented
meat processing, where 2–3% salt is typically incorporated into the
product, muscle fibres and proteins undergo major structural changes
due to electrostatic interactions between proteins and both sodium
and chloride ions. The meat proteins, actin and myosin, can dissolve
in the presence of low concentrations of salt and this accounts for the
increased solubilisation of proteins found in whole muscles treated
with salt. Salt additionally results in protein conformational changes,
probably by altering hydrophobic and electrostatic interactions that
stabilise the protein structure. These changes collaborate in the bind-
ing and retention of water inside the tissues (Pighin et al. 2008).
8.1.3.5 Nitrite and Nitrate Compounds The effects of nitrite in cured

products have been studied for many years and can be summarised
as follows: formation of the typical cured meat colour; inhibition of
the growth of spoilage and pathogenic bacteria, including Clostridium
botulinum; contribution to the development of the typical cured
meat flavour; and delaying the onset of oxidative rancidity (Flores
and Toldrá 1993). Development of the typical pink colour is a conse-
quence of the conversion of nitrite to nitric oxide (NO) and its reac-
tion with muscle myoglobin to form pink nitrosomyoglobin (Adams
and Moss 2000). Nitrate may also be used as a source of nitric oxide
and, due to the slower rate of degradation of nitrate compared to
nitrite, it is often used to extend the stability of cured meat colour.
Nitrate is reduced to nitrite by bacteria having nitrate reductase activ-
ity (e.g., Micrococcaceae). These bacteria may be naturally present in
the meat or be added as starter culture (Martin 2001). Nitrite is a
strong oxidant and reacts with endogenous or added reductants. Thus,
the addition of reductants, such as ascorbic acid, can be used to reduce
the formation of nitric oxide and nitrosamines in cured meats (Adams
and Moss 2000).
Residual nitrite levels in foodstuffs are important, partly because of
their potential to react with secondary amines to form carcinogenic
nitrosamines (Honikel 2008), and partly because of their contribution
as a source of nitrite in human nutrition (Lee et al. 1976). Therefore,
many countries, including Thailand, have regulations limiting the
residual amount of nitrate and nitrite permitted in cured meats. The

maximum allowable concentrations of nitrite and nitrate in nham are
125 and 500 mg kg−1, respectively (TISI 2003). However, in practice
the concentration of nitrite can be reduced further by using lactic acid
bacterial starter cultures.
8.1.3.6 Erythrobate Sodium erythorbate is a food additive used pre-

dominantly in meats and poultry. It is structurally related to vitamin C
and is an antioxidant. When used in processed meats, it increases the
rate at which nitrate reduces to nitric oxide and thus promotes a faster
cure (Mancini et al. 2007). It is used in nham to aid the development
of pink–red colour and to help maintain the colour during storage. As
an antioxidant, it also helps maintain flavour stability and prevents the
formation of carcinogenic nitrosamines (Sindelar and Milkowski 2011).
8.1.3.7 Phosphate Compounds Sodium tripolyphosphate is widely used

in nham. Phosphates function by sequestering metal ions and pro-
moting the dissociation of the actomyosin complex, bringing about
an increase in water-holding capacity. Another important reason for
using phosphate is for their ability to increase meat pH and slow dis-
colouration by stabilising vitamin C. Several polyphosphates are used
in the meat industry and they differ from one another in their proper-
ties. Tripolyphosphate is soluble and dissolves rapidly but its effect on
water-holding capacity in meat is slow because of the need for conver-
sion (enzymic and acid hydrolysis) to pyrophosphate. For this reason,
the use of a mixture of polyphosphates with different chain lengths
is more effective (Dušek et al. 2003). Addition of phosphate not only
raises the pH of meat mixtures, causing a shift away from the pI of
myofibrillar proteins, but also increases the total negative charge on
the myofibrillar proteins (Pearson and Gillett 1996). Both actions
contribute to increased water-holding capacity.
Phosphate compounds also provide pH buffering capacity in meat
products. This helps to prevent an increase in pH during the early
stages of meat fermentation but may also extend the time before the
drop in pH occurs (Bacus 1984).
The effectiveness of phosphates as antimicrobial agents in meat
products depends on the type of phosphate, the amount used, the
specific food product and conditions under which they are used (Sofos
1986). Wager and Busta (1985) reported that pyrophosphate was more
inhibitory towards microorganisms than tripolyphosphate or longer
chain polyphosphates in sausages. Pohlman et al. (2002) claimed
that trisodium phosphate not only maintained the redness, but that
it could also reduce the number of E. coli, Salmonella spp., coliforms
and aerobic bacteria in ground beef. However, the mechanism(s) by
which phosphate or polyphosphates might exert antibacterial effects is
not at all clear. The amount of added phosphates in nham is limited to
3 g kg−1 (expressed as P2O5) due to concerns that excessive phosphate
intake might pose health risks (Dušek et al. 2003).
8.1.3.8 Other Ingredients Fresh bird chilli is usually included in

nham and sugar and monosodium glutamate are frequently added to
enhance sweetness and umami taste but the amounts of these ingre-
dients is quite variable in different nham formulations. Wiriyacharee
et al. (1993b) found that the inclusion of 0.05% of pepper and 1% of
minced bird chilli increased the production of lactic acid. Ingolf and
Skjelkvale (1982) suggested that spices increased the use of carbohy-
drates by lactic acid bacteria.
Another ingredient that some nham manufacturers have recently
started to include is ang kak or red rice to enhance the red colouration
(Rojsuntornkitti et al. 2010). However, this may introduce additional
risks as red rice may not have the same inhibitory effects towards
Clostridium botulinum that nitrite has.
8.1.4 Physico-Chemical Changes during Fermentation
8.1.4.1 Sugars Utilisation Initial total concentrations of sugars found

in nham by Piluk (2009) were (g kg−1 wet weight) as follows: glu-
cose, 13.2 ± 3.2; fructose, 8.5 ± 2.7; sucrose, 4.5 ± 1.2; and maltose,
1.0 ± 0.3. Glucose was present in rice starch and fructose in garlic
fructo-oligosaccharides. These sugars could serve as carbohydrate
substrates for lactic acid bacteria during nham fermentation. Valyasevi
et al. (2006) reported that glucose, fructose and sucrose were preferen-
tially utilised at higher rates than maltose during nham fermentation.
Valyasevi et al. (2006) showed that cooked rice and garlic were
the major sources of fermentable sugars for lactic acid production. In
natural fermentations, sucrose and rice were the main carbon sources
with sucrose being used up rapidly during the first 36 h and none was
detectable after 96 h of fermentation. Rice starch is utilised following
hydrolysis to maltose and glucose. In fermentations inoculated with
Lb. plantarum, fructans from garlic were hydrolysed to free fructose.
During 24–72 h of fermentation the concentration of free fructose
present was 15–130 times greater in the fermentation with Lb. plan-
tarum than in the natural fermentation. The ratio of free sugar fer-
mented to the increase in organic acid was about one, suggesting that
all the glucose and fructose were converted to lactic acid. At the end
of fermentation, when the pH was 4.6, free sugars were still detect-
able. It is noteworthy that Palludan-Müller et al. (2002) also observed
utilisation of garlic fructans by Lb. plantarum.
The average amounts of total glucose and fructose utilised in
nham inoculated with Lb. plantarum BCC 9546 after fermentation
at 30°C to pH 4.6 was estimated to be 12.7 ± 2.4 g kg−1 wet weight,
representing ~60% of the initial total glucose + fructose in the rice
starch and garlic fructans (Piluk 2009). The amount of monosac-
charides utilised in this study is similar to the amount of dextrose
required to achieve the final product pH in other fermented sau-
sages. It was noted that total glucose was depleted at a higher rate
and to a greater extent than total fructose during nham fermenta-
tion, suggesting that rice starch was hydrolysed to a greater extent
than garlic fructans.
8.1.4.2 Acidification of Nham The pH of nham decreased from an

initial value of ~6.2 to ~4.6 within 60–72 h in nham naturally
fermented or within 48 h in nham inoculated with starter culture
(Visessanguan et al. 2007). During fermentation, lactic acid bacteria
convert sugars primarily to lactic acid, which is the main compo-
nent responsible for the pH decrease (Plumed-Ferrer et al. 2008).
Of the organic acids detected in nham, lactic acid represented
80–90% of the total (Visessanguan et al. 2004). Other organic
acids detected in decreasing amounts were acetic acid, oxalic acid
and pyruvic acid. Fermentation with a mixed culture of Lb. planta-
rum and Debaryomyces hansenii formed the greatest amount of lac-
tic acid followed by single culture fermentation with Lb. plantarum
and uninoculated natural fermentation. Acetic acid was found at the
highest concentration in the defined mixed culture fermentation. In
addition to the effects noted below, acidification has a preservative

effect due to inhibition of spoilage and pathogenic bacteria (Bover-
Cid et al. 2001a).
8.1.4.2.1 Acid-Induced Changes in Proteins Development of nham

characteristics involves subtle changes associated with microbial
growth and acidification of meat. These include not only desirable
changes, but also undesirable changes in colour and drip formation
that is associated with loss in quality. Overall, meat acidification
and pH lowering has a more significant impact on quality attributes
related to texture and colour than on other variables associated with
changes in protein and lipids. Generally, changes related to composi-
tion and properties of proteins are of much greater significance than
those related to lipids.
The effect of pH on proteins in nham possibly includes acid dena-
turation and/or proteolysis of myofibrillar and sarcoplasmic protein
fractions. Although these fractions are less abundant than the alka-
line soluble fraction, acid-induced changes on myofibrillar and sarco-
plasmic protein fractions have a markedly greater impact on texture,
colour and drip than changes in the alkaline soluble proteins. In par-
ticular, the myofibrillar proteins were the most important in relation
to hardness, adhesiveness and resilience while sarcoplasmic proteins
were the more important in relation to springiness. With increasing
fermentation time, nham became more rigid, elastic, cohesive and
less adhesive (Visessanguan et al. 2004). As with other restructured
meat products, texture formation in nham probably involves changes
in both adhesion and cohesion, mainly induced by the slow lowering
of pH.
Visessanguan et al. (2007) suggested that myofibrillar and sarco-
plasmic proteins are also involved in colour. Pork meat colour is highly
correlated with precipitation of sarcoplasmic proteins. The precipi-
tated proteins mask the red colour of the sarcoplasm and the muscle
becomes pale (Goldspink and McLoughlin 1964). The pale colour of
pork muscle is related to the soluble sarcoplasmic protein concentra-
tion (Joo et al. 1999) in that the lightness of pork loin appeared to
decrease with increasing sarcoplasmic protein solubility. In addition,
myofibrillar proteins may also affect paleness of meat (Young and
West 2001).
8.1.4.2.2 Effects of Acid on Textural Development Texture formation

in nham is closely associated with fermentation as the mechanism of
binding of protein components in nham is an acid-initiated reaction
(Visessanguan et al. 2004). Slow decrease in pH gradually induced
aggregation of proteins, leading to the formation of an ordered pro-
tein structure and contributing to firmness (Fretheim et al. 1985).
Additionally, acid solubilisation of collagen is presumed to be involved
(Aktaş and Kaya 2001). The weak acids promote disruption of non-
covalent intermolecular bonds that reinforce the collagen fibril struc-
ture, enhancing swelling and decreasing the denaturation temperature.
Myofibrillar proteins play a critical role in meat texture as they are
responsible for the cohesive structure and the firm texture of meat
products. It is likely that acid-induced gelation of these proteins causes
the formation of nham texture (Visessanguan et al. 2004). Fermented
nham showed a marked increase in all texture attributes of texture pro-
file analysis (TPA) (Bourne 1978) compared to non-fermented nham
and with increasing fermentation time nham became more rigid, elas-
tic and cohesive and less adhesive (Visessanguan et al. 2004). Nham
inoculated with Lb. curvatus exhibited faster and higher development
of all TPA texture attributes compared to nham naturally fermented
(Visessanguan et al. 2006a).
8.1.4.3 Colour Development The red colour of uncooked cured meats

is largely due to nitrosomyoglobin, formed by the reaction of nitric
oxide with myoglobin. Nitrosomyoglobin is spontaneously formed
in nham during the mixing process and accounts for ~90% of total
heme pigment (Panya et al. 2003). Nitrosomyoglobin is unstable and
discolouration can be rapid, as was observed during the final stages
of slow-fermenting nham (Panya et al. 2003). The disappearance of
nitrosomyoglobin was accompanied by an increase in metmyoglobin
and it was hypothesised that nitrosomyoglobin was converted to the
more stable metmyoglobin. A rapid increase in peroxide value was
coincident with a substantial decrease in nitrosomyoglobin, suggest-
ing a possible causal relationship (Baron et al. 1998; Hornsey 1956),
and contributed to a loss of colour in nham during prolonged fermen-
tation. Increasing peroxide value, indicating increased lipid oxidation
during nham fermentation, was observed with all nham formula-
tions. An increase in metmyoglobin occurred concomitantly with
an undesirable increase in lipid oxidation (Faustman and Casssens

1990).
The characteristic colour of meat is a function of meat pigments
and light-scattering properties. Acidification of meat proteins affects
colour by increasing the light-scattering properties and, thus, meat
becomes more opaque and paler. The extent of light scattering is
also related to the structure of the muscle tissues. The shrinkage of
the myofilament lattice increases the light reflection from meat and
changes in colour may also result from precipitation of soluble pro-
teins of the sarcoplasm (McLoughlin and Goldspink 1963).
Visessanguan et al. (2004) reported that lightness (L*) of naturally
fermented nham decreased during the first 12 h and then continuously
increased to a maximum at 36 h before remaining relatively constant
thereafter. Redness (a*) increased during the first 12 h of fermenta-
tion and then did not change during the later stages of fermentation.
Yellowness (b*) decreased slightly during the first 24 h but was more
pronounced at 36 h. The colour of nham inoculated with Lactobacillus
curvatus during fermentation indicated an increase in both lightness
and redness but a decrease in yellowness. Inoculation with Lb. cur-
vatus accelerated both the increase in L* values and the decrease in
b* values during fermentation of nham (Visessanguan et al. 2006b).
Myoglobin and its derivatives are the most important compounds in
relation to the redness and yellowness of nham. Depending on pro-
cessing conditions, various forms of myoglobin can be interconverted
and this may affect the colour of meat (Hornsey 1956; Baron et al.
1998). An increase in whiteness can be due to changes in the light-
scattering property of proteins (Young and West 2001).
8.1.4.4 Drip Formation Drip formation in nham is mainly associ-

ated with loss in water and water-holding capacity of meat and can
be assayed using three parameters, weight loss, released water and
expressible water. Weight loss is mainly associated with loss in weight
during fermentation, while released water is generally referred to as the
water retained in the casing and at the surface. In contrast, expressible
water is the water released when pressure is applied. The weight of fer-
menting nham generally decreases as the fermentation time increases.
Weight loss, released water and expressible water of nham of different
formulations ranged from 0.5% to 3%, 1% to 5%, and 5% to 15% of the
wet weights, respectively (Visessanguan et al. 2006a). Visessanguan

et al. (2006a) observed that as the fermentation time proceeded, nham
inoculated with Lb. curvatus had greater weight loss than unioculated
nham. In addition, increasing the proportion of minced pork in nham
generally resulted in higher weight loss and water release during fer-
mentation (Visessanguan et al. 2005). Expressible water in nham was
positively influenced by total acidity, TCA-soluble peptides and con-
tent of alkaline soluble proteins. The results indicated that increase in
TCA-soluble peptides, an indication of the extent of proteolysis, was
associated with weight loss and water release.
8.1.4.5 Proteolysis Proteolysis is one of the most important bio-

chemical changes that occur during ripening of fermented sausages
(Hughes et al. 2002; Kaban 2009; Spaziani et al. 2009). It influences
both the texture and the development of flavour due to the forma-
tion of low molecular weight compounds, including peptides, amino
acids, aldehydes, organic acids and amines, which are important
flavour compounds or precursors of flavour compounds. It is a con-
sequence of the activities of both endogenous muscle enzymes and
microbial enzymes, and the relative contribution of each depends on
the product and conditions during ripening (Fernández et al. 2000).
Proteolysis of myofibrillar and sarcoplasmic proteins occurs in
nham during fermentation. Proteolysis is primarily mediated by
endogenous meat enzymes, such as calpains and cathepsins, but exog-
enous microbial enzymes from lactic acid bacteria may play a minor
role (Fadda et al. 1998, 1999a, b; Toldrá et al. 2001). Both Lb. cur-
vatus and Lb. sake display proteinase activity on sarcoplasmic pro-
teins. Skeletal muscle contains a large number of proteases. Among
these, the synergistic action of calpains and lysosomal cathepsins are
responsible for proteolysis at pH values below 6.0 (Kim et al. 1995).
The initial hydrolysis of muscle proteins has been attributed mainly
to endogenous cathepsin, followed by the action of bacterial enzymes
which actively degraded oligopeptides into small peptides and free
amino acids (Molly et al. 1997).
Visessanguan et al. (2004) reported that proteolysis of both myofi-
brillar and sarcoplasmic proteins occurred during nham fermentation.
Degradation of proteins resulted in an increase in peptides and free
amino acids, which may have contributed to the flavour and aroma of
nham. The level of starter culture inoculation affected the rate of pro-
teolysis. Mixtures inoculated with 106 cfu g−1 of Lb. curvatus showed
the fastest and largest decrease in both myofibrillar and sarcoplasmic
protein fractions, followed by mixtures inoculated with 104 cfu g−1
and nham naturally fermented. It is possible that both myofibrillar
and sarcoplasmic proteins were degraded or became insoluble due to
acid-induced denaturation (Visessanguan et al. 2006a).
Collagen is the main component of the skin. During nham fermen-
tation, most proteolytic enzymes have little activity against native col-
lagen, although they readily degrade denatured collagen (Visessanguan
et al. 2004). However, Berge et al. (2001) reported a direct effect of
lactic acid on collagen in causing a swelling of perimysium, a connec-
tive tissue surrounding bundles of muscle fibres.
8.1.4.6 Lipolysis and Lipid Oxidation Changes in lipid composition

and the fatty acids profile during nham fermentation were investi-
gated by Visessanguan et al. (2006b). Total lipids were in the range
of 2–3% of dry matter. The extracted lipid of the initial nham mix
consisted mainly of triacylglycerols, accounting for more than 75%
of the total lipid, followed by phospholipids and trace amounts of
diacylglycerols and free fatty acids. During fermentation, triacylg-
lycerols, diacylglycerols and phospholipids tended to decrease with
a corresponding increase in free fatty acids. Free fatty acids content
increased from 0.3% (w/dry wt) at the beginning to 3% of the total
lipids content at the end of fermentation, indicating lipolysis of lip-
ids. In both total and neutral lipid fractions, the major fatty acids
found, in quantitative order, were oleic (C18:1), linoleic (C18:2) and
palmitic (C16:0), which together accounted for 90% of the total fatty
acids. Increases in fatty acids in total and non-polar lipid fractions
were observed with a corresponding decrease in the quantity of fatty
acids of phospholipids during the fermentation. Nham inoculated
with 106 cfu g−1 of Lb. curvatus had higher concentrations of free fatty
acids than nham inoculated with 104 cfu g−1 or uninoculated nham
(Visessanguan et al. 2006a).
Lipolysis is believed to play a central role in aroma formation but
little information is available on the lipolytic activity of lactobacilli
during sausage fermentation. Lb. plantarum DSMZ 12028 iso-
lated from chourico, a traditional Portuguese dry fermented sausage,
exhibited extracellular lipase activity and the production of lipase

appeared to be significant under conditions relevant to sausage ripen-
ing (de Fátima Silva Lopes et al. 1999). However, lipases from many
lactobacilli often display little or no activity under conditions found in
fermented sausages (Demeyer et al. 2000).
Peroxide value was used to assess lipid oxidation in nham. Per
oxide values increased with increasing fermentation time and nham
inoculated with 106 cfu g−1 of Lb. curvatus exhibited a faster and larger
increase in peroxide value than nham inoculated with 104 cfu g−1 or
uninoculated nham (Visessanguan et al. 2006a). As the fermentation
proceeded, thiobarbituric acid-reactive substance (TBARS) values
decreased. Overall, the oxidation levels observed in nham was rela-
tively high. Nevertheless, the levels observed would not be sufficient
to cause objectionable odour or tastes in any of the nhams tested.
8.1.4.7 Volatile Compounds in Nham The flavour characteristics of

nham are thought to derive from a combination of added garlic,
microbial activities and auto-oxidation, with the relative importance
of each varying from formulation to formulation.
Rotsatchakul et al. (2009) used dynamic headspace sampling gas
chromatography mass spectrometry to identify volatile compounds
formed in nham naturally fermented by indigenous microorgan-
isms, in nham inoculated with Lactobacillus curvatus starter culture,
and in a non-fermented nham mix prepared with added antibiot-
ics to inhibit microbial growth. A total of 113 volatile compounds
were identified, including 9 hydrocarbons, 12 aldehydes, 9 ketones,
19 alcohols, 25 esters, 29 S-containing compounds, 2 acids, 1 terpene,
1 furan, 2 organic halogen-containing compounds and 4 miscella-
neous compounds. S-containing compounds were the most abundant
of the volatile compounds in all samples, presumably due to the use of
garlic in the formulations. A striking increase in new volatiles, mainly
esters, alcohols and S-containing compounds was observed with
increasing incubation time. Prominent aroma components identified
by their odour active values were ethyl butanoate (buttery, ripe fruit
note) and S-containing compounds from garlic. Other aroma impact
compounds were ethyl 2-methylbutanoate (fruity), ethyl 2-methylpro-
panoate (fruity), 3-hydroxy-2-butanone (yoghurt-like), octanal (fatty-
fruity) and hexanal (green). Ethyl butanoate, ethyl 2-methylbutanoate,
ethyl 2-methylpropanoate and octanal were found only in the micro-

bially fermented samples, whereas hexanal and 3-hydroxy-2-butanone
were found in the non-fermented sample.
The intensities of ‘fermented’, ‘acidic’, ‘garlic’ and ‘overall fla-
vour of nham’ aromas, evaluated by 10 trained panelists, increased
sharply on the first day. Acidic, garlic and overall flavour of nham
aroma intensities then remained almost constant. However, the inten-
sity of fermented aroma in Lb. curvatus nham continued to increase
on the second day. This was concomitant with an increase in ethyl
butanoate (odour active value, 9072). In comparison to the naturally
fermented sample, Lb. curvatus fermented nham exhibited a faster
reduction in pH, a higher fermented aroma on the second day and
higher concentrations of ethyl butanoate and 3-hydroxy-2-butanone.
The other potent aroma compounds in nham were dimethyldisulfide
(strong onion, cabbage like) and dimethyltrisulfide (onion-garlic like).
S-containing compounds generally have low threshold values and it
can be extrapolated that the other S-containing compounds detected
in nham may also play an important role in nham aroma. An earlier
study by Valyasevi et al. (2006) suggested that major volatile com-
pounds formed in nham were di-2-propenyl disulfide, methyl thiirane
and methyl-2-propenyl disulfide, but none of these were detected by
Rotsatchakul et al. (2009) in their samples.
The most noticeably increased esters in nham were methyl butano-
ate (sweet, ethereal fruity, odour threshold, 60–76 ppb), ethyl acetate
(ethereal, sharp, wine-brandy like, threshold, 5 ppb) and methyl hex-
anoate (methyl caproate, ethereal fruity, pineapple-apple, threshold,
70 ppb). The naturally fermented nham had a higher level of total
esters than the Lb. curvatus fermented sample. This is presum-
ably because the indigenous microorganisms had higher esterase
activity than L. curvatus. The Lb. curvatus nham contained much
more ethyl butanoate (ethereal, fruity, buttery and ripe fruit notes,
threshold, 1 ppm) but less methyl butanoate than the naturally fer-
mented nham. At the end of the fermentation, the concentrations
of ethyl caproate, ethyl acetate, ethyl 2-hydroxypropanoate, ethyl
2-methylpropanoate and ethyl 2-methylbutanoate in Lb. curvatus
nham were higher than those in the naturally fermented sample.
The amount of esters in the non-fermented sample was low and
remained unchanged throughout the fermentation. Stahnke (1994)
showed that ethyl esters, associated with fruity aromas, can mask
rancid odours in fermented sausage.
8.1.4.8 Biogenic Amine Formation Biogenic amines in sausages are

of concern due to their toxicological effects on the nervous system,
blood pressure and the intestinal system, with typical symptoms of
biogenic amine poisoning including migraine, headache and increased
blood pressure (Joosten 1988). Histamine, putrescine, cadaverine,
tyramine, tryptamine, β-phenylethyl amine, spermine and spermi-
dine are considered to be the most important biogenic amines in foods
(Shalaby 1996). In addition to their possible direct toxigenic effects,
some biogenic amines (particularly putrescine and cadaverine) may
react with nitrite to form potentially carcinogenic nitrosamines (Silla
Santos 1996). These problems may be more severe in consumers having
reduced mono- and diamine oxidase activity, the enzymes responsible
for biogenic amines detoxification (Bardócz 1995).
The major biogenic amines found in nham by Tosukhowong et al.
(2011) were tyramine, cadaverine, putrescine, spermidine and sperm-
ine. Cadaverine was detected in stored pork meat under all storage
conditions tested but was undetectable in freshly slaughtered meat
while tyramine was present in pork meat stored at 4°C and above
and spermine was present only in pork meat stored at 30°C for 6 h.
No other biogenic amines (putrescine, histamine, phenylethylamine
or tryptamine) were detected in any meat samples. Spermine and
spermidine are naturally occurring polyamines in fresh meat at levels
of 20–40 and <5 mg/kg−1, respectively. Their concentrations in pork
meat remain nearly constant or tend to decrease with storage.
Formation of biogenic amines during the fermentation period was
not observed in nham manufactured from fresh pork meat, but there
was a marked increase in cadaverine, putrescine, tyramine and his-
tamine in nham processed from stored meat. This phenomenon has
been observed in the production of other types of sausages (Bover-
Cid et al. 2000) and emphasises the role that the hygienic status
of the pork meat has on biogenic amines accumulation in nham.
Cadaverine was the major biogenic amine produced in nham fer-
mented from stored pork meat. Taken together with the presence of
Enterobacteriaceae during the early period of nham fermentation,
this suggests that cadaverine formation may be related to the presence
of lysine decarboxylase-positive Enterobacteriaceae. The slow devel-

opment of lactic acid bacteria during the onset of fermentation may
have allowed these bacteria to grow and produce lysine decarboxylase,
which is responsible for cadaverine formation and which can continue
to function in the absence of viable cells (Bover-Cid et al. 2000).
The formation of putrescine was observed only in nham processed
from pork meat stored at elevated temperatures, even for only a short
period. This suggests that putrescine might be considered as an indi-
cator of the use of poorly stored meat in fermented sausages. Similarly,
Ruiz-Capillas and Jiménez-Colmenero (2004) reported that putres-
cine formed in different meat products reflected the microbial popu-
lations of the raw materials used and the hygienic conditions of the
manufacturing process. However, not all putrescine originates from
spoilage bacteria as Arena and Manca de Nadra (2001) described
putrescine production by Lactobacillus hilgardii X1B.
The formation of tyramine seems to be related to the presence of
lactic acid bacteria and enterococci. Strains of lactobacilli belonging
to the species Lb. brevis and Lb. curvatus have been shown to produce
tyrosine decarboxylase (Lucas et al. 2003). Tyramine is quantitatively
the most important biogenic amine produced in fermented meats
(Bover-Cid et al. 2008; Latorre-Moratalla et al. 2008). The toxic
level of tyramine is 100–800 mg kg−1 (Silla Santos 1996) and, there-
fore, the tyramine concentrations (2.7–125 mg kg−1) found in nham
samples do not represent an obvious hazard (Valyasevi et al. 2008).
However, McCabe (1986) reported that the consumption of 6 mg of
tyramine can produce a slight reaction, whereas 10–25 mg can cause
severe headache and hypertensive crisis in patients under treatment
with monoamine oxidase inhibitor drugs. Thus, there may be a risk
that the tyramine levels observed in nham could provoke some symp-
toms in such individuals consuming more than 100 g of sausage.
Histamine content of nham varied considerably in samples pro-
cessed from different types of stored pork meat, but it was unde-
tectable in nham prepared from fresh meat. Histamine is the only
biogenic amine subject to legal regulations in some fish species, with
an upper limit of 100 mg kg−1 in Europe (European Commission
2005). However, there are no regulations regarding the amounts of
biogenic amines permitted in fermented sausages. It has been sug-
gested that 100 mg kg−1 of histamine be the upper limit for potential
risk to healthy individuals (Brink et al. 1990). Levels of 0–41 mg kg−1

of histamine in nham were below this limit.
It is possible that biogenic amine formation in nham could be
restricted by use of appropriate starter cultures known to lack amino
acid decarboxylases, such as Lb. plantarum BCC 9546 (Tosukhowong
et al. 2011). Moreover, the use of a mixed starter culture of Lb. plan-
tarum BCC 9546 and tyraminogenic Lb. brevis BCC 26756 yielded
nham with a lower tyramine content than nham inoculated solely
with Lb. brevis BCC 26756. These results indicate that addition of
a starter culture having functional properties could be an important
means of reducing biogenic amine formation during nham fermenta-
tion. Bover-Cid et al. (2001b) and González-Fernández et al. (2003)
also pointed out that achieving a rapid pH decrease with amino acid
decarboxylase-negative starter cultures would limit the growth of
decarboxylase-positive bacteria and result in lower biogenic amine
accumulation in sausages.
8.1.5 Safety Considerations and Recommendations
Like many, if not all, traditional food fermentations, nham is pro-

cessed under non-sterile conditions and is thus exposed to the risk
of contamination by undesirable microorganisms, such as Salmonella
spp., Staphylococcus aureus and Listeria monocytogenes, and such
organisms are frequently observed in nham and especially in nham
with a pH higher than 4.6 (Paukatong and Kunawasen 2001). There
is also great variation in the hygienic practices under which tradi-
tionally fermented nham is produced. The safety of nham depends
largely on careful selection of raw ingredients, especially of the raw
pork, safe manufacturing practices, proper fermentation and correct
handling and cooking by the consumer. Since raw pork may con-
tain parasites and harmful pathogens, cooking prior to consump-
tion is usually recommended. However, this is not common practice
among nham consumers since they prefer the flavour and texture of
uncooked nham.
8.1.5.1 Physical Hazards Although this kind of hazard is uncommon,

there is always a risk during processing that pieces of broken knife,
blade, metal packaging clips or other undesirable physical entities
enter the product. This hazard is usually controlled by proper inspec-

tion and maintenance of equipment and, with large manufacturers,
the installation of a metal detector at the final product inspection.
8.1.5.2 Chemical Hazards The risk is contamination of the product

by hazardous chemicals, either added intentionally or accidently or
produced during the fermentation. Additives, especially nitrate and
nitrite, usually added in the form of sodium or potassium salts, is one
of the major concerns in cured meat products, including nham. They
are added to inhibit the growth of spoilage bacteria and pathogens
and, especially, to inhibit growth and toxin production by Clostridium
botulinum, and to generate the red colour of cured meats. Added
nitrate is reduced to nitrite by nitrate-reducing microorganisms
present in the raw materials or added in the starter culture. Nitrite
is unstable, especially in acidic conditions, and further breaks down
to compounds, such as nitrous acid (HNO2), which has antibacte-
rial activity, and nitric oxide (NO), which reacts with myoglobin in
meat to form nitrosomyoglobin that contributes to the desirable red
colour in the final product (Møller and Skibsted 2002). However, at
high temperatures or under acidic conditions, nitrite can also react
with secondary amines in the product to produce nitrosamines. These
n-nitroso compounds are potent carcinogens in primate animal mod-
els and cause a variety of tumours (Bogovski and Bogovski 1981). In
humans, Paik et al. (2001) suggested a link between changes in meat
curing practices and a dramatic decline in gastric cancer incidence
and mortality in the United States in the 1920s. Prior to 1923, the
amounts of nitrite in US cured meat products was extremely high
and variable, up to 1400 mg kg−1 in frankfurters and 960 mg kg−1 in
ham with average levels of 185 mg kg−1. Since 1925, when the United
States Department of Agriculture authorised the use of sodium nitrite
in cured meats at a maximum level of 200 mg kg−1, the level of nitrite
in cured meats was significantly reduced. By 1937, the average content
of detectable nitrite in cured meats had decreased to 57 mg kg−1, a
69% decrease (Paik et al. 2001).
Exposure to high concentrations of nitrite may lead to acute symp-
toms of methemoglobinemia in which nitrite entering the blood
stream interacts with deoxyhemoglobin and changes it to methemo-
globin, which lacks oxygen-binding capacity. The symptoms, related
to the body’s lack of oxygen, include pale and purple lips and body
(cyanosis), metabolic acidosis, difficulty breathing, hypotension and
shock, and in the worst case brain damage and death may occur.
Incidences of nitrite poisoning in Thailand are usually due to acci-
dental use of curing salt with high nitrite content by an inexperienced
producer. In May–June 2007, a total of 13 children were hospitalised
in Ayuthaya Province with symptoms of nitrite poisoning, with pale
and purple lips, after eating sausages without the Thai-FDA authori-
sation seal on the package and that had been produced by an unau-
thorised producer. Samples of the sausage were analysed and found
to contain very high levels of nitrite, up to 3140–3540 mg kg−1. The
product was recalled and the producer was suspended from production
(Hongchumphon 2007; Noimoh and Areechokchai 2007). Another
case of nitrite poisoning in 2007 involved 24 cases in Chiang Rai
Province and was due to fried chicken made in a cooking class in
which a high percentage sodium nitrite mixture was mistaken for a
low percentage mixture (Hongchumphon 2007).
A survey during 2005–2006 of 156 samples of commercial nham
from 29 brands produced by 18 manufacturers in 7 provinces, all
with the appropriate FDA seal, showed that these samples contained
acceptable levels of curing agents, in the range of 0–110 mg kg−1
for nitrate and 0–65 mg kg−1 for nitrite (Valyasevi et al. 2008). The
Thailand Industrial Standards Institute specifies that the maximum
permissible concentrations in nham are 500 mg kg−1 for nitrate and
125 mg kg−1 for nitrite (expressed as sodium salt; Nham standard,
TISI 1219-2547, 2003). Thus, the survey samples had well below the
maximum permitted levels. Possible risks due to nitrosamines and
other amines in nham were assessed in the same survey. Nitrosamines
were not detected in any sample and biogenic amines also did not
exceed levels considered hazardous in any sample (Valyasevi et al.
2008).
8.1.5.3 Biological Hazards Since nham is made from raw meat and
is usually consumed without cooking, risks due to biological haz-
ards such as parasites and bacterial pathogens are of great concern.
Although heating is known to be an efficient and economically effec-
tive way to get rid of most of the biological hazards, a significant por-
tion of consumers still prefer the uncooked product.
8.1.5.3.1 Parasites A nematode parasite, Trichinella spiralis, is the

major concern in nham since many farm and wild animals, especially
pigs, are the main reservoirs of this parasite. The disease, caused by
Trichinella sp., is referred to as trichinellosis or trichinosis. The symp-
toms of the parasitic infection are related to migration of the parasite
from the intestinal epithelium into the muscles of the infected host.
The first symptoms appear between 12 h and 2 days after ingestion of
infected meat. The symptoms may include nausea, vomiting, sweat-
ing, diarrhoea, facial edema and fever. At a later stage of infection,
intense muscular pain, difficulty in breathing, weakening of pulse
and blood pressure, heart damage, and various nervous disorders may
occur. Death may occur due to heart failure, respiratory complica-
tions, or kidney malfunction. Treatments usually aim to alleviate
symptoms. Complete eradication of the parasite from the patient’s
body is uncertain since drugs may not reach their target embedded in
the host’s muscles.
The incidence of, and deaths due to, trichinosis in Thailand are in
steady decline following changes in the pork production industry, from
the use of free-roaming swine, earth-based enclosures, and untreated
food garbage as feed by small-scale producers, to large-scale produc-
tion units with better control of the hazards through good production
practices. Surveillance reports of trichinosis in Thailand (Chuxnum
2012) showed a median rate of infection of ~0.08 per 100,000 popu-
lation during 2003–2012. Most of the cases were clustered in north-
east Thailand and were traced to consumption of raw or under-cooked
wild or free-roaming pigs or other wild animals.
Earlier incidences of trichinosis prompted an investigation of the
pork supply in northern Thailand in 2009. A survey found no parasite
in any of 156 pork samples collected from standard slaughter houses in
five provinces in the north of Thailand, indicating that the pork sup-
ply was parasite-free in these main markets (Srichareuan et al. 2009).
The source of pork used in the production of nham is crucial since
the salt content and fermentation time used in nham production are
not sufficient to completely destroy encysted parasites in pork muscle.
Several methods for treating Trichinella sp. infected pork meat and
products are described in the U.S. Code of Federal Regulations
(2012). Treatments include specific cooking temperatures and times,
freezing temperatures and times and curing methods (Table 8.1).
Table 8.1 Freezing Times Required to Destroy Trichinella spp. in Meat of Differing Thicknesses
TEMPERATURE THICKNESS OF THE MEATa
FAHRENHEIT CELSIUS ≤15 CM 15–68 CM
5 −15 20 days 30 days
−10 −23.3 10 days 20 days
−20 −28.9 6 days 12 days
Source: Modified from the U.S. Code of Federal Regulation number 318.10. 2012. Prescribed
Treatment of Pork and Products Containing Pork to Destroy Trichinae. Washington: U.S.
Government Printing Office.
a Thickness of the layers of meat or products arranged either in separate pieces or in containers and
spaced to ensure free circulation of air between layers.
In the case of nham, freezing is probably the most appropriate

method. Freezing may be done on the pork meat before processing,
after stuffing, or after completion of fermentation. Freezing can also
help the producer to devise a production plan to maintain the sup-
ply of nham. Currently, production of nham relies on a supply of
fresh pork meat, which is unavailable on certain religious days, and is
made-to-order as it is considered perishable and over-fermentation is
undesirable. Since well-controlled freezing technology is increasingly
available and affordable, a freezing regime (Table 8.2) may be the best
approach to ensuring the absence of trichinellae in nham.
8.1.5.3.2 Bacterial Pathogens Their high water, nutrient and min-

eral content make raw meats a superb habitat for bacterial growth and
hygienic post-mortem handling is crucial for safety of meat products.
Table 8.2 Freezing Times Required to Destroy Trichinella spp.
MAXIMUM INTERNAL TEMPERATURE
FAHRENHEIT CELSIUS MINIMUM TIME (H)
0 −17.8 106
−5 −20.6 82
−10 −23.3 63
−15 −26.1 48
−20 −28.9 35
−25 −31.7 22
−30 −34.5 8
−35 −37.2 0.5
Source: Modified from the U.S. Code of Federal Regulation number 318.10.
2012. Prescribed Treatment of Pork and Products Containing Pork to
Destroy Trichinae. Washington: U.S. Government Printing Office.
In tropical countries, such as Thailand, this is a challenging task.

High humidity and high temperatures support rapid growth and per-
sistence of bacteria, both in meat products and in the environment.
Obviously, good hygienic standards in the slaughter house and refrig-
eration of meats during transportation are required to minimise bac-
terial contamination and maintain populations at a low level.
Although nham is sometimes regarded as a product similar to
European fermented sausages, it has several unique aspects. The
major difference is the very low fat content of nham compared with
European fermented meats. This is a consequence of trimming any
visible fat from the meat and not adding any fat to the mixture. The
nham mixture is then fermented in air-impermeable packaging, cre-
ating anaerobic conditions throughout, in contrast to other type of
sausages which are cased in air permeable materials such as intestines
or synthetic casings in which aerobic bacteria may grow and produce
toxin on the aerobic surface. Additionally, nham is ready to eat after
3–5 days of fermentation without any of the drying, smoking or rip-
ening processes usually followed with European fermented sausages.
These properties set nham apart from other types of fermented sau-
sages and extrapolation of biological hazards reported in other types
of sausages to nham should be undertaken with care.
Several surveys of nham have revealed the presence of con-
tamination by a variety of bacterial pathogens. Salmonella spp. and
Staphylococcus aureus were the most commonly found pathogens in
nham with Salmonella spp. being detected in 12–23% of samples
and S. aureus in 15–26% (Paukatong et al. 1999; Wongsommart
et al. 1994). Paukatong et al. (1999) also reported Listeria monocyto-
genes contamination in 12% of 60 nham samples examined. Neither
Escherichia coli O157:H7 nor Yersinia enterocolitica were found in
this study. However, 83% of 80 nham samples collected from the
north, north-east and central Thailand were found to harbour faecal
coliform bacteria and 16% were positive for Clostridium perfringens
(Wongsommart et al. 1994). Wiboonchat (2003) also detected C.
perfringens in 20% of 30 commercial nham samples but none were
positive for Campylobacter jejuni.
One of the reasons for the high incidence of bacterial pathogens
in nham is that, with the exception of product sold in temperature-
controlled cabinets in supermarkets, nham is frequently sold to the
retailer by the manufacturer before the fermentation is fully complete.

In fact, shipping immediately after stuffing without any fermentation
step is a common practice for nham sold in fresh produce markets. The
reason for this practice is to allow the fermentation to occur during
transportation and shelving at ambient temperature, thus obviating
the need for refrigeration and saving costs and time. However, ambi-
ent temperatures vary widely, from 20–25°C in winter to 30–35°C in
summer, and the practice increases the risk of incompletely fermented
product being consumed, since it relies on the consumer to determine
when the product is ready to be eaten. We suspect that a significant
part of the pathogen-contaminated nham reported earlier is associ-
ated with the sale of incompletely fermented product. According to
the Thailand Industrial Standards Institute, nham should have a pH
value of no higher than 4.6 before consumption. Paukatong et al.
(1999) reported that 32% of 60 nham samples randomly collected
from retailers had incomplete fermentations and pH values higher
than 4.6, and all the samples contaminated with Salmonella spp., S.
aureus and L. monocytogenes were products in this group. In another
study (Chokesajjawatee et al. 2009), 61% of 155 nham samples had
pH >4.6 and high pH value was found to be one of the main predic-
tors for incidence of S. aureus in nham, with the probability of nham
with pH higher than 4.6 containing S. aureus being 0.76.
S. aureus is ubiquitous in the environment and some level of con-
tamination is probably unavoidable. In nham, the numbers of S.
aureus increased slightly during the first day of fermentation. This is
not surprising since the pH of nham in the initial phase of fermenta-
tion is relatively high and within the range for growth of S. aureus and
other potential spoilage bacteria. Additionally, S. aureus is relatively
salt tolerant and is able to grow at the water phase salt concentrations
in nham. However, at later stages of the fermentation, the number
of S. aureus decreased with the decrease in pH. The numbers of the
pathogen continued to reduce after the pH reached 4.6 and after the
fermentation was stopped by refrigeration at 4°C (Chokesajjawatee
et al. 2009). Paukatong et al. (1999) made similar observations on
nham inoculated with 104 –106 cfu S. aureus g−1 and noted that even
after 7 days of fermentation and a final pH lower than 4.5, nham
inoculated with S. aureus at levels higher than 104 cfu g−1 still tested
positive for S. aureus.
Paukatong et al. (1999) also studied the fate of different ini-
tial numbers of Listeria monocytogenes added to nham and observed
that, while the numbers reduced during the fermentation, surviving
L. monocytogenes remained detectable after 7 days of fermentation when
the initial contamination was >103 cfu g−1.
The inhibition of the growth of spoilage and pathogenic bacteria
in nham depends on the rapid creation of acidic conditions and the
production of organic acids. The toxicity of organic acids depends on
the concentration of the uncharged acid, which is a function of the
pH value and the pKa of the specific acid. The major organic acid in
nham is lactic acid (pKa, 3.86) with smaller amounts of acetic acid
(pKa, 4.78). Hence, at pH 4.6, near the final pH of nham, most (84%)
of the lactic acid is present as non-toxic lactate ion while ~59% of
acetic acid is present as toxic undissociated acetic acid. It is evident
from the observations cited above that the combination of pH value
and organic acid concentrations in nham are not sufficient to ensure
the elimination of S. aureus and L. monocytogenes. These results sug-
gest limited efficacy of the natural fermentation process to control
undesirable bacteria and emphasises, therefore, the crucial impor-
tance of careful selection of raw materials with minimal initial levels
of contamination.
Even with good quality raw materials it is still important to obtain
a rapid and reliable fermentation to minimise the numbers of unde-
sired bacteria in nham. Petchsing and Woodburn (1990) showed
that, in the absence of a starter culture, inoculated Escherichia coli
and S. aureus remained unchanged or slightly increased during nham
fermentation and that the fermentation itself was incomplete with a
final pH of only 4.9. With 1.5% commercial starter culture added, no
viable S. aureus or E. coli were recovered after 36 h and 96 h, respec-
tively, and a complete fermentation to pH <4.6 occurred within 96 h.
8.1.5.3.3 Clostridium botulinum A major safety concern with a

product such as nham, where an anaerobic, nutrient rich environment
is created and then fermented at ambient temperatures, is the pos-
sibility of the growth of and toxin formation by Clostridium botuli-
num before the pH value is lowered to levels where it cannot grow.
The initial restriction of growth of C. botulinum primarily depends
on the added salt but there is a risk that incomplete mixing may leave
some parts without salt and, additionally, some time is required for
the salt to diffuse and establish a uniform concentration through-
out the ingredients, as shown with acid in shrimp cocktails by Lerke
(1973). Hence, there may be regions and periods of time at the start
of the fermentation when C. botulinum may multiply unimpeded.
Nevertheless, there do not appear to be reports of botulism poisoning
from consumption of nham.
Although botulinum toxin is heat labile and is destroyed by cook-
ing, the fact that nham is commonly eaten uncooked would appear
to make it a particularly high-risk product. Growth of C. botuli-
num is inhibited in the presence of ~8% salt (aw, ~0.95) at pH values
of 7, 6 and 5 (ComBase 2013) and by nitrite. The aw of fermented
nham is within the range 0.97–0.95 depending on the formulation
(Wiriyacharee 1990) and, thus, the aw on its own does not preclude
growth of C. botulinum. The effective inhibitory concentrations of
nitrite are much influenced by other food ingredients but initial con-
centrations of sodium nitrite greater than 100 mg kg−1 are generally
used commercially (Adams and Moss 2000). The maximum permit-
ted added concentration of NaNO2 or KNO2 in comminuted meats
in the United States is 156 mg kg−1 (Sindelar and Milkowski 2011)
and the maximum allowable concentrations of nitrite and nitrate in
nham are 125 and 500 mg kg−1, respectively (TISI 2003). It is not
clear whether these concentrations and the conditions existing in
nham preclude the growth of C. botulinum or not.
8.1.6 Development of Nham Starter Culture Technology
Inherently, the quality and consistency of naturally fermented nham

cannot be fully controlled and, as nham is generally consumed with-
out cooking, the presence of any pathogens is of concern. The levels of
pathogens can be reduced by the use of appropriate starter cultures to
achieve a rapid and consistent acidification to pH of ≤4.6 (Paukatong
and Kunawasen 2001; Petchsing and Woodburn 1990; ). The first
starter culture for nham fermentation was developed by Wiriyacharee
et al. (1990). The culture comprised a defined mixture of Lactobacillus
plantarum NHI 1100, Pediococcus cerevisiea and Micrococcus varians
ATCC 15306 and has been successfully and extensively utilised in
the industrial production of nham. To further improve control of
the process and to maintain product diversity, Valyasevi et al. (2001)

have continued development of starter cultures for nham fermenta-
tion. Selected isolates from the dominant groups were evaluated for
their ability to rapidly ferment and produce a consistent product with
desirable sensory characteristics. This resulted in the identification of
a number of Lb. plantarum strains with desirable properties and some
of these starter cultures are presently being tested by various nham
manufacturers (Smitinont et al. 2001; Valyasevi et al. 2001). Besides
nham, the cultures can be applied to other kinds of fermented meat
products, such as fermented chicken sausages, fermented pork spare
rib and so on. Three formulations of the starter cultures are being
marketed.
Kingcha et al. (2012) investigated the possibility of using bacterio-
cin-producing lactic acid bacterial starters to control Listeria mono-
cytogenes in nham. A strain of Pediococcus pentosaceus, BCC 3772,
producing pediocin PA-1/AcH was shown to reduce the growth of
L. monocytogenes in nham without affecting other properties and it
was suggested that the use of such strains in nham starter cultures
could contribute to improving the microbiological safety of the fer-
mented product.
8.1.6.1 Safety Recommendations To ensure safety of nham consump-

tion, the following recommendations should be followed.
8.1.6.1.1 Recommendations for Producers

1. Carefully select raw materials, especially raw pork, to mini-
mise contamination from unwanted chemical or biological
hazards. The pigs should be from farms with good agricultural
practices and be slaughtered under good hygienic practices.
After slaughtering, the pork meat should be processed as soon
as possible to minimise growth of unwanted microorganisms.
2. Strictly follow good manufacturing practice and obtain the
Thai-FDA authorisation seal. The Thai-FDA seal helps reas-
sure the consumer that the manufacturer operates good man-
ufacturing practices.
3. Freeze the raw meat or the stuffed product for the specified
time and temperature to destroy any parasites that may exist
in the meat.
4. Carefully read and strictly follow the instructions on the cur-

ing salt package. Since there are curing salt mixtures with
different concentrations of sodium nitrite (or nitrate) avail-
able on the market, the instructions on the package should be
strictly followed to avoid over- or under-dosing of the addi-
tive. Special attention should be paid when a new batch or
new supplier of the additive is used.
5. Ensure complete fermentation by incubating the product until
pH ≤4.6, according to the nham standard, before marketing.
6. Use a starter culture to ensure complete fermentation.
7. For additional safety, gamma irradiation can be used to
destroy parasites and pathogenic bacteria. According to the
nham standard, the allowable limit is no higher than 7 kGray.
At this irradiation level, total lactic acid bacteria were reduced
from 107–109 to 104 –106 cfu g−1 and no pathogenic bacteria
were detected (Paukatong et al. 1999; Valyasevi et al. 2008).
The irradiation is usually done after the fermentation is com-
pleted for convenience and the additional benefit of delaying
the occurrence of over-fermentation and increasing the prod-
uct shelf life.
8.1.6.1.2 Recommendations for Consumers

1. Buy nham from a trustworthy producer. Look for the Thai-
FDA seal on the product’s package for assurance that the
product was produced under conditions of good manufactur-
ing practice.
2. Thoroughly cook nham before consumption. The heat from
cooking can destroy parasites and pathogenic bacteria in
nham.
3. Avoid eating incompletely fermented nham that does not
have the usual sour taste.
4. If the product is to be consumed without cooking, freezing
before consumption may help reduce the risk from parasites,
unless the producer has already frozen it during production.
Storage in a refrigerator can reduce risk from pathogenic bac-
teria without affecting the organoleptic properties. If avail-
able, irradiated nham is the safest option.
8.1.7 Future Developments
To date, owing to the lack of detailed scientific understanding of the

fermentation and lack of technological know-how, research on nham
has concentrated on product development to ensure the quality and
safety of the product. Thus, scientific research and development on
nham has yet to develop its full potential and there are many quality,
safety and health issues for future research.
1. Over-fermentation is the most pressing problem that affects

the physical and chemical properties of nham. The continuous
and excessive decrease in pH of nham during storage leads
to unacceptable characteristics of nham, including drip loss,
discolouration, softening and off-flavour, eventually leading
to rejection by consumers.
2. With the current interest in the possible health-promoting
effects of microorganisms, opportunities exist for the incor-
poration of such microbes into fermented foods, including
nham. Accurate data on the health-promoting activities of
probiotic bacteria or prebiotic substrates in vivo would greatly
aid the development of a completely new generation of fer-
mented functional foods tailored to contain clinically proven
health-promoting microorganisms and/or nutrients. Future
prospects to exploit the potential health-promoting proper-
ties of such foods will continue to expand as we gain a greater
understanding of the microorganisms concerned and how
they, or their metabolites, can directly interact in a positive
way with the human host.
3. Excessive salt intake has adverse health effects. As part of the
on-going pressure to reduce salt intake, the development of
reduced sodium nham presents a challenge to allow consum-
ers to enjoy a healthier product. To reduce the sodium content
while ensuring the highest possible level of safety and quality,
it will be necessary to consider alternatives to sodium-con-
taining preservatives and/or treatments that allow markedly
lower salt content. However, a particular problem with low-
salt meat products may be that not only the perceived saltiness,
but also the intensity of the characteristic flavour is decreased.
Thus, any alteration in the salt content of meat products may

require ingredient reformulation or manipulation, and in
particular more savoury/umani taste to compensate for the
reduced saltiness. Through the use of these flavours and by
adding more aromatics, such as herbs and spices, a reduction
in salt can be achieved.
4. Some basic aspects of the fermentation are still not clear. For
example, what is the source of fermentable carbohydrate? Is it
primarily rice starch or garlic fructans or both?
5. Botulism would appear to be a potential hazard and a careful
evaluation of the risk, with challenge tests, is needed.
8.2 Bruneian Belutak (Fermented Meat Sausage)

Junaidah Abu Bakar
Belutak is a slightly acidic, salty and chewy fermented sausage made

from small intestines of cows or buffaloes stuffed with a mixture of
meat trimmings, salt and sugar. The sausage is allowed to ferment
overnight followed by sun drying the next day. The sun-dried product
has a shelf life of about 3 months at 30°C (Abu Bakar 2000).
8.2.2 History, Places of Production, How Consumed and Role in Diet
Belutak was originally produced by the water village people of

Kampung Ayer. Kampung Ayer is considered to be the world’s largest
water village, home to more than 20,000 inhabitants who live in an
intricate network of stilt houses built over Sungai Brunei, the largest
river in Brunei. Sixteenth century European explorers called it the
‘Venice of the East’.
In the early years in the history of Brunei, the people in the water vil-
lage were traditionally fishermen and their daily diet consisted mainly
of seafood and rice. Meat was a commodity not easily available and
was seldom eaten except on special occasions. Hence, when the oppor-
tunity arose every part of the animal, mostly water buffalo, was eaten
or made into other products. The absence of cool storage also made the
people more innovative in their desire to preserve any left-over meat
and remnants. Belutak is still more popular and better known to the
people of the water village than to those who live on land.
Belutak is normally consumed as a side dish during family meals.
The sausage-like product is sliced into about 5 cm portions and fried
with chillies and onions. Because of the high salt content, it is always
eaten along with rice and vegetables. Belutak is a popular product in
the water village, providing a source of protein as a luxury item for the
villagers. However, it plays a minimal role in the daily diet and is like
a delicatessen item rather than a staple food. Although its popular-
ity has not greatly increased, the production of belutak is still being
done on a domestic scale home and as a cottage industry. Belutak also
appeals to students who study overseas and miss home food. These
students normally have families who originate from the water village,
although some may now live on land.
Traditional belutak preparation is carried out at home and its produc-

tion is very much dependent on personal consumption or, in the case
of small manufacturers, in response to demand.
To prepare belutak, 1 kg of cow or water buffalo small intestine is
used as casing. Excess tissue and fat are trimmed off and the intes-
tines are washed thoroughly inside and outside under running water.
The cleaned intestine is turned inside out and a knot is tied at one
end. About 2 kg of meat trimmings, including lean meat, fat and
tendons, are washed and drained well before being mixed with salt
and sugar. Some recipes use equal amounts of salt and sugar. The
sugar and salt are thoroughly mixed with the trimmings, and some-
times optional ingredients, such as onions and chillies, are added for
extra flavour. Added chillies give a brownish colour to the finished
product. The meat–salt–sugar mixture is then stuffed into the casing
and the open end of the casing is tied with plastic string. The sausage
is then allowed to ferment at room temperature (28–30°C) for 24 h
followed by at least 5 days of sun drying (Figure 8.3). The belutak
is ready to eat at the end of the drying process. Belutak is normally
sliced and fried with chillies and onions and consumed along with
rice. Due to the high salt content, only a small amount of belutak is
eaten at one time.
Meat trimmings ~2 kg Small intestines ~1 kg

(water buffalo or cow)
Salt Trim excess fat and tissue

Sugar
Onion, chillies
(optional)
Mix Clean and wash
Stuff into cleaned intestines
Incubate 24 h, room temperature (28–30ºC)
Sundry, 5 days
Belutak
Wrap in polythene
Figure 8.3 Production of belutak (Bruneian fermented sausage).
Traditionally, belutak was sold without packaging and just hung

up with a string but during the late twentieth century polythene bags
started to be used as packaging. Nowadays, a drying oven at 70–80°C
is used to dry the belutak, thus speeding up the drying process and
increasing production. Some belutak is now also vacuum packed and
frozen, which increases its shelf life to more than 3 months. A regular
size belutak (280–300 g) costs BND 6.50 (US$5.1). Currently, there
are several small-scale producers who manufacture belutak. Their
product is available in local supermarkets in the frozen foods section.
Thus, belutak is now readily available compared to earlier times when
it was only produced at home by a few water villagers.
8.2.4 Characteristics of the Fermentation
There have been few scientific investigations on belutak and the

changes in the substrate or the involvement of bacteria have not been
documented. However, the inclusion of sugar as a source of ferment-
able carbohydrate and a final pH value of ~4.4 suggests that a lac-
tic fermentation occurs. Microorganisms found on belutak included
Bacillus spp., Alcaligenes sp., Pseudomonas sp., Lactobacillus spp. and
Staphylococcus spp. (Abu Bakar 2000; Petra and Abu Bakar 1999).
Table 8.3 Proximate Chemical Composition of Belutak (Bruneian Fermented Sausage)

CONCENTRATION (G [KG EDIBLE PORTION, WET WEIGHT]−1)
Moisture 200–505
Protein 190–295
Carbohydrate 185–560
Fat 14–26
Fibre 8–46
pH 4.3–4.5
The composition of belutak is very variable, reflecting the different

formulations used to make it and the properties of the meat (Table
8.3). Belutak is consumed as a delicatessen item and, thus, its contri-
bution as a nutrient source is negligible. Also, the high salt content
means that only a small amount of it is eaten at one time.
The relatively high salt content combined with drying should render
belutak a safe product, with little risk of growth of bacterial pathogens
in the final product. However, the risks of contamination by patho-
gens at earlier stages or possible pathogen growth before the estab-
lishment of high salt concentrations throughout the meat emphasises
the need for good hygienic practices during its preparation.
8.2.7 Future
Belutak has received very little attention among researchers and there
are opportunities to establish the details of its preparation and the
diversity of recipes used, its composition and the microbiology of the
process. Potentially, this would then allow the establishment of some
standard procedures and possible expansion of its manufacture.
References
Abu Bakar, J. 2000. An Introduction to some Fermented Food of Brunei Darus
salam. Brunei Darassulam: Universiti Brunei Darassulam.
Adams, M.R. and M.O. Moss. 2000. Food Microbiology. Second ed. Cambridge:
Royal Society of Chemistry.
Aktaş, N. and Kaya, X. 2001. Influence of weak organic acids and salts on the
denaturation characteristics of intramuscular connective tissue. A differ-
ential scanning calorimetry study. Meat Science 58:413–419.
Arena, M.E. and M.C. Manca de Nadra. 2001. Biogenic amine production by
Lactobacillus. Journal of Applied Microbiology 90:158–162.
Bacus, J.N. 1984. Utilization of Microorganisms in Meat Processing. Letchworth,
England: Research Studies Press.
Bardócz, S. 1995. Polyamines in food and their consequences for food
quality and human health. Trends in Food Science and Technology

6:341–346.
Baron, C.P., Skibsted, L.H. and H.J. Anderson. 1998. Interaction between the
antioxidative heme protein iron (II) nitrosylmyoglobin and peroxide. In
IFT Annual Meeting’s Abstracts, USA. 20–24.
Baumgartner S., T.G. Dax, W. Praznik and H. Falk. 2000. Characterization of
high-molecular weight fructan isolated from garlic (Allium sativum L.).
Carbohydrate Research 328:177–183.
Berge, P., Ertbjerg, P., Larsen, L.M., Astruc, T., Vignon, X. and A.J. Møller.
2001. Tenderization of beef by lactic acid injected at different times post
mortem. Meat Science 57:347–357.
Bernbom, N., Ng, Y.Y., C. Paludan-Müller and L. Gram. 2009. Survival and
growth of Salmonella and Vibrio in Som-fak, a Thai low-salt garlic-con-
taining fermented fish product. International Journal of Food Microbiology
134:223–229.
Bogovski, P. and S. Bogovski. 1981. Animal species in which N-nitroso com-
pounds induce cancer. International Journal of Cancer 27:471–474.
Bourne, M.C. 1978. Texture profile analysis. Food Technology 32:62–66.
Bover-Cid, S., M. Izquierdo-Pulido and M.C. Vidal-Carou. 2000. Influence
of hygienic quality of raw materials on biogenic amine production during
ripening and storage of dry fermented sausages. Journal of Food Protection
63(11):1544–1550.
Bover-Cid, S., M.J. Miguélez-Arrizado and M.C. Vidal-Carou. 2001a. Bio
genic amine accumulation in ripened sausages affected by the addition of
sodium sulphite. Meat Science 59:391–396.
Bover-Cid, S., M. Izquierdo-Pulido and M.C. Vidal-Carou. 2001b. Effect
iveness of a Lactobacillus sakei starter culture in the reduction of biogenic
amine accumulation as a function of the raw material quality. Journal of
Food Protection 64(3):367–373.
Bover-Cid, S., M.J. Miguélez-Arrizadoa, B. Becker, W.H. Holzapfel and M.C.
Vidal-Caroua. 2008. Amino acid decarboxylation by Lactobacillus curva-
tus CTC273 affected by the pH and glucose availability. Food Microbiology
25:269–277.
Brink, B., C. Damink, H.M.L.J. Joosten and J.H.J. Huis in’t Veld. 1990. Occur
rence and formation of biological active amines in foods. International
Chokesajjawatee, N., S. Pornaem, Y.G. Zo, S. Kamdee, P. Luxananil,

S. Wanasen and R. Valyasevi. 2009. Incidence of Staphylococcus aureus
and associated risk factors in Nham, a Thai fermented pork product. Food
Chuxnum, T. 2012. Trichinosis. Annual Epidemiological Surveillance Report
2012, pp. 72–73. Bangkok: Ministry of Public Health, Thailand.
ComBase. 2013. Predictive model for proteolytic Clostridium botulinum. www.
combase.cc/index.phb/en/ (accessed 3 March 2013).
Corzo-Martínez, M., N. Corzo and M. Villamiel 2007. Biological properties
of onions and garlic. Trends in Food Science and Technology 18:609–625.
de Fátima Silva Lopes, M., Leitão, A.L., Regalla, M., Marques, J.J.F.,
Carrondo, M.J.T. and M.T.B. Crespo. 1999. Processing of extracellular
lipase of Lactobacillus plantarum involvement of a metalloprotease. FEMS
Microbiology Letters 176(2):483–487.
Demeyer, D., M. Raemaekers, A. Rizzo, A. Holck, A. De Smedt, B.T. Brink, B.
Hagen et al. 2000. Control of bioflavor and safety in fermented sausages:
First results of a European project. Food Research International 33:171–80.
Dušek, M., F. Kvasnička, L. Lukáškovš and J. Krátká. 2003. Isotachophoretic
determination of added phosphate in meat products. Meat Science
65:765–769.
European Commission. 2005. Commission Regulation (EC) No 2073/2005
of 15 November 2005 on microbiological criteria for foodstuffs. Official
Journal of the European Union L338:1–26.
Fadda, S., G. Vignolo, A.P.R. Holgado and G. Oliver. 1998. Proteolytic activity
of Lactobacillus strains isolated from dry-fermented sausages on muscle
sarcoplasmic proteins. Meat Science 49:11–18.
Fadda, S., Y. Sanz, G. Vignolo, M.C. Aristoy, G. Oliver and F. Toldrá. 1999a.
Characterization of muscle sarcoplasmic and myofibrillar protein hydro-
lysis caused by Lactobacillus plantarum. Applied Environmental and Micro
biology 65(8):3540–3546.
Fadda, S., Y. Sanz, G. Vignolo, M.C. Aristoy, G. Oliver and F. Toldrá. 1999b.
Hydrolysis of pork muscle sarcoplasmic proteins by Lactobacillus cur-
vatus and Lactobacillus sake. Applied Environmental and Microbiology
65(2):578–84.
Faustman, C. and R.G. Cassens. 1990. The biological basis for discoloration
in fresh meat: A review. Journal of Muscle Food 1(3):217–243.
Fernández, M., J.A. Ordóñez, J. M. Bruna, B. Herranz and L. de la Hoz. 2000.
Accelerated ripening of dry fermented sausages. Trends in Food Science
and Technology 11:201–209.
Flores, J. and F. Toldrá. 1993. Curing. In Encyclopaedia of Food Science, Food
Technology and Nutrition, edited by R. Macrae, R.K. Robinson and M.J.
Sadler, p. 600. New York: Academic Press.
Fretheim, K., B. Egelandsdal, O. Harbitz and K. Samejima. 1985. Slow lower-
ing of pH induces gel formation of myosin. Food Chemistry 18:169–178.
Goldspink, G. and J.V. McLoughlin. 1964. Studies on pig muscle. 3. The effect
of temperature on the solubility of the sarcoplasmic proteins in relation to
colour changes in post-rigor muscle. Irish Journal of Agriculture Research

3:9–16.
González-Fernández, C., E.M. Santos, I. Jaime and J. Rovira. 2003. Influence
of starter cultures and sugar concentrations on biogenic amine contents
in chorizo dry sausage. Food Microbiology 20:275–284.
Hongchumphon, N. 2007. Food Poisoning. Annual Epidemiological Surveillance
Report 2007. Ministry of Public Health, Thailand, pp. 103–105.
Honikel, K.-O. 2008. The use and control of nitrate and nitrite for the process-
ing of meat products. Meat Science 78:68–76.
Hornsey, H.C. 1956. The color of cooked cured pork. I. Estimation of
nitric oxide-heme pigments. Journal of the Science and Food Agriculture
7:534–540.
Hornsey, H.C. 1959. The color of cooked cured pork. III. Distribution and
relationship of pigments, pH and cysteine. Journal of the Science and Food
and Agriculture 10:114–124.
Hughes, M.C., J.P. Kerry, E. K. Arendt, P.M. Kenneally, P.L.H. McSweeney
and E.E. O’Neill. 2002. Characterization of proteolysis during ripening
of semi-dry fermented sausages. Meat Science 62:205–216.
Ingolf, F.N. and R. Skjelkvale. 1982. Effect of natural spices and oleoresins on
Lactobacillus plantarum in fermentation of dried sausage. Journal of Food
Science 47:1618–1621.
Joo, S.T., R.G. Kauffman, B.C. Kim and G.B. Park. 1999. The relationship
of sarcoplasmic and myofibrillar protein solubility to color and water-
holding capacity in porcine longissimus muscle. Meat Science 53:291–297.
Joosten, H.M.L.J. 1988. The biogenic amine contents of Dutch cheese and
their toxicological significance. Netherlands Milk and Dairy Journal
42(1):25–42.
Judprasong, K., T. Siriporn, P. Puwastien and S. Pongtorn. 2011. Investigation
of Thai plants for potential sources of inulin-type fructans. Journal of
Food Composition and Analysis 24:642–649.
Kaban, G. 2009. Changes in the composition of volatile compounds and in
microbiological and physicochemical parameters during pastrima pro-
cessing. Meat Science 82:17–23.
Khieokhachee, T., W. Praphailong, C. Chowvalitnitithum, S. Kunawasen,
S. Kumphati, S. Chavasith, V. Bhumiratana and R. Valyasevi. 1997.
Microbial interaction in the fermentation of Thai pork sausage. In
Proceeding of the 6th ASEAN Food Conference. 24–27 November 1997,
Singapore, pp. 312–318.
Kim, K., Y. Homma, Y. Ikeuchi and A. Suzuki. 1995. Cleavage of connec-
tin by calpain and cathepsin D. Bioscience Biotechnology and Biochemistry
59(5):896–899.
Kingcha, Y., A. Tosukhowong, T. Zendo, S. Roytrakul, P. Luxananil, K.
Chareonpornsook, R. Valyasevi, K. Sonomoto and W. Visessanguan.
2012. Anti-listeria activity of Pediococcus pentosaceus BCC 3772 and appli-
cation as starter culture for nham, a traditional fermented pork sausage.
Food Control 25(1):190–196.
Krairojananan, P., S. Tanasupawat, W. Noonpakdee and R. Valyasevi. 1997.

Catalases of bacteria isolated from Thai fermented foods. Food Bio
technology 12:221–238.
Kumar, M. and J.S. Berwal. 1998. Sensitivity of food pathogens to garlic (Allium
sativum). Applied and Environmental Microbiology 84:213–215.
Kunawasen, S. 2000. Molecular typing of lactic acid bacteria isolated dur-
ing nham fermentation. M.Sc. Thesis, Mahidol University, Bangkok,
Thailand.
Latorre-Moratalla, M.L., T. Veciana-Nogués, S. Bover-Cid, M. Garriga,
T. Aymerich, E. Zanardi, A. Ianieri. et al. 2008. Biogenic amines in tradi-
tional fermented sausages produced in selected European countries. Food
Laxananil, P., R. Promchai, S. Wanasen, S. Kamdee, P. Thepkasikul, V.
Plengvidhya, W. Visessanguan and R. Valyasevi. 2009. Monitoring
Lactobacillus plantarum BCC9546 starter culture during fermentation
of nham, a traditional Thai pork sausage. International Journal of Food
Lee, C.H. 1997. Lactic acid fermented foods and their benefits in Asia. Food
Control 8:259–269.
Lee, S.H., R.G Cassens and O.R. Fennema. 1976. Effect of muscle type on
residual nitrite in cured meat. Journal of Food Science 41:100–101.
Lerke, P. 1973. Evaluation of potential risk of botulism from seafood cocktails.
Applied Microbiology 25:807–8210.
Losso, J.N. and S. Nakai. 1997. Molecular size of garlic fructooligosaccha-
rides and fructopolysaccharides by matrix-assisted laser desorption ion-
ization mass spectrometry. Journal of Agricultural and Food Chemistry
45(11):4342–4346.
Lucas, P., J. Landete, M. Coton, E. Coton and A. Lonvaud-Funel. 2003.
The tyrosine decarboxylase operon of Lactobacillus brevis IOEB9809:
Characterization and conservation in tyramine-producing bacteria.
FEMS Microbiology Letters 229:65–71.
Mancini R.A., M.C. Hunt, M. Seyfert, D.H. Kropf, K.A. Hachmeister, T.J.
Herald and D.E. Johnson. 2007. Comparison of ascorbic acid and sodium
erythorbate: Effects on the 24 h display colour of beef lumbar vertebrae
and longissimus lumborum packaged in high-oxygen modified atmo-
spheres. Meat Science 75(1):39–43.
Martin, M. 2001. Meat curing technology. In Meat Science and Applications,
edited by S. Damodaran and A. Paraf, pp. 491–508. New York: Marcel
Dekker.
McCabe, B.J. 1986. Dietary tyramine and other pressor amines in MAOI
regimens: A review. Journal of the American Dietetic Association 86(8):
1059–1064.
McLoughlin, J. and G. Goldspink. 1963. Postmortem changes in the color of
pig Logissimus dorsi muscle. Nature 19:584–585.
Møller, J.K.S. and Skibsted, L.H. 2002. Nitric oxide and myoglobins. Chemical
Reviews 102:1167–1178.
Molly, K., D. Demeyer, G. Johansson, M. Raemaekers, M. Ghistelinck and

I. Geenen. 1997. The importance of meat enzymes in ripening and fla-
vour generation in dry fermented sausages. Food Chemistry 59:539–545.
Noimoh, T. and D. Areechokchai. 2007. Outbreak investigation of nitrite
toxicity after eating sausage. Weekly Epidemiological Surveillance Report,
Ministry of Public Health, Thailand 38:361–365.
Osburn, W.N., R.W. Mandigo and K.M. Eskridge. 1997. Pork skin connec-
tive tissue gel utilization in reduced-fat bologna. Journal of Food Science
62:1176–1182.
Paik, D.C., D.V. Saborio, R. Oropeza and H.P. Freeman. 2001. The epide-
miological enigma of gastric cancer rates in the US: Was grandmother’s
sausage the cause? International Journal of Epidemiology 30:181–182.
Paludan-Müller, C., Valyasevi, R., Huss, H.H. and Gram L. 2002. Genotypic
and phenotypic characterization of garlic-fermenting lactic acid bacteria
isolated from som-fak, a Thai low-salt fermented fish product. Journal of
Panya, A., S. Riebroy, A. Assavanig, S. Benjakul and W. Visessanguan. 2003. Pale
color formation in nham, a Thai fermented pork sausage during fermenta-
tion. In Proceedings of the 4thAgro-Industry Conference. May 30-June 1, 2002.
Bangkok, Thailand.
Paterson, B.C., F.C. Parrish and M.H. Stromer. 1988. Effects of salt and pyro-
phosphate on the physical and chemical properties of beef muscle. Journal
of Food Science 53:1258–1265.
Paukatong, K.V. and S. Kunawasen. 2001. The hazard analysis and criti-
cal control points (HACCP) generic model for the production of Thai
pork sausage (nham). Berliner Und Munchener Tierarztliche Wochenschrift
114:1–4.
Paukatong, K., T. Smitinont, W. Prapailong and S. Kunawasen. 1999. Final report
on development of quality control system, quality assurance and stan-
dard for nham industry and nham products. National Center for Genetic
Engineering and Biotechnology, Pathumthani, Thailand. (In Thai).
Pearson, A.M. and T.A. Gillett. 1996. Casings, extenders, and additives. In
Processed Meats, edited by A.M. Pearson and T.A. Gillett, p. 448. New
York: Chapman and Hall.
Petchsing, U. and M.J. Woodburn. 1990. Staphylococcus aureus and Escherichia
coli in nham (Thai-style fermented pork sausage). International Journal of
Petra, S.F. and J. Abu Bakar, J. 1999. Studies on belutak and budu kupang-
traditional fermented products of Brunei Darussalam. B.Sc. dissertation,
Biology Department, University of Brunei Darussalam.
Phithakpol, B., W. Varanyanond, S. Reungmaneepaitoon and H. Wood.
1995. The Traditional Fermented Foods of Thailand. Bangkok: Institute
of Food Research and Product Development, Kasetsart University,
Thailand.
Pighin, D.G., A.M. Sancho and C.B. Gonzalez. 2008. Effect of salt addition
on the thermal behavior of proteins of bovine meat from Argentina. Meat
Science 79:549–556.
Piluk, S. 2009. Controlled fermentation of nham inoculated with Lactobacillus

plantarum BCC 9546 by limitation of carbon sources. M.Sc. Thesis, Food
Science and Technology, Maejo University, Chiang Mai, Thailand.
Plengvidhya, V., N. Chokesajjavatee, P. Thepkasikul and M. Rattanaprasert.
2008. Studying the effect of starter culture on microbiological and
chemical changes during nham fermentation. In Final report submit-
ted to National Center for the Genetic Engineering and Biotechnology,
Pathumthani, Thailand.
Plumed-Ferrer, C., K.M. Koistinen, T.L. Tolonen, S.J. Lehesranta, S.O.
Kärenlampi, E. Mäkimattila, V. Joutsjoki, V. Virtanen, A. von Wright.
2008. Comparative study of sugar fermentation and protein expression
patterns of two Lactobacillus plantarum strains grown in three different
media. Applied and Environmental Microbiology 74(17):5349–5358.
Pohlman, F.W., M.R. Stivarius, K.S. McElyea, Z.B. Johnson and A.L. Waldroup.
2002. Reduction of E. coli, Salmonella typhimurium, coliforms, aerobic bac-
teria and improvement of ground beef color using trisodium phosphate or
cetylpyridinium chloride before grinding. Meat Science 60:349–356.
Rojsuntornkitti, K., N. Jittrepotch, T. Kongbangkerd and K. Kraboun. 2010.
Substitution of nitrite by Chinese red broken rice powder in Thai tradi-
tional fermented pork sausage (nham). International Food Research Journal
17:153–161.
Rotsatchakul, P., W. Visessanguan, T. Smitinont and S. Chaiseri. 2009. Changes
in volatile compounds during fermentation of nham (Thai fermented
sausage). International Food Research Journal 16:391–414.
Ruiz-Capillas, C. and F. Jiménez-Colmenero. 2004. Biogenic amines in
meat and meat products. Critical Reviews in Food Science and Nutrition
44:489–499.
Sallam, K.I., M. Ishioroshi and K. Samejima. 2004. Antioxidant and antimi-
crobial effects of garlic in chicken sausage. Lebensmittel-Wissenschaft und-
Technologie 37:849–855.
Shalaby, A.R. 1996. Significance of biogenic amines to food safety and human
health. Food Research International 29(7):675–690.
Silla Santos, M.H. 1996. Biogenic amines: Their importance in food. Inter
national Journal of Food Microbiology 29:213–231.
Sindelar, J.J. and A.L. Milkowski. 2011. Sodium nitrite in processed meat and
poultry meats: A review of curing and examining the risk/benefit of its
use. American Meat Science Association, White Paper Series, No. 3.
Smitinont, T., W. Praphailong, S. Khumphati, S. Chaiseri, P. Rojsajakul and
K. Kiatpathomchai. 2001. Profiles of flavour compounds in traditional
Thai fermented pork sausage. In Proceedings of the Seventh ASEAN Food
Conference, 19–22 Novermber 2001, Philipines.
Sofos, J.N. 1986. Use of phosphate in low-sodium meat products. Food
Technology 40(3):52–69.
Spaziani, M., M.D. Torre and M.L. Stecchini. 2009. Changes in physicochem-
ical, microbiological and textural properties during ripening of Italian
low-acid sausages. Proteolysis, sensory and volatile profiles. Meat Science
81:77–85.
Srichareuan, W., N. Akephobnarong, D. Keartivechphaisan, S. Melarbphiboon

and N. Chaiprechawit. 2009. Prevalence of Trichinella spiralis infec-
tion in swine in the north part of Thailand. Presented at the Research
fair, National Agricultural Fair 2009, Kasetsart University, Bangkok,
Thailand.
Stahnke, L.H. 1994. Aroma components from dried sausages fermented with
Staphylococcus xylosus. Meat Science 38:39–53.
Swetwiwathana, A., Leutz, U. and Fischer, A. 1999. Role of garlic on growth
and lactic acid production of starter cultures. Fleischwirtschaft International
1(99):124–128.
Tanasupawat, S. and K. Komagata. 1995. Lactic acid bacteria in fermented
foods in Thailand. World Journal of Microbiology and Biotechnology 11(3):
253–256.
TISI. 2003. Naem (Fermented ground pork), TIS 1219-2547. Thailand Industrial
Standard Institute, Bangkok, Thailand.
Toldrá, F., Y. Sanz and M. Flores. 2001. Meat fermentation technology.
In Meat Science and Applications, edited by S. Damodaran and A. Paraf,
pp. 537–561. New York: Marcel Dekker.
Tosukhowong, A., W. Visessanguan, L. Pumpuang, P. Tepkasikul, A. Panya and
R. Valyasevi. 2011. Biogenic amine formation in nham, a Thai fermented
sausage, and the reduction by commercial starter culture, Lactobacillus
plantarum BCC 9546. Food Chemistry 129(3):846–853.
Urbach, E., C. Schindler and S.J. Giovannoni. 1998. A PCR fingerprinting
technique to distinguish isolates of Lactococcus lactis. FEMS Microbiology
Letters 162:111–115.
U.S. Code of Federal Regulation number 318.10. 2012. Prescribed Treatment
of Pork and Products Containing Pork to Destroy Trichinae. Washington:
U.S. Government Printing Office.
Valyasevi, R., N. Chokesajjawatee, W. Visessanguan, P. Luxananil, S.
Nuanualsuwan, K. Budsaba, A. Tosukhowong et al. 2008. Final report on
quantitative microbial risk assessment of Staphylococcus aureus and quan-
titative chemical risk assessment of volatile N-nitrosamines in Thai fer-
mented pork sausage (nham). National Center for Genetic Engineering
and Biotechnology, Pathumthani, Thailand. (In Thai).
Valyasevi, R., P. Jungsiriwat, T. Smitinont, W. Praphailong and C.
Chowalitnitithum. 2001. Final report on improvement of starter culture
for nham fermentation. National Center for the Genetic Engineering
and Biotechnology, Pathumthani, Thailand. (In Thai).
Valyasevi, R. and R.S. Rolle. 2002. An overview of small-scale food fermen-
tation technologies in developing countries with special reference to
Thailand: Scope for their improvement. International Journal of Food
Valyasevi, R., S. Sornprajak and S. Tanasupawa. 1997. Identification of Lacto
bacillus plantarum, Lactobacillus pentosus and Pediococcus pentosaceus using
random amplified polymorphic DNA fingerprint. Proceedings of the 6th
ASEAN Food Conference, 24–27 November 1997, pp. 305–311. Singapore:
Singapore Institute of Food Science and Technology.
Valyasevi, R., T. Smitinont, W. Praphailong, C. Chowvalitnitithum, S. Kunawasen

and V. Chavasith. 1999. The microbiology and development of starter cul-
tures for nham, the Thai pork sausage. In Proceedings of the Seventeenth
International Conference of the International Committee on Food Microbiology
and Hygiene (ICFMH), pp. 709–711. Veldhoven: International Committee
on Food Microbiology and Hygiene.
Valyasevi, R., T. Smitinont, W. Visessanguan, S. Lertsiri, A. Tongta, M.
Nopharat, P. Tepkasikul, C. Kittikhun and T. Dol-kanokphob. 2006.
The fermentation rate and production of metabolic compounds during
the fermentation of cultured nham. Final report submitted to National
Center for Genetic Engineering and Biotechnology, Pathumthani,
Thailand. (In Thai).
Varnam, A.H. and J.P. Sutherland. 1995. Meat and Meat Products Technology,
Chemistry and Microbiology. London: Chapman and Hall.
Visessanguan, W., S. Benjakul, A. Panya, C. Kittikun and A. Assavanig. 2005.
Influence of minced pork and rind ratios on physico-chemical and sen-
sory quality of nham—a Thai fermented pork sausage. Meat Science
69:355–362.
Visessanguan, W., S. Benjakul, S. Chaiseri, P. Tepkasikul, C. Kittikhun, S.
Riebroy. 2007. Final report on physico-chemical and biochemical changes
in nham during fermentation. National Center for Genetic Engineering
and Biotechnology, Pathumthani, Thailand.
Visessanguan, W., S. Benjakul, S. Riebroy, M. Yarchai and W. Tapingkae. 2006b.
Changes in lipid composition and fatty acid profile of nham, a Thai fer-
mented pork sausage, during fermentation. Food Chemistry 94:580–588.
Visessanguan, W., S. Benjakul, S. Riebroy and P. Thepkasikul. 2004. Changes in
composition and functional properties of proteins and their contributions
to nham characteristics. Meat Science 66:579–588.
Visessanguan, W., S. Benjakul, T. Smitinont, C. Kittikun, P. Thepkasikul and
A. Panya. 2006a. Changes in microbiology, biochemical and physico-
chemical properties of nham inoculated with different inoculum lev-
els of Lactobacillus curvatus. Lebensmittel-Wissenschaft und-Technologie
39:814–826.
Wager, M.K. and F.F. Busta. 1985. Inhibition of Clostridium botulimun 52A
toxicity and protease activity by sodium acid pyrophosphate in meat sys-
tems. Applied and Environment Microbiology 50:16–20.
Wiboonchat, S. 2003. Detection of Campylobacter jejuni and Clostridium
perfringens contaminated in nham (Thai fermented pork) and effects
of ingredients on pathogen contamination. M.Sc. Thesis, Mahidol
University, Bangkok.
Wiriyacharee, P. 1990. The systematic development of a controlled fermenta-
tion process using mixed bacterial starter cultures for nham, a Thai semi-
dry sausage. Ph.D. Thesis, Massey University, New Zealand.
Wiriyacharee, P., L. Rujanakraikarn and A. Kuntiya 1993b. Nham product
development using mixed bacterial starter cultures 1. Suitable source
of carbohydrate for nham production. Journal of Agriculture 9(1):51–60.
(In Thai).
Wiriyacharee, P., L. Rujanakraikarn and P. Chavalee 1993a. Nham product

development using mixed bacterial starter cultures: Effect of non-gluti-
nous and glutinous rice on lactic acid production. Journal of Agriculture
9(1):61–74. (In Thai).
Wiriyacharee, P., L. Rujanakraikarn, P. Vibulseth, S. Sitthipojna, A.
Hunpongkittikul and S. Chusangjan. 1990. Final report submitted
to National Center for the Genetic Engineering and Biotechnology,
Pathumthani, Thailand.
Wongsommart, D., N. Reungprapan, C. Boonwongwiroj, N. Tongsakulpanit
and C. Rungrojjanarak. 1994. Parasites and food poisoning bacteria in
nham. Bulletin of the Department of Medical Sciences Ministry of Public
Health, Thailand 36:163–171.
Xiong, Y.L. 1997. Structure-function relationships of muscle proteins. In Food
Proteins and their Applications, edited by. S. Damodaran and A. Paraf,
Yang, G.C., P.M. Yasaei and S.W. Page. 1993. Garlic as anti-oxidants and free
radical scavenges. Journal of Food and Drug Analysis 1(4):337–364.
Young, O.A. and J. West. 2001. Meat color. In Meat Science and Applications,
edited by S. Damodaran and A. Paraf, pp. 39–69. New York: Marcel
Dekker.
9
S oya S auce
S A R D J O N O AT M O K O
Contents
9.1 Indonesian Soya Sauce (Kecap) 359

9.1.2 History 360
9.1.3 Places of Production, How Consumed and Role in
Diet 360
9.1.5.1 Characteristics of Koji Starter Moulds 365
9.1.5.2 Characteristics of Pediococcus halophilus 366
9.1.5.3 Characteristics of Saccharomyces rouxii 366
Fermentation 367
9.1.7.1 Koji Fermentation 367
9.1.7.2 Moromi Fermentation 368
9.1.11 Future Prospects 370
9.1.12 Research Needs and Opportunities 370
References 370
9.1 Indonesian Soya Sauce (Kecap)

Kecap or Indonesian soya sauce is a viscous, brown-black liquid sea-

soning produced from fermented soya bean with the addition of
coconut palm sugar. There are two kinds of kecap, kecap manis (sweet
soya sauce) and kecap asin (salty soya sauce). Kecap manis has a sweet
taste and a meat-like flavour. Almost all kecap currently sold in the
359
Indonesian market is kecap manis. It is manufactured by fermenting

soya beans using yellow-green Aspergillus spp., adding coconut palm
sugar to the moromi filtrate, and boiling the mixture. Kecap is pro-
duced by a two-step fermentation. The first fermentation is an aerobic
mould fermentation, producing a product called koji, and the second
is a submerged fermentation in brine, called moromi fermentation.
Kecap is used as an all-purpose seasoning agent and also as a table
condiment. It may be added to foods during or after preparation.
Foods become browner in colour and acquire a stronger flavour and
taste, including especially a meat-like umami taste and flavour.
9.1.2 History
Soya sauce originated in China more than 2000 years ago (Huang
2000) and then spread to Japan and other Asian countries. It is not
known precisely when kecap was first produced in Indonesia. The
word ‘kecap’ comes from the Taiwanese word ‘kôe chiap’, meaning
fish sauce. Formerly, kecap was produced by Chinese families who
passed the technology down the generations. They brought the tech-
nology from their forebears in China and then modified the taste to
meet the preference by the Javanese for a sweet taste in their food.
Historically, kecap manis was especially popular in Java and was pro-
duced by hundreds of small producers, including Indonesians, and
several large manufacturers, making products with different tastes
and of varying qualities.
Kecap is produced across Indonesia and especially in Java island. The

scale of manufacture ranges from small backyard producers to large-
scale kecap factories. Different manufacturers produce kecap with
different tastes, so that they supply different sectors of the market.
Some add spices during the cooking stage in kecap production to fur-
ther distinguish their product, while others add coconut palm sugar
only without any spices.
Kecap is a condiment or flavouring agent that is used to add fla-
vour and colour to foods, such as barbequed ‘tumis’, a kind of cooked
vegetable using spices and a small amount of oil in its preparation,
S oya S au c e 3 61
and many other dishes. The development of the culinary industries in

Indonesia is helping to promote the further development of the kecap
industry.
In traditional methods, kecap is produced using black soya bean as

raw material but some of the larger-scale manufacturers now use
imported defatted soya bean mixed with roasted wheat as raw mate-
rial, as used for soya sauce production in Japan. Steps in the tradi-
tional production are as follows (Figure 9.1): local black soya bean
is cleaned and soaked for 5–10 h in water at ambient temperature.
The weight of the beans increases 1.8–2.0 times during the soaking.
Next, the hydrated soya beans are cooked until they become soft. The
cooked soya beans are then poured on to bamboo trays in layers about
2 cm deep to cool them quickly. Traditionally, koji fermentation was
Black soya beans
Soak in water
Boil
Drain, spread on bamboo trays
Mold starter
Incubate ambient temperature 36–48 h
Koji
Brine
Moromi (22–24% NaCl)
Store ambient temperature, 4–6 months
Filter
Filtrate
Coconut sugar
Spices as required or none
Boil to concentrate to degrees Brix ~75
Bottle
Kecap
Figure 9.1 Traditional process for the production of Indonesian kecap.

carried out naturally and no starter was used. Consequently, several

species of fungi present in the environment grew during the fermen-
tation. Because of the high risk of growth of mycotoxin-producing
moulds, some traditional manufacturers have switched to using koji
starter for fermentation. Generally, a mixed culture of Aspergillus ory-
zae and Aspergillus sojae are used as starter but some manufacturers use
a mixed culture of A. oryzae, A. sojae and Rhizopus oryzae. The cooked
soya beans on the bamboo trays are inoculated with 0.1% koji starter
and then incubated at room temperature (Figure 9.2). Usually, the
fermentation conditions, including humidity and temperature, are not
controlled. However, the fermenting materials are stirred one to three
times during the fermentation in order to decrease the mass tempera-
ture, which rises due to metabolic heat generated during the fermen-
tation, and also to dissipate carbon dioxide produced by the mould. It
is important to remove CO2 because accumulation of CO2 can inhibit
growth of the mould. Care in this step at the 24–30 h fermentation
period is also important to control the growth of undesirable fungi.
The product of this fermentation is called koji and comprises mould-
covered beans with a high content of protease enzymes. Usually, koji
is harvested after 40–48 h incubation and before spores are formed.
Rectangular koji rooms are sometimes used in large factories and
some factories have specially designed koji rooms with automatic con-
trol of the fermentation conditions.
Figure 9.2 Koji fermentation on bamboo trays. (Courtesy of Sardjono.)

S oya S au c e 363
Figure 9.3 Traditional moromi fermentation in porcelain jars. Photograph taken in Penang, 1974.
In some traditional factories, harvested koji may be sun-dried in

order to store it for use when black soya bean is out of season and
very expensive. Generally, koji is not dried but is mixed directly with
three times its volume of brine until the final concentration of brine
is 22–24% w/v to produce moromi mash. Moromi fermentation was
traditionally carried out in earthenware jars, sometimes opened to
sunlight during the day (Figure 9.3), but big factories now use rectan-
gular concrete vats or tanks up to 20 m3 capacity (Figure 9.4). Moromi
fermentation is not inoculated. Generally, lactic acid bacteria grow
first, produce acids and when the moromi pH has dropped to 5.0–4.5
osmophilic yeasts grow. Moromi fermentation is continued for 4.5–5.0
months at ambient temperature to get the desired intense flavour and
Figure 9.4 Moromi fermentation tanks with capacities up to 20 m3 for each tank. Fermentation
is carried out at ambient temperatures for 4–5 months. (Courtesy of Sardjono.)
colour. After fermentation, the moromi is filtered to remove the beans

and coconut palm sugar is then added to the filtrate and the mixture
boiled until it becomes a viscous liquid, with degrees Brix about 75.
Several factories add sodium benzoate as a preservative but some
do not, as the high sugar concentration in kecap prevents the growth
of most potential spoilage organisms.
9.1.5 Microbiology
Formerly, almost all of the traditional kecap producers in Java did

not use a starter for their koji fermentations and, consequently, many
different mould strains grew during the fermentation. More than 12
species of mould were isolated by Sardjono et al. (1995), including
Aspergillus sojae, A. oryzae, A. versicolor, A. niger var. niger, A. flavus,
A. alliaceus, Eurotium chevalieri, E. amstelodami, Fusarium monili-
forme, Penicillium echinulatum, Rhizopus oryzae and Curvularia lunata.
Because many toxigenic fungi may grow in the traditional koji fermen-
tation, it is necessary to pay particular attention to the fermentation
and much effort has been directed towards improving the traditional
methods by introducing defined mixed culture starters for small-scale
producers.
The large-scale kecap manufacturers do use koji starters for their
mould fermentations. Most use a mixed culture of A. oryzae and A. sojae
but some also include Rhizopus oligosporus or R. oryzae. Sardjono et al.
(2004a, b) isolated a strain of A. oryzae from koji that was able to degrade
aflatoxin, which raised the possibility of developing a starter that would
eliminate aflatoxin in the event of growth of aflatoxin-producing fungi.
The microbiology of the kecap moromi or brine fermentation is
basically similar to that in soya sauce fermentation. Generally, owing
to non-aseptic operating conditions, there is much microbial contam-
ination of koji, including Bacillus spp., yeasts and proteolytic bacteria,
but they have little effect on the fermentation and most, with the
exception of salt-tolerant, spore-forming proteolytic bacteria, die dur-
ing moromi fermentation. The microorganisms that grow in moromi
fermentation are limited to specific lactic acid bacteria and halophilic
yeasts, which are resistant to high salt concentrations. Generally,
kecap manufacturers in Indonesia do not use starters for their moromi
fermentations but rely on indigenous lactic acid bacteria and yeasts.
S oya S au c e 365
In pilot-scale moromi fermentations, the concentrations of lactic

acid bacteria increased up to 107–108 cfu mL −1 and then decreased
after 8 weeks fermentation time (unpublished observations). Some
were Lactobacillus spp. but most were Pediococcus spp. Unexpectedly,
concentrations of proteolytic bacteria increased initially and then
remained unchanged for the duration of the fermentation. Similar
bacteria were also observed in moromi from a big kecap factory.
The role(s) of these bacteria remains to be elucidated but they may
be involved in the degradation of protein during moromi fermenta-
tion. Yeasts, mainly Saccharomyces rouxii, started to grow after about
5 weeks fermentation, when the pH of the moromi had decreased to
4.5, and their concentrations increased slowly over the next 9 weeks
(Figure 9.5). The metabolic activities of the microorganisms are
important in relation to flavour development in moromi. Apart from
producing lactic and acetic acids, lactic acid bacteria convert amino
acids into flavour components through transaminase reactions and
yeasts produce ethanol that reacts with organic acids to form esters,
several of which are flavour components of kecap.
9.1.5.1 Characteristics of Koji Starter Moulds

9.1.5.1.1 Aspergillus oryzae A. oryzae is a highly variable mould:
aspergilla of some isolates are predominantly uniseriate (with phialides
only) while others are predominantly biseriate with metulae and
8
Concentration (log cfu/mL)
1
0 2 4 6 8 10 12 14 16
Time (weeks)
Figure 9.5 Changes in concentrations of microbes in moromi fermentation during the prepara-
tion of Indonesian kecap. ◻, proteolytic bacteria; ○, lactic acid bacteria; Δ, yeasts. (From Sardjono,
unpublished data.)
phialides covering most of the surface of the vesicle; conidia smooth

to finely roughened, globose to ovoid; colonies spread broadly on all
media and are usually greyish-yellow to olive in colour; mycelium
white. During koji fermentation, the relative humidity of the fermen-
tation room must be high (around 95–97%) to keep the mould growing
without sporulation. Spore formation will also occur when the water
activity of a medium or koji decreases to 0.86 (Sardjono et al. 1998).
When sporulation occurs, enzyme production ceases and so it is very
important to keep the fermentation conditions favourable for growth
and enzyme production. Excess oxygen also stimulates spore forma-
tion. A. oryzae is closely related to A. flavus and it is suggested that it is
a domesticated form of A. flavus (Pitt and Hocking 1997).
9.1.5.1.2 Aspergillus sojae Conidial heads radiate to loosely columar;

aspergilla predominantly uniseriate; conidia spherical, rough walled.
This species is distinguished by its olive brown colour on Czapek yeast
extract agar and is also distinguished from A. parasiticus by its olive
brown conidial colour (Klich and Pitt 1988). The minimum water
activity for its growth is 0.78 and it starts to sporulate when the water
activity decreases to 0.86. Excess oxygen during fermentation of koji
also stimulates sporulation (Sardjono et al. 1998). A. sojae is possibly a
domesticated form of A. parasiticus (Pitt and Hocking 1997).
9.1.5.2 Characteristics of Pediococcus halophilus P. halophilus is a fac-

ultatively anaerobic Gram-positive coccus. The optimum water activ-
ity for growth is 0.99–0.94, corresponding to a salt concentration
of 2–10% (w/w) and the minimum water activity allowing growth
is 0.81, corresponding to 24% salt. The optimum temperature for
growth is 25–30°C. P. halophilus converts 1 mol of glucose to 1.7 mol
of l-lactic acid, 0.28 mol of acetic acid and 0.17 mol of formic acid
(Fukushima 1989).
9.1.5.3 Characteristics of Saccharomyces rouxii S. rouxii is responsible

for alcoholic fermentation during the brine fermentation of kecap. It
is a salt-tolerant yeast and can grow in media with a salt content of
24% and a water activity of 0.81. In the presence of 18% salt, it grows
only in the pH range 4–5. The range of temperature for growth is
20–35°C.
S oya S au c e 367
Soya bean is preferred as raw material because of its high protein con-
tent. Two kinds of soya bean are used for kecap fermentation, black
soya bean and yellow soya bean. Utilisation of yellow soya bean for raw
material is not as common as black soya bean because the growth of
starter moulds on yellow soya bean is poor compared to that on black
soya bean. Several big factories now use imported defatted soya bean
meal as raw material. In this process, roasted wheat is added equiva-
lent to 30–35% of the defatted soya bean. The addition of roasted
wheat is important in order to create voids and to promote aeration in
the otherwise tightly packed meal. This is needed to stimulate growth
and enzyme production by the mould.
9.1.7 Chemical Composition and Changes during Fermentation
9.1.7.1 Koji Fermentation During koji fermentation, A. oryzae pro-

duces amylases and proteases, while A. sojae produces protease and
glutaminase enzymes. Amylases degrade starch to sugars, provid-
ing fermentable carbohydrates for yeasts and lactic acid bacteria in
moromi fermentation. Proteases produced include alkaline proteases
with pH optima 8.3–10.5 and formed in the largest amounts; neu-
tral proteases (optimum pH 6–7); and acid proteases (optimum pH
3–4). Acid carboxy peptidase, with optimum pH 3–4, and amino
peptidase, optimum pH 5–8.5, are also produced by koji moulds
(Fukushima 1989).
Soya beans contain glutamine as well as glutamic acid in polypep-
tides. Glutamine is easily converted to pyroglutamic acid, which does
not contribute to flavour. Glutamine in soya bean is mainly converted
to glutamic acid, the most important flavour component in kecap,
by glutaminase produced by A. sojae. In experiments using local yel-
low soya beans and a mixed starter of A. oryzae and A. sojae, con-
centrations of 15 free amino acids increased during koji fermentation
while concentrations of l-asparagine were unchanged. Present at the
highest concentration was l-glutamine, followed by l-glutamic acid,
l-threonine, l-tryptophan, l-phenylalanine, l-leucine, l-lycine and
l-tyrosine (Sardjono, unpublished data). Aspartic acid and glutamic
acid are important contributors to the umami taste of kecap, while
other amino acids serve as precursors of flavour compounds formed

during moromi fermentation.
The total nitrogen content of the raw materials is relatively
unchanged during koji fermentation but the amount of formol nitro-
gen almost doubles by the end of the fermentation. Total carbohy-
drate decreases by 20–25% due to utilisation by the moulds and the
concentration of reducing sugars increases slightly.
9.1.7.2 Moromi Fermentation Soluble amino nitrogen increases dur-

ing moromi fermentation, reaching 5.6–6.5% (w/w) after 4–5 months.
Aspartic acid and glutamic acid are important contributors to the flavour
of kecap while other amino acids may contribute flavour components
through deamination reactions, producing keto-acids, and transami-
nase reactions catalysed by transaminase enzymes produced by lactic
acid bacteria. Total acidity increases from 0.35 to 1.0–1.2% (as lactic
acid). Many flavour components are formed during moromi fermen-
tation, including alcohols (2-methyl-1-propanol, 3-methyl-1-butanol
and propylene glycol), acids (acetic, propanoic, 2-methylpropanoic and
butanoic) and furans (2-furan methanol) (Sardjono, unpublished data).
Kecap is not used as a nutrient source but as a seasoning, especially

to add umami or meaty flavour to dishes. Hence, is it unlikely to
make a significant contribution to nutrient intake. Typical proximate
analysis, based on the analysis of several commercial kecap samples,
is sugar, 43–45%; water, 30–34%; ash, 6–8%; salt, 3.8–4%; protein,
1.7–1.9%; fat, 0.12–0.13%; and Fe, 25–27 mg kg−1; specific gravity
1.23–1.38 g mL −1. For controlling the final product, the factories use
75 degrees Brix as a parameter. The Indonesian National Standard for
kecap is still in the process of being revised.
The main safety concern with kecap is the possibility of the produc-
tion of mycotoxins by moulds during koji fermentation and, especially,
the formation of aflatoxins. Previously, small-scale manufacturers
usually carried out a natural koji fermentation without the use of a
S oya S au c e 369
starter culture and some kecap sold in the markets were contami-
nated with aflatoxins. However, the concentrations observed did not
exceed 15 μg kg−1 (Sardjono et al. 1992, 1995), which is within or
close to the range of permitted concentrations in foodstuffs (Adams
and Moss 2000). Utilisation of koji starter by small manufacturers is
increasing in order to obtain more control over the fermentation. The
Laboratory of Biotechnology, Department of Food and Agricultural
Product Technology, Gadjah Mada University, Yogyakarta, prepares
koji starter for kecap manufacturers.
The author has co-operated with one kecap factory to improve the
process and increase the capacity of the factory. The main objective
was focused on improving koji fermentation and preparing koji starter
using selected moulds. Koji fermentation is traditionally carried out
on bamboo trays but for this factory we developed a bioreactor, the
kojiroom, with automatic control of temperature, relative humidity,
gas composition (O2 and CO2) and with provision to mechanically
turn the koji mass with vertical and horizontal screws (Figure 9.6).
Figure 9.6 Bioreactor for koji fermentation. (Courtesy of Sardjono.)

This bioreactor proved to be very successful and the factory now uses
eight bioreactor units. Usually, big kecap factories use purpose-built
koji rooms with control of temperature, relative humidity and aera-
tion for their koji fermentation and sophisticated systems, such as the
kojiroom, have not yet been generally adopted by the industry.
9.1.11 Future Prospects
The kecap industry is growing fast in Indonesia, supported by culinary

innovation and population growth. One constraint on future devel-
opment is the difficulty of increasing production capacity to meet the
increasing market demand with current processes and technologies.
Hence, innovation and research in the areas of process design and pro-
cess technology is needed, especially to increase the process efficiency of
the moromi fermentation. Currently, the high price of raw materials is
also hampering investment for improving kecap factories in Indonesia.
9.1.12 Research Needs and Opportunities
1. What are the characteristics of the proteolytic bacteria

observed in moromi fermentation and do they have a role in
fermentation?
2. What are the relative contributions of salt concentration,
water activity (mainly derived from sugar and salt), pH value,
concentrations of organic acids, and other factors to the pre-
vention of growth of spoilage organisms in kecap? How may
the use of sodium benzoate as a preservative be avoided and
the concentration of salt minimised?
3. What is the effect of added sugar on the perception of fla-
vours/aromas generated in the fermentation?
4. The development of process designs and process technologies
more efficient than currently used methods and suitable for
small-scale Indonesian manufacturers.
References
Adams, M.R. and M.O. Moss. 2000. Food Microbiology. Second ed., p. 286.
Cambridge: Royal Society of Chemistry.
S oya S au c e 3 71
Fukushima, D. 1989. Industrialization of fermented soysauce production cen-

tering around Japanese shoyu. In Industrialization of Indigenous Fermented
Foods, edited by K.H. Steinkrauss. New York: Marcel Dekker.
Huang, H.T. 2000. Science and Civilisation in China. Volume 6 Biology and
Biological Technology. Part V Fermentations and Food Science, pp. 359–359.
Cambridge: Cambridge University Press.
Klich, M.A. and J.I. Pitt. 1988. A Laboratory Guide to the Common Aspergillus
Species and Their Teleomorphs. New Orleans: USDA and Australia: CSIRO.
Pitt, J.I. and A.D. Hocking. 1997. Fungi and Food Spoilage. Second ed. London:
Blackie Academic.
Sardjono, E.S. Rahayu and S. Naruki. 1995. Mycoflora and aflatoxin in soybean
and koji for kecap production. Journal of Biosain (PAU Biotechnology,
ITB, Bandung, Indonesia) 1(2):11–15.
Sardjono, K. Rahayu and S. Sudarmadji. 1992. Growth and aflatoxin produc-
tion by Aspergillus flavus in mixed culture with Aspergillus oryzae. ASEAN
Food Journal 7(1):30–33.
Sardjono, S. Raharjo, E.S. Rahayu and K. Rahayu. 2004a. Indigenous proteo-
lytic Aspergillus isolated from koji and its ability for aflatoxin B1 degrada-
tion. Agritech (FTP-UGM, Yogyakarta, Indonesia) 24(3):139–145.
Sardjono, S. Raharjo, E.S. Rahayu and K. Rahayu. 2004b. Detoxification of afla-
toxin B1 by extracellular enzymes of Aspergillus oryzae KKB4. Indonesian
Food and Nutrition Progress 11(1):30–34.
Sardjono, Z. Yang and W. Knol. 1998. Comparison of fermentation profiles
between lupine and soybean by Aspergillus oryzae and Aspergillus sojae
in solid-state culture systems. Journal of Agricultural and Food Chemistry
46(8):3376–3380.
10
B acillus Fermentati ons
J . DAV I D O W E N S A N D K A P T I
R A H AY U K U S WA N T O
Contents
10.1 Thai Thua Nao 374

10.1.2 Places of Production, Scale of Production,
Method of Consumption and Role in Diet 375
10.1.4 Environment Offered by the Beans in the Basket 379
10.1.4.1 Temperature 379
10.1.4.2 Gaseous Atmosphere 380
10.1.5.1 Sources of Microbes in the
Fermentation 382
10.1.5.2 Ecology of the Fermentation 382
10.1.6 Characteristics of Microorganism(s) 386
10.1.6.1 Bacillus subtilis 386
10.1.6.2 Production of Poly γ-Glutamic Acid 386
10.1.6.3 Genomic Information 388
10.1.6.4 Nutrition 388
10.1.6.5 Vitamins 388
10.1.6.6 Responses to Environmental Conditions 389
10.1.6.7 Heat Resistance of Bacillus subtilis Spores 390
Fermentation 390
10.1.8.1 Proximate Analysis 392
10.1.8.2 Fermentation Yield 392
10.1.8.3 Texture 393
10.1.8.4 Proteins 393
10.1.8.5 Lipids 394
10.1.8.6 Slime 395
3 73
10.1.8.7 Ammoniacal Aroma 396

10.1.8.8 Oligosaccharides 398
10.1.8.9 Volatile Compounds 399
10.1.9.1 Functional Food Attributes 400
10.1.10.1 Food Poisoning 401
10.1.10.2 Biogenic Amines 402
10.1.12 Future Prospects and Research Requirements 403
10.2 Indonesian Cabuk (Fermented Sesame Press Cake) 404
10.2.2 History 405
10.2.3 Places of Production, Method of Consumption
and Role in Diet 405
Processing and Fermentation 407
References 408
10.1 Thai Thua Nao

J. David Owens
Thua nao is a whole soya bean product fermented by Bacillus spp.

The fermented beans are brownish in colour, covered with a sticky
slime and possess a strong ammoniac smell. Thua nao is sold either
as balls of ground freshly fermented material or as dried thin discs
(Figure 10.1).
Thua nao is an example of a proteolytic alkaline fermentation.
Other examples include Japanese natto (Kiuchi and Watanabe 2004),
Nepalese kinema (Sarkar et al. 2002), African daddawa (Odunfa
1986), Nigerian kpaye (Omafuvbe et al. 1999; Omafuvbe and
Adenowo 2002) and Nigerian ugba (Njoku et al. 1990). All involve
BACILLUS f erm en tati o ns 375
Figure 10.1 Thua nao. (a) Fresh material wrapped in banana leaf. (b) Dried material. (Chiang
Mai, 1998. Courtesy of J.D. Owens.)
the fermentation of a proteinaceous, usually leguminous/pulses, sub-

strate by Bacillus spp. and a major metabolic activity is the hydrolysis
and oxidation of proteins, accompanied by the release of free amino
acids, peptides and ammonia.
10.1.2 Places of Production, Scale of Production, Method of Consumption

and Role in Diet
Thua nao is produced only in northern Thailand, in the provinces

of Chiang Rai, Chiang Mai, Phayao, Maehongsorn, Phrae, Nan,
Lamphun and Lampang (Leejeerajumnean 2000). It is used as a
flavouring agent in ways similar to that of shrimp paste in southern
Thailand. The moist fresh thua nao is mixed with salt and sometimes
also with garlic, onion and red peppers, wrapped in small packets
in banana leaf and cooked by steaming or roasting over an open fire
before selling or eating. The dried product is added directly to soups,
curries and other dishes as a flavouring agent and, in some areas, is
also an item of diet in its own right (Sundhagul et al. 1972).
Production of thua nao is still done using traditional methods in

the home and by small-scale manufacturers. Thua nao is on sale in
markets in northern Thailand but there is no information on the over-
all scale of its production.
Leejeerajumnean (2000) observed the methods used by a small-scale

manufacturer in Fang district, Chiang Mai, from December 1996 to
January 1997 and December 1997 to January 1998 and the following
description is based on her observations (Figure 10.2).
The process began with washing whole soya beans in water at
least twice. The beans were then boiled in aluminium pots over
wood fires for 7 h until the beans were thoroughly cooked and soft.
During cooking, water was added as required to keep the beans
Soya bean seeds (20 kg)

(Glycine max, variety SJ or SJ4)
Wash
Boil ~7 h
Drain
Put hot seeds (2–5 kg wet weight) into bamboo basket lined with fern leaves
and polypropylene mesh and having a central aeration tube of fern leaves
Cover top and sides of basket with thick plastic sheet
Incubate, ambient temperature (28–32ºC), 3 days

(outside or inside a building if it rained)
Grind
Mold into balls ~3 cm diameter
Press between plastic sheet and glass sheet to make

disks ~10 cm diameter and ~2 mm thick
Dry in the sun, 1 days
Thua nao
Figure 10.2 Small-scale production of thua nao as practised at Fang, Chiang Mai, December
1996/January 1997. (Adapted from Leejeerajumnean, A. 2000. Bacillus fermentation of soybeans:
Characterization of traditional thua nao manufacture. PhD thesis, Department of Food Science and
Technology, University of Reading, Reading, UK.)
Figure 10.3 Polypropylene mesh used to line bamboo basket. The strips are ~4 mm wide.
submerged. The cooked beans were drained and the hot beans
transferred to pre-prepared bamboo lattice or perforated plastic
baskets. The bamboo baskets were first lined with woven polypro-
pylene mesh (to contain the beans in the loosely constructed basket;
Figures 10.3 and 10.4) and then with fresh fern fronds (Thelypteris
subelata (Bak.) K. Iwats). A cylinder of fern fronds was also put in
the centre of each basket. The surface of the beans was covered with
fern fronds and the woven polythene mesh. Finally, each basket
was covered, except for the base, with a polythene bag to maintain
a humid atmosphere in the beans and reduce heat loss. Each basket
contained 3–5 kg wet weight of cooked beans. Although fern fronds
were used on this occasion, leaves of other plants, such as banana
and teak, may also be used (Leejeerajumnean 2000; Sundhagul
et al. 1972).
Figure 10.4 Bamboo baskets containing fermenting thua nao. The basket in the foreground
showing polypropylene mesh lining; the one at the rear covered with polyethylene sheet. (Chiang Mai
1997. Courtesy of A. Leejeerajumnean.)
By December 1997, the factory had changed to using plastic bas-

kets (25 × 25 × 33 cm, perforated with 0.5 cm diameter holes) instead
of bamboo ones but otherwise the process was similar.
The baskets of beans were placed outside at ambient temperature
(28–32°C) or kept inside if it rained. The fermentation was com-
plete in 3 days, when the beans appeared pale brown in colour and
were covered with a slightly sticky material and smelled strongly of
ammonia. The fermented beans were ground into a paste with a disc
mill and the paste was moulded by hand into balls ~3 cm in diameter.
These were pressed between a plastic and a glass sheet to produce
discs of ~10 cm diameter and ~2 mm thickness, which were dried
in the sun for 1 day (Figure 10.5). The manufacturer stated that the
dried thua nao could be kept for up to 6 months, although he some-
times had problems with mould growth on product made during the
rainy season
Sundhagul et al. (1972) described the production of thua nao in
two northern provinces: Lampang and Lampham. The method was
essentially as observed by Leejeerajumnean (2000) but usually only
1–2 kg of dry beans was processed and cooking was completed in
3–4 h. The beans were considered cooked when they were soft and
could be easily crushed between the fingers. The cooked beans were
transferred to a bamboo basket lined with banana leaf and the top was
also covered with banana leaf. They were allowed to ferment for 3–4
days, when they were thoroughly soft and easily crushed to a paste
between the fingers. The fermented beans were mashed into a paste
Figure 10.5 Thua nao discs drying in the sun. (Chiang Mai 1997. Courtesy of A. Leejeerajumnean.)
BACILLUS f erm en tati o ns 3 79
with salt and sometimes also with onion, garlic and red pepper. The
paste was wrapped in small packets in banana leaf and then steamed
or roasted for ~30 min. The cooked thua nao would keep for about
2 days under normal conditions and the thua nao was normally sold
in this form. If longer storage was required, sun-dried discs were pre-
pared as described by Leejeerajumnean (2000).
10.1.4 Environment Offered by the Beans in the Basket
Properties of the beans themselves and the changes during the fer-
mentation are dealt with later, and only the temperature and oxygen
and carbon dioxide concentrations in the fermenting beans are dealt
with in this section.
10.1.4.1 Temperature The beans are put into the baskets while still
hot and consequently the initial temperature of the beans is higher in
the first filled basket than in the later filled baskets. The initial tem-
peratures are sufficiently high, and the rate of cooling sufficiently slow
(Figure 10.6) to ensure that any contaminating vegetative bacteria or
75
70
65
60
Temperature (°C)
55
50
45
40
35
0 10 20 30 40 50 60 70
Time (h)
Figure 10.6 Temperature profiles of fermenting thua nao. Approximately 5 kg (wet wt) hot, cooked
soya beans were put in a basket of 25 cm diameter and 33 cm depth. Depth of beans was 25 cm and
ambient temperature, 28–32°C. Position in basket: ______, centre, 18 cm above b ottom; - - - - -, centre,
8 cm above bottom; . . . . . . . ., 3 cm from edge, 13 cm above bottom. (Adapted from Leejeerajumnean,
A. 2000. Bacillus fermentation of soybeans: Characterization of traditional thua nao manufacture. PhD
thesis, Department of Food Science and Technology, University of Reading, Reading, UK.)
fungi and fungal spores would be killed, leaving bacterial endospores

as the only possible survivors.
Temperatures in the centre of the baskets fell to a minimum of
34–37°C before increasing again, presumably due to the production
of metabolic heat by the growing bacteria. The beans reached maxi-
mum temperatures of 52–59°C, depending on their location in the
basket (Figure 10.6). Thereafter, the temperature of the beans slowly
fell, but was still ~40°C at the end of the fermentation after 72 h
incubation.
10.1.4.2 Gaseous Atmosphere Leejeerajumnean (2000) monitored the

gas composition in baskets of fermenting beans using a system that
recycled the assayed gas back to the sampling point at the centre of
the basket and thus avoided perturbing the conditions in the basket.
Oxygen concentration decreased slowly at first and then increasingly
rapidly to reach ~12% (v/v) after 24 h incubation. Thereafter, the con-
centration increased to reach ~18% by the end of the fermentation at
72 h (Figure 10.7). Changes in the concentration of carbon dioxide
complemented those of oxygen, increasing to 12% at 24 h and decreas-
ing to 2.5% at the end of the fermentation.
25
20
15
Gas (%)
10
0
–2 4 10 16 22 28 34 40 46 52 58 64 70
Time (h)
Figure 10.7 Oxygen and carbon dioxide concentrations in fermenting thua nao. Gas was sampled
from the middle of ~5 kg wet weight beans in a perforated (5 mm diameter holes) plastic container
of 25 cm diameter and 33 cm depth (depth of beans was 25 cm). ◻, oxygen; Δ, carbon dioxide.
(Adapted from Leejeerajumnean, A. 2000. Bacillus fermentation of soybeans: Characterization
of traditional thua nao manufacture. PhD thesis, Department of Food Science and Technology,
University of Reading, Reading, UK.)
It is evident that the conditions in the baskets remained aerobic at

all times and that oxygen was able to diffuse into the centre of the
bean mass and that carbon dioxide diffused out.
10.1.5 Microbiology
The microflora of thua nao has not been extensively studied but
Sundhagul et al. (1972), Leejeerajumnean (2000) and Chukeatirote
et al. (2006) monitored a few fermentations and isolated some domi-
nant bacteria from a small number of thua nao samples. Initial numbers
of bacteria, presumably entirely spores, in the cooked beans were in
the range 103–105.8 cfu (g wet wt)−1 and these grew to final concentra-
tions of 108–1010 cfu (g wet wt)−1 (Figure 10.8). The dominant bacteria
isolated were Bacillus spp., including mainly Bacillus subtilis but also
Bacillus pumilus, Bacillus megaterium and Bacillus cereus. Apart from
bacilli, lactic acid bacteria grew and reached populations of 107 cfu
(g wet wt)−1 at 72 h. The lactic acid bacteria were not characterised but
did not include enterococci. Yeasts were not detected but some moulds
appeared at the end of the fermentation (Leejeerajumnean 2000).
It is obvious that thua nao fermentation is similar to other proteo-
lytic alkaline fermentations conducted around the world. The natural
10
8
Bacteria (log cfu/g)
0
–2 4 10 16 22 28 34 40 46 52 58 64 70
Time (h)
Figure 10.8 Concentrations of bacteria during traditional thua nao fermentation. ◻, total via-
ble count; Δ, bacterial spores; ⚪, lactic acid bacteria. (Adapted from Leejeerajumnean, A. 2000.
Bacillus fermentation of soybeans: Characterization of traditional thua nao manufacture. PhD the-
sis, Department of Food Science and Technology, University of Reading, Reading, UK.)
fermentations are all characterised by the presence of a diversity of

Bacillus species, though B. subtilis is usually predominant. B. subtilis
strains in thua nao are quite diverse (Inatsu et al. 2002, 2006), in con-
trast to the single strains used to make Japanese natto.
10.1.5.1 Sources of Microbes in the Fermentation The soya beans for

thua nao production are boiled for 3–7 h. Hence, while it is evi-
dent that all vegetative bacteria and fungi and fungal spores must be
killed, it is possible that bacterial endospores might survive. Sarkar
(2000) demonstrated that spores of B. subtilis DK-WI survived soak-
ing (20–25°C, 12–20 h) and cooking (~90 min) treatments in the
preparation of kinema and the cooked soya bean grits carried a pop-
ulation of 102–103 cfg g−1. Evidently, processes in which the cooking
allows Bacillus spores to survive but eliminates other microbes pro-
vide an environment where Bacillus spp. face little or no competition,
aiding their domination of the fermentation. However, where very
long cooking times are used, it seems likely that spores may also be
eliminated and in these cases the cooked raw material is liable to be
contaminated with a great diversity of microbes and the domination
of the fermentation by bacilli must involve factors other than initial
inoculation levels.
10.1.5.2 Ecology of the Fermentation Steinkraus (1991, 1996) states

that ‘… an alkaline pH in combination with ammonia control the
fermentation’ and ‘in all these fermentations, the alkaline pH helps
B. subtilis dominate the fermentations by hydrolysing protein to
amino acids and peptides and releasing ammonia, which makes the
substrate unsatisfactory for invasion by microorganisms that might
spoil the product’. This hypothesis has been promulgated by other
authors, such as Wang and Fung (1996) and Parkouda et al. (2009).
However, no evidence to support the hypothesis was provided then or
since. Leejeerajumnean et al. (2000) investigated the sensitivity of 40
bacterial strains to alkaline pH (OH− concentration), ammonium ion
(NH4 +) and ammonia (NH3). The strains included 26 Bacillus strains
and one Enterococcus sp. from alkaline fermented foods, Bacillus pas-
teurii (an alkaliphilic, ammonia-requiring species), and eight Gram-
positive bacteria and four Gram-negative bacteria not specifically
associated with alkaline fermented foods. All the strains grew at
pH 7, 8 and 9 (OH− concentrations of 10−4, 10−3 and 10−2 mmol L −1),

though some, such as Escherichia coli K12, grew much more slowly
at pH 9 than at pH 7 or 8. All strains also grew in the presence of
931 mmol L −1 of NH4+ (7.05 g (NH4)2SO4 L −1), the highest concen-
tration tested, at pH 7. However, NH3, as an uncharged molecule
able to cross the cell membrane in a manner analogous to uncharged
weak organic acids (Klocke et al. 1972; Warren 1962), was inhibi-
tory to many of the strains. All 40 bacterial strains grew in the pres-
ence of 5 mmol L −1 NH3 at pH 7.0 and 8.0 (Table 10.1), but at higher
concentrations, the growth of some of the bacteria was inhibited. At
500 mmol L −1, the highest concentration tested, only B. pasteurii and
B. pumilus grew. Cultures that failed to show growth after 2 days of
incubation still contained viable cells able to grow on nutrient agar,
suggesting that the minimum growth-inhibiting concentrations of
NH3 were bacteriostatic rather than bacteriocidal. However, it is also
possible that the survival was due to the presence of spores.
Bacteria particularly sensitive to NH3 (inhibited by 25 mmol L −1) or
especially tolerant to NH3 (growth in the presence of 300 mmol L −1)
included both Gram-positive and Gram-negative strains and there
was no evidence for sensitivity or tolerance being associated with any
specific bacterial genera.
Bacteria isolated from alkaline food fermentations included strains
relatively sensitive to NH3 (inhibited by 50 mmol L −1) and strains that
were relatively tolerant to NH3 (grew in the presence of 300 mmol L −1).
In fact, the range of tolerance shown by Bacillus strains isolated
from alkaline fermented foods is no different from that of bacteria
not isolated from such ferments (Table 10.1). It is also the case that
the initial pH (6.4–6.8) and total ammonia levels (NH4+ + NH3,
1–30 mmol kg−1 wet weight; Allagheny et al. 1996; Sarkar et al. 1993;
Figure 10.9) of cooked soya beans are too low to have any differential
effect on the initial microflora. In the final fermented products, the
maximum total ammonia concentrations are 120–300 mmol kg−1 wet
weight (Allagheny et al. 1996; Sarkar et al. 1993; Figure 10.9), cor-
responding to NH3 concentrations at pH 8.0–8.5 of 6–45 mmol kg−1
wet weight. It is possible that NH3 concentrations are higher in the
surface layer of microbial growth than values averaged over the entire
bean, but this concentration effect may be counteracted by the near-
neutral pH value during the phase of maximum microbial growth
Table 10.1 Minimum Inhibitory Concentrations of NH3 for Bacteria from Alkaline Fermented
Foods and Some Other Bacteria in a Chemically Defined Medium, pH 8.0 or 9.0 at 37°C
MINIMUM INHIBITORY
CONCENTRATION OF
NH3 (MMOL L−1) SPECIES STRAINa SOURCE
25 Enterococcus faecalis NCTC 00775
Escherichia coli K12 NCTC 10538
Listeria innocua NCTC 11288
50 Bacillus subtilis T1, T2, T19, Thua nao
T22, T33, T39
Bacillus subtilis N3, N4, N5 Natto, Japan
Bacillus subtilis DK-W1 Kinema, Sikkimb
Bacillus cereus T31, T38 Thua nao
Bacillus megaterium T3, T4, T21, Thua nao
T34, T37, T40
Pseudomonas aeruginosa NCTC 10299
150 Bacillus subtilis T5 Thua nao
Bacillus subtilis DA2 Laboratory soya bean
fermentationc
Bacillus cereus NCIMB 9373
300 Bacillus subtilis T20, Thua nao
Bacillus subtilis N1, N2 Natto, Brazil
Bacillus subtilis DA1 Laboratory soya bean
fermentationc
Bacillus subtilis NCIMB 3610
Bacillus cereus T41 Thua nao
Bacillus licheniformis ATCC 39302
Enterococcus faecium DK-C1 Kinema, Sikkimb
Micrococcus luteus NCDO 0982
Staphylococcus aureus NCDO 0949
Salmonella typhimurium NCIMB 10248
500 Bacillus subtilis T36 Thua nao
Proteus morganii NCIMB 00067
>500 Bacillus pumilus NCIMB 9369
Bacillus pasteurii NCIMB 8841
Source: Adapted from Leejeerajumnean, A., J.M. Ames and J.D. Owens. 2000. Journal of Applied
a NCTC, National Collection of Type Cultures; NCIMB, NCIMB Ltd. Aberdeen AB21 9YA, Scotland;
ATCC, American Type Culture Collection; NCDO, National Collection of Dairy Organisms.
b Sarkar et al. (1993).
c Allagheny et al. (1996).
7.8 140
7.6 120
Ammonia (mmol/kg wet wt)

7.4
100
7.2
80
pH value
7.0
60
6.8
40
6.6
6.4 20
6.2 0
0 12 24 36 48 60 72
Time (h)
Figure 10.9 pH value and ammonia concentration during thua nao fermentation. ◻, pH value;
Δ, ammonia concentration. (Adapted from Leejeerajumnean, A. 2000. Bacillus fermentation of
soybeans: Characterization of traditional thua nao manufacture. PhD thesis, Department of Food
Science and Technology, University of Reading, Reading, UK.)
(Figure 10.9). Thus, it is evident that NH3 concentrations are unlikely

to be high enough during the fermentation to affect the growth of
spoilage or pathogenic bacteria. This conclusion is supported by many
reports of the isolation of a great diversity of bacteria from traditional
alkaline fermented foods (Leejeerajumnean 2000; Nout et al. 1998;
Ogbadu and Okagbue 1988; Sarkar et al. 1994). Hence, it must be
concluded that the ecology of these fermentations is not controlled by
alkaline pH in combination with ammonia.
The question remains as to what leads to the almost invariable
dominance of natural fermentations of cooked legume seeds, exposed
to diverse sources of contaminating microbes, by Bacillus spp. and
especially B. subtilis. Possibly, it is a consequence of the survival of
their spores on the dry seed substrates, the survival of their spores and
the elimination of vegetative bacteria during the cooking of the beans,
and the rapid germination and growth of the spores on the cooled,
cooked beans. Tamang (2003) recorded concentrations of spores dur-
ing the preparation of kinema (cfu g−1): raw soya beans, 103.8; soaked
soya beans, 103.5; cooked soya beans, 101.8; fresh kinema, 108.6; kinema
3-day-old, 108.6. B. subtilis was found in all stages and the implication
is that it survived the cooking stage, though this was not explicitly
claimed by the author. Similarly, Sarkar (2000) observed that boiling
soya beans in an open cooker for 90 min killed all members of the
microflora except B. subtilis spores, reducing the total microbial popu-

lation from 105–107 to 102–103 cfu g−1. He concluded that B. subtilis
spores on the soya beans survived the soaking and cooking treatments
to initiate and effect the fermentation of kinema. In contrast, in other
Bacillus fermentations, such as that for ugba, which involves two boil-
ing stages, the second boiling completely eliminated the microflora
(Njoku et al. 1990). The authors considered the leaves used for pack-
aging the cooked beans to be the major source of organisms respon-
sible for fermentation. Similarly, in traditional natto manufacture, the
rice straw in which the cooked beans are wrapped is considered to
be the source of B. subtilis (Steinkraus 1986). It is also likely that the
practice of putting the hot beans directly into the fermentation con-
tainer serves to eliminate heat-sensitive microbes, leaving only heat-
resistant bacterial endospores.
10.1.6 Characteristics of Microorganism(s)
10.1.6.1 Bacillus subtilis B. subtilis is a Gram-positive spore-forming

bacterium with relatively large motile cells.
Inatsu et al. (2002) characterised B. subtilis strains from fermented
soya bean foods from Southeast Asia, including eight from thua nao,
and compared them with natto strains. Natto starter strains of B. sub-
tilis produce poly γ-glutamic acid and have a requirement for biotin
(Kubo et al. 2011). Production of poly γ-glutamic acid in these strains
is regulated by a quorum-sensing system and is genetically unstable
(Inatsu et al. 2002).
Of the eight strains from thua nao, all produced protease and
amylase but only three produced poly γ-glutamic acid and only four
required biotin (however, the agar medium used to identify biotin
requirement may not be adequate to identify a requirement in all cases
[Graffham et al. 1995]). The insertion sequence, IS4Bsu1 harboured
by natto strains, was present in all eight of the thua nao strains in vari-
ous copy numbers. It was also observed that the nucleotide sequences
of IS4Bsu1 in four thua nao isolates (only four were examined) were
identical with that of natto strains (Inatsu et al. 2002).
10.1.6.2 Production of Poly γ-Glutamic Acid According to Sundhagul

et al. (1972), ‘the beans were considered properly fermented
when they were covered with sticky, viscous, colourless material

…’. However, Dajanta et al. (2012) observed only slightly sticky
material on samples of thua nao collected from six markets in
Chiang Mai. Of the nine thua nao B. subtilis strains examined by
Leejeerajumnean (2000), four produced slime but to a far lesser
extent than two natto strains (Figure 10.10). Similarly, Inatsu et al.
(2002) observed poly γ-glutamic acid formation only in four of
eight strains from thua nao.
The slime produced by B. subtilis is composed of poly γ-glutamic
acid (the amide bond is between the amino group and the γ-carboxyl
group rather than with the α-carboxyl group, as in proteins), made
of l- and d-glutamic acid residues, and polysaccharide and may con-
stitute 2% of the dry weight of natto (Kiuchi and Watanabe 2004;
Teng et al. 2004). Poly γ-glutamic acid is a water-soluble polymer
with a molecular mass of 105–106 (Ashiuchi 2013). It is a major
capsular component in virulent Bacillus anthracis and plays a role in
defence against phagocytosis (Tran et al. 2000). In B. anthracis, the
formation of poly γ-d-glutamic acid is promoted by elevated carbon
dioxide/bicarbonate concentration (Thorne 1993). In B. subtilis, poly
γ-glutamic acid production is under the control of a quorum-sensing
system and is synthesised at high cell densities, such as occur in the
biofilms on the surface of the beans in thua nao and natto (Vlamakis
et al. 2013).
Figure 10.10 Slime production by Bacillus subtilis N5 (isolated from natto). (Courtesy of
J.D. Owens.)
10.1.6.3 Genomic Information B. subtilis strain 168 was the first Bacillus
and the first Gram-positive bacterium to have its genome sequenced and
has long been the model organism for Firmicutes (low-G + C Gram-
positive bacteria) (Barbe et al. 2009; Sonenshein et al. 2001). The genome
of one natto strain has been sequenced (Nishito et al. 2010). Analysis of
the genome leads to the suggestion that B. subtilis is adapted to life as
an epiphyte and that its preferred niche is the phyloplane and the rhi-
zoplane, though it is also able to grow in soil and in the gastrointestinal
tract (Earl et al. 2008). Among the properties associated with growth
on surfaces is the formation of biofilms (i.e., surface-associated multicel-
lular aggregates where cells are embedded in a matrix of extracellular
polymers). The matrix is composed of protein and polysaccharide and
may include poly γ-glutamate, though it is not an essential component of
B. subtilis biofilms (Vlamakis et al. 2013). B. subtilis also typically pro-
duces a diversity of surfactant and antibiotic compounds, which permit
smooth development on planar surfaces (Barbe et al. 2009). These fea-
tures are of relevance to the growth of B. subtilis on the surface of beans
in fermented products such as thua nao and natto.
10.1.6.4 Nutrition Bacillus subtilis is nutritionally undemanding and

can grow in minimal medium on glucose and ammonium (Logan and
de Voss 2009). It utilises a relatively restricted range of sugars, amino
acids and other compounds as carbon and energy sources and ammo-
nium, nitrate, amino acids, some purines, urea, uric acid, allantoin
and some peptides as sole nitrogen sources.
B. subtilis produces a diversity of extracellular proteases, amylase,
lipases, cellulase, β-mannanase, xylanase, β-galactosidase and phy-
tase (Eggert et al. 2003; Ferrari et al. 1993; Wongputtisin et al. 2012).
10.1.6.5 Vitamins B. subtilis strains generally grow in minimal

media without added vitamins (Logan and de Voss 2009) but 56 of 59
strains collected from rice straw that had been used to initiate natto
fermentations and which produced poly γ-glutamic acid on steamed
soya beans uniformly required biotin for growth (Kubo et al. 2011).
The authors suggested that biotin auxotrophism was tightly linked to
natto fermentation. It is not known whether or not B. subtilis strains
associated with thua nao or other alkaline food fermentations are
biotin-requiring.
10.1.6.6 Responses to Environmental Conditions

10.1.6.6.1 Temperature The optimum growth temperature is
20–30°C with a minimum of 5–20°C and a maximum of 40–55°C.
Three B. subtilis strains examined by Warth (1978) had optimum
growth temperatures of 46°C. The maximum growth temperatures
for these strains were 53–54°C and at only slightly higher tempera-
ture (55–57°C) did the vegetative cells die with a D value of 10 min.
Maximum growth rates observed were 2.8–3.2 h−1, corresponding to
doubling times of 14.8–13 min.
10.1.6.6.2 pH Value The pH range for growth is 5.5–8.5 (Logan

and de Voss 2009).
10.1.6.6.3 Oxygen Bacillus subtilis is primarily an aerobic organ-

ism but can grow anaerobically with nitrate as the terminal electron
acceptor or by fermentation (Logan and de Voss 2009). Folmsbee
et al. (2004) showed that B. subtilis strain 168 (ATCC 23857) and
three Bacillus mojavensis strains (closely related to B. subtilis; Logan
and de Voss 2009) required four deoxyribonucleosides or DNA for
growth under strictly anaerobic conditions. It was suggested that
B. subtilis strain 168 lacks a ribonucleotide reductase, required to
reduce ribonucleotides to deoxyribonucleotides, able to function
under anaerobic conditions.
10.1.6.6.4 Carbon Dioxide CO2 concentrations of up to 40% (v/v)

had no effect on the rate of growth or total amount of growth pro-
duced by B. subtilis T36 (Leejeerajumnean 2000) and B. subtilis DA2
appeared to be unaffected by 50% CO2 and also grew well in the
presence of 80% CO2, although it is possible that growth was slightly
restricted (Allagheny et al. 1996).
10.1.6.6.5 Ammonia The range of tolerance shown by B. subtilis

strains isolated from thua nao to ammonia is no different from that
of bacteria not isolated from such ferments, with some strains being
inhibited by 50 mmol L −1 and one strain tolerating 300 mmol L −1
(Table 10.1). Most strains exhibited their normal cellular morphol-
ogy in the presence of high concentrations of NH 3 but B. subti-
lis N5 formed long curled or helical filaments in the presence of
50 mmol L −1 NH 3 (Leejeerajumnean 2000), similar to those

described by Hashimoto et al. (1994).
10.1.6.7 Heat Resistance of Bacillus subtilis Spores If B. subtilis spores

are to survive the cooking of soya beans for some hours, they need
to exhibit adequate heat resistance. Some reported decimal reduc-
tion times are listed in Table 10.2. The vast majority of strains have
D100°C in the range 2–12 min. However, strain B692, isolated from
heated milk (Warth 1978) has a much greater heat resistance with a
D111°C of 11 min. Spores with a D100°C in the range 2–12 min would
not appear to be sufficiently heat resistant to survive even 1 h of
boiling and provide a final population of 103.5 cfu g−1 (Figure 10.8)
in the cooked soya beans unless the initial levels were extraordi-
narily high.
It is possible that the traditional process selects strains that are par-
ticularly heat resistant. However, there are no published reports of the
heat resistance of Bacillus strains from thua nao or other traditional
alkaline fermented foods.
Information on the properties of soya beans is provided in other chap-

ters (see Chapter 1) and is not repeated here. Thua nao is made from
whole soya beans and, hence, includes the testa, which has a high fibre
content.
The major activity of the microorganisms in the fermentation is the

hydrolysis of proteins and the utilisation of the released amino acids.
The oxidation of amino acids is accompanied by the release of ammo-
nium and, for most amino acids, also of hydroxyl ions. For example,
glycine:

( )
CH 2 NH3 + COO− + 1.5Ο 2 → 2CO2 + NH4 + + OH−
Thus, oxidation of amino acids is inevitably accompanied by a

rise in the pH of the substrate and the pH value of fermenting thua
Table 10.2 Reported Values for Heat Resistance of Spores of Bacillus subtilis Strains
STRAINa TEMPERATURE (°C) CONDITIONS D VALUE (MIN) SOURCE
A 100 Butterfield’s buffer, pH 7.2 (KH2PO4/Na2CO3) 11.6 Ababouch et al. (1995)
9372 100 Purified water ~5.3 Srimani et al. (1980)
9372 100 Ion exchanged water, pH 7.2 5 Srimani and Loncin (1980)
NCIMB 8054 95 Phosphate buffer, pH 7.0 8 Molin and Snygg (1967)
112 Phosphate buffer, pH 7.0 1 Molin and Snygg (1967)
168 98 45 mmol L−1 K phosphate buffer, pH 6.0 ~1.2 Jagannath et al. (2003)
45 mmol L−1 K phosphate buffer, pH 7.0 6.3 Jagannath et al. (2003)
45 mmol L−1 K phosphate buffer, pH 8.0 10 Jagannath et al. (2003)
168 98 Whole milk, pH 6.4 2.4 Jagannath and Tsuchido (2003)
Skim milk, pH 6.7 2.4 Jagannath and Tsuchido (2003)
From canned asparagus 100 McIlvaine buffer, pH 6.1 (Na2HPO4/citric acid) ~4.2 Condon and Sala (1992)
100 McIlvaine buffer, pH 7.1 ~16.6 Condon and Sala (1992)
100 Homogenised meat balls, pH 7.0 2 Condon and Sala (1992)
Homogenised lentils, pH 7.1 2.1 Condon and Sala (1992)
BACILLUS f erm en tati o ns
P 99 50 mmol L−1 phosphate buffer, pH 7.0 10 Warth (1978)

168 97 50 mmol L−1 phosphate buffer, pH 7.0 10 Warth (1978)
B692, from heated milk 111 50 mmol L−1 phosphate buffer, pH 7.0 10 Warth (1978)
Various 100 Various buffers 1.9–31.6 Cited by Jagannath et al. (2003)
a NCIMB, NCIMB Ltd., Aberdeen, UK; 168 is presumably Marburg strain; other numbers are institution numbers.
3 91
nao typically rises from an initial value of 6.6–6.8 to 7.6–8.0 or

higher (Figure 10.9). In some Bacillus fermentations of soya bean,
an initial decline in pH has been observed (Sarkar et al. 1993;
Figure 10.11). Presumably, this is a consequence of the utilisation
of non-protein compounds for growth, but there are no data on
what these might be.
10.1.8.1 Proximate Analysis The proximate composition of market

samples of thua nao (Table 10.3) is similar to that of whole soya
beans (Chapter 1, Table 1.5). As has been pointed out elsewhere
(see Chapter 1), proximate analysis data give little or no informa-
tion on the chemical changes that occur during fermentation and,
especially with aerobic fermentations where a substantial amount of
the substrate may be utilised by the microorganisms, can be quite
misleading.
10.1.8.2 Fermentation Yield There appears to be no data avail-

able for thua nao but Allagheny et al. (1996) measured the yield of
cooked soya beans fermented by B. subtilis DA2 at 35°C. The major
loss of dry matter occurred during the first 24 h of fermentation,
5.5
5.0
Nitrogen (mol/kg initial dry matter)
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
–0.5
0 12 24 36 48 60 72
Time (h)
Figure 10.11 Changes in amounts of nitrogenous compounds during Bacillus subtilis T36 fer-
mentation of soya bean cotyledons in air at 35°C. Values are presented relative to the initial mass
of the cooked cotyledons. ◻, protein-N (total-N−TCA soluble-N); Δ, trichloroacetic acid (TCA) sol-
uble-N (ammonia, amino acids, small peptides and some proteins); ⚪, ammonia-N. (Adapted from
Allagheny, N. et al. 1996. International Journal of Food Microbiology 29:321–333.)
Table 10.3 Proximate Composition of Commercial Fresh Thua Nao

COMPONENT CHEMICAL COMPOSITION (G [KG DRY WEIGHT]−1)
Crude protein 390–420
Fat 170–250
Fibre 120–280
Carbohydrate 220–260
Reducing sugar 27–77
Ash 47–57
Source: Adapted from Dajanta, K., E. Chukeatirote and A. Apichartsrangkoon.
2012. Chiangmai Journal of Science 39:562–574 and Sundhagul,
M., P. Smanmathuroj and W. Bhodacharoen. 1972. Thai Journal of
Agricultural Science 5:43–56.
Note: Moisture contents were 560–650 g (kg fresh thua nao)−1 and 110 g
(kg dried thua nao)−1.
and approximated to 16% of the initial dry matter. After 72 h

incubation, by which time the soya bean cotyledons were over-
fermented and smelled strongly of ammonia, 22% of the initial
dry matter had been lost. These values compare with a 14% loss
during the fermentation of African locust beans at 37°C for 72 h
(Ikenebomeh et al. 1986). These values appear very high relative to
the amount of bacterial biomass formed. If B. subtilis cells have a
volume of 5 μm3, then a population of 1010 cfu g−1 would constitute
a volume of 5 × 10 −2 cm3 g−1, and a mass of ~1 × 10 −2 g (g thua nao
fresh weight) −1. It is difficult to reconcile a final biomass concentra-
tion of 10 mg g−1 with a loss of dry matter during the fermentation
of ~0.15 g g−1.
10.1.8.3 Texture During fermentation, the beans become softer

and, in fact, one criterion for judging when the fermentation is
complete is to squeeze the beans between the fingers. The fermen-
tation is considered complete when the beans are thoroughly soft
and easily crushed to a paste between the fingers (Sundhagul et al.
1972).
10.1.8.4 Proteins The major metabolic activity in fermentation

is the hydrolysis of proteins with the accumulation of free amino
acids, some of which are oxidised to generate energy, with the
release of free ammonium, and some, presumably, are assimilated
by the bacteria. However, while there are numerous articles that

report on the proteolytic activity in Bacillus fermentations, there
are relatively few reports on the actual extent of protein hydrolysis
and its products.
Dajanta et al. (2011a) measured concentrations of free amino
acids of 10 g (kg dry matter) −1 in soaked, dehulled and boiled soya
beans and this increased to 76 g (kg dry matter) −1 in naturally fer-
mented thua nao after 72 h at 42°C. These authors noted that the
concentrations in soya beans cooked by autoclaving at 121°C for
40 min and in the thua nao prepared from them by pure culture
B. subtilis fermentation were twice the concentrations observed in
the beans and thua nao prepared by the traditional process. In con-
trast, Sarkar et al. (1997a), with soaked and autoclaved (121°C for
15 min) whole beans, measured total free amino acid concentra-
tions of only 2.0 g (kg dry matter)−1 in the unfermented beans and
82 g (kg dry matter)−1 in the fermented product. Leejeerajumnean
(2000) monitored ammonia concentrations during traditional thua
nao fermentations (Figure 10.9). Ammonia concentration increased
from an initial value of 0.5 to 5.0 g (kg dry matter)−1 after 72 h at
28–32°C.
Measurement of the concentrations in the starting material
and final product does not show how the total amounts of mate-
rials have changed. No estimates have been made for thua nao
but Allagheny et al. (1996) attempted to estimate the amount of
protein degraded during Bacillus fermentation of soya bean coty-
ledons (Figure 10.11). After 24 h incubation, ~32% of the protein-
N (estimated as total-N–trichloroacetic acid soluble-N) had been
degraded and at 72 h, this figure was 41%. This compares with
50–60% of soya bean proteins being hydrolysed in natto (Teng
et al. 2004). If one assumes that TCA soluble-N is primarily amino
acids and ammonium, it can then be estimated (using a conversion
factor of N × 5.7) that the amount of amino acids increased from
5.0 g (kg initial dry matter) −1 to 94 g at 24 h and 120 g (kg initial
dry matter)−1 at 72 h.
10.1.8.5 Lipids The only detailed investigation into the fates of

lipids during Bacillus fermentations of soya beans is that by Sarkar
et al. (1996). They measured a crude lipid content of 190 g (kg dry
matter)−1 in cooked soya bean grits and 240 g (kg dry matter)−1 in
product fermented by B. subtilis DK-W1 at 37°C for 48 h, an increase
of 26%. If the increase was entirely a consequence of the oxidation
of carbohydrate and amino acids to CO2, it would imply that 21% of
the total dry matter had been oxidised. This figure is similar to the
16–22% loss of material observed by Allagheny et al. (1996).
The concentration of free fatty acids increased by a similar pro-
portion (27%) from 6.95 g (kg dry matter)−1 in the cooked soya bean
grits to 8.85 g (kg dry matter)−1 in the fermented material. This would
suggest that none had been utilised by the bacteria and that neither
had any fats been hydrolysed by bacterial lipases. The relative absence
of lipase activity in Bacillus soya bean fermentations has been com-
mented on by other authors (Chukeatirote et al. 2006; Kiuchi and
Watanabe 2004; Omafuvbe et al. 2000). B. subtilis is known to be able
to produce lipases but Eggert et al. (2003) observed that the expres-
sion of lipase synthetic gene, lipA, was repressed by high amino acid
concentrations.
Sarkar et al. (1996) also assayed the plant phytosterols, campes-
terol, stigmasterol and β-sitosterol, and observed a 58% increase dur-
ing the fermentation, from 0.80 g (kg dry matter)−1 in the cooked grits
to 1.27 g (kg dry matter)−1 in the fermented grits. It is difficult to
reconcile this large increase with the smaller increases observed in
concentrations of crude lipid and free fatty acids. It is possible that
steryl glycosides were not hydrolysed in the saponification procedure
used (Lagarda et al. 2006) but were hydrolysed by bacterial glycosi-
dases and that this accounts for the apparent increase in phytosterols
by fermentation.
The concentrations of crude lipid, free fatty acids and phytosterols
in the cooked beans were similar to those in raw soya beans, indi-
cating that the soaking (22–25°C for 16 h) and cooking (121°C for
15 min) procedures had no effect on them.
It may be concluded that B. subtilis DK-W1 had negligible interac-
tion with the soya bean lipid components during the fermentation but
may have produced glycosidases that hydrolysed steryl glycosides.
10.1.8.6 Slime While the growth of B. subtilis on soya beans would

appear to conform to the definition of a biofilm, the composition
of the matrix on soya bean does not correspond to the usual matrix
composition of B. subtilis biofilms. Typically, B. subtilis biofilms have

an exopolysaccharide and protein matrix and poly γ-glutamic acid is
important primarily in submerged biofilms (Vlamakis et al. 2013). In
contrast, the matrix on natto is composed mainly of γ-polyglutamic
acid and fructan (Teng et al. 2004).
10.1.8.7 Ammoniacal Aroma Fresh thua nao, like most other tra-

ditional Bacillus fermented products, smells strongly of ammonia
(Dajanta et al. 2012; Sundhagul et al. 1972). In contrast, natto does
not and, in fact, strains of bacteria that do produce a strong ammo-
niacal aroma are excluded from use in natto production (Kubo et al.
2011). Cooking thua nao largely eliminates the ammonia and so it is
not a problem in practice.
Many authors refer to ammonia production being a consequence
of proteolysis but this is, of course, not strictly true. The breakdown
and utilisation of proteins proceeds step-wise. The first step is the
hydrolysis of peptide bonds by protease enzymes (proteolysis), with
the formation of peptides and free amino acids. Only with the deami-
nation of amino acids as they are oxidised to provide the energy for
growth is ammonia released. Ammonium ion (NH4 +) is not volatile
while ammonia (NH3) is. Thus, the presence or absence of an ammo-
nia aroma depends on the total ammonium + ammonia concentra-
tion and the pH value. Ammonium has a pKa of 9.25, and, at the
total (NH4 + + NH3) concentrations found in Bacillus fermentations
of soya bean cotyledons, an ammonia aroma was detectable at pH
values above about 8.2 and not detectable at values below about 7.9
(Allagheny et al. 1996).
10.1.8.7.1 Control of Ammonia Formation There are no reports of

attempts to restrict ammonia formation in thua nao but Allagheny
et al. (1996) conducted several experiments with African locust beans
and soya beans. These authors showed that the rise in pH value and the
consequent ammonia aroma could be prevented by adding a humec-
tant mid-way (15 h) through fermentation. With NaCl (0.5–1.5 mol
[29–88 g] kg−1 wet wt), enzymic activity was also inhibited and the
soya bean cotyledons did not soften during the fermentation. With
glycerol (1.7 mol [156 g] kg−1 wet wt), the cotyledons softened and the
pH remained at 6.8, ensuring that no ammonia aroma was detectable.
Allagheny et al. (1996) also investigated the effects of modified atmo-

spheres on pH and ammonia aroma. They showed that the amount
of growth (assessed visually) and the production of ammonia aroma
were unaffected by 50% CO2 (with 50% O2). With 80% CO2 (20%
O2), growth was little affected but the rise in pH was reduced and no
ammonia aroma was present. However, the main determinant of pH
rise and ammonia aroma was the amount of O2 provided initially. By
restricting access to O2, it was possible to have good proteolysis without
the increase in pH value that leads to volatilisation of ammonia. They
suggested that the fermentation be conducted in a sealed container
with an initial supply of 130–175 mL air (g cooked cotyledons)−1.
Leejeerajumnean (2000) further investigated the role of increased
atmospheric CO2 concentrations and showed that, while the final pH
of beans fermented in an atmosphere of CO2:air of 40:60 (% v/v) was
lower than that of beans fermented in air, the growth of B. subti-
lis was very similar in air or in the presence of 12.5–40% (v/v) CO2
(Figure 10.12). In fact, growth was faster in the presence of added
CO2 than in air, though the final populations were practically iden-
tical. The concentrations of free ammonium (0.10–0.12 mol kg−1
11.0 8.4
10.5 8.2
10.0 8.0
9.5 7.8
9.0
Bacteria (cfu/g)
7.6
8.5
pH value
7.4
8.0
7.2
7.5
7.0 7.0
6.5 6.8
6.0 6.6
5.5 6.4
0 6 12 18 24 30 36
Time (h)
Figure 10.12 Effect of carbon dioxide concentration on growth of Bacillus subtilis T36 and pH
value of fermenting soya beans incubated in mixtures of carbon dioxide and air at 42°C. ◻, bacterial
concentration in air; Δ, bacterial concentration in mixtures of CO2 (12 or 25 or 40% v/v) and air; ○,
pH in air; ×, pH in CO2:air, 12:88; ◇, pH in CO2:air, 25:75; +, pH in CO2:air, 40:60 (% v/v). (Adapted
from Leejeerajumnean, A. 2000. Bacillus fermentation of soybeans: Characterization of traditional
thua nao manufacture. PhD thesis, Department of Food Science and Technology, University of
Reading, Reading, UK.)
wet wt) after 36 h incubation were also similar in all the ferments
and in ferments where CO2 was replaced by N2. Thus, the observed
effects were not due to differences in O2 concentration but due to CO2
restricting the rise in pH by acting as a pH buffer. These observations
suggest that excessive pH rise and ammonia aroma might be pre-
vented by incubating Bacillus fermentations in closed containers and
allowing the CO2 produced by respiration to accumulate and buffer
the pH change in the beans.
Leejeerajumnean (2000) also evaluated phosphate (pKa, 7.2) as a
pH buffer agent. Potassium phosphate buffer, pH 6.5, at the rate of
0.1 mol (kg wet wt soya beans)−1, added to the beans prior to ster-
ilisation by autoclaving, was effective in restricting the rise in pH in
the fermenting beans and the development of an ammoniacal aroma,
without having any effect on the proteolytic activity or the amount of
ammonium formed.
Japanese natto does not have a strong ammonia aroma, and it seems
that this is achieved by the selection of appropriate starter cultures
(Kubo et al. 2011) and stopping the fermentation by refrigeration or
freezing before excessive rise of pH value. It is also possible that the
accumulation of CO2 in the incubation rooms (up to 15% [v/v]; Teng
et al. 2004) serves as a pH buffer in the fermenting natto.
In conclusion, it does seem that there are various approaches that
might be further investigated to limit the development of strong
ammoniacal aromas in thua nao. Carrying out the fermentation in a
closed container appears easy but carries the risk that, if the container
does not contain sufficient air, anaerobic conditions might develop,
with the concomitant risk of growth of anaerobic pathogens.
10.1.8.8 Oligosaccharides Soya beans contain substantial amounts of

oligosaccharides, sucrose (63 g kg−1 dry weight), raffinose (19 g) and
stachyose (43 g; Sarkar et al. 1997b). Data are not available on their
fates during thua nao fermentation but Sarkar et al. (1997b) showed
that during the laboratory preparation of kinema, there was a 73–88%
loss of oligosaccharides during the soaking (16 h at 22–25°C) and
cooking (121°C for 15 min) stages, leaving only 17 g sucrose, 2.3 g
raffinose and 10.5 g stachyose (kg dry weight)−1 in the soya bean grits
prior to fermentation. In fermented beans (18 h at 45°C), concentra-
tions of each were less than 1 g (kg dry weight)−1. B. subtilis is able to
hydrolyse all three oligosaccharides (Sarkar and Tamang 1995) and,

presumably, the bacteria utilised the resultant monosaccharides for
growth.
10.1.8.9 Volatile Compounds Several authors have analysed the vola-

tile compounds in thua nao (Dajanta et al. 2011b; Leejeerajumnean
et al. 2001) or other Bacillus-fermented soya bean products (Kanno
and Takamatsu 1987; Owens et al. 1997). All the studies observed
that fermentation led to a large increase in the concentration of volatile
compounds and that these included a great diversity of compounds.
During a traditional thua nao fermentation, the concentration of total
volatile compounds increased from 35 μg kg−1 wet weight in the cooked
soya beans to 3500 μg kg−1 wet weight in 72 h fermented thua nao
(Leejeerajumnean et al. 2001). Dajanta et al. (2011b) observed consid-
erably higher concentrations of volatile compounds in laboratory-pre-
pared thua nao, namely 15,850 μg kg−1 wet weight (calculated assuming
a moisture content of 65%) in thua nao fermented with B. subtilis
TN51 and 13,900 μg kg−1 wet weight in naturally fermented thua nao.
Volatile compounds formed in the largest amounts in natural fermen-
tations tended to be ketones, including 3-hydroxy-2-butanone (acet-
oin), 5-methyl-3-hexanone, 2-butanone and 3-methyl-2-pentanone,
as well as 2-methylbutanoic acid, benzaldehyde, pyrazines, dimethyl
sulphide and 2-pentylfuran (Dajanta et al. 2011b; Leejeerajumnean
et al. 2001). However, there is no information on which volatile com-
pounds are important in the aroma of thua nao or which compounds
are desirable to obtain the most highly appreciated thua nao. In the
case of chungkook-jang, a Korean Bacillus-fermented soya bean prod-
uct, trimethyl pyrazine, butanoic acid and methyl pyrazine were posi-
tively correlated with overall acceptance while 3-hydroxy-2-butanone
and 2,3-butanediol were negatively correlated (Yoon et al. 2013). It is,
of course, possible that the aroma compounds are less important than
the amino acid profile in determining the flavour attributes of, and
consumer preferences for, thua nao.
Sun drying of traditionally prepared thua nao resulted in the loss
of 65% of the volatile compounds (from 3500 to 1250 μg kg−1 wet
weight), including important aroma compounds. Two commercial
dried thua nao samples had even lower concentrations of volatile com-
pounds (380 μg kg−1 wet weight; Leejeerajumnean et al. 2001).
Apart from losses during sun drying, which could be expected,

Owens et al. (1997) observed that concentrations of volatile com-
pounds can undergo large changes during fermentation. In soya beans
laboratory-fermented with B. subtilis DA2, concentrations of volatile
compounds were substantially higher after 18 h incubation at 35°C
(8690 μg kg−1 wet weight) than after 36 h incubation (2280 μg kg−1
wet weight). This was mainly due to the disappearance of 3-hydroxy-
2-butanone and some pyrazines.
The nutritional value of thua nao is essentially that of the soya bean
starting material, as there is little difference in proximate composition
between thua nao (Table 10.3) and soya beans (Chapter 1, Table 1.5).
The protein digestibility of natto in humans is reported to be similar
to that of boiled whole soya beans and soya bean curd (JSFNS 1984).
Thus, although it is commonly suggested that the availability of free
amino acids rather than unhydrolysed proteins confers nutritional
benefits, there does not appear to be any clear evidence to support this
for normally healthy people. There is, of course, a significant loss of
nutrients during processing and fermentation (see Section 10.1.8.2).
Since thua nao is consumed primarily as a flavouring agent and condi-
ment, it does not generally make a large contribution to diet.
It is possible that the bacteria hydrolyse steryl glucosides during
fermentation, leading to an increase in the concentration of sterols
(see above), but this is probably of little dietary significance as the gly-
cosides are normally hydrolysed in the gut anyway (Wikipedia n.d.a).
10.1.9.1 Functional Food Attributes Apart from the possible benefits

of consuming soya beans (see Chapter 1), there is evidence that activi-
ties of the bacteria and/or the consumption of bacterial endospores
may confer some health benefits. Humans cannot synthesise vitamin
K and depend on dietary intake or synthesis by intestinal bacteria.
Vitamin K occurs in two biologically active forms: K1 (phytoquinone,
having a phytyl side-chain) in green vegetables and K 2 (menaquinone,
having a side-chain made up of a variable number of isoprenoid resi-
dues), which is synthesised by bacteria (Wikipedia n.d.b). The con-
centration of MK-7 (vitamin K 2 with seven isoprenoid residues) is
relatively high in natto and consumption of natto leads to increased

serum plasma levels (Homma et al. 2006). A large study in Japan sug-
gested that habitual consumption of natto was associated with higher
bone density in the elderly and that this was primarily mediated by
vitamin K (Fujita et al. 2011; Kaneki et al. 2001). It was concluded
that the intake of foods, such as natto, containing vitamin K should
be encouraged for the elderly to prevent osteoporosis.
However, patients taking warfarin for the prevention of thrombo-
embolisms must restrict their intake of vitamin K-rich foods since the
anticoagulant action of warfarin depends on a vitamin K-dependent
carboxylation reaction. In addition to limiting their intake of spin-
ach, kale and broccoli, consumption of natto is strongly prohibited.
Homma et al. (2006) investigated ways of reducing the vitamin K
content of natto and suggested that boiled natto did not cause a
marked increase in plasma vitamin K. They suggested that warfarin
patients may be able to eat boiled natto without negative effects.
There are no data on vitamin K levels in thua nao but they might be
expected to be similar to levels in natto and to confer similar benefits
on elderly consumers.
A second area where consumption of Bacillus-fermented foods may
confer benefits relates to the probiotic potential of B. subtilis (Hong
et al. 2005), particularly in relation to immune function. Rhee et al.
(2004) showed that the combination of Bacteroides fragilis and B. subti-
lis consistently promoted the development of gut-associated limphoid
tissue in rabbits and that sporulation by the bacillus was essential for
the effect. Current views suggest that B. subtilis has an intestinal life
cycle involving spore injestion, vegetative cell growth, and sporula-
tion under putative anaerobic conditions (Tam et al. 2006). However,
B. subtilis does not appear to persist in the gut and establish a per-
manent existence. Hence, it is possible that regular consumption of
Bacillus-fermented foods, such as thua nao, containing large numbers
of spores, which would survive cooking, may confer benefits.
10.1.10.1 Food Poisoning Thua nao contains large numbers of

Bacillus subtilis spores, most of which can be expected to survive nor-
mal cooking, and are therefore consumed. While it seems that the
consumption of B. subtilis spores poses no problems, and may even be

beneficial (see above), B. subtilis has been implicated in a number of
food poisoning outbreaks, characterised by a short incubation period
(10 min to 14 h, median 2.5 h), followed by nausea, vomiting, stom-
ach cramps and diarrhoea in about half the cases (Kramer and Gilbert
1989; Logan 2011). Implicated foods have contained 105–109 (median
5.0 × 106) cfu g−1. The short incubation period suggests a toxigenic
illness, and three peptide heat-stable toxins, lichenysin, amylosin and
pumilacidin, associated with human illness have been characterised
(Constantin 2008; Apetroaie-Constantin et al. 2009).
Thua nao is normally eaten after cooking and so any heat-labile
toxin present would be destroyed. In contrast, natto is generally eaten
uncooked but, as there is no association between eating natto and food
poisoning, it must be concluded that natto strains do not produce toxin,
at least under the conditions of the fermentation. There are no reports on
investigations into toxin production by thua nao or natto Bacillus strains.
Apart from the possible toxicity of B. subtilis, a major concern is
the possibility of growth and toxin production by B. cereus strains.
Putative B. cereus bacteria are commonly isolated from thua nao
(e.g., Leejeerajumnean 2000) and other Bacillus-fermented products
(Ouoba et al. 2004; Sarkar et al. 2002). Nout et al. (1998) noted some
very high levels of putative B. cereus in samples of kinema, with 3 of 28
samples having 106 –109 cfu g−1. Most of the isolated B. cereus strains
produced B. cereus diarrhoeal toxin on cooked rice and a lower num-
ber also produced it on cooked soya beans. Nevertheless, none of the
market samples of kinema, containing confirmed B. cereus diarrhoeal
toxin-producing strains, contained detectable levels of the toxin. Nout
et al. (1998) showed that B. cereus readily grew and produced B. cereus
diarrhoeal toxin on cooked soya beans. However, B. cereus diarrhoeal
toxin is quite heat sensitive (inactivated by 56°C × 5 min; Kramer and
Gilbert 1989) and so will normally be eliminated by cooking.
10.1.10.2 Biogenic Amines Most biogenic amines, produced by micro-

bial decarboxylation of amino acids, are pharmacologically active and
can cause intoxications, with symptoms including nausea, respira-
tory distress, hot flushes, headache and rashes (Kim et al. 2012). Kim
et al. (2012) found levels of β-phenylalanine and tyramine in samples
of Japanese and Korean natto that exceeded suggested upper limits
for human foods (100–800 mg kg−1 for tyramine and 30 mg kg−1 for

β-phenylalanine). They showed that B. subtilis strains isolated from
the nattos were capable of producing these amines and concluded that
the amines in natto were bacterially produced. There are no reports on
the concentrations of biogenic amines in thua nao.
Natto is produced in highly controlled and reproducible conditions

to create a consistent product (Teng et al. 2004), in contrast to thua
nao which is still largely produced by entirely traditional methods. It
is not likely that thua nao production will quickly move to Japanese
natto-type process. However, since the traditional process effectively
sterilises the beans, it would be relatively easy to introduce the use of
selected starter cultures which, combined with adequate hygienic pro-
cedures, would ensure a pure culture fermentation and reproducible
product. Since the starter would be bacterial spores, it could be pro-
duced and distributed as a very stable powder or aqueous suspension.
Such a material would be relatively cheap to produce and distribute.
10.1.12 Future Prospects and Research Requirements
The putative health benefits that accrue from regular consumption of

natto (see Section 10.1.9.1) possibly/probably also occur with regular
consumption of thua nao. If this is true, then a case can be made
for promoting the consumption of thua nao, especially as the Thai
population ages. However, thua nao is still a largely traditionally
produced product and has received relatively little research attention.
Consequently, there are many questions that remain to be answered
to put its production on a more reliable footing and that would be
required to be answered if its consumption was to be encouraged.
1. Is slime production a characteristic desired by consumers?
Although Sundhagul et al. (1972) described thua nao as a
slimy product, it is evident from subsequent descriptions that
this is a highly variable characteristic, both among thua nao
samples and among the bacteria isolated from it.
2. What are the characteristics of thua nao that distinguish it
from natto, and what are those that are similar to natto? Since
thua nao is consumed in a way different from natto, one needs

to know what the differences are that matter to the consumer.
3. What are the important aroma and flavour compounds and
how may their production be promoted and controlled? What
is the role of the amino acids profile in determining flavour
characteristics?
4. How may the pungent ammonia aroma be best restricted?
5. Is it possible to select single-strain or mixed-strain starter
cultures to produce a consistent product with the desired
characteristics?
6. What hygiene procedures are required to ensure that a starter
culture reproducibly dominates the microflora to produce a
consistently high-quality product?
7. How may thua nao be preserved without losing important
aroma compounds? Do consumers prefer fresh, dried or
refrigerated thua nao? If there is a consumer preference for
fresh, this suggests that sun drying of thua nao results in the
loss of desired aroma compounds.
8. What is the yield of thua nao per kg soya bean starting mate-
rial, and how may it be maximised? It is difficult to reconcile
reported yields with the reported extent of bacterial growth
and this conundrum needs to be resolved.
10.2 Indonesian Cabuk (Fermented Sesame Press Cake)

Cabuk is a fermented food made from black sesame (Sesamum indi-

cum) seed press cake. It is made in Wonogiri and Sukoharjo, Central
Java and is not found in other places in Indonesia. The product has
a distinctive ammoniacal aroma, is black-brownish in colour and is
slightly sticky with an oily appearance. It is used as a flavouring ingre-
dient in vegetable soups or as a snack after being wrapped in wilted
banana leaf, steamed and roasted.
A similar fermented sesame seed product, ogiri-saro, is made
in Sierra Leone. The product is formed into balls about the size of
table tennis balls and wrapped in wilted leaf of Alchornea cordifolia
(Christmas leaf) or banana. Ogiri-saro is brown, soft and sticky with

an ammoniacal odour and is usually used as a flavouring ingredient in
soups (Steinkraus 1996).
10.2.2 History
Sesame oil has long been produced by pressing the seeds and the
remaining press cake used as animal feed or allowed to naturally fer-
ment to produce cabuk. Serat Centhini (Anonymous 1814), a Javanese
historical text, mentions that during the sixteenth century, cabuk was
widely consumed in Central Java, especially in the areas surround-
ing Solo and Surakarta. The product is still available in Wonogiri,
Sukoharjo and some places in Solo.
10.2.3 Places of Production, Method of Consumption and Role in Diet
The Sukoharjo district offers good conditions for the growth of sesame
and productivity in 2002 was up to 300 kg seeds ha−1, with a protein
content of 18% and a fat content of 48% (Sri Handayani et al. 2003).
For consumption, cabuk is mixed with spices, such as garlic, chilli and
salt, wrapped in banana leaf, steamed and sometimes roasted for addi-
tional flavour. It is used as a flavouring agent in vegetable soups or as
a snack. As a condiment, cabuk does not play a significant role in diet.
To extract oil, seeds are dried and the oil expressed by pressing.
The seed coats are separated by winnowing using a bamboo tray.
Alternatively, sesame seeds may be soaked overnight, washed, dehu-
lled by crushing in a wooden mortar, and then steamed for 30 min.
The hot seeds are placed in a jute sack, tied with a rope and then
placed in a second sack. The bagged sesame is placed in a wicker bas-
ket and pressed with a heavy stone for 30–60 min to extract the oil.
Water or lime water is added to the press cake and the mixture is
ground to a thick paste. The paste is moulded into balls, steamed for
~30 min, cooled and the balls placed on a bamboo tray and covered
with banana leaf. They are then left to ferment at ambient tempera-
ture (25–30°C) for 3–4 days (Figure 10.13).
Black sesame seed
Dry
Dehull
Press to separate oil

Oil
Presscake
Add water or lime solution and grind to a paste
Mold into balls
Steam for 30 min
Cool and cover with the banana leaf and bamboo mat
Incubate, room temperature (25–30ºC), 4–6 days
Cabuk
Figure 10.13 Production of cabuk in Wonogiri, Central Java.
10.2.5 Microbiology
The predominant microorganisms are aerobic, endospore-forming,

Gram-positive rods identified as Bacillus subtilis (Iino et al. 1997).
Presumably, vegetative cells are killed during the heat treatments
and only bacterial endospores survive and grow during the fer-
mentation. Wibowo and Muslimatun (1992) also found yeasts,
Micrococcus sp. and Staphylococcus sp. in the later stages of the fer-
mentation but considered them to be contaminants, present at rel-
atively low concentrations. The total bacterial count ranged from
10 6 to 109 cfu g−1.
Sesame (Sesamum indicum L.) is probably the most ancient oilseed

cultivated by mankind. The crop has been grown for centuries owing
to its high content of excellent quality oil (42–54%) and protein
(22–25%). The press cake has a lower oil content and proportionately
increased protein content (Table 10.4).
Most of the proteins present in sesame seeds are storage proteins
found as albumins (8.9%), globulins (67.3%), glutelins (6.9%) and
Table 10.4 Composition of Raw Sesame Seed, Sesame Seed Press Cake and Cabuk (Fermented
Sesame Press Cake)
CHEMICAL COMPOSITION (G KG−1 WET WT)
SESAME SESAME CABUK SOURCE OF
COMPONENT SEED PRESS CAKE SUKOHARJO SAMPLEa WONOGIRI (1) WONIGIRI (2)
Moisture 47 n.d.
b 125 110 110
Crude protein 200 300–350 250 260 260
Crude fat 520 230–300 210 220 220
Carbohydrate 150 n.d. 380 380 380
Crude fibre 31 85–97 32 30 34
Ash 52 32–35 0.5 1.0 0.8
Note: Sesame seed and cabuk data adapted from Iino et al. (1997); press cake data from Sri
Handayani et al. (2003).
a Location in Central Java, Indonesia.
b No data.
prolamins (1.3%) (Rivas et al. 1981). The amino acid composition of
the sesame seed protein is unique among seed proteins due to its high
content of sulphur-containing amino acids (methionine and cysteine)
and low content of lysine (Johnson et al. 1979). Because of its amino
acids profile, sesame has been recommended as a protein supplement
for legumes (Boloorforooshan and Markakis 1979; Brito and Núñez
1982).
Sesame seed oil is mainly composed of the unsaturated fatty acids,
linoleic (~46%) and oleic (39%), with only minor amounts of satu-
rated fatty acids, palmitic (8.5%), stearic (5.5%) and arachidic (0.9%)
(Nzikou et al. 2009).
In addition, sesame is also a rich source of niacin, folic acid, vita-
min E, calcium and phosphorus. The lignans and gamma-tocopherol
in sesame seed were shown by Yamashita et al. (1992) to promote
vitamin E synthesis in rats.
10.2.7 Chemical Composition and Changes during Processing and

Fermentation
During fermentation, the temperature of the paste rises from an ini-

tial 25–30°C to about 42°C. A major chemical change during fer-
mentation is an increase in pH value due to hydrolysis and utilisation
of protein, accompanied by a fivefold increase in the concentration
of free amino acids (Muslimatun 1990) and an increase in ammonia
content. Some hydrolysis of fats to free fatty acids occurs, which con-
tributes to the flavour of the final product.
The nutritional value of cabuk is essentially similar to that of sesame

press cake (Table 10.4). The crude protein content is ~25% and fat
content ~22%. The protein is of good quality and the fat is composed
mainly of unsaturated fatty acids. Hence, cabuk has a relatively good
nutritional profile.
In December 2010, the price of cabuk was Rp 20,000 (approxi-

mately US$2) for a pack of about 250 g wet weight. Most cabuk is
produced at household level and is consumed directly by the fam-
ily and at present there is little prospect of it being produced on a
larger scale.
References
Ababouch, L.H., L. Grimit, R. Eddafry and F.F. Busta. 1995. Thermal inactiva-
tion kinetics of Bacillus subtilis spores suspended in buffer and oils. Journal
of Applied Bacteriology 78:669–676.
Allagheny, N., Z.A. Obanu, G. Campbell-Platt and J.D. Owens. 1996. Control
of ammonia formation during Bacillus subtilis fermentation of legumes.
International Journal of Food Microbiology 29:321–333.
Anonymous. 1814. Serat Centhini. A Javanese encyclopaedia of traditional
songs and foods in Java.
Apetroaie-Constantin, C., R. Mikkola, M.A. Andersson, V. Teplova, I.
Suominen, T. Johansson and M. Salkinoja-Salonen. 2009. Bacillus subtilis
and B. mojavensis strains connected to food poisoning produce the heat
stable toxin amylosin. Journal of Applied Microbiology 106:1976–1985.
Ashiuchi, M. 2013. Microbial production and chemical transformation of poly
γ-glutamate. Microbial Biotechnology 6:664–674.
Barbe, V., S. Cruveiller, F. Kunst, P. Lenoble, G. Meurice, A. Sekowska, D.
Vallenet, T. Wang, I. Moszer, C. Medigue and A. Danchin. 2009. From
a consortium sequence to a unified sequence: The Bacillus subtilis 168
reference genome a decade later. Microbiology-SGM 155:1758–1775.
Boloorforooshan, M. and P. Markakis. 1979. Protein supplementation of navy
beans with sesame. Journal of Food Science 44:390–392.
Brito, O.J. and N. Núñez. 1982. Evaluation of sesame flour as a complementary

protein source for combination with soy and corn flours. Journal of Food
Science 47:457–461.
Chukeatirote, E., C. Chainun, A. Siengsubchart, C. Moukamnerd, P.
Chantawannakul, S. Lumyong, N. Boontim and P. Thakang. 2006.
Microbiological and biochemical changes in thua nao fermentation.
Research Journal of Microbiology 1:38–44.
Condon, S. and F. J. Sala. 1992. Heat resistance of Bacillus subtilis in buffer and
foods of different pH. Journal of Food Protection 55:605–608.
Constantin, C. 2008. Cereulide producing Bacillus cereus and amylosin pro-
ducing Bacillus subtilis and Bacillus mojavensis: Characterization of strains
and toxigenicities. Dissertation, Department of Applied Chemistry
and Microbiology, Division Microbiology, Faculty of Agriculture and
Forestry, University of Helsinki. https://helda.helsinki.fi/bitstream/han-
dle/10138/20842/cereulid.pdf ?… (accessed 30 July 2014).
Dajanta, K., A. Apichartsrangkoon, E. Chukeatirote and R.A. Frazier. 2011a.
Free-amino acid profiles of thua nao, a Thai fermented soybean. Food
Dajanta, K., A. Apichartsrangkoon and. E. Chukeatirote. 2011b. Volatile profiles
of thua nao, a Thai fermented soy product. Food Chemistry 125:464–470.
Dajanta, K., E. Chukeatirote and A. Apichartsrangkoon. 2012. Nutritional and
physicochemical qualities of thua nao (Thai traditional fermented soy-
bean). Chiangmai Journal of Science 39:562–574.
Earl, A.M., R. Losick and R. Kolter. 2008. Ecology and genomics of Bacillus
subtilis. Trends in Microbiology 16:269–275.
Eggert, T., U. Brockmeier, M.J. Dröge, W.J. Quax and K.-E. Jaeger. 2003.
Extracellular lipases from Bacillus subtilis: Regulation of gene expres-
sion and enzyme activity by amino acid supply and external pH. FEMS
Microbiology Letters 225:319–324.
Ferrari, E., A.S. Jarnagin and B.F. Schmidt. 1993. Commercial production
of extracellular enzymes. In Bacillus subtilis and Other Gram-Positive
Bacteria, edited by A.L. Sonenshein, J.A. Hoch and R. Losick, pp. 917–
937. Washington, DC: ASM Press.
Folmsbee, M.J., M.J. McInerney and D.P. Nagle. 2004. Anaerobic growth of
Bacillus mojavensis and Bacillus subtilis requires deooxyribonucleosides or
DNA. Applied and Environmental Microbiology 70:5252–5257.
Fujita, Y., M. Iki, J. Tamaki, K. Kouda, A. Yura, E. Kadowaki, Y. Sato, J.-S.
Moon, K. Tomioka, N. Okamoto and N. Kurumatani. 2011. Association
between vitamin K intake from fermented soybeans, natto, and bone min-
eral density in elderly Japanese men: The Fujiwara-kyo Osteoporosis Risk
in Men (FORMEN) study. Osteoporosis International 23:705–714.
Graffham A.J., M.H. Gordon, A. Westby and J.D. Owens. 1995. Nutrition of
Hashimoto, M., T. Hamamoto, M. Kitada, M. Hino, T. Kudo and K. Horikoshi.
1994. Characteristics of alkali-sensitive mutants of alkaliphilic Bacillus sp.
strain C-125: They show cellular morphological abnormalities. Bioscience,
Biotechnology and Biochemistry 58:2090–2092.
Homma, K., N. Wakana, Y. Suzuki, M. Nukui, T. Daimatsu, E. Tanaka,

K. Tanaka, Y. Koga, Y. Nakajima and I. Nakazawa. 2006. Treatment of
natto, a fermented soybean preparation, to prevent excessive plasma vita-
min K concentrations in patients taking warfarin. Journal of Nutritional
Science and Vitaminology 52:297–301.
Hong, H.A., L.H. Duc and S.M. Cutting. 2005. The use of bacterial spore
formers as probiotics. FEMS Microbiology Reviews 29:813–835.
Iino, H., K.R. Kuswanto, M. Kozaki and S. Okada. 1997. Microorganisms of
fermented black sesame pressed cake ‘cabuk’. Bulletin of Doctoral Course of
Science for Living Systems, Showa Women’s University 6:107–110.
Ikenebomeh, M.J., R. Kok and J.M. Ingram. 1986. Processing and fermentation
of the African locust bean (Parkia filicoidea Welw.) to produce dawadawa.
Journal of the Science of Food and Agriculture 37:273–282.
Inatsu, Y., K. Kimura, and Y. Itoh. 2002. Characterization of Bacillus subti-
lis strains isolated from fermented soybean foods in Southeast Asia:
Comparison with B. subtilis (natto) starter strains. Japan Agricultural
Research Quarterly (JARQ) 36 (3):169–175.
Inatsu, Y., N. Nakamura, Y. Yuriko, T. Fushimi, L. Watanasiritum and S.
Kawamoto. 2006. Characterization of Bacillus subtilis strains in thua nao,
a traditional fermented soybean food in northern Thailand. Letters in
Jagannath, A., I. Nakamura and T. Tsuchido. 2003. Modelling the combined effects
of temperature and sodium chloride on the thermal inactivation of Bacillus
subtilis spores in a buffer system. Journal of Applied Microbiology 95:135–141.
Jagannath, A. and T. Tsuchido. 2003. Validation of a polynomial regression
model: The inactivation of Bacillus subtilis spores in milk. Letters in Applied
Johnson, L.A., T.M. Suleiman and E.W. Lucas. 1979. Sesame protein: A review
and prospectus. Journal of the American Oil Chemists Society. 56:463–468.
JSFNS. 1984. Rice, soybean and fish—Scientific approach to staple foods. In
Japanese Society of Nutrition and Food Science, edited by M. Fujimaki, G. Inoue
and T. Tanaka. Tokyo: Kouseikan Publishing. Cited by Liu 1997, p. 67.
Kaneki, M., S.J. Hedges, T. Hosoi, S. Fujiwara, A. Lyons, N. Ishida, M.
Nakagawa, M. Takechi, Y. Sano, Y. Mizuno, S. Hoshino, M. Miyao, S.
Inoue, K. Horiki, M. Shiraki, Y. Ouchi and H. Orimo. 2001. Japanese
fermented soybean food as the major determinant of the large geographic
difference in circulating levels of vitamin K2: Possible implications for
hip-fracture risk. Nutrition 17:315–321.
Kanno, A. and H. Takamatsu. 1987. Changes in the volatile components
of ‘natto’ during manufacturing and storage. Nippon Shokuhin Kogyo
Gakkaishi 34:330–335.
Kim, B., B.Y. Byun and J.-H. Mah. 2012. Biogenic amine formation and bacte-
rial contribution in natto products. Food Chemistry 135:2005–2011.
Kiuchi, K. and S. Watanabe. 2004. Industrialization of Japanese natto. In
Industrialization of Indigenous Fermented Foods, edited by K.H. Steinkraus,
Klocke, R.A., K.K. Andersson, H.H. Rotman and R.E. Forster. 1972.
Permeability of human erythrocytes to ammonia and weak acids.
American Journal of Physiology 222:1004–1013.
Kramer, J.M. and R.J. Gilbert. 1989. Bacillus cereus and other Bacillus species.
In Foodborne Bacterial Pathogens, edited by M.P. Doyle, pp. 21–70. New
York: Marcel Dekker.
Kubo, Y., A.P. Rooney, Y. Tsukakoshi, R. Nakagawa, H. Hasegawa and
K. Kimura. 2011. Phylogenetic analysis of Bacillus subtilis strains applica-
ble to natto (fermented soybean) production. Applied and Environmental
Lagarda, M.J., G. García-Llatas and R. Farré. 2006. Analysis of phytos-
terols in foods. Journal of Pharmaceutical and Biomedical Analysis
41:1486–1496.
Leejeerajumnean, A. 2000. Bacillus fermentation of soybeans: Characterization
of traditional thua nao manufacture. PhD thesis, Department of Food
Science and Technology, University of Reading, Reading, UK.
Leejeerajumnean, A., J.M. Ames and J.D. Owens. 2000. Effect of ammonia on
the growth of Bacillus species and some other bacteria. Journal of Applied
Leejeerajumnean, A., S.C. Duckham, J.D. Owens and J.M. Ames. 2001.
Volatile compounds in Bacillus-fermented soybeans. Journal of the Science
of Food and Agriculture 81:525–529.
Logan, N.A. 2011. Bacillus and relatives in foodborne illness. Journal of Applied
Logan, N.A. and P. de Voss. 2009. Bacillus. In Bergey’s Manual of Systematic
Bacteriology, Volume Three, The Firmacutes, second ed., edited by P. de
Voss, G.M. Garrity, D. Jones, N.R. Krieg, W. Ludwig, F.A. Rainey, K.-H.
Schleifer and W.B. Whitman, pp. 21–128. New York: Springer.
Molin, N. and B.G. Snygg. 1967. Effect of lipid materials on heat resistance of
bacterial spores. Applied Microbiology 15:1422–1426.
Muslimatun, S. 1990. Microflora during fermentation of cabuk. B.Sc. dis-
sertation. Faculty of Agricultural Technology, Gadjah Mada University,
Yogyakarta.
Nishito, Y., Y. Osana, T. Hachiya, K. Popendorf, A. Toyoda, A. Fujiyama, M.
Itaya and Y. Sakakibara. 2010. Whole genome assembly of a natto pro-
duction strain Bacillus subtilis natto from very short read data. BMC
Genomics 11:243. www.biomedcentral.com/1471-2164/11/243 (accessed
5 March 2014).
Njoku, H.O., J.N. Ogbulie and C. Nnubia. 1990. Microbiological study of
the traditional processing of African oil bean (Pentaclethra macrophylla
Bentham) for ugba production. Food Microbiology 7:13–26.
Nout, M.J.R., D. Bakshi and P.K. Sarkar. 1998. Microbiological safety of
kinema, a fermented soya bean food. Food Control 9:357–362.
Nzikou, J.M., L. Matos, G. Bouanga et al. 2009. Chemical composition of
the seed of sesame (Sesamum indicum L.) grown in Congo-Brazzaville.
Advance Journal of Food Science and Technology 1:6–11.
Odunfa, S.A. 1986. Dawadawa. In Legume-Based Fermented Foods, edited by

N.R. Reddy, M.D. Pierson and D.K. Salunkhe, pp. 173–189. Boca Raton:
CRC Press.
Ogbadu, L. and R.N. Okagbue. 1988. Bacterial fermentation of soya bean for
‘daddawa’ production. Journal of Applied Microbiology 65:353–356.
Omafuvbe, B.O. and A.O. Adenowo. 2002. Biochemical changes in ferment-
ing mesquite beans (Prosopis africana Taub.) for kpaye production. Science
Focus (Nigeria) 1:79–84.
Omafuvbe, B.O., O.O. Shonukan and S.H. Abiose. 2000. Microbiological and
biochemical changes in the traditional fermentation of soybean for ‘soy-
daddawa’—Nigerian food condiment. Food Microbiology 17: 469–474.
Omafuvbe, B.O., S.H. Abiose and O.O. Adaraloye. 1999. The production of
‘Kpaye’—A fermented condiment from Prosopis africana (Guill and Perr)
Taub. seeds. International Journal of Food Microbiology 51:183–186.
Ouoba, L.I.I., B. Diawara, W.K. Amoa-Awua, A.S. Traoré and P.L. Møller.
2004. Genotyping of starter cultures of Bacillus subtilis and Bacillus pumi-
lus for fermentation of African locust bean (Parkia biglobosa) to produce
Soumbala. International Journal of Food Microbiology 90:197–205.
Owens, J.D., N. Allagheny, G. Kipping and J.M. Ames. 1997. Formation of
volatile compounds during Bacillus subtilis fermentation of soya beans.
Journal of the Science of Food and Agriculture 74:132–140.
Parkouda, C., D.S. Nielsen, P. Azokpota, L.I.I. Ouoba, W.K. Amoa-Awua,
L. Thorsen, J.D. Hounhouigan, J.S. Jensen, K. Tano-Debrah, B. Diawara
and M. Jakobsen. 2009. The microbiology of alkaline-fermentation of
indigenous seeds used as food condiments in Africa and Asia. Critical
Reviews in Microbiology 35:139–156.
Rhee, K.-J., P. Sethupathi, A. Driks, D.L. Lanning and K.L. Knight. 2004.
Role of commensal bacteria in development of gut-associated lymphoid
tissues and preimmune antibody repertoire. The Journal of Immunology
172:1118–1124.
Rivas, N.R., J.E. Dench, and J.C. Caygill. 1981. Nitrogen extractability of
sesame (Sesamum indicum L.) seed and the preparation of two protein
isolates. Journal of the Science of Food and Agriculture 32:565–571.
Sarkar, P.K. 2000. Ecology of soybean acidifications and some biochemical
changes during controlled kinema production. Journal of Food Science and
Sarkar, P.K., B. Hasenack and M.J.R. Nout. 2002. Diversity and functionality
of Bacillus and related genera isolated from spontaneously fermented soy-
beans (Indian kinema) and locust beans (African soumbala). International
Sarkar, P.K. and J.P. Tamang. 1995. Changes in the microbial profile and proxi-
mate composition during natural and controlled fermentation of soy-
beans to produce kinema. Food Microbiology 12:317–325. Cited by Sarkar
et al. (1997a).
Sarkar, P.K., J.P. Tamang, P.E. Cook and J.D. Owens. 1994. Kinema—A tra-
ditional soybean fermented food: Proximate composition and microflora.
Sarkar, P.K., L.J. Jones, G.S. Craven and S.M. Somerset. 1997b. Oligosaccharide
profiles of soybeans during kinema production. Letters in Applied
Sarkar, P.K., L.J. Jones, G.S. Craven, S.M. Somerset and C. Palmer. 1997a.
Amino acid profiles of kinema, a soybean fermented food. Food Chemistry
59:69–75.
Sarkar, P.K., L.J. Jones, W. Gore, G.S. Craven and S.M. Somerset. 1996.
Changes in soya bean lipid profiles during kinema production. Journal of
the Science of Food and Agriculture 71:321–328.
Sarkar, P.K., P.E. Cook and J.D. Owens, 1993. Bacillus fermentation of soy-
beans. World Journal of Microbiology and Biotechnology 9:295–299.
Sonenshein, A.L., J.A. Hoch and R. Losick, eds. 2001. Bacillus subtilis and
Other Gram-Positive Bacteria. Washington, DC: ASM Press.
Sri Handayani, Isti Astuti, Kusnandar and Achmad Nur Chamdi. 2003. The
potency of products based on sesame on agricultural sustainability in
Sukoharjo district, Central Java. Seminar on Agricultural Sustainability
in Indonesia, Yogyakarta, September 17.
Srimani, B. and M. Loncin. 1980. Determination of decimal reduction
time of Bacillus stearothermophilus and Bacillus subtilis spores at 100°C.
Lebensmittel–Wissenschaft und Technologie 13:190–192.
Srimani, B., R. Stahl and M. Loncin. 1980. Death rates of bacterial spores
at high temperatures. Lebensmittel–Wissenschaft und Technologie 13:
186–189.
Steinkraus, K.H. 1986. Fermented foods, feeds and beverages. Biotechnology
Advances 4:219–243.
Steinkraus, K.H. 1991. Alkaline fermented foods and their relation to similar
food in other parts of the world. In Traditional African Foods: Quality and
Nutrition, edited by A. Westby and P.J.A. Reilly, pp. 87–92. Stockholm:
International Foundation for Science.
Steinkraus, K.H. 1996. Indigenous fermented foods involving an alkaline fer-
mentation. In Handbook of Indigenous Fermented Foods, second ed., edited
by K.H. Steinkraus, pp. 349–362. New York: Marcel Dekker.
Sundhagul, M., P. Smanmathuroj and W. Bhodacharoen. 1972. Thua-nao: A
fermented soybean food of northern Thailand. I. Traditional processing
method. Thai Journal of Agricultural Science 5:43–56.
Tam, N.K.M., N.Q. Uyen, H.A. Hong, L.H. Duc, T.T. Hoa, C.R. Serra, A.O.
Henriques and S.M. Cutting. 2006. The intestinal life cycle of Bacillus
subtilis and close relatives. Journal of Bacteriology 188: 2692–2700.
Tamang, J.P. 2003. Native organisms in the fermentation of kinema. Indian
Journal of Microbiology 43:127–130.
Teng, D.-F., C-S. Lin and P.-C. Hsieh. 2004. Fermented whole soybeans and
soybean paste. In Handbook of Food and Beverage Technology, edited by
Y.H. Hui, L. Meunier-Goddik, Å.S. Hansen, J. Josephsen, W.-K. Nip,
P.S. Stanfield and F. Toldrá, pp. 533–569. New York: Marcel Dekker.
Thorne, C.B. 1993. Bacillus anthracis. In Bacillus subtilis and Other Gram-
Positive Bacteria, edited by A.L. Sonenshein, J.A. Hoch and R. Losick,
pp. 113–124. Washington, DC: ASM Press.
Tran, L.-S.P., T. Nagai and Y. Itoh. 2000. Divergent structure of the ComQXPA
quorum-sensing components: Molecular basis of strain-specific com-
munication mechanism in Bacillus subtilis. Molecular Microbiology 37:
1159–1171.
Vlamakis, H., Y. Chai, P. Beauregard, R. Losick and R. Kolter. 2013. Sticking
together: Building a biofilm the Bacillus subtilis way. Nature Reviews |
Wang. J. and D.Y.C. Fung. 1996. Alkaline-fermented foods: A review with
emphasis on pidan fermentation. Critical Reviews in Microbiology
22:101–138.
Warren, K.S. 1962. Ammonia toxicity and pH. Nature 195:47–49.
Warth, A.D. 1978. Relationship between the heat resistance of spores and the
optimum and maximum growth temperatures of Bacillus species. Journal
of Bacteriology 134:699–705.
Wibowo, D. and S. Muslimatun. 1992. Cabuk, fermented black sesame pressed
cake. Poster Abstract, World Food Conference, Toronto.
Wikipedia. n.d.a. Vitamin K. https://en.wikipedia.org/wiki/Vitamin K
(accessed 23 February 2014).
Wikipedia. n.d.b. Phytosterol. https://en.wikipedia.org/wiki/Phytosterols
(accessed 5 March 2014).
Wongputtisin, P.C. Khanongnuch, W. Khongbantad, P. Niamsup and S.
Lumyong. 2012. Screening and selection of Bacillus spp. for fermented
corticated soybean meal production. Journal of Applied Microbiology
113:798–806.
Yamashita, K., Y. Nohara, K. Katayama and M. Namiki. 1992. Sesame seed
lignans and gama-tocopherol act synergistically to produce vitamin E
activity in rats. Journal of Nutrition 122:2440–2446.
Yoon, K.-H., Y.-I. Chang and G.-H. Lee. 2013. Characteristic aroma com-
pounds of cooked and fermented soybean (chungkook-jang) inoculated
with various bacilli. Journal of the Science of Food and Agriculture 93:85–92.
Food and Culinary Science
Brings Together Current Knowledge and State-of-the-Art Information on

Indigenous Fermented Foods:
Fermented foods and beverages span a range of root crops, cereals, pulses,
vegetables, nuts, fruits, and animal products. Southeast Asia has a long history
of utilizing fermentation in the production and preservation of foods, and is
widely recognized for its prominent use. Indigenous Fermented Foods of
Southeast Asia examines some indigenous fermented foods of Thailand,
Vietnam, Indonesia, Malaysia, and the Philippines, focusing on the chemical,
microbiological, and technological factors associated with their manufacture,
quality, and safety. This text establishes a need for an adequate understanding
of the fermentation process to ensure safe and reliable practices, as well as
the consistent production of a quality product.
The authors describe the production, microbiology, biochemistry, nutritional

value, and dietary roles of a wide variety of indigenous fermented foods of
Southeast Asia. Emphasizing the microbiological and biochemical processes
in fermentations and examining the factors that influence the development of
the characteristic microflora and chemical changes induced, they accurately
describe each process and critically evaluate the roles of microbes in the
fermentation. The classification of products is based on their microbial ecology
(i.e., the predominant microbes involved), and the text includes examples of
every major category of fermented food. The book covers tempe, starter
cultures, sweet/sour/alcoholic rice and cassava fermentations, alcoholic
fermentations, soy sauce, Bacillus fermentations, and lactic acid bacterial
fermentations of vegetables, durian fruit, rice noodles, meats, and sea foods.
This book answers a series of basic questions addressing:

• Dominant/desired microbes
• Suitable factors in processing and the environment
• Commonly present microbes
• Compounds utilized as major carbon and energy sources
• Sources of fermentable carbohydrates
• Main biochemical activities and chemical changes
• True yield of product per kilogram of initial raw materials
• Possible hazards associated with a product
• How possible hazards may be minimized or eliminated
• Research needs and opportunities
Indigenous Fermented Foods of Southeast Asia evaluates the state of

scientific knowledge of the fermentations and identifies specific questions
that need to be answered in order to promote the reproducibility, safety, and
future prospects of these fermented foods.
K11988

Indigenous Fermented Foods of Southeast Asia - J. David Owens

Uploaded by

Document Information

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

Indigenous Fermented Foods of Southeast Asia - J. David Owens

Uploaded by

Copyright:

Available Formats

Indigenous

Indigenous Fermented Foods of Southeast Asia (2014)

Boca Raton London New York

CRC Press is an imprint of the

No claim to original U.S. Government works

International Standard Book Number-13: 978-1-4398-4481-6 (eBook - PDF)

C h a p t e r 5 L a c t i c V e g e ta b l e and F r u i t F e r m e n tat i o n s 185

C h a p t e r 10 B a c i llu s F e r m e n tat i o n s 373

Natural fermentation precedes human history, and since ancient

selected, tailor-made, or improved processes that provide sustain-

preclude upgrading the technology to produce a safer and more consis-

to the problem of how to obtain a desired microbial flora and product

Table I.1 Some Indigenous Fermented Foods Produced in Southeast Asia

3. SWEET, SOUR, ALCOHOLIC SOLID-SUBSTRATE FUNGAL FERMENTATIONS

(tape ketela) roots cassava roots with whitish fungal

5. LACTIC ACID BACTERIAL VEGETABLE AND FRUIT FERMENTATIONS

acetate, ethanol, CO2 appetiser Malaysia

6. LACTIC ACID BACTERIAL CEREAL FERMENTATIONS

7. LACTIC ACID BACTERIAL FISH AND SEA FOOD FERMENTATIONS

8. LACTIC ACID BACTERIAL MEAT FERMENTATIONS

9. LACTIC ACID BACTERIAL DAIRY FERMENTATIONS

10. MIXED LACTIC ACID BACTERIAL AND YEAST FERMENTATIONS

NAME OF APPEARANCE AND MODE OF COUNTRIES

11. BACILLUS FERMENTATIONS

12. ACETIC ACID BACTERIAL FERMENTATIONS

solution. Dessert Thailand (1996)

J. David Owens graduated in microbiology and chemistry from

invented the Indirect Method for the conductimetric assay of micro-

Mary Astuti Nipa Chokesajjawatee

Kapti Rahayu Kuswanto Renu Pinthong

1.1.7.1 Changes during Hydration of Beans 44

1.2.4.1 Preparation of Pigeon Pea Tempe 80

in areas with unfertile land, such as in Wonogiri (southeast Central

1.1.2 History of Tempe

1.1.2.1 Origin of Usar Tempe Inoculum The question of where tempe

The use of an inoculum to ‘steer’ a fermentation in the desired

1.1.2.2 Soya Beans in Indonesia Soya beans originated in north China

1.1.3 Places and Scale of Production, How Tempe Is

1.1.3.1 Distribution of Tempe Producers Indonesia comprises 33 prov-

The amount of soya beans processed into tempe varies between

1.1.3.2 Consumption of Tempe in Indonesia Generally, tempe is con-

Figure 1.5 Tempe crisps. (Courtesy of J.D. Owens.)

Figure 1.6 Deep-fat-fried tempe. (Courtesy of J.D. Owens.)

Table 1.1 Consumption of Protein Food Sources in Indonesia in 2012

In Indonesia, tempe is one of the cheapest sources of protein. Tempe

1.1.4 Traditional and Current Production Methods

There are many variations in the methods for making tempe in

Table 1.2 Some Traditional Methods for Preparing Tempe

in Indonesia conducted by Saono et al. (1986) revealed seven variant

Figure 1.9 Mechanical dehulling machine. (Courtesy of M. Astuti.)

Figure 1.10 Powdered ragi tempe inoculum. (Courtesy of M. Astuti.)

Although details of preparation methods vary from one place

1.1.4.1 Soaking Some producers use a slow method which includes

Figure 1.11 Bags of tempe on incubation rack. (Courtesy of M. Astuti.)

Figure 1.12 Soaking soya beans. (Courtesy of M. Astuti.)

1.1.4.2 Cooking Cooking can be done by steaming or by boiling

Figure 1.13 Boiling soya beans in small-scale industry. (Courtesy of M. Astuti.)

1.1.4.3 Dehulling Dehulling soya beans is now generally performed

1.1.4.4 Washing Washing with clean water removes dirt attached to

1.1.4.5 Draining and Cooling The cooked kernels are drained to

the surface and sufficiently cool to allow fungal growth. Spreading

1.1.4.6 Inoculation Cooled beans at approximately room temperature

1.1.4.7 Packaging It is important to pack the inoculated kernels prop-

1.1.4.7.1 Trunk of Banana Trees Banana tree trunk (called debog)

1.1.3 Places and Scale of Production, How Tempe Is

1.1.7.6 Overall Losses of Dry Matter Prior to the Mould Fermen