Advances in Agronomy v.69

Agronomy
D VA N C E S I N
VOLUME 69
Advisory Board
Martin Alexander Ronald Phillips

Cornell University University of Minnesota
Kenneth J. Frey Larry P. Wilding

Iowa State University Texas A&M University
Prepared in cooperation with the

American Society of Agronomy Monographs Committee
Diane E. Stott, Chairman
John Bartels Linda S. Lee Wayne F. Robarge
Jerry M. Bigham David Miller Dennis E. Rolston
Jerry L. Hatfield Matthew J. Morra Richard Shibles
David M. Krell John E. Rechcigl Jeffrey Volenec
Donald C. Reicosky
Agronomy DVANCES IN
VO L U M E 69
Edited by
Donald L. Sparks
Department of Plant and Soil Sciences
University of Delaware
Newark, Delaware
San Diego San Francisco New York Boston London Sydney Tokyo
COPYRIGHT PAGE SUPPLIED BY AP
Contents
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
THE MEASUREMENT AND INTERPRETATION OF SORPTION

AND DESORPTION RATES FOR ORGANIC COMPOUNDS
IN SOIL MEDIA
Joseph J. Pignatello
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
II. The Nature of Elementary Sorption Processes in Soils. . . . . . . . . . . . 3
III. Slow Sorption and Desorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
IV. Sorption Kinetic Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
V. Experimental Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
VI. Sorption Kinetics and Bioavailability . . . . . . . . . . . . . . . . . . . . . . . . . . 56
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
ENVIRONMENTAL INDICATORS OF AGROECOSYSTEMS

O. H. Smith, G. W. Petersen, and B. A. Needelman
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
II. Agroecosystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
III. Monitoring and Assessment Endpoints . . . . . . . . . . . . . . . . . . . . . . . . 77
IV. Environmental Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
V. Soil Organic Matter as a Candidate Environmental Indicator. . . . . . . 85
VI. Indicator Ranking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
VII. Conclusions and Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
GROWTH PROMOTION OF PLANTS INOCULATED WITH

PHOSPHATE-SOLUBILIZING FUNGI
M. A. Whitelaw
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
II. Soil Phosphorus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
III. Phosphate-Solubilizing Soil Microorganisms. . . . . . . . . . . . . . . . . . . . 106
IV. Liquid Medium Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
v
vi CONTENTS
V. Plant Growth Promotion by Phosphate-Solubilizing Fungi . . . . . . . . 133

VI. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
HYDROLOGICAL FACTORS FOR PHOSPHORUS TRANSFER

FROM AGRICULTURAL SOILS
P. M. Haygarth, A. L. Heathwaite, S. C. Jarvis, and T. R. Harrod
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
II. Temporal Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
III. Spatial Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
IV. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
CASSAVA, Manihot esculenta Crantz, GENETIC

RESOURCES: THEIR COLLECTION, EVALUATION,
AND MANIPULATION
Nagib M. A. Nassar
I. Wild Taxa of Cassava Manihot Species . . . . . . . . . . . . . . . . . . . . . . . . . 180
II. Broadening the Genetic Base of Cassava, M. esculenta Crantz, and
Development of Interspecific Hybridization . . . . . . . . . . . . . . . . . . . . 198
III. Development and Selection for Apomixis in Cassava,
M. esculenta Crantz. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
IV. Production of Polyploid Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
V. Protein Contents in Cassava Cultivars and Its Hybrid with
Wild Manihot Species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
Contributors
Numbers in parentheses indicate the pages on which the authors’ contributions begin.
T. R. HARROD (153), Soil Survey and Land Research Centre, Cranfield Univer-
sity, North Wyke, Okehampton, Devon EX20 2SB, United Kingdom
P. M. HAYGARTH (153), Institute of Grassland and Environmental Research,
North Wyke, Okehampton, Devon EX20 2SB, United Kingdom
A. L. HEATHWAITE (153), Department of Geography, University of Sheffield,
Sheffield S10 2TN, United Kingdom
S. C. JARVIS (153), Institute of Grassland and Environmental Research, North
Wyke, Okehampton, Devon EX20 2SB, United Kingdom
NAGIB M. A. NASSAR (179), Departamento de Genética e Morfologia, Univer-
sidade de Brasília, Brasília 70919, Brazil
B. A. NEEDELMAN (75), Department of Agronomy, Pennsylvania State Uni-
versity, University Park, Pennsylvania 16802
G. W. PETERSEN (75), Department of Agronomy, Pennsylvania State Univer-
sity, University Park, Pennsylvania 16802
JOSEPH J. PIGNATELLO (1), The Connecticut Agricultural Experiment Sta-
tion, New Haven, Connecticut 06511
O. H. SMITH (75), Department of Agronomy, Pennsylvania State University,
University Park, Pennsylvania 16802
M. A. WHITELAW (99), School of Wine and Food Sciences, Charles Sturt Uni-
versity, Wagga Wagga, New South Wales 2678, Australia
vii
This Page Intentionally Left Blank
Preface
Volume 69 contains five excellent reviews dealing with crop and soil sciences.
Chapter 1 is a comprehensive and timely review on the measurement and inter-
pretation of sorption and desorption rates for organic compounds in soil media.
Topics covered include the nature of elementary sorption processes in soil, hin-
dered sorption and desorption processes, sorption kinetic models, experimental
methods, and sorption kinetics and bioavailability. Chapter 2, by O. H. Smith and
co-workers, is an excellent overview of environmental indicators of agroecosys-
tems. Soil organic matter content is discussed in detail as a candidate environ-
mental indicator. A ranking scheme is proposed for the use of multiple indicators
in decision-making applications. Chapter 3, by M. A. Whitelaw, is an interesting
treatise on plant growth as affected by phosphate-solubilizing soil microorgan-
isms. The author provides a discussion on soil phosphorus, studies on P-solubiliz-
ing soil microorganisms, aspects of liquid medium studies, and plant growth pro-
motion by phosphate-solubilizing fungi. Chapter 4, by P. M. Haygarth et al., is a
fine review on hydrological factors affecting phosphorus (P) transfer from agri-
cultural soils. The authors review current knowledge to define the spatial and tem-
poral controls on P transfer from agricultural soils via the various hydrological
pathways. Chapter 5, by N. M. A. Nassar, provides a thorough treatment of Cas-
sava, Manihot esculenta Crantz, genetic resources. Topics that are discussed in-
clude wild taxa, the genetic base and development of interspecific hybrids, devel-
opment and selection for apomixis, production of polyploid types, and protein
content in cultivars.
Many thanks to the authors for their first-rate contributions.
Donald L. Sparks
ix
THE MEASUREMENT
AND INTERPRETATION OF SORPTION
AND DESORPTION RATES FOR
ORGANIC COMPOUNDS IN SOIL MEDIA
Joseph J. Pignatello
The Connecticut Agricultural Experiment Station
New Haven, Connecticut 06511
I. Introduction
II. The Nature of Elementary Sorption Processes in Soils
A. Intermolecular Interactions Available to Organic Molecules
B. Properties of Soil Components and Mechanisms of Sorption
C. Thermodynamic Driving Force for Sorption
D. Rates of Elementary Processes
III. Slow Sorption and Desorption
A. Uptake and Release Profiles
B. Retardation Mechanisms
IV. Sorption Kinetic Models
A. Models Based on Bond Energetics
B. Driving Force Models
C. Diffusion Models
D. Stochastic Models
V. Experimental Methods
A. Batch Techniques
B. Column Techniques
C. Stirred-Flow Cell Technique
D. Zero-Length Columns
VI. Sorption Kinetics and Bioavailability
A. Assimilation of Chemicals in Soil Systems
B. Coupled Sorption–Biodegradation Kinetic Models
References
Sorption controls the physical and biological availability of chemicals in soil. Most
organic molecules undergo primarily weak physisorption interactions and the dri-
ving force for sorption is the hydrophobic effect. Sorption and desorption rates,
therefore, are governed mainly by molecular diffusion through the fixed interstitial
pores of particle aggregates and through the three-dimensional pseudophase of soil
organic matter. Retardation in the fixed pore system is due to tortuosity, chro-
matographic adsorption to pore walls, and, in the smallest pores, steric hindrance.
1
Advances in Agronomy, Volume 69
Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.
0065-2113/00 $30.00
2 JOSEPH J. PIGNATELLO
Soil organic matter, which has the strongest affinity for most organic compounds,
may exist in rubbery and glassy phases and retards sorption and desorption by its
viscosity and by the presence of internal nanopores, which detain molecules and
may sterically inhibit their migration. Soot carbon and/or ancient organic matter
may be present in some soils but their roles are yet unclear. Desorption rates are
correlated with the size and shape of the diffusant. Hysteresis is commonly observed
but a satisfactory explanation for it has yet to emerge. Mathematical models based
on bond energetics, driving force theory, diffusion, and stochastic analysis are dis-
cussed. These models have been used to describe batch experiments and have been
coupled to advection–dispersion transport equations for use in flowing water sys-
tems. Diffusion models are the most realistic but also the most difficult to apply be-
cause diffusion is highly dependent on the geometry and composition of the sor-
bent. Soil heterogeneity impedes the mechanistic interpretation of rates. Particles
span an extremely wide range of sizes. The appropriate diffusion length scale is of-
ten uncertain. The diffusion coefficient is expected to be concentration dependent
in any diffusing medium in which sorption is nonlinear. Furthermore, the diffusant
may alter the structure of soil organic matter. Bioavailability can be rate limited by
desorption. Cells are believed to access only dissolved molecules, but organisms
may affect sorption kinetics indirectly by steepening the concentration gradient or
by altering soil properties through bioactivity. Coupled sorption–biodegradation
models are necessary whenever nonequilibrium conditions prevail during exposure.
Models coupling Monod or first-order biodegradation kinetics with “two-site,”
driving-force, or diffusion models have been employed. Some have been used in con-
junction with the advection–dispersion transport model. © 2000 Academic Press.
I. INTRODUCTION
Sorption is fundamental to the fate of organic chemicals in soil environments.

In order to assess the influence of sorption, it is important to understand the nature
of the bonding forces between the sorbing molecule (sorbate) and the solid (sor-
bent), the thermodynamic driving forces responsible for establishing the position
of equilibrium, and the rates of association and dissociation of sorbate with time.
The primary focus of this chapter is on sorption kinetics. Sorption kinetics is an
important field of investigation in soil and environmental science because non-
equilibrium sorption conditions often apply as other fate processes, such as vapor-
or liquid-phase transport, uptake by organisms, and chemical reactions, are taking
place. In fact, sorptive equilibrium may take as long as many months. Recent re-
views (Alexander, 1995; Linz and Nakles, 1997; Pignatello and Xing, 1996) dis-
cussed the rate-limiting effects of sorption on mass transport and bioavailability
and the ramifications thereof for the management and risk assessment of chemi-
cals in soils and aquatic sediments.
Soil particles are typically porous or have phases, such as soil organic matter
(SOM), that are penetrable by the sorbate. Hence, sorption usually consists of at
SORPTION AND DESORPTION RATES 3
least three steps: (i) transport from the bulk fluid (vapor or liquid) to the vicinity
of the external surface of the particle, (ii) transport through the pore structure or
interstices of the particle, and (iii) formation of a “bond” at the “site” of sorption.
This chapter will begin by discussing the properties of sorbate and sorbent and the
nature of the sorption bond; in addition to providing a brief review of mechanisms,
the purpose of this introductory material is to make the reader aware of the com-
plexity of the sorption process, an essential prerequisite to understanding the ki-
netics of sorption. The chapter will then give an overview of the current state of
our knowledge about the mechanisms that retard sorption and desorption. Next,
mathematical approaches to describing sorption/desorption kinetics are discussed,
followed by a discussion of the experimental techniques for measuring rates. The
last section will address sorption kinetics in relation to bioavailability. The math-
ematical models discussed in this chapter will be presented only in their essential
features to save space and spare the reader unnecessary mental toil; consequently,
it is incumbent on an investigator to consult the original works before embarking
on their use.
It should be noted that a full understanding of the mechanisms that retard sorp-
tion has not been attained. As a result, there is plenty of opportunity for advance-
ment in the field. It is still not generally possible to predict a priori the entire up-
take or release curve for any given soil–chemical system. The goal of this chapter
is to lay a foundation for understanding the causes of sorption/desorption rate lim-
itations. In this chapter, we will deal with systems in a hydrated state.
II. THE NATURE OF ELEMENTARY SORPTION

PROCESSES IN SOILS
Defined broadly, the terms sorption and desorption refer to bulk mass-transfer
phenomena in which molecules leave the fluid phase and become associated with
an immobile phase and vice versa. The terms imply nothing about the nature of
the interaction nor about the transport of the sorbate molecules once in the con-
fines of the immobile phase. We speak of solution–solid and vapor–solid sorption.
Sorption of organic compounds may be broadly divided into the following cate-
gories (Fig. 1):
• Adsorption (A in Fig. 1) refers to association of molecules at the solid–fluid in-
terface. The interface may exist on the external surface of the particle facing the
bulk fluid or on the surface of a pore facing the fluid contained in that pore.
• Absorption (B in Fig. 1) occurs when molecules penetrate the solid surface and
intermingle with its three-dimensional molecular or atomic matrix. In natural
soils, SOM is the only component that is penetrable in this manner. Polluted soils
Figure 1 Different types of sorption available to organic molecules. A, Adsorption; B, absorption

in SOM or NAPL phase; C, capillary condensation; D, dissolution in water film; E, adsorption to wa-
ter film.
may contain additional absorptive phases in the form of nonaqueous phase liq-
uids (NAPLs)—solvents, oils, tars, and so on.
• Condensation (C in Fig. 1) refers to a phase change from the vapor or solution
state to a nonaqueous liquid or solid state. Condensation may occur on any sur-
face when the concentration is above the solubility or vapor pressure. However,
it is facilitated in small pores (50 nm): As a pore width decreases there is a pro-
gression from monolayer adsorption to capillary condensation owing to the ef-
fect of surface tension, which reduces the vapor pressure below the value of the
pure liquid in accordance with the Kelvin equation (Ruthven, 1984). Water com-
petes effectively with organics for condensation in pores of minerals because
such surfaces are ordinarily polar; however, recent studies of aquifer sediments
suggest that capillary condensation of compounds such as benzene may occur
even from aqueous solution and even at concentrations lower than their bulk wa-
ter solubility (Corley et al., 1996).
• Association with water films: Depending on the relative humidity, unsaturated

soils contain liquid water in pores and as coatings of surfaces. When organic va-
pors contact unsaturated soils, dissolution in (D in Fig. 1) and adsorption on (E
in Fig. 1) water films may occur (Kim et al., 1998; Ong and Lion, 1991; Petersen
et al., 1995). Molecules in such states are technically sorbed because they are re-
moved from the surrounding vapor state.
A. INTERMOLECULAR INTERACTIONS AVAILABLE

TO ORGANIC MOLECULES
Organic compounds can undergo chemisorption, physisorption, or ion pair for-

mation (ion exchange) with natural particles.
• Chemisorption involves significant atomic or molecular orbital overlap with the
solid phase; that is, the formation of a covalent or coordination bond. Examples
relevant to this chapter include “inner-sphere” coordination complexes between
carboxylate, phenolate, amine, or sulfhydryl groups and metal ions; i.e., IMn+ –
ZR, where IMn+ is a structural or adsorbed metal ion. Such bonds have both
ionic and covalent character. Sorption accompanied by formation of a true co-
valent bond (such as a C–C bond) is seldom reversible and thus is not consid-
ered relevant to this chapter.
• Physisorption involves weak intermolecular attractive forces between atoms and
molecules, including “van der Waals,” hydrogen (H-) bonding, and charge trans-
fer.
Van der Waals force encompasses the following interactions (Castellan, 1971;
Israelachvili, 1992): (i) dipole–dipole forces, resulting from mutual attraction
between permanent dipoles; (ii) induced dipole–induced dipole (dispersion)
forces, resulting from the synchronization of electronic motion in each molecule
producing momentary dipole moments in each; (iii) Dipole-induced dipole, re-
sulting from the attraction of a permanent dipole with the dipole it induces in its
neighbor. Forces (i–iii) involve no appreciable molecular orbital overlap, are
randomly oriented in space, and are only a few kilojoules per mole in energy.
Force (ii) is available to all atoms and molecules. The total van der Waals ener-
gy is the sum of all individual interactions between the sorbate and the site, and
it depends on the distance of approach, the sorbate size, and the polarizabilities
and polarities of both sorbate and site.
H-bonding (Schuster et al., 1976) involves interactions between acids and
bases of the type –AH...:B–, where A and B are ordinarily N, O, or S atoms. H-
bonding is a combination of the dipole–dipole force and a small degree of mo-
lecular orbital overlap. It is oriented in space (A–H–B angle ⬃15) and ranges
in strength from 10 to 25 kJ/mole.
Charge-transfer interactions (often referred to as donor–acceptor interactions)

may occur when an electron-poor acceptor (A) encounters an electron-rich donor
(D) and forms a complex in which one resonance structure represents transfer of
an electron (Foster, 1969):
A D S {A...D } A...D+}. (1)
Charge-transfer complexes most relevant to soil systems are n r and r
types, where n refers to a nonbonding lone-pair electrons and refers to an aro-
matic ring or other extended -conjugated system (Foster, 1969). Haderlein et
al. (1996) proposed n r charge-transfer complexation between permanent
charges on clays (donor) and electron-deficient polynitroaromatic rings (accep-
tor). r charge-transfer bonds are possible between appropriate functional
groups on sorbate and SOM (Müller-Wegener, 1987).
• Ion-exchange force involves electrostatic attraction between an organic anion or
cation and a charged group on the sorbent. It may be augmented by physisorp-
tion forces. For minerals, this type of sorption is best described as a concentra-
tion enhancement of the organic ion in the water near the surface, accompanied
by depletion of the original (usually inorganic) ion. Ion exchange may also oc-
cur at charged sites in SOM, usually carboxylate or phenolate groups.
B. PROPERTIES OF SOIL COMPONENTS

AND MECHANISMS OF SORPTION
1. Mineral Surfaces
Two principal types of surface exist on natural minerals:

1. The hydroxylated surface consists of -OH groups protruding into solution
from the topmost layer of metal ions (IMn+ –OH). It exists on all hydrous oxides
of Si, Fe, and Al and on the edges of layer silicate clays. It has variable positive or
negative charge density, depending on mineral, pH, and ionic strength. Regardless
of charge, it is strongly hydrophilic; adsorption of water on this surface is more
energetic than adsorption of nonpolar organic molecules (Curthoys et al., 1974),
and it is believed that at ordinary humidities one or more layers of ordered water
(“vicinal water”) are strongly under the influence of the surface.
2. The siloxane surface consists of oxygen atoms bridging underlying Si4+ ions.
It exists on the faces of many layer silicate clays. It has permanent negative charge,
depending on the degree of isomorphic substitution in the underlying lattice. The
charged sites are closely associated with metal or organic cations and the surface
in the vicinity of the charge is strongly hydrophilic. The neutral regions between
charges are hydrophobic or only weakly hydrophilic (Chen, 1976; Jaynes and
Boyd, 1991).
Figure 2 Depiction of sorption. (a) Sorption to mineral surfaces: A1, solvent-separated phys-
isorption; A2, physisorption with direct interaction with the surface; A3, chemisorption by coordination
with underlying metal ion. (b) Sorption to SOM: B1, adsorption to the SOM-coated mineral surface;
B2, adsorption to the extended organic surface; B3, absorption in the random network polymer phase.
Although not fully understood, several different modes of adsorption are be-
lieved occur on soil minerals (Fig. 2a). A1 in Fig. 2 refers to physisorption in which
the sorbate is separated from the surface by solvent molecules (Goss, 1992). This
occurs on hydroxylated surfaces for compounds that cannot displace adsorbed wa-
ter. This type of adsorption might be best described as a concentration enhance-
ment of the solute in the “vicinal water.” A2 refers to physisorption in which the
sorbate is in direct contact with surface atoms. Direct contact occurs on neutral
siloxane surfaces, as well as on hydroxylated surfaces, provided water is scarce or
the compound’s H-bond ability is sufficiently great that it can displace tightly
bound water. A3 refers to chemisorption through inner-sphere coordination with
lattice or adsorbed metal ions. This mechanism requires appropriate coordinating
functional groups on the molecule. D refers to pore condensation as discussed in
reference to Fig. 1.
2. Soil Organic Matter
It is well established that sorption of hydrophobic compounds out of aqueous

solution or at high relative humidity is dominated by the SOM fraction unless that
fraction is very small (Schwarzenbach et al., 1993). For example, sorption of chlo-
rinated benzenes and polycyclic aromatic hydrocarbons (PAHs) to nonporous in-
organic oxides is so weak that it is expected to be insignificant when the fraction
of soil organic carbon ( foc ) is ⬃0.0001 (Mader et al., 1997)! Situations in which
the predominance of SOM does not necessarily hold include (i) very dry condi-
tions, when capillary condensation or adsorption can be important, and (ii) when
chemisorption is important.
SOM consists of plant and microbial material in various stages of decomposi-
tion. Materials bearing little physical and chemical resemblance to their precursor
biological polymers are called humic substances and make up the bulk of SOM
(Hayes et al., 1989). Knowledge about humic substances is mainly inferred from
studies of humic and fulvic acids, which are humic substances isolated from nat-
ural waters or extracted from soil with dilute base or polar solvents. Humic and
fulvic acids are a refractory mixture of polyanionic macromolecules ranging from
hundreds (Novotny et al., 1995) to hundreds of thousands of grams per mole (see
Pellegrino and Piccolo (1999), however). Bearing in mind that each humic
macromolecule may be unique, a hypothetical structure has been proposed on the
basis of physical, spectroscopic, and fragmentation-identification studies (Schul-
ten and Schnitzer, 1993) (Fig. 3a). It has both aliphatic and aromatic subunits and
an abundance of oxygen functional groups. In solution, the macromolecules coil
up in a random fashion and aggregate to form a spheroidal, water-swollen phase
of entangled humic macromolecules (Fig. 3b). The density of the particle increas-
es gradually from edge to center (Hayes and Himes, 1986; Swift, 1989).
The unextractable SOM—typically more than half the total—is called humin.
Humin is separated from minerals only by drastic treatment such as hydrofluoric
acid digestion which dissolves the minerals (Preston et al., 1989). Humin may be
separated into lipid-like and humic-like components (Rice and MacCarthy, 1990).
Little is known about humin, even though it may have a greater affinity for organic
compounds than whole SOM (Xing and Pignatello, 1997). The bulk of humin may
consist simply of humic acid-like molecules of higher molecular weight and
stronger affinity for mineral surfaces. Humin is more hydrophobic and more con-
densed than humic or fulvic acids.
In the native state, SOM is usually bound to mineral particles on a scale rang-
ing from a monolayer organic film to a discreet organic phase. The nature of SOM
as a sorbent of organic compounds—obviously crucial to its role in sorption ki-
netics—is controversial. SOM has been modeled as a coating on mineral surfaces,
an extended organic surface, or a random network polymer phase. These are de-
picted in Fig. 2b.
As a coating, SOM is regarded to enhance the surface affinity for organic mol-
ecules by making it more “hydrophobic,” similar to the effect of alkyl chains at-
tached to the surface of silica gel used in reverse-phase liquid chromatography. On
such a surface, the sorbate may be under the simultaneous influence of the miner-
al and the organic matter. Mayer (1999) provides evidence, however, that even in
low-organic carbon (OC) aquifer sediments SOM exists in multilayer patches
rather than as monolayers on the surface. The extended organic surface concept
regards SOM to be an impenetrable adsorptive surface. The external surface area
of SOM measured by N2 adsorption at 77 K using the B.E.T. equation is on the or-
der of ⬃100 m2 /g (Chiou et al., 1993), which appears to be too low to account for
the high affinity of SOM for organic compounds, implying that little impenetrable
surface exists.
The preponderance of evidence points to SOM behaving as a random network
polymer phase that provides a three-dimensional hydrophobic environment for or-
ganic molecules. The “surface” of such a phase is expected to be diffuse rather than
sharply defined due to more extensive solvation of the outer polar regions of the hu-
mic polymers that face the solvent (Hayes and Himes, 1986; Swift, 1989). If true,
a long-lived surface-adsorbed state would be disfavored. Instead, the sorbate is ex-
pected to penetrate the phase and intermingle with the humic strands, much the
same way in which small molecules interact with synthetic organic polymers
(Rogers, 1965; Vieth, 1991; Frisch and Stern, 1983). The structure of lignin, the
woody component of plant material and probably the main precursor of terrestrial
humic substances, is also considered a “random network polymer” (Goring, 1989).
According to the polymer phase concept, sorption is attributed to dissolution
(absorption) of the hydrophobic solute in the liquid-like, organophilic phase in or-
der to escape the polar environment of water (Chiou, 1989). Unlike a liquid, how-
ever, the sorption potential of SOM is not uniform (Pignatello, 1998, 1999; Xing
et al., 1996; Xing and Pignatello, 1997; Young and Weber, 1995). Sorption
isotherms tend to be nonlinear in the sense that sorption diminishes with increas-
ing concentration. The isotherm can be fit to the Freundlich equation,
qe KeCen, (2)
where qe and Ce are the equilibrium sorbed and solution concentrations, Ke is the
sorption coefficient, and n is a constant 1. Moreover, sorption in bisolute and
multisolute systems is competitive. These behaviors indicate a more specific
mechanism than ideal solid-phase dissolution and can be reconciled by consider-
ing SOM as a composite of “rubbery” and “glassy” polymers. Accordingly, the
properties of SOM vary continuously from rubbery-like phases that have an ex-
panded, flexible, and highly solvated structure to glassy-like phases that have a
condensed, rigid, and less solvated structure (Pignatello, 1998, 1999; Xing et al.,
1996; Xing and Pignatello, 1997). The glassy character has been suggested to in-
crease with diagenetic alteration in the following natural progression: SOM r
kerogen r coal and shale (Huang and Weber, 1997; Young and Weber, 1995).
The nature of sorption is postulated to change along the rubbery–glassy con-
Figure 3 Soil organic matter. (A) Hypothetical structure of a humic macromolecule (reprinted
from Schulten and Schnitzer, 1993, with permission from Springer-Verlag). (B) Three-dimensional de-
piction of a natural organic matter colloid in aqueous solution. The colloid is an approximately spher-
ical polymer mesh of entangled humic macromolecules that is swollen with water (water molecules not
shown). The mass density increases toward the center. Some negative charges on the humic strands
form ion pairs with metal cations, whereas others are balanced by counterions in solution. Cross-link-
ing between strands is illustrated for the divalent cations Ca2+ and Mg2+. (Reprinted from Pignatello,
1998.)
tinuum in the same fashion as sorption of gases and organic molecules in poly-
mers. In highly rubbery regions of SOM sorption occurs by solid-phase dissolu-
tion, whereas in glassy regions sorption occurs by a combined mechanism of sol-
id-phase dissolution and site-specific, “hole-filling” processes. The holes are
postulated to be nanometer-size pores made up of rigid humic segments, in which
the guest molecules undergo an adsorption-like interaction with the pore walls.
The sorption isotherm is thus given by the “dual-mode” equation (Eq. 3) (Pigna-
tello, 1999; Xing et al., 1996; Xing and Pignatello, 1997), in which total sorption
(qe ) is contributed by solid-phase dissolution (qD) and the sum of multiple site-se-
lective processes (qL), each of which follows a Langmuir relationship:
Figure 3 Continued
Figure 4 Rubbery–glassy polymer concept of soil organic matter. The perspective is intended to
be three-dimensional. The rubbery and glassy phases both have dissolution domains in which sorption
is linear and noncompetitive. The glassy phase, in addition, has pores of subnanometer dimension
(“holes”) in which adsorption-like interactions occur with the walls, giving rise to nonlinearity and
competitive sorption. The binding is analogous to host–guest inclusion complexes in chemistry.
(Reprinted from Xing and Pignatello, 1997.)
n
b QC
qe = qD + qL = K D Ce + ∑ 1 +i bi Ce , (3)
i =1 i e
where KD is the (linear) dissolution domain coefficient made up of inseparable

terms representing the rubbery phase and the dissolution domain of the glassy
phases, Ce the equilibrium solution concentration, and bi and Qi are the Langmuir
affinity and capacity constants for the ith unique site in the hole-filling or Lang-
muir domain. The dual-mode model is depicted in Fig. 4.
Gas adsorption studies confirm the existence of internal nanoporosity in SOM
samples which increases the total surface area by at least two orders of magnitude
(Xing and Pignatello, 1997; de Jonge and Mittelmeijer-Hazeleger, 1996). The
nanoporosity is correlated with the degree of nonlinearity in the isotherms and the
degree of competition between compounds of like structure (Xing and Pignatello,
1997). Conditions that favor the rubbery state—increased temperature, the pres-
ence of cosolvents such as methanol, and high concentrations of cosolute—tend
to make the isotherm more linear. The degree of nonlinearity follows the order ex-
pected on the basis of the glassy character of the material: humic acid humin.
As will be shown, there is increasing evidence that the mass transfer rates depend
on the rubbery–glassy character of SOM.
3. Carbonaceous Materials Other Than SOM
Soils may contain forms of carbon not usually classified as SOM. These include
ancient materials such as kerogen, coal, and shale, and “black carbon” (also known
as “soot”), which refers to incompletely combusted organic material. Such mate-
rials are widely distributed in the environment and, because they are hydrophobic,
are expected to have a high affinity for organic compounds (Kuhlbusch, 1998;
McGroddy et al., 1996). The nature of these materials as sorbents of organic com-
pounds is not well-known. Coal appears to have properties quite like glassy poly-
mers—“internal microporosity” (Larson and Wernett, 1988) and demonstrable
glass-to-rubber transition temperatures (above 300°C) (Lucht et al., 1987). Soots
are expected to have some impenetrable hydrophobic surface. If this is true, sorp-
tion may occur by adsorption and condensation in fixed pores, as occurs in famil-
iar inorganic materials. However, they may also have tar-like phases which behave
more like absorption domains. PAHs, being products of incomplete combustion
themselves, may become occluded in the interstices of soot particles during their
formation in a way that makes them extremely unavailable (Gustafsson et al.,
1997).
Sorption of chemicals by NAPLs occurs by simple liquid-phase dissolution
analogous to organic solvents such as hexane and octanol. The partitioning be-
tween the fluid phase and NAPLs is therefore governed by Raoult’s law (water–
NAPL) or Henry’s law (vapor–NAPL); that is, the fluid-phase concentration is
proportional to the mole fraction of contaminant in the NAPL times the solubility
or vapor pressure, respectively, of a pure reference state (Schwarzenbach et al.,
1993).
C. THERMODYNAMIC DRIVING FORCE FOR SORPTION
Upon sorption from solution, an organic molecule exchanges one set of inter-
actions with the solvent for another set of interactions with the sorbent. The mo-
lar free energy change at constant temperature and pressure encompasses free en-
ergy changes in sorbate–sorbent, sorbate–solvent, sorbent–solvent, and solvent–
solvent interactions of all components involved in the sorption process, including
displaced molecules such as water from the surface, and any reorganization oc-
curring on the surface. For nonpolar and weakly polar compounds capable of in-
teracting only by nonspecific physisorption mechanisms, sorption from water to
mineral surfaces (Goss, 1997), as well as to SOM (Chiou, 1989), is driven princi-
pally by the hydrophobic effect. The hydrophobic effect results from the gain in
free energy when a molecule possessing hydrophobic surface area is transferred
out of the polar medium of water. The hydrophobic effect plays the same domi-
nant role in aqueous solubility. Surface tension studies show that the hydrophobic
effect is due almost entirely to the H-bonding cohesive energy of water (van Oss
et al., 1988; van Oss and Good, 1988). It is thought that water molecules form an
ordered cage around the hydrophobic portions of the solute, costing enthalpy, en-
tropy, or both (Schwarzenbach et al., 1993; van Oss et al., 1988). H-bonding and
dipolar interactions with the sorbent will increase the thermodynamic driving force
for sorption only if such interactions with the sorbent are stronger than those with
the solvent.
D. RATES OF ELEMENTARY PROCESSES
It has been shown that sorption of organic compounds to soil particles usually
involves the weak physisorption interactions. In solution, such interactions are
practically instantaneous. For example, the lifetime of the H2O...NH2CH3 hydro-
gen bond in water is only 1.2
1011 s (Emerson et al., 1960). Van der Waals in-
teractions are even shorter lived. The situation on the surface is more complex,
however. Consider the elementary collision of a gas molecule with an unhindered
plane surface having a number of “sites” of identical energy. The energy profile
versus distance from the surface is illustrated in Fig. 5. As the adsorbate ap-
proaches the surface it descends into a potential energy well of depth Q. The in-
stantaneous rate of adsorption is proportional to the pressure p and the concentra-
tion of vacant sites Sv. The instantaneous desorption rate is proportional to the
concentration of occupied sites So. In the Arrenhius formulation, the rate expres-
sions are
Rate of sorption Aae(Ea*/RT )pSv , (4)
Rate of desorption Ade(Ed*/RT )So , (5)
Figure 5 Energy diagram for a physisorbing molecule approaching the surface.

where Ea* and Ed* are the adsorption and desorption activation energies (kJ mole1),
R is the universal gas constant, and Aa and Ad are the preexponential constants re-
flecting the “sticking probability.”
Physisorption of a molecule from the vapor state to an unhindered surface is
considered to be unactivated (Ea* near zero) (Adamson and Gast, 1997). Phys-
isorption from solution may be slightly activated due to reorganization of solvent
molecules around the sorbate and the surface. Desorption, however, is always ac-
tivated because it requires “ascension” from the potential energy well. Assuming
the principle of microscopic reversibility (that desorption follows the reverse path-
way of sorption),
Ed* Q Ea*. (6)
The Q is approximately equal to the isosteric heat of adsorption H (Ruthven,
1984). Values of H for vapor sorption on dry, nonporous silica gel (which rep-
resents a hydroxylated mineral surface) range from 36 to 63 kJ/mol for small non-
polar molecules, such as n-hexane, tetrachloromethane, and benzene, as well as
polar molecules containing ether, cyano, nitro, ketone, and ester groups (Curthoys
et al., 1974). Goss (1994) found that H for C1 –C10 compounds on moist sur-
faces of silica, quartz sand, and clays ranged from 28 to 50 kJ/mol. Adamson and
Gast (1997) estimate the mean time of stay on the surface, , of a molecule at 25C,
assuming Ea* 0 and a sticking probability of unity. For Q 37.6 kJ/mol,
106 s and for Q 83.7 kJ/mol, 102 s. Thus, we may expect that phys-
isorption of typical compounds on a plane surface should occur in minutes or less.
This is generally borne out experimentally (Adamson and Gast, 1997; Ruthven,
1984).
The same conclusion of rapid equilibrium does not apply, however, when the
elementary reaction involves forces stronger than physisorption. For example, in
ligand exchange, Ea* may be appreciable due to such effects as orbital rehy-
bridization and displacement of other ligands (OH, H2O, or organic ligand) from
the inner coordination sphere of the metal ion (McBride, 1994). The Ed* will reflect
the energy of the coordination bond, which is especially strong when the molecule
coordinates through two or more functional groups (the “chelate effect”). There is
little quantitative information available on the forward or reverse rates of ligand-
exchange reactions on soils or appropriate model sorbents.
Also, the conclusion of rapid equilibrium does not apply to sorption in a steri-
cally hindered environment such as the interstices of SOM. There, from the stand-
point of the sorbate molecule as it tries to weave its way around the humic strands,
each elementary jump from site to site is sterically hindered. Depending on the
flexibility of the humic matrix, a cooperative movement in the position of humic
macromolecules may be required in order to allow the sorbate molecule to pass.
Thus, an elementary jump is expected to involve the making and breaking of nu-
merous interactions simultaneously and therefore may be highly activated.
III. SLOW SORPTION AND DESORPTION
A. UPTAKE AND RELEASE PROFILES
Evidence for slow (“nonequilibrium”) sorption phenomena has come from

many sources—the extended tailing of breakthrough curves in soil column ex-
periments, the formation of so-called “resistant” fractions in batch experiments,
and the unexpected persistence of pesticides in the environment (Alexander, 1995;
Pignatello, 1989). Since the late 1980s there have been many reports of chemical
residues in samples collected from the field that appear to desorb extremely slow-
ly, as well as reports of laboratory-spiked compounds that undergo sorption or de-
Figure 6 Sorption of phenanthrene in suspensions of Pahokee peat soil and derivatives. (a) Up-
take by the peat soil showing slow approach to equilibrium. Apparent Koc is the ratio of sorbed to so-
lution concentration referenced to organic carbon. (b) Normalized desorption of 2 g/g phenanthrene
from the peat soil a in the presence of Tenax adsorbent beads after 3- to 100-day aging periods. (c)
Same as b from peat humin (a is from White and Pignatello, unpublished; b and c are reprinted from
White et al., 1999).
sorption over weeks or months. This subject has been reviewed by Pignatello
(1989) and Pignatello and Xing (1996).
Figure 6a shows that phenanthrene, a three-ring PAH, required in excess of 110
days to reach equilibrium in a shaken suspension of a peat soil containing mostly
organic matter (6.9% ash). Figure 6b shows the desorption of phenanthrene after
various contact (“aging”) periods of 3–100 days (White et al., 1999). The de-
sorptions were carried out in the presence of Tenax polymeric adsorbent, which
rapidly sorbs phenanthrene as it leaves the soil, approximating conditions of zero
concentration infinite bath and maximizing the driving force for desorption. One
can see that the desorption rate slows with an increase in the aging period. It is
worth noting that after only 3 days of sorption approximately 20% of phenanthrene
strongly resists desorption over the subsquent 90 days in the presence of the infi-
nite sink. Similar findings—that short-term contact can lead to formation of a
strongly resistant fraction—have been reported by others (Kan et al., 1997, 1998).
The following are observations pertaining to the resistant fraction:
1. Desorption is highly temperature dependent, being significantly enhanced by

heating. For example, the apparent desorption activation enthalpy for aged 1,2-di-
bromoethane was 66 kJ/mol (Steinberg et al., 1987) and that of aged chloroben-
zenes, polychlorinated biphenyls (PCBs), and PAHs ranged from 60 to 70 kJ/mol
(Cornelissen et al., 1997b).
2. Desorption is accelerated by addition of cosolvents but only slightly by ad-
dition of surfactants (Deitsch and Smith, 1995).
3. Desorption is accelerated by breaking up particles (Steinberg et al., 1987;
Ball and Roberts, 1991b).
4. Resistant fractions may be formed in soils containing no appreciable miner-
al matter (e.g., Fig. 6), in strictly inorganic porous particles (Farrell and Reinhard,
1994a,b; Werth and Reinhard, 1997), and perhaps even in colloidal-size particles
(Maguire et al., 1995; Schlebaum et al., 1998).
B. RETARDATION MECHANISMS
Slow kinetics has been exhibited by aliphatic and aromatic hydrocarbons, halo-
genated aliphatic and aromatic hydrocarbons, and agricultural chemicals. Gener-
ally, only physisorption interactions are open to them. Chemical and biological
transformations, although quite possible, are irreversible in the sense that the by-
products cannot easily revert to starting compound and would not be identified as
starting compound by the analyst using modern techniques. Therefore, the only
reasonable explanation for slow kinetics for such compounds is mass transfer re-
sistance—the resistance of the matrix to molecular diffusion.
Diffusion is the tendency of molecules to migrate against a gradient in concen-
tration (more correctly, a gradient in chemical potential) so as to achieve maxi-

mum entropy. Soil particles are characteristically porous and contain minerals
such as SOM that can absorb small molecules within their interstices. These ma-
terials can provide resistance to diffusing molecules in many ways. Most suggest-
ed mechanisms for hindering the sorption process can be grouped into the follow-
ing: pore diffusion (PD), and intraorganic matter diffusion (IOMD).
Variations exist within each group, and in some cases there is some overlap. Al-
though investigators have argued the merits of one compared to the other, it is like-
ly that both operate, depending on soil properties. Resistant fractions can be gen-
erated in purely inorganic sorbents such as porous silica gel (Farrell and Reinhard,
1994b; Werth and Reinhard, 1997) and in pure organic materials such as low-ash
peat soils (White et al., 1999).
1. Pore Diffusion
PD attributes slow rates to hindered diffusion of molecules through the fixed in-
traparticle pore system. Fixed pores are more or less permanent and unaffected in
shape by the presence of the diffusant. Porosity exists in cracks, lattice disconti-
nuities, along grain boundaries, and in the interlayers of expandable clays. Pore
sizes are classified by IUPAC according to their aperture (d ):
Macropores: d 50 nm,
Mesopores: 50 d 2 nm,
Micropores: d 2 nm.
In addition, there is a class of pores in the ⬃0.3- to 1-nm range referred to in the
literature as “ultramicropores” or “nanopores.” For perspective, the C–C bond is
⬃0.15 nm long and CCl4 is ⬃0.5 nm in diameter. Researchers have different views
on the nature of the pores and the root causes of hindered sorption.
The pore surface may be organic or inorganic. In most PD models it is assumed
that molecules instantaneously equilibrate locally between the pore liquid phase
and the surface (“local equilibrium”). Diffusion in pores may be hindered with re-
spect to diffusion in a bulk fluid by any or all of the following mechanisms: (i) tor-
tuosity, a term encompassing elongation of diffusion paths relative to a straight
line, variations in pore diameter, and the degree of pore connectivity as reflected
by the presence of “dead-end” pores; (ii) sorption to pore walls analogous to a
“chromatographic effect”; and (iii) steric interference from pore walls, especially
in pores approaching the diffusant diameter. These will be discussed in more de-
tail in Section IV,C. In addition, diffusion in small pores may be hindered by the
viscous nature of water near hydrophilic surfaces where water molecules are
strongly under the influence of the surface.
Although the concept of pore diffusion is long known, Wu and Gschwend
(1986) appear to be among the first to employ it to describe intraparticle diffusion

of chemicals in soils and sediments. Ball and Roberts (1991b) used the PD mod-
el to describe sorption of trichloroethane (TCE) and tetrachlorobenzene in aquifer
solids over long periods. They often obtained superior fits by including an instan-
taneously sorbing fraction of up to 30% of total. In these (Ball and Roberts, 1991b;
Wu and Gschwend, 1986) and other studies (Kleineidam et al., 1999) the results
were consistent with the nominal particle radius as the length scale over which dif-
fusion occurs. By contrast, other studies (Carroll et al., 1994; Cornelissen et al.,
1998b; Farrell and Reinhard, 1994b; Pignatello et al., 1993; Pignatello and Xing,
1996; Steinberg et al., 1987) found little or no dependence of diffusion rates on
nominal soil particle radius, suggesting the appropriate diffusion length scale may
be much smaller than the nominal particle radius, perhaps as small as 10 –100 nm.
The diffusion length scale likely depends on the micromorphology of the soil par-
ticles in the sample.
Pignatello (1990b) observed enhanced release of a portion of strongly resistant
fractions of halogenated hydrocarbons by acidification of the suspension to pH
3. This suggested that some SOM, in particulate or coating form, had been shield-
ed by mineral grains that were subsequently dispersed when the materials ce-
menting them were acid dissolved. The results of Holmén and Gschwend (1997)
on PAH transport in aquifer sand columns support this idea. They suggested that
diffusion in porous oxide coatings on quartzite sand particles controls the rate of
diffusion. The coatings, which consisted of fine-grained iron oxide and alumi-
nosilicate clay particles, had porosities of 0.4 or 0.5, thicknesses up to ⬃200 m,
and OC contents (0.7–1.6%) higher than the quartz substrate. Since the retarda-
tion of PAH transport was less than expected based on calculated Koc values, they
inferred that only a fraction of SOM was accessed during a compound’s pass
through the column. The flow velocities, however, were quite high–0.5–115 cm/
h for a 7-cm column or 1.7–400 column pore volumes per day.
Farrell and Reinhard (1994b) and Werth and Reinhard (1997) desorbed TCE va-
pors from unsaturated silica gel columns or soils preequilibrated with TCE at fair-
ly high relative pressures and 100% humidity. They observed biphasic kinetics
(fast and slow phases). The small, highly resistant fraction of TCE that was formed
was attributed to hindered diffusion in “hydrophobic micropores.” Corley et al.
(1996) suggested that the resistant fraction of TCE and other volatile organic com-
pounds (VOCs) might be associated with a neat VOC phase formed by capillary
condensation in micropores or small mesopores during the sorption step.
In their study of chlorinated benzenes and biphenyls in freshwater sediment
(2.8–6.3% OC), Lick and coworkers (Borglin et al., 1996; Lick and Rapaka, 1996;
Tye et al., 1996) proposed that sorption/desorption rates are controlled by diffu-
sion in the pore network of flocs. Flocs result from the aggregation of sediment
grains suspended in water. Their size and density is a function of sediment con-
centration, fluid shear force, and water chemistry. Consistent with their mecha-
nism, the effective diffusion coefficient depended on the floc size and porosity, sed-
iment OC content, and (linear) partition coefficient of the sorbate.
Another location where diffusion might be hindered is the interlayers of ex-
pandable clays. The interlayer gap is typically 1 nm—small enough to provide
steric hindrance to diffusion or even size exclusion of some compounds. An im-
portant question that has not been satisfactorily answered regards availability of
clay interlayers in natural soils to pesticides and other chemicals. The small
amount of published information suggests that diffusion in the interlayer, when it
is accessible, is relatively fast. Sawhney and Gent (1990) sorbed TCE and 1,2-di-
bromoethane vapors onto various expandable and nonexpandable clays under dry
conditions. Desorption from the (expandable) smectite gave among the fastest
rates, and X-ray analysis did not support penetration of the interlayer. In desorp-
tion of TCE from moist packed columns, Farrell and Reinhard (1994b) found that
montmorillonite gave the smallest resistant fraction among many model and nat-
ural sorbents, but they, too, argued that interlayer penetration had not occurred.
Huang et al. (1996) found that sorption of phenanthrene to bentonite was complete
in a few hours, but no evidence of interlayer processes could be found. The re-
maining literature on the subject is confined to organoclays. Organoclays have
quaternary ammonium ions (e.g., hexadecyltrimethyl ammonium) as exchange-
able cations that are believed to provide an organophilic phase, or surface, with
high affinity for hydrophobic compounds. Studies of organoclays [e.g., naph-
thalene and di-uron (Nzengung et al., 1997) and carbon tetrachloride and 1,2-
dichlorobenzene (Deitsch et al., 1998)] indicate that sorption equilibrium appears
to be complete in hours and is much faster than sorption to SOM in the form of
peat particles (Deitsch et al., 1998). Moreover, the solute–sorbent aging time did
not significantly affect the rate of desorption (Deitsch et al., 1998). Questions re-
main, however, about how much sorption in these organoclays occurred in the in-
terlayer versus on the edges (Nzengung et al., 1997).
2. Intraorganic Matter Diffusion
a. General Considerations
Since neutral organic compounds tend to associate predominantly with the SOM
fraction, it is natural to consider whether SOM is the principal cause of hindered
diffusion. SOM can hinder diffusion in at least two ways. First, even as a “rub-
bery” organic gel, SOM represents a highly viscous fluid that impedes molecular
diffusion compared to water. Diffusion coefficients of small molecules in rubbery
polymers compared to water are several orders of magnitude smaller and depend
more strongly on the size and shape of the diffusant (Berens, 1989; Rogers, 1965).
In the solid state, humic acid is believed to be a more rubbery form of organic mat-
ter than the SOM from which it was extracted (Xing and Pignatello, 1997). The
Figure 7 Diffusion coefficient at 30C for gases and organic vapors in glassy () or rubbery ()
polyvinyl chloride. The rubbery state was obtained by adding phthalate ester plasticizers. (Redrawn
from Fig. 9 of Berens, 1989, with permission.)
diffusion coefficients of toluene, n-hexane, and acetone is pressed humic acid disks
range from 108 to 109 cm2 s1 (Chang et al., 1997), about the same as those in
rubbery polymers at the same temperature, and may be compared to values of ap-
proximately 105 cm2 s1 in water.
Second, glassy SOM offers a much greater impediment to diffusion than rub-
bery SOM because it is more rigid and condensed and it contains holes (nanopores)
in which organic molecules can momentarily be detained (Pignatello, 1998; Pig-
natello and Xing, 1996; Xing and Pignatello, 1997). Figure 7 shows that the dif-
fusion coefficient of gases and organic molecules in glassy polyvinyl chloride
(PVC) is smaller than that in rubbery PVC for a given molecular diameter. Fur-
thermore, they sharply diverge as the molecular size of the diffusant increases
(Berens, 1989). Hole filling (and hole emptying) becomes an increasingly activat-
ed process as steric constraints at the hole increase. Figure 6c shows that desorp-
tion of phenanthrene from peat humin—the insoluble organic matter after humic
acid is removed—is slower than that from the original peat SOM for a given
aging period, reflecting the more glassy character of the humin compared to the
native SOM (White et al., 1999). It has also been shown (White and Pignatello,
2000) that pyrene, a four-ring PAH, not only acts thermodynamically as a com-
peting co-solute toward phenanthrene but also increases the rate of phenanthrene
desorption, presumably by blocking nanopore sites ordinarily available to phenan-
threne. This strongly suggests that the presence of nanopores impedes molecular
diffusion inside SOM.
Similar conclusions about the effect of SOM structure on diffusion rates have
been reached by Weber and coworkers in their studies of hydrophobic compound
sorption on soils and model materials (Weber and Huang, 1996). They proposed a
three-domain model of soil. The domains fill up in the following order:
Domain I: exposed inorganic surface
Domain II: “amorphous” SOM (equivalent to rubbery)
Domain III: “condensed” SOM (equivalent to glassy)
Domain I, which is minor for hydrophobic compounds, is filled in minutes
(Huang et al., 1996). The conclusion that domain III fills slowest is based on find-
ings that the Freundlich exponent of phenanthrene (n) decreases with approach to
equilibrium, especially in the first hours. The nonlinearity is assumed due to sorp-
tion in domain III. Similar changes in n with time were reported by Xing and Pig-
natello (1996) for 2,4-dichlorophenol, metolachlor, and 1,2-dichlorobenzene in
two soils, including the 93% SOM peat soil. The decrease in linearity is due to the
combined effects of increasing contribution from the glassy SOM with time and
the intrinsic concentration dependence of diffusion in the glassy state (see Section
IV,C,4). In the glassy state, diffusivity increases with sorbate concentration due to
the following: (i) the decline in hole-filling sorption (see Eq. 3)—i.e., as the holes
fill up, there is less impedance for subsequent molecules as they pass through. This
is confirmed by the competition experiments between phenanthrene and pyrene
mentioned above. (ii) At high enough concentrations the sorbate can “plasticize”
the polymer—that is, bring about its conversion to a more rubbery state.
b. Structure–Activity Relationships
On the assumption that IOMD is the important limiting mechanism, many re-
searchers have tried to relate the desorption rate parameter to molecular structure.
Carroll et al. (1994) found that the effective diffusion coefficient (Deff) of PCBs
in a sediment decreased with molecular size; about an order of magnitude decline
in Deff occurred from monochlorinated to trichlorinated biphenyls. Brusseau and
coworkers (Brusseau, 1993; Hu et al., 1995; Piatt and Brusseau, 1998) studied the
transport of various compounds in packed soil columns. Through analysis of solute
breakthrough curves they obtained a desorption mass transfer coefficient () for
the noninstantaneous fraction (see discussion of the “two-site” model in sections
IV,A and V,B). Since the residency time of the solutes in the columns was only a
few minutes to a few hours, their results apply to short-timescale phenomena; ap-
plicability to longer timescale sorption requires caution. They found a linear log–
log relationship between and Kow:
log a log Kow b, (7)
where a and b are regression constants. This constitutes a linear free energy rela-
tionship (LFER) between sorption rate and sorption strength since log is pro-
portional to activation energy, E*d, and log Kow is proportional to log Ke, which in
Figure 8 Linear free-energy relationship between the desorption mass transfer coefficient () and
the first-order molecular connectivity index (1Xv) for PAHs, alkyl benzenes, chlorinated benzenes, and
alkenes in two sandy aquifer samples (SB13-5 and SB-13-9) taken from a single bore hole at different
depths. (Reprinted with permission from Piatt and Brusseau, 1998. Copyright 1998 American Chemi-
cal Society.)
turn is proportional to the thermodynamic free energy of sorption, Gsorp. The

slope of Eq. (7) was negative, which means that the rate of desorption decreases
with increasing affinity for the sorbent.
Brusseau and coworkers interpreted the LFER in terms of a polymer diffusion
concept. Thus, increasing molecular size results simultaneously in increasing hy-
drophobicity and decreasing mobility in the viscous organic phase. Such interpre-
tation has also been given for diffusion of small- and medium-size molecules in
polymers (Rogers, 1965). Brusseau (1993) and Piatt and Brusseau (1998) actual-
ly obtained slightly better LFERs between and the molecular connectivity index
X—a measure of topological size and degree of branching—than between and
Kow. Figure 8 presents such a correlation for hydrophobic compounds in two soils.
They argued that diffusion through SOM is not just dependent on molecular po-
larity or hydrophobicity but also on size and shape. Such findings are consistent
with diffusion limitations in an organic phase. They are not, however, inconsistent
with diffusion limitations in fixed pore systems, as numerous studies of reference
materials have shown (Kärger and Ruthven, 1992).
3. Effect of Soil Heterogeneity on Sorption Kinetics
When dealing with a heterogeneous mixture of particles, the rate of sorptive up-
take will be dominated by the faster-sorbing particles at short times and the slow-
er-sorbing particles at long times. Pedit and Miller (1994, 1995) showed that bet-
ter fits to the pore diffusion model could be obtained by incorporating different
size classes into the model (see Section IV,C,6). The size classes not only have dif-
ferent diffusive path lengths but also may have different equilibrium sorption ca-
pacities. Kleineidam et al. (1999) studied sorption to sands and gravels in south-
west Germany and Switzerland. The samples, fragments of Triassic and Jurassic
sedimentary rock, were separated according to size and lithographic type. They
found that the rates of phenanthrene uptake depended on both particle size and
properties. In general, dark-colored particles had the highest OC contents, lowest
porosities, and highest sorption capacities while giving the slowest kinetics (e.g.,
10% equilibrium in 500 days). The lighter-colored particles were just the opposite
and showed the fastest kinetics (e.g., equilibrium in a few to 100 days). Most of
the OC in these samples was ancient.
4. Hysteresis
Hysteresis refers to the apparent asymmetry (nonsingularity) of the sorption/

desorption process. There is reference in the literature to asymmetry in the
isotherm, where the curve defining the relationship between sorbed and fluid-
phase concentrations is different depending on whether it is determined in the for-
ward (sorption) or the reverse (desorption) direction. There is also reference to
nonsingularity in the rate parameters for sorption and desorption. Hysteresis has
been observed in many soil–chemical systems but its causes have not been satis-
factorily explained.
Provided sorption is reversible and true thermodynamic equilibrium is attained,
isotherms constructed from the sorptive and desorptive directions are expected to
be superimposable. Figure 9 shows two examples of isotherm hysteresis by
phenanthrene—in a riverine sediment (Fig. 9a; Kan et al., 1998) and in a shale
material (Fig. 9b; Huang and Weber, 1997). In the former, a single sample was sub-
jected to numerous desorption cycles, while in the latter, each sample was de-
sorbed only once. As exemplified by Fig. 9a, the desorption curve often appears to
intersect the ordinate at a nonzero value, indicating the presence of a “strongly re-
sistant” fraction. Aside from method artifacts or chemical transformations (Rao
and Davidson, 1980), there are several possible causes of isotherm hysteresis:
1. Formation of metastable states: Metastability plays an important role in the
condensation/evaporation of gases in mesopores. The “hysteresis loop” common-
ly observed in gas adsorption isotherms is caused by formation of metastable films
during uptake that abruptly coalesce to the condensed phase triggered by nucle-
ation (Gregg and Sing, 1982). Hysteresis has also been observed in absorption of
gases (e.g., CO2 and small hydrocarbons) by glassy, but not rubber, polymers
(Kamiya et al., 1989, 1992). In this case the cause is believed to be slow volume-
structural relaxation; that is, the microvoid volume which increases on sorption
does not instantly relax to the original value on desorption. A mechanism involv-
ing metastable states in the context of sorption of dilute chemicals from soil solu-
tion, however, has not been articulated.
2. Insufficient time allowed for equilibrium: Nonattainment of equilibrium due
Figure 9 Hysteretic isotherms of phenanthrene in two soils. Experiments were done in decant-re-
seal batch cycles with replacement of most of the fluid after each cycle. (a) Lula sediment. Adsorption:
four cycles lasting 1–4 days each; desorption: 49 cycles lasting 1– 59 days each. (b) Norwood shale.
Adsorption: 28 days; desorption, 14 days. [(a) Reprinted with permission from Kan et al., 1998. Copy-
right 1998 American Chemical Society. (b) Reprinted with permission from Huang and Weber, 1997.
Copyright 1997 American Chemical Society.]
to rate-limited diffusion can lead to an underestimation of equilibrium sorbed con-

centration in the sorption direction and an overestimation in the desorption direc-
tion. A likely explanation for hysteresis in many cases, this is a vexing problem
experimentally because true equilibrium can require very long times and may be
concentration dependent.
3. Changes in the properties of the sorbent on sorption such that desorption
takes place from a different molecular environment than sorption. For SOM it may
be hypothesized that some sorbed molecules experience a conformational rear-
arangement of the local humic matrix, resulting in encagement or at least an en-
hancement of the activation barrier for subsequent escape. An analogy has been
made between sorbate-induced changes in the conformation of humic molecules

and substrate-induced changes in the conformation of enzymes (Pignatello and
Xing, 1996). Rearrangement has been observed in computational simulations of
pollutant molecules [e.g., atrazine (Schulten, 1995) and pentachlorophenol (Schul-
ten, 1996)] interacting with the hypothetical humic acid macromolecule shown in
Fig. 3a.
In order to explain isotherm hysteresis, Kan et al. (1998) proposed that total
sorption includes reversible and irreversible components. The term irreversible,
rather than implying permanent immobilization, is intended to mean that mole-
cules leave a site by a different microscopic pathway than that by which they en-
ter because of some kind of change of state taking place in the meantime (Adam-
son and Gast, 1997). Such behavior has been discussed in regard to adsorption of
surfactants and polymers on oxides (Adamson and Gast, 1997, pp. 404–405) but
without resolution of the cause. According to Kan et al. (1998), the “irreversible”
compartment has a fixed maximum capacity for sorbate and fills in one or more
steps in response to the solution-phase concentration. They proposed that the SOM
matrix rearranges to trap the sorbate. Huang and Weber (1997) suggested that, in
addition to nonattainment of equilibrium, hysteresis may be contributed by sor-
bate-induced expansion of condensed SOM to form pores that may “have no ex-
its” once configurational changes in humic molecules occur.
There are numerous examples of kinetic hysteresis, in which sorption appears
to be faster than desorption (Connaughton et al., 1993; Farrell and Reinhard,
1994b; Harmon and Roberts, 1994; Pignatello et al., 1993). Harmon and Roberts
(1994), for example, found the diffusion coefficient to be two to five times small-
Figure 10 Sorption and desorption rate curves for a hypothetical case. The cumulative mass
gained or lost (M) relative to the mass gained or lost after infinite time (M) is shown for different
Freundlich n values. The abscissa is the square root of dimensionless time. (Reprinted with permission
from Lin et al., 1994. Copyright 1994 American Chemical Society.)
er for desorption than sorption. “Thermodynamic” and “kinetic” hysteresis may

have the same underlying cause; in studies of TCE and benzene vapor uptake by
soil grains using an intragrain diffusion model, Lin et al. (1994) suggested that
much of the diffusion asymmetry can be explained simply by nonlinearity of the
isotherm. The results of the hypothetical case appear in Fig. 10. Note that the ef-
fect of nonlinearity is relatively minor unless the Freundlich exponent is less than
about 0.75. Also, Farrell and Reinhard (1994b) found that the “slow fraction” of
TCE remaining after N2 sparging was not well simulated by taking into account
only equilibrium nonlinearity. A common assumption in many studies is that the
rate parameter pertaining to sorption or desorption is single valued when in fact,
because of the heterogeneous nature of soils, it is more likely to take on a range of
values, depending on position along the uptake or release curve. Because most
studies to date have focused on the behavior of the bulk of the chemical (first 80%
sorbed or desorbed), much useful information has been missed.
IV. SORPTION KINETIC MODELS
A. MODELS BASED ON BOND ENERGETICS
The simple rate laws in Eqs. (4) and (5) seldom apply to real particles for two
reasons. First, diffusion (mass transfer) is intrinsic to sorption kinetics because
most sites are located in pores or within the SOM matrix and thus not directly
accessible by molecules in the bulk fluid phase. Second, sites vary energetically
because soils are heterogeneous. Nevertheless, kinetic models based on bond en-
ergetics, particularly those modified to account for soil heterogeneity, serve a pur-
pose because, unlike diffusion models, they do not require knowledge about par-
ticle geometry. Only the essential features are presented for the models that
follow: Readers are urged to consult the original papers for details about their ap-
plication.
The Langmuir kinetic model, reviewed by Adamson and Gast (1997), posits a
collection of sites of uniform energy. Combining Eqs. (4) and (5) (since sorption
and desorption events occur concurrently) and recognizing that the exponentials
are constant at constant temperature,
dSo
dt [ ] [ ]
= A a e ( − Ea / RT ) pSv − A d e ( − Ed / RT ) So
* *
= ka′ p( ST − So ) − kd′ So (8)
where total sorption ST Sv So, and ka are the adsorption and desorption rate
constants. Equation (5) may be put into a relevant soil–water frame of reference
by replacing p with aqueous concentration C [M L3]; replacing So with sorbed

concentration q [M M1]; and letting Q [M M1] be the capacity constant—the
sorbed concentration when all sites are occupied. This gives
dq ka
= C (Q − q ) − kd q (9)
dt
where ka is in units of [L3M1T1] and kd is in units of [T1]. The Langmuir
kinetic model is the basis for the familiar Langmuir isotherm since at equilibrium
(dq/dt 0) Eq. (9) reduces to
bQCe k
qe = , where b = a (10)
1 + bCe kd
Site nonuniformity has been dealt with customarily by including multiple sites.
The two-site model (Hu and Brusseau, 1998) envisions an instantly reversible site
S1, comprising a fraction f of the total sites, and a slower kinetic site S2:
Ke k2
C S1 S2 (11)
k−2
The equilibrium expressions are as follows.

S1e f KeCen, (12)
S2e (1 f )KeCen, (13)
where Ke is determined on the basis of total sorbed concentration, q. The overall
rate is the sum of the rates for each of the sites;
∂q ∂S1 ∂S
= + 2 (14)
∂t ∂t ∂t
Sorption at site 1 is governed by the equilibrium expression in Eq. (12), whereas
site 2 follows a first-order reversible rate law in which the forward rate is propor-
tional to S1 and the reverse rate is proportional to S2.
∂S2
= k2 S1 − k−2 S2 . (15)
∂t
After differentiation and substitution to eliminate S1, and realizing that k2 /k2
(1 f )/f, the overall rate law becomes
∂q ∂C
= n fK e C n −1 + k2 fK e C n − (1 − (16)
∂t ∂t
A Dutch group (Cornelissen et al., 1997a; 1998; 1999; Ten Hulscher et al., 1999)
used a multicompartmental model (either two or three compartments) for desorp-
tion of polychlorinated benzenes and PAHs from sediments in the presence of
Tenax TA polymeric beads as a third-phase adsorbent sink. The initial sorbed con-
centration q0 fiq0, where fi is the fraction in the ith compartment. Desorption
in each compartment was regarded to occur in a first-order manner:
dSi
= − ki Si , (17)
dt
Si Si,0eki t. (18)
Since the beads sorbed PAHs faster than does soil, the rate of resorption by the soil
was assumed negligible. Thus,
qt
= ft e − kt t + fs e − ksi t + fvsi e − kvsi t (19)
q0
where fr, fsl, and fvsl refer to rapid, slow, and very slow fractions of initial chemi-
cal demarcated, somewhat arbitrarily, on the basis of discontinuities in the de-
sorption curve.
One could add as many different types of sites as one wanted. The “multireac-
tion” model tested by Xue and Selim (1995) on alachlor sorption considers up to
four sorption domains: an equilibrium domain Se obeying a Freundlich isotherm
(n, Ke), a reversible kinetic domain S1, a “consecutive” irreversible kinetic domain
S2 accessible only from S1, and a “concurrent” irreversible domain Sirr accessible
only from solution.
B. DRIVING FORCE MODELS
Several models are based on the idea that the rate is related to the degree that
the system has reached equilibrium. The reversible model reviewed by Travis and
Etnier (1981) assumes the rate is proportional to the difference between the equi-
librium amount sorbed and the amount already sorbed:
dq
= R ( qe − q ) = R ( K e C n − q ) (20)
dt
where R is a rate parameter [T1]. Equation (20) is the reversible linear model
or reversible nonlinear model, depending on the value of n. The reversible linear
model is identical in form to the Langmuir model at very low concentration.
The film resistance model (Eq. 21) assumes mass transfer resistance of mole-
cules across a stagnant boundary layer (“film”) at the interface. The rate of sorp-
tion is controlled by the difference between bulk solution concentration C and the
concentration C* in equilibrium with the surface,
dq
= FR (C − C*) (21)
dt
where is the solids concentration [ML3] and is the volumetric water content
[L3L3].
One can see that if C* is related to q by the linear Freundlich isotherm then the
film resistance model is of the same form as the reversible linear model. On the
other hand, if C* is related to q by the Langmuir isotherm then it is of the same
form as the Langmuir model.
The second-order driving force model (Hendricks and Kuratti, 1982) regards
the rate to be proportional to solution concentration times the difference between
the equilibrium amount sorbed and the amount sorbed at time t:
dq
= C(qe − q ) (22)
dt
where is now in units of [L3M1T1]. In sorption of a dye to Dowex 50 ion-
exchange resin or to activated carbon, Hendricks and Kuratti (1982) found this
model to be superior to the reversible model, the Langmuir model, and two other
models conceptually similar to the second-order and reversible models.
The Fava–Eyring model (Fava and Eyring, 1956) is a nonlinear driving force
model in which the rate is related to (t), defined as the distance from equilibrium
divided by the initial distance from equilibrium; that is, (t) (q qe)/(q0 qe),
where q0 is the initial amount sorbed. The rate expression is given by Eq. (23),
where a and b are constants. The hyperbolic sine term accounts for diminishing
affinity for the sorbate with increasing loading:
d
= 2 sinh(b ) (23)
dt
C. DIFFUSION MODELS
1. General Considerations
Soil particles are typically porous and contain highly viscous sorptive phases
(i.e., SOM). Any mechanistic-based depiction of sorption would have to take dif-
fusion of one kind or another into account. The form that the diffusion model takes
is critically dependent on the geometry of the diffusing medium and the boundary
conditions. To understand diffusion in heterogeneous systems such as soils we
must consider studies of model sorbents. These studies are covered well by Kärg-
er and Ruthven (1992) for fixed-pore sorbents and Frisch and Stern (1983) and Vi-
eth (1991) for the organic solid state. The mathematics of diffusion is dealt with
by Crank (1975).
Transport diffusion is the nonequilibrium migration of molecules along the con-
centration gradient. Self-diffusion is scrambling of molecules due to their Brown-
ian motions under equilibrium (no gradient) conditions and may be approximated
by adding a tiny amount of radiolabeled tracer. Transport and self-diffusion are not
necessarily equal. The fundamental diffusion equations are known as Fick’s first
and second laws, which are given in Eqs. (24) and (25), respectively, for one-di-
mensional diffusion in the z direction and the general case:
∂c
J = − D(c) ; J = − D(c) grad c (24)
∂z
∂c ∂2c ∂c
= D(c) 2 ; = D(c) div(grad c) (25)
∂t ∂z ∂t
where J(J) is the flux (E L2 T1], c is the total local volumetric concentration in
the diffusing medium [M L3], and D(c) is the diffusion coefficient, or diffusivi-
ty [L2 T1]. Equation (26) gives Fick’s second law in radial coordinates for a v-
dimensional sample (v 1 for a slab, v 2 for a cylinder, and v 3 for a sphere),
∂c 1 ∂ ⎛ ( v −1) ∂c ⎞
= ( v −1) ⎜r D(c) ⎟ (26)
∂t r ∂r ⎝ ∂r ⎠
where r is the thickness (slab) or radius (cylinder, sphere). The equations apply as
long as the sample is isotropic and homogeneous and not appreciably changed by
the penetrant. The equations for the slab and cylinder assume no edge effects.
In Eqs. (24)–(26), D is expressed as a function of concentration, although in many
cases it may be constant at a given temperature; D is concentration dependent when
1. The diffusant alters the sorbent properties of the solid. For example, high
loadings of an organic diffusant may cause an organic solid to soften and swell
(Frisch and Stern, 1983; Lyon, 1995; Lyon and Rhodes, 1993; Vieth, 1991). This
increases diffusivity because it makes the macromolecular chains more flexible,
allowing the diffusant to pass more easily.
2. Sorption is nonlinear with concentration. Since the gradient in chemical po-
tential is related to the logarithm of concentration, the transport diffusivity D is re-
lated to the self-diffusivity D (also called “corrected diffusivity”) (Kärger and
Ruthven, 1992) by
d ln p
D =D (27)
d ln c
where p is the pressure of diffusant in the external fluid. It can be seen from
Eq. (27) that D D when c is linear in p. If the isotherm is Langmuir-type, then
D D (1 )1, where is the fractional coverage, qeQ1. Sorption is always
linear at infinite dilution; thus, D r D as c r 0.
Diffusion is an activated process analogous to any elementary chemical reac-
tion. In the Arrhenius formulation,
D DoeE*D /RT, (28)
* is the diffusion activation energy and D is the preexponential constant.
where ED o
Analytical solutions to the diffusion equation are available for a variety of sim-
ple situations that might be encountered, including
• sorption/desorption in an infinite bath of constant or variable concentration;
• sorption/desorption in a finite bath; and
• “evaporation” at the surface to an infinite bath.
The solutions are available for plane-sheet, cylinder, and sphere geometries in
Crank (1975) and for cube geometry in Kärger and Ruthven (1992). They are too
cumbersome to present here. Obviously, soil particles are not perfectly round;
however, since the diffusion curves for the cylinder and cube cases are quite sim-
ilar to that of the spherical case, except at long times, it is common practice to use
the simpler spherical expression, whereupon the radius corresponds to that of a
sphere having the same volume to external surface ratio.
Analytical solutions are useful; however, the researcher must be aware that they
come at the expense of many assumptions and simplifications. Some of the com-
mon ones include the following:
1. Diffusion obeys Fick’s laws.
2. The diffusion coefficient is concentration independent.
3. The sample consists of particles of uniform size.
4. The diffusant concentration is uniform throughout the particle at equilibrium.
5. Local equilibrium exists in a finite element of the particle.
Assumption (1) may be invalid if the properties of the sorbent change in re-
sponse to sorption, as could occur for SOM. Assumption (2) is not always valid
but may be regarded to be valid over a narrow concentration range. Assumptions
(3) and (4) are typically invalid for soils. Assumption (5) requires that, e.g., steric
hindrance at the entrance/exit to sites is not rate limiting.
2. Diffusion in Fixed-Pore Systems
Soil particles are aggregates of smaller grains that are cemented together with
organic or inorganic matter (Greenland and Hayes, 1981). Figure 11 shows a rea-
Figure 11 Soil particle aggregate with spherical-equivalent radius Ra made up of smaller grains
of minerals of radius (rg ) and organic matter (rom). Illustrated is a micropore structure within grains and
macropore/mesopore structure between grains.
sonable depiction of a soil aggregate having an equivalent spherical radius Ra. The
aggregate consists of SOM particles, SOM coatings, mineral particles, and min-
eral particles coated with finer grains. It has a macropores/mesopores between the
grains and a micropore network within individual grains.
a. Macropores and Mesopores

Diffusion is similar in macropores and mesopores in many respects and it is con-
venient to discuss the two sizes simultaneously. At relative humidities most rele-
vant to the environmental (i.e., ⬃50%) the following conditions hold: (i) at least
a monolayer of water is present on external mineral surfaces, (ii) micropores and
mesopores are filled with water, and (iii) macropores are empty or only partially
filled with water. As the humidity approaches 100%, the bulk Ke of a chemical is
close to that observed under saturated conditions (Chiou, 1989).
Diffusion of small molecules in macropores and mesopores is affected princi-
pally by tortuosity and sorption. For a particle of radius R, consider a volume ele-
ment of that particle in which cp [M L3] is the local pore fluid concentration and
sp [M L3] is the local sorbed concentration (i.e., c cp (1 )sp, where

is the particle porosity). The governing equation in radial coordinates is
∂cp ∂sp ⎡ ∂ 2 cp 2 ∂cp ⎤

+ (1 − ) = Dp ⎢ 2 + ⎥ (29)
∂t ∂t ⎢⎣ ∂R R ∂R ⎦⎥
and Dp is the local pore diffusivity which includes diffusion along the pore sur-
face and diffusion in the pore fluid. Equation (29) assumes Dp to be concentration
independent. By further assuming instantaneous local equilibrium, Eq. (29) be-
comes
∂cp ⎡ ∂ 2 cp 2 ∂cp ⎤
= Deff ⎢ 2 + ⎥ (30)
∂t ⎢⎣ ∂R R ∂R ⎦⎥
If the isotherm is linear —sp K cp, where K [dimensionless] is the local sorption
coefficient—the effective diffusivity is given by
Dp
Deff = (31)
+ (1 − )nK c
Pore diffusivity taking into account surface diffusion can be expressed as
⎛ 1 − ⎞
Dp = Dpf + ⎜ ⎟ K Dps (32)
⎝ ⎠
where Dpf is the pore fluid diffusivity and Dps is the pore surface diffusivity. Sur-
face diffusion may be important in pores with a large surface to volume ratio. Sur-
face diffusivities are generally determined by the difference between observed
pore diffusivity and the estimated value of pore fluid diffusivity. Surface diffusiv-
ity tends to increase with diffusant concentration and temperature. Surface diffu-
sivities have been determined for gases in activated carbons and mesoporous glass-
es (Kapoor et al., 1989; Kärger and Ruthven, 1992). The contribution of surface
diffusion in water-filled soil pores, however, is not well established.
Pore fluid diffusivity is reduced with respect to bulk fluid diffusivity by a tortu-
osity factor (1) that reflects deviation from straight-line paths and pore inter-
connectedness and by a steric parameter (1) that reflects steric hindrance by
the pore walls:
Dpf Db /. (33)
Theoretical models predict to be proportional to 1 (Currie, 1960; Kärger and
Ruthven, 1992; Wakao and Smith, 1962). Currie (1960) studied H2 diffusion in
beds of glass beads, sand, carborundum, several soils, and sodium chloride packed
at various bulk densities. The beds had porosities between 0.18 and 0.65 consist-
ing primarily of macropores and mesopores. Currie found the exponent of to
vary with the material and with . Columns of high porosity gave an exponent of
close to 1, but this exponent became less negative with decreasing porosity to
values as high as 0.38. For lower porosities and smaller pore sizes, tortuosity ap-
pears to be greater than predicted on the basis of 1. For example, for intra-
particle diffusion of chlorinated hydrocarbons in aquifer sediments, was 102 or
103 times greater than predicted (Ball and Roberts, 1991b; Grathwohl and Rein-
hard, 1993). This can be attributed in part to incomplete interconnectedness of
pores and in part to steric hindrance of diffusion.
Steric effects begin to show when the ratio () of the minimum critical molec-
ular diameter to the pore diameter reaches ⬃0.1 and become severe as ap-
proaches unity. Most molecules of interest are ⬃1.5 nm in their smallest di-
mension. Hence, steric effects are important in micropores and smaller mesopores.
Many theoretical models have been presented for steric hindrance that fit specific
data reasonably well (Kärger and Ruthven, 1992; Lee et al., 1991). Satterfield et
al. (1973) found 2.0 (0.09 0.5) for nonsorbing sorbates in micro-
porous (3.2 nm diameter) silica–alumina beads. For sorbing compounds, the ef-
fects of tortuosity and steric hindrance are difficult to separate. The diffusivities of
Cd2+ ion vs the larger SeOe2 ion in aluminas were consistent with a greater steric
effect in micropores than in mesopores (Papelis et al., 1995).
Combining Eqs. (31–33), we get
−1 Dpf + (1 − ) K Dps
Deff = (34)
+ (1 − ) K
To simplify modeling it has been customary to (i) neglect surface diffusion;

(ii) combine tortuosity and steric hindrance into a single “effectve tortuosity”
parameter (/ e ); and (iii) relate the local sorption coefficient to the bulk
equilibrium sorption cofficient to the bulk equilibrium sorption coefficient
(K Ke , where [M L3] is the particle density inclusive of intraparticle
porosity).
b. Micropores
In the narrow confines of a micropore the diffusant is always under the influ-
ence of the surface; therefore, it is meaningless to distinguish between fluid and
sorbed concentrations. The following is the governing equation in a spherical par-
ticle of radius r :
∂s ⎡ ∂2 s 2 ∂s ⎤
= D ⎢ 2 + ⎥, (35)
∂t ⎣ ∂r r ∂r ⎦
Figure 12 (a) Schematic of biporous diffusion model: microporous spherical grains inside a meso-
porous or macroporous spherical aggregate. (b) Coupled pore–intraorganic diffusion model: (i) completely
mixed aggregate of uniform mineral microparticles and uniform spherical SOM microparticles or (ii) a
porous mineral particle in which SOM microparticles are uniformly distributed throughout the internal sur-
face. (Reprinted from Yiacoumi and Tien, 1994, with permission from the American Geophysical Union.)
where s(r) is the local concentration and D is the effective micropore diffusivity.
Equation (35) is a restatement of Eq. (26) for a sphere and concentration-inde-
pendent D .
c. Dual-Resistance Models
One may consider the case in which diffusion takes place in a biporous particle
such as the one shown in Fig. 12a, which depicts a macroporous (or mesoporous)
aggregate (Ra) made up of individual microporous grains (r ). This model com-

bines Eqs. (29) and (35) and has the following boundary conditions:
s(r, 0) = cp ( R, 0) = 0 (36a)
∂s ∂cp
(0, t ) = (0, t ) = 0 (36b)
∂r ∂R
s(r , t ) = Kcp ( Ra , t ) (36c)
r
3
sp ( R, t ) = s ( R, t ) = 3 ∫ s ( r )r
2
dr (36d)
r 0
(36e)
where C(t) [ML3] is the external solution concentration.

An analytical solution of the biporous dual-resistance model is found in Ruck-
enstein et al. (1971) and Lee (1978) for linear sorption under two types of well-
mixed batch conditions: (i) a step change in surface concentration, in which the
external fluid concentration is kept constant, and (ii) variable surface concentra-
tion, in which diffusion occurs from a finite medium. The mathematics are simi-
lar for two other conceptual systems: particle-scale independent macropore and
micropore networks (not realistic for natural particles) and a particle with in-se-
ries macropore and micropore networks. Arocha et al. (1996) describe a numeric
solution to cases in which sorption in the macroporous and microporous systems
is nonlinear and obeys different Freundlich equations; they applied this nonlinear
biporous dual-resistance model to toluene vapor sorption to dry soil crumbs and
Na–montmorillonite. The diffusivity of toluene in the micropore zone was ap-
proximately 1012 cm2 s1, which is consistent with values reported for zeolites
(Kärger and Ruthven, 1992).
3. Diffusion in Organic Matter
Rubbery SOM, in which diffusant molecules are considered to be dissolved, is

analogous to a highly viscous fluid. Equation (35) applies if the particle is spher-
ical. However, as discussed previously, SOM may have both a dissolution domain
and a Langmuir domain in which adsorption-like interactions occur, with an over-
all isotherm given by the dual-mode model [Eq. (3); for polymers, it is common-
ly assumed that n 1]. In this case, diffusion is more complex.
The model of Vieth and Sladek (1965) assumes that diffusion occurs only in the
dissolution domain (DD) and that molecules at Langmuir sites are totally immo-
bile (DL 0). For diffusion through a planar sheet of polymer,
∂s ∂s
J = Deff = − DD D (37)
∂z ∂z
∂( sD + sL ) ∂2 s
= DD 2 (38)
dt ∂z
Analogous equations can be written for other geometries. If we assume local equi-
librium exists between sD and sL , then
(bsL0 / K D )sD
sL = (39)
1 + ( b / K D ) sD
Therefore, we can eliminate sL in Eq (38). After derivatizing we find that

−1
⎡ K′ ⎤
Deff = ⎢1 + 2⎥ (40)
⎣ (1 + sD ) ⎦
where K bsoL /KD, and b/KD. Note that Deff is a function of the local disso-
lution domain concentration and thus changes as sorption or desorption progresses.
The model of Paul and Koros (1976) considers dual but independent diffusivi-
ties. This means that molecular jumps are allowed within the dissolution domain
and between sites in the Langmuir domain, but cross-jumps are ignored. For dif-
fusion through a planar sheet of polymer,
∂s ∂s ∂s
J = Deff = − DD D − DL L (41)
∂z ∂z ∂z
where s is the total local volumetric concentration in the polymer. Assuming local
equilibrium and performing the mathematics in the analogous manner, it can be
shown that
⎡ K ′( DL / DD ) ⎤
⎢1 + (1 + s )2 ⎥
Deff = DD ⎢ D ⎥ (42)
⎢1+ K′ ⎥
⎢⎣ (1 + sD )2 ⎥⎦
The models of Barrer (1984) and Fredrickson and Helfand (1985) allow cross-
jumping between the Henry and Langmuir domains. The expression for effective
diffusivity then becomes
( DDL + DLL ) K ′ ⎡ sD K ′ sD K ′ 2 ⎤
DDD + + DDL ⎢1 − + 3 0⎥
(1 + sD ) 2
⎣ (1 + sD )sL0
(1 + sD ) sL ⎦
Deff = (43)
K′
1+
(1 + sD )2
where DDD and DLL are the intradomain diffusivities, and DDL is the cross-domain
diffusivity.
For SOM, it must be assumed that there are many different kinds of sites. This
makes the solution very complicated, indeed. Horas and Nieto (1994) derived an
expression for Deff for the general case in which the energy is distributed:
∞ ∞
∂
Deff = f ∫ ∫ l ij2
∂S
[ ]
Si vij ( Si0 − S j ) exp( − Eij / k T ) dEi dE j (44)
−∞ −∞
where f is the fraction of jumps in the direction of the macroscopic flux, Si is the
concentration of occupied sites of type i, Sio is the total number of sites of type i,
ᐍij is the average distance between sites i and j, vij is the vibrational frequency of
the molecule at site i, Eij is the potential energy barrier between sites i and j, T is
the temperature (oK), and k is the Boltzman constant (1.38044
1016 erg/deg).
The derivation of Eq. (44) requires many assumptions, some of which are de-
batable when applied to soils: (i) Only one molecule can occupy each site; (ii) the
values of ᐍij , f, and vij are constant for a given system; (iii) the approximate jump
frequency of molecules from site i to site j is a function of vij , the probability of
finding a vacancy at j, and Eij (this follows from statistical mechanics and transi-
tion-state theory); and (iv) Eij is assumed to be equal to the thermodynamic depth
of the potential energy well of site i (see Fig. 5), which is equivalent to saying that
the energy level of the transition state is the same throughout the sorbent. Equa-
tion (44) has been integrated for the case in which the energy distribution is Gauss-
ian. The solution is not presented here due to space limitations. Horas and Nieto
have shown that it simplifies to Eq. (42) or Eq. (43) when the corresponding as-
sumptions intrinsic to those models are employed.
4. Combined Pore Diffusion/Organic Matter Diffusion Model
Yiacoumi and coworkers (Yiacoumi and Rao, 1996; Yiacoumi and Tien, 1994,
1995) developed a model that includes both pore and intraorganic matter process-
es simultaneously. It is mathematically equivalent for either of the situations de-
picted in Fig. 12b in which the particle aggregate of radius Ra is (i) a completely
mixed aggregate of uniform mineral microparticles and uniform spherical SOM
microparticles (radius rom) or (ii) a porous mineral particle in which SOM mi-
croparticles are uniformly distributed throughout the internal surface. The model
takes into account pore diffusion (Dp) composed of pore fluid (Dpf) and pore sur-
face (Dps) terms; adsorption on the mineral surface, assuming linearity and local
equilibrium (sp Kmf cp); pore surface diffusion (Dps); sorption in SOM, assum-
ing SOM is a homogeneous Henry’s law partition medium (som Komcp); and ra-
dial diffusion in SOM (Dom). The model also assumes that mass transfer across
the bulk solution–macroparticle interface, as well as across the pore liquid–SOM
interface, is never rate limiting.
Figure 13 Sorption predicted by the combined pore–intraorganic matter diffusion model of Yia-
coumi and coworkers. (Top) Approach to equilibrium in completely mixed batch experiment. (Bottom)
Column breakthrough curve. (Bulk solution concentrations: cb, time ; cb0, time zero; ce, equilibrium;
cin, input eluent concentration. is dimensionless time.). The refers to the ratio of diffusion time
scales in pores vs OM (Eq. 43). Small reflects dominance of intraorganic matter diffusion resistance.
(Reprinted from Yiacoumi and Tien, 1994, with permission from the American Geophysical Union.)
The governing equation is
∂cp ∂N ⎛ ∂cp2 2 ∂cp ⎞

[ + (1 − )Kmf ] ∂t
+
∂t
[
= Dpf + (1 − ) K mf Dps ⎜ 2 +
⎝ ∂R
]
R ∂R ⎟⎠
(45)
Equation (45) is identical in form to Eq. (29) except that it includes a term for up-
take of solute in pore liquid by SOM:
∂N ∂c 3f ⎛ ∂c ⎞
= fom om = om Dom ⎜ om ⎟ (46)
∂t ∂t rom ⎝ ∂r ⎠ r = r
om
where fom is the fraction of organic matter and c̄om is the average sorbed concen-
tration in SOM. The solution to Eqs. (45) and (46) for typical initial and boundary
conditions was found by applying the method of Laplace transforms. The Laplace
transform and the numerical techniques for inverting it are described in the origi-
nal papers by Yiacoumi and coworkers cited previously. The solution was then
used to provide aqueous solute concentration as a function of time in batch ex-
periments and aqueous solute concentration as a function of both time and posi-
tion in columns.
On the basis of calculations, Yiacoumi and Tien (1994) examined the influence
of pore diffusion resistance vs intraorganic matter diffusion resistance. The ratio
of diffusion timescales in pores vs OM is
Dom / rom 2
= . (47)
( Dp / )/ Ra 2
In batch systems (see Fig. 13, top), they found that , the dimensionless time for
equilibrium, increased as intraorganic matter resistance increased in importance
(i.e., as decreased). In column experiments (Fig. 13, bottom), the breakthrough
curves became more spread out (less like piston flow) as the intraorganic matter
resistance increased in importance. The mentioned effects were more pronounced
at low fomKom (i.e., at low SOM concentrations or for compounds with low affin-
ity for SOM). They also applied their model to data existing in the literature on
short-term sorption desorption studies. When the model was applied to the data of
Nkedi-Kizza et al. (1989) on transport of atrazine and diuron in columns of Eustis
soil, Yiacoumi and Tien concluded that diffusion resistance in SOM rather than the
pore system was rate controlling. When applied to batch desorption of tetra-
chlorobenzene from Charles River and North River sediments by Wu and
Gschwend (1986), the results were inconclusive in this respect.
5. Mixtures of Particle Sizes
Particle size distributions in soil span many orders of magnitude. For a distri-
bution of sphere sizes whose probability density function is (a), where a is Ra or
r , the average concentration is obtained by mass balance as
∞
⎡ Mt ⎤
∫ ( a ) ⎢ M
3
⎥ ( a) a da
⎛ Mt ⎞ ⎛ q − q0 ⎞ ⎣ ∞⎦
⎜ ⎟ =⎜ ⎟ = 0
∞
(48)
⎝ M∞ ⎠ mix ⎝ q0 − q∞ ⎠ mix
∫ (a)a
3
da
0
where the quantity [Mt /M](a) is the functional expression for the fractional mass
of chemical taken up or released for each discrete particle size a. For narrow par-
ticle size fractions, Eq. (48) may be approximated as
⎛ Mt ⎞ ∞ ⎡M ⎤
⎜ ⎟
⎝ M∞ ⎠ mix
=1− ∑ Xi ∑ ⎢ M t ⎥ ( a ) (49)
i n =1 ⎣ ∞ ⎦
where Xi is the mass fraction of particles having geometric average radius a R̄ai
or r̄ i. If experiments are done on pooled size fractions, the Sauter mean radius is
recommended (Ball and Roberts, 1991b; Pedit and Miller, 1995):
−1
⎡ n ⎤⎡ n F ⎤
as = ⎢∑ Fi ⎥ ⎢∑ i ⎥ (50)
⎣i =1 ⎦ ⎣i =1 ai ⎦
where Fi and ai are the mass fraction and geometric mean radius of each of the n
subfractions pooled.
Particle size can have an enormous effect on the diffusion curves (Fig. 14). The
approach to equilibrium in a finite bath at fixed D is shown for uniform particles
whose radii span a range of 100 arbitrary units. The small, medium, and large par-
ticles reach 95% equilibrium at approximately 0.14, 14, and 1400 units of time,
Figure 14 Effect of spherical particle radius on uptake or release of a solute in a mixed finite bath.
Mt /Minf is mass of solute taken up or released divided by mass taken up or released at equilibrium. Ra-
dius (left to right: 1, 10, and 100) and time are in arbitrary units. Equations in Chapter 6 of Crank (1975)
were used. Final mass taken up was 50% of total present.
respectively. Multiple-particle class models better represented long-term sorption

rates than did single-particle class models (Pedit and Miller, 1995). Some re-
searchers have used a two-compartment model, with one instantaneous and the
other diffusion limited (Ball and Roberts, 1991b; Lorden et al., 1998; Pignatello
et al., 1993) (see Section V,C,2,c); in this case, the instantaneous compartment may
partially account for the smaller particles if the nominal particle radius is indeed
the appropriate length scale for diffusion.
D. STOCHASTIC MODELS
Researchers are well aware of the high degree of spatial variability of soils and
the difficulty of applying models designed for homogeneous systems. Several in-
vestigators have therefore attempted to model sorption kinetics stochastically. In
such models it is assumed that sorption rate can be described with an array of first-
order rate constants k that are continuously distributed according to a probability
density function (PDF).
A useful PDF is the gamma function (Chen and Wagenet, 1995; Connaughton
et al., 1993; Gustafson and Holden, 1990; Pedit and Miller, 1994, 1995), which
expresses the frequency distribution of k as
∞
k −1e −k −1 − x
f (k ) =
()
; () = ∫x e dx (51)
0
where and are the shape and scale parameters, respectively, of the PDF curve,
and () is a mathematical statistical function. The mean of k is 1 and the vari-
ance of k is ¹⁄₂ 1. The shape of the PDF ranges from Gaussian at ⬃10 to
positively skewed (i.e., high frequency of small k values) at ⬃1. The fraction
of initial mass remaining after time t in any compartment having rate constant k is
simply ekt. Thus, the mass remaining after t for all compartments is
∞
Mt ⎛ ⎞
= ∫ f ( k )e − kt dt = ⎜ ⎟ (52)
Mi 0
⎝ + t⎠
The values of and are obtained by regression. The time t1y necessary for
the concentration to reach the fraction y of the initial concentration is t1y
[y1/ 1].
The gamma model has been applied to sorption/desorption of naphthalene in a
freshly equilibrated soil and a field soil long contaminated with coal tar (Con-
naughton et al., 1993), spiked lindane in a soil (Pedit and Miller, 1994), diuron in
an aquifer sand (Pedit and Miller, 1995), and TCE during transport of its vapors in
an aquifer soil column (Lorden et al., 1998). Interestingly, Gustafson and Holden
Figure 15 TCE vapor elution curves from an aquifer sand at 90% relative humidity and under
“low concentration, low flow” (LC–LF) conditions (Lorden et al., 1998) showing the superiority of the
gamma stochastic model (GS) over the two-site first-order (TSFO) model and the two-site spherical
diffusion (TSSD) model. The latter two assumed one of the sites to be instantaneous with a sorption
distribution coefficient KD determined experimentally; when KD was included as a third fitting para-
meter, both TSFO and TSSD fit better but still not as good as the GS model. (Reprinted with permis-
sion from Lorden et al., 1998. Copyright 1998 American Chemical Society.)
(1990) successfully applied the gamma PDF to field and laboratory “dissipation”
data of 45 different pesticide–soil systems, where dissipation means all fate
processes including sorption.
Pedit and Miller (1994, 1995) investigated a log-normal PDF, which is given by
1 ⎡ 1 ⎛ ln k − ⎞ 2 ⎤
f (k ) = exp ⎢− ⎝ ⎥ (53)
(2 )1 / 2 k ⎣ 2 ⎠ ⎦
where the arithmetic mean of k is k̄ exp( 2 /2) and the variance of k is V

k̄ 2exp(2 1). The log-normal PDF simulated lindane uptake by soil more accu-
rately than did the gamma PDF.
Not surprisingly, stochastic models often outperform other models, given the
same number of adjustable parameters. They outperform the single-site first-order
model (Pedit and Miller, 1994), the two-site first-order model (Connaughton et al.,
1993; Lorden et al., 1998), and the two-site diffusion model (one instantaneous
and one diffusion controlled) (Lorden et al., 1998; Pedit and Miller, 1994). Figure
15 shows the superiority of the gamma model in describing the tailing of TCE elu-
tion curves observed by Lorden et al. (1998).
The value of a stochastic approach depends on if and how well it can predict be-
havior under different conditions or different soil–chemical systems than the one
in which the model parameters were obtained. Lorden et al. (1998) demonstrated
predictive capability of the gamma function for elution of TCE vapors at different
flow rates and TCE pressures. The ability of stochastic models to translate to dif-
ferent soil–chemical systems has not been established, however. Also, these mod-
els do not offer much mechanistic insight.
V. EXPERIMENTAL METHODS
In a review such as this, it is worth discussing the methodology used in sorp-

tion/desorption rate studies. The sorption time frame may be transient or steady
state. In the transient time frame one observes a changing concentration gradient,
whereas in the steady-state time frame one establishes a stable concentration gra-
dient and measures the mass of chemical entering or leaving the system. The tran-
sient time frame is the one most often used for soils.
Popular techniques for measuring transient processes include batch, column,
zero-length column (ZLC), and stirred-flow cell (SFC). Batch techniques measure
phenomena on the intraparticle scale since the external surfaces are exposed to a
mixed fluid of uniform concentration. Column techniques measure phenomena on
intra- and interparticle scales concurrently. The column’s effluent concentration
profile depends on both dispersion (longitudinal mixing) and sorption/desorption
processes. Column techniques avoid abrasion of particles and can be more realis-
tic. ZLC and SFC are hybrid techniques that attempt to combine the advantages of
each.
A. BATCH TECHNIQUES
In solution–solid systems, agitation (stirring or shaking) is usually required to

achieve uniform conditions, whereas in vapor–solid systems bulk gas diffusion is
rapid enough to maintain uniformity at all but the fastest uptake rates, provided the
sorbent bed is sufficiently thin. Sorption may be carried out under infinite or finite
source conditions, which require different approaches. Under infinite source con-
ditions, the fluid-phase concentration is not reduced by sorption and uptake has to
be measured by analysis of the solid. This approach is commonly used in vapor–
solid studies in which uptake is followed gravimetrically. When the source is fi-

nite, the bulk fluid-phase concentration changes, providing a convenient way to
follow uptake, assuming no chemical transformations occur.
1. Measurement Techniques
Batch solution–solid experiments are commonly performed by mixing the com-

ponents in sealed flasks and, at predetermined times, measuring solution concen-
tration after first separating the phases. In estimating Ke, the optimal degree of
sorption is about 55% (McDonald and Evangelou, 1997); the same degree of final
sorption would apply to an uptake kinetic experiment if the objective is to quanti-
fy changes in Ke with time. Although this methodology appears simple to manage,
the researcher must be aware of the following potential complications:
1. “Bottle losses”: Adsorption on glass walls can be important if it is more
than a few percent of sorption to the soil. It is seldom an issue except for highly
hydrophobic compounds and may be corrected by running controls. Diffusion
through polymer materials is greater for small molecules and occurs regardless of
whether the chemical comes into contact with the material via the solution or va-
por phase (the chemical potentials in the two phases are equal). Little or no loss
through Teflon cap liners was reported for lindane during 84 days (Miller and Ped-
it, 1992) or for phenanthrene over several weeks (White et al., 1999). Polymer ves-
sels and soft polymer stoppers or cap liners (e.g., “butyl” rubber or “phenolic”)
should be avoided. Polymer materials can be avoided altogether by using flame-
sealed glass ampules (Ball and Roberts, 1991a) or screw-cap vials with foil
stretched tightly over the lip of the vial before the cap with its liner is screwed on
(Huang et al., 1997; Xia and Ball, 1999). Piercing the septum liner with a needle
to remove sample will expose hydrophobic septum material (usually silicone rub-
ber) to the flask contents, resulting in noticeable extraction even within hours; thus,
most researchers use the “bottle point” method in which replicates are sacrificed
at time points.
2. Phase distribution artifacts, such as headspace partitioning of volatile com-
pounds and sorption to colloidal particles not removed by centrifugation or fil-
tration: The former can be assessed by considering Henry’s law. The latter is
potentially significant for highly hydrophobic compounds (log Kow ⬃4) or com-
pounds that otherwise interact strongly with colloids (Schwarzenbach et al., 1993).
Assay of the liquid phase may not distinguish between colloid-bound and truly dis-
solved fractions. Moreover, one cannot assume the concentration of nonsettling
colloids will remain constant throughout an experiment. The problem of sorption
to colloids can be avoided by retaining the soil in dialysis tubing (Allen-King et
al., 1995), provided a correction can be made for sorption to the tubing.
3. Artifacts caused by agitation: Shear forces generated by vigorous agitation
may change the particle size distribution and expose additional surface area, thus
affecting rates. Stir bars grind particles (Wu and Gschwend, 1986). The minimum
degree of agitation necessary to achieve rapid mixing on the timescale of sorption
is recommended; slow end-over-end tumbling seems to be the best technique.
4. Method of analyte introduction: Strongly sorbing compounds tend to sorb to
the first surfaces encountered and may not readily redistribute to other particles;
therefore, attention must be paid to the way in which initial contact between par-
ticle and chemical is achieved. Homogeneous particle coverage requires fast dis-
persion of particles into a large volume of a solution. When spiking bulk soils in
preparation for desorption studies the method of Karickhoff et al. (1979) is rec-
ommended: First, the compound is deposited on the walls of the flask by evapo-
rating a solution in a volatile carrier (e.g., CH2Cl2) and then the soil suspension is
added, followed by agitating for 24–48 h.
5. Biodegradation: Sterilization may be accomplished by heat, irradiation, or
chemical treatment. Wolf et al. (1989) compared sterilization techniques based on
plate counts of bacteria and fungi. The following were completely effective on
three soils: double or triple autoclaving, propylene oxide (48-h exposure), 60Co -
irradiation (0.05 MGy), and HgCl2 (500 g/g). The remaining treatments were
less effective and followed the order: once-autoclaving oven-drying NaN3 ⬃
CHCl3 ⬃ microwave treatment control ⬃ antibiotics. Autoclaving is generally
avoided because of concerns it can affect soil chemistry, particularly SOM struc-
ture. Propylene oxide may react chemically with SOM, potentially affecting sorp-
tion of the compound of interest. [The same holds for formaldehyde, which was
not tested by Wolf et al. (1989) but is widely known to be effective.] 60Co -irra-
diation equipment is not widely available. Mercuric chloride had the least effect
of all treatments on soil CEC, pH, and extractable metal ions. Xing et al. (1996)
showed that HgCl2 had no effect on the 48-h Freundlich sorption parameters of
metolachlor. Mercury presents a disposal problem, however. Sodium azide (NaN3)
is a popular bacteriostat that binds to cytochromes, inhibiting terminal electron
transport. However, its efficacy cannot be taken for granted since it may not in-
hibit enzymatic transformations that require little energy, and because bioactivity
may be revived if dilution of the NaN3 occurs in subsequent steps [e.g., as it did
in the protocol of Wolf et al. (1989)].
To separate the solid and liquid phases, centrifugation is not suitable for sam-
pling frequencies more often than about one per 10 minutes. Filtration through a
microporous filter (e.g., 0.2 m) can increase sampling frequency to perhaps one
per minute. Filtration has been employed in experiments at temperatures other than
ambient in order to minimize temperature effects during sample handling (Xing
and Pignatello, 1997). Wu and Gschwend (1986) used an air-stripping apparatus
which constantly recirculated air through the soil suspension and then through a
detector. This technique, which is suitable for sorption occurring over a few sec-
onds to 48 h, avoids the colloid problem because activities are measured in the gas
phase and related to solution-phase concentration via Henry’s law. Similar appa-
ratuses are described by Brusseau et al. (1990) and Benzing et al. (1996). For semi-
volatile compounds Karickhoff and Morris (1985) designed a gas purge cell in
which air was passed through glass frits in contact with the suspension and the va-
pors were collected in external traps containing Tenax, a polymeric hydrophobic
resin. Harmon and Roberts (1994) describe a purge apparatus that fits around a
glass ampule after the flame-sealed top is broken off. The apparatus of Benzing et
al. (1996) can operate in both recirculate and purge modes.
A popular way to monitor desorption is by use of the successive-dilution tech-
nique, in which a known volume of the supernatant is replaced with “clean” wa-
ter. The mathematical model that one uses needs to take into account the stepped
boundary conditions that this technique entails.
Pignatello and coworkers (Pignatello, 1990a,b; White et al., 1999) and Cornel-
lison et al. (1997a,b, 1998a) used an in situ trap of Tenax polymeric adsorbent
beads. Tenax has a high affinity for nonpolar compounds, even C1 and C2 halo-
genated hydrocarbons (Pignatello, 1990b). The beads are readily separated from
the soil after centrifugation because they float. Soil particles have little tendency
to adhere to the beads and are easily rinsed off. Carroll et al. (1994) used XAD-4
resin as an in situ trap; however, K2CO3 had to be added prior to sampling to in-
crease the solution density in order to make the beads float.
The purge gas or in situ polymer trap rapidly lowers the solution concentration
to near zero. Hence, diffusion kinetics can be modeled using equations that de-
scribe “diffusion with surface evaporation” (Crank, 1975). These equations take
forms that depend on the extent of diffusion and the sorbent shape. If the particle
is a sphere of radius Ra and initially at uniform concentration, the surface bound-
ary condition uses the linear driving force assumption
∂c
−D = if (cs − co ) (54)
∂R
where cs is the concentration just within the sphere, co is the concentration required
to maintain equilibrium with the solution, and if is the interfacial mass transfer
coefficient [L T1]. Provided that enough time has passed so that the rate of purg-
ing from the solution is equal to the rate of desorption, the governing equation for
the fractional amount of substance leaving the particle at any time is
Mt ∞ (
6 L2 exp −2n Dt / a 2 ),
=1− ∑
M∞ n =1 n
2
{
2n }
+ L( L − 1)
(55a)
where L a/D and the n values are the roots of
n cot n + L − 1 = 0 . (55b)
In their studies of tetrachloroethane (PCE) desorption from aquifer sediments,

Harmon and Roberts (1994) purged the suspension only intermittantly during de-
sorption; therefore, they used a model combining boundary conditions for diffu-
sion with surface evaporation and diffusion to a solution of finite volume.
O’Dell et al. (1992) describe a static batch reactor in which nearly saturated
soil containing a herbicide (imazethapyr) was placed in syringe barrels. At pre-
determined times the soil solution was displaced with saturated CaSO4 when the
syringe was placed under vacuum. By including a tracer in the CaSO4, they ver-
ified that the recovered soil solution was uncontaminated with the displacing so-
lution.
Direct analysis of the solid is sometimes carried out. Sorbed concentration is
conveniently measured by solvent extraction after correcting for solute in the wa-
ter removed along with the solid. However, resistant sorbed fractions may not be
completely recovered unless extraction is fairly vigorous. Sawhney et al. (1988),
Pignatello (1990a), and Huang and Pignatello (1990) found that complete solvent
extraction of even weakly sorbing compounds required several hours at 75C. Sim-
ilarly, Ball et al. (1997) found that a 16-h methanol extraction at 70C was required
to remove volatile organic contaminants in field aquifer sediments. A water-
miscible solvent (e.g., acetone, methanol, or acetonitrile) will more effectively
penetrate SOM and intraparticle pores than an immiscible solvent. Harmon and
Roberts (1994) found that extraction of [14C]PCE from aquifer sediments by non-
miscible scintillation cocktail required tens of days.
Tognotti et al. (1991) describe a single-particle technique in which the particle
is suspended in a vapor-laden gas stream within an electrodynamic thermo-
gravimetric analyzer (EDTGA). Uptake or release of toluene or carbon tetrachlo-
ride was measured by the voltage required to keep the particle stationary in the
EDTGA as a function of time. To determine the effect of temperature, a CO2 laser
was used to heat the particle.
B. COLUMN TECHNIQUES
Glass columns with perforated Teflon end plates and Teflon adjustable plunger
are commercially available. Stainless-steel preparatory-scale liquid chromatogra-
phy columns have also been widely used. Column packing techniques apparently
have not undergone systematic investigation. Typically, columns are packed by
layering dry material a little at a time in an attempt to create a uniform bed. The
investigator should be aware, however, that dry soil has a tendency to segregate
by particle size, making it difficult to scoop up a representative sample from the
container from which the soil is taken. Moistening the soil (e.g., to 5 or 10% wa-
ter by weight), if appropriate for the experiment, helps alleviate this problem. Pig-
natello et al. (1993) filled the column with air-dry soil, allowing about one-fourth
of the volume to be free of soil; the soil was then homogenized by manually rock-
ing and rotating the column axially for several minutes. After packing, the column
is then slowly saturated from the bottom with eluent. Most investigators use dilute
(e.g., 0.01 M) CaCl2 as the eluent to help prevent dispersion of clays and mobi-
lization of colloids. The column is often “conditioned” by passing through many
pore volumes of eluent prior to injecting the solute solution.
Eluent is pumped through the column with a constant-volume displacement
pump such as the kind used in liquid chromatographic systems. A three-way valve
is installed before the column to allow switching between reservoirs containing
eluent with or without solute. The effluent may be collected with a fraction col-
lector for later analysis or passed through a UV-visible, scintillation counter, flu-
orescence, or other type of flowthrough detector. For vapor-elution studies, Lor-
den et al. (1998) describe an apparatus that diverts flow to a gas chromatograph
through a programmable automatic switching valve.
Injection of the solute may occur according to a step function or a pulse func-
tion. For a step function, the concentration of solute is increased abruptly from an
initial value (usually zero) to a second value, and the solute “breakthrough curve”
at the terminus is observed until the concentration reaches the stepped value. Des-
orption may be observed at some later time by switching the inflow to clean elu-
ent. For pulse injection, the solute is introduced over a finite period and the efflu-
ent concentration is monitored at the terminus.
Many reviews are available that discuss contaminant transport in soil columns
(Brusseau and Rao, 1989; Parker and van Genuchten, 1984; van Genuchten and
Wagenet, 1989). Solute transport when no decomposition occurs is described by
the advection–dispersion (A–D) equation with sorption:
∂C ∂q ∂2C ∂C
+ = Dh 2 − v (56)
∂t ∂t ∂x ∂x
where is bulk porosity [L3 L3], is bulk solids density [M3 L3], Dh is the hy-
drodynamic dispersion coefficient [L2 T1], v is the linear average flow velocity
[L T1], and x is longitudinal distance [T1]. The term of focus, !q/!t, represents
the rate of change in average sorbed concentration as seen by the flow regime.
In addition to the complicating effects of dispersion, another limitation of col-
umn studies is that the degree of sorption that occurs depends on the residency time
of the solute in the column and hence the flow rate. That is, slow sorption states
may not be fully accessed by solute molecules in a fast-moving fluid. For exam-
ple, Brusseau (1992), using the two-site model, showed more than an order of mag-
nitude decrease in the sorption rate parameter for several compounds with a de-
crease in flow rate from ⬃45–90 to ⬃5 cm/h (13 to 0.7 pore volumes per hour).
Likewise, Lorden et al. (1998) showed a decrease in the characteristic time for
sorption of TCE vapors with increasing flow rate. It is difficult to achieve flow rates
of less than one pore volume per day; however, batch studies show that a sub-
stantial fraction of sorption may require weeks or months to complete. Obvious-
ly, in the above-cited studies the equilibrium and kinetic fractions of the two-site
formalization are accommodating to a shift in timescale of elution.
One must also be aware that significant amounts of injected chemical may reach
slow sites but not desorb on the timescale of the experiment. For all practical
purposes, such amounts are “irreversibly” sorbed and thus missed by the model.
For example, Spurlock et al. (1995) carried out delayed elution experiments with
monuron and fenuron, relatively polar compounds with high water solubilities, in
a soil (3.4% OC). They found that the elution curves were affected little by the
length of the prior contact period (8, 80, or 240 days). However, appreciable
amounts (10–30%) remained sorbed after the effluent concentration declined to
C/Co 0.01. This residual sorbed herbicide increased with contact time, became
increasingly greater than predicted on the basis of the equilibrium law, and was
uniform along the length of the column.
2. Transport Models with Sorption Kinetic Term
a. Two-Region Model
In this model (van Genuchten and Wierenga, 1976, 1977; van Genuchten et al.,
1977), the pore space is divided into mobile water (m) and immobile water (im)
regions, with advection and dispersion occurring only in the mobile region. The
A–D equation is thus
∂Cm ∂S ∂Cim ∂S ∂ 2 Cm ∂Cm

m + f m + im + (1 − f ) im = m Dh 2 − m v (57)
∂t ∂t ∂t ∂t ∂x ∂x
where f is the fraction of sorption occurring in the mobile region. Suppose sorp-
tion is locally instantaneous and obeys the Freundlich equation Sm f KeCm n and
Sim (1 f )KeCim. Upon differentiating Sm and Sim with respect to t and substi-
n
tuting into Eq. (57), one obtains
(m + fnKcCim n −1 ) ∂C∂tm + (im + (1 − f )nKc Cim n −1 ) ∂C∂tim =
(58)
The relationship between Cm and Cim is established by assuming that the dri-
ving force for mass transfer between the regions is simply proportional to the dif-
ference in their respective liquid-phase concentrations; thus,
(im + (1 − f )Kc nCim n −1 ) ∂C∂tim = (Cm − Cim ) (59)
where is a mass transfer coefficient. An analytical solution to Eqs. (58) and (59)
for the linear sorption case is available (van Genuchten and Wierenga, 1976). For
the nonlinear case the solution must be obtained numerically. If Ke is known, the
fitting parameters are , f, and im. A reasonable assumption is that f and im are
equal (Nkedi-Kizza et al., 1984) or im may be obtained by fitting the model to the
elution curve of a nonsorbing tracer (Spurlock et al., 1995).
b. First-Order Models
A widely used model is the two-site (equilibrium and kinetic) one, described in
Section IV,A (Eqs. 11–15) (Cameron and Klute, 1977; Nkedi-Kizza et al., 1984;
van Genuchten and Wagenet, 1989). The appropriate expression for !q/!t is Eq.
(16). The A–D equation becomes
⎞ ∂C ∂2C
⎛
⎝
1 + f nK e C n −1
⎠ ∂t
( )
+ f K e C n − (1 − f )q = Dh 2 −
∂x
(60)
Ordinarily, Ke is determined in separate batch experiments or calculated by

moments analysis (Valocchi, 1985), and Dh is determined by eluting a nonsorbing
tracer, such as 3H2O. Thus, Eq. (60) has two fitting parameters, k2 and f. An an-
alytical solution is available for the linear case only (van Genuchten and Wagenet,
1989).
c. Diffusion Models
Diffusion concepts have been used directly or in conjunction with other mod-
els. Pignatello et al. (1993) and Lorden et al. (1998) used a two-compartment mod-
el having a rapid equilibrium compartment (S1, f ) and a second slow compartment
(SD, 1 f ) governed by radial diffusion so that q S1 SD:
Ke D
C S1 ⇔ SD (61)
The model regards the diffusion domain as being a collection of idealized equal
spheres (radius, a) in which diffusion obeys a Fickian diffusion law Eq. (62a)
(analogous to Eq. 26) with boundary conditions as given in Eq. (62b):
∂s D ∂ ⎛ 2 ∂s ⎞
= 2 ⎜r ⎟ (62a)
∂t r ∂r ⎝ ∂r ⎠
∂s(r )
=0 and s(r = a, t ) = (1 − f ) K e C (62b)
∂r r = 0
The volume-averaged sorbed concentration in the diffusion domain is given by

a
3
a 3 ∫0
SD = s(r, t )r 2 dr (63)
where is the particle density (g/cm3). Curve fitting requires numerical tech-
niques.
A two-region diffusion concept was proposed by Nkedi-Kizza et al. (1982) for
inorganic solutes and later used by Young and Ball (1994) to simulate break-
through of PCE in aquifer sediments. In this conceptualization, the “immobile
zone” is the intraparticle porosity () and its associated sorption capacity, and the
“mobile zone” is the interparticle porosity and its associated sorption capacity. The
same two-region A–D equation (Eq. 57) applies, except that im is replaced by ,
and Cim is replaced by the volume-averaged concentration inside the particles C̄im
(spherical, Ra):
3 Ra
Cim =
Ra3 ∫0 Cim ( R, t ) R 2 dR (64)
The term !C̄im(R,t)/!t is expressed in terms of intraparticle radial diffusion laws

(Eq. 30; cp C̄im) and the series of equations solved numerically. This model as-
sumes that particles are uniform in size, porosity, and chemistry; that Ke is the same
in the mobile and immobile regions; and that Cm cim at the particle/mobile–wa-
ter interface. The was determined (Ball et al., 1990) by mercury porisimetry
which measures intraparticle porosity down to pore widths on the order of ⬃10
nm (mesopore to macropore range).
Piatt and Brusseau (1998) analyzed elution curves using the two-site model and
then related the S2 site first-order desorption rate constant k2 (see Section IV,A,
Eq. 11) to an organic matter diffusivity Dom, a particle shape factor ", and a dif-
fusion length scale, l:
k2 "Dom /l 2 (1 f ). (65)
A solution for solute transport through a bed of uniform porous pellets for a step
function change in solute concentration in the influent is provided by Yiacoumi
and Rao (1996) and Yiacoumi and Tien (1994). Diffusion occurs within the pellet
pore system and SOM microparticles contained therein, as described in Section
IV,C,5. The solute uptake per pellet is given by
t
⎛ ∂cp ⎞
M = ∫ 4 Ra2 Dp ⎜ ⎟ dt (66)
0 ⎝ ∂R ⎠ R
a
where the (!cp /!R)Ra term is from Eq. (45). Upon considering solute uptake on a
per-gram sorbent basis, the rate of sorption is given by
∂ ⎧⎪ 3(1 − ) Dp ⎛ ∂cp ⎞ ⎫⎪
t
∂q ∂ ⎡ 1− ⎤
Ra ∫o ⎝ ∂R ⎠ R ⎪
= M = ⎨ ⎜ ⎟ dt ⎬ (67)
∂t ∂t ⎢ 4 R3 ⎥ ∂t ⎪
⎢ a ⎥ ⎩ a ⎭
⎣ 3 ⎦
where is the particle density (g/cm3). Equation (67) is then substituted into the
A–D equation (Eq. 56). The combined equations were solved by Yiacoumi et al.
by the method of Laplace transforms. This gives a complex expression (not repro-
duced here) relating the observed bulk solute concentration in the Laplace domain
C̃ as a function of the following: initial bulk solute concentration Co (if other than
zero), the input concentration Cin, diffusivities (Dpf, Dps, and Dom), partition
coefficients (Kmf and Kom ), pellet and SOM microparticle radii (Ra and rom), bulk
moduli (, , and ), flow velocity (v), dispersion coefficient (Dh), and distance
along the column. Inversion of the Laplace transform was performed by a Fouri-
er series approximation algorithm (Yiacoumi and Tien, 1992).
C. STIRRED-FLOW CELL TECHNIQUE
The SFC is known in engineering as a “continuous-flow stirred tank reactor.”

Eluent is passed through a chamber of volume VR[L3] containing a well-mixed
suspension of soil and out to a fraction collector, adsorbent cartridge, or flow-
through detector (Deitsch and Smith, 1995; Eick et al., 1990; Miller et al., 1989;
Seyfried et al., 1989; Sparks, 1989; Zhang and Sparks, 1993). Frits are positioned
at either end of the cell to hold in particles. As a hybrid of batch and column tech-
niques, the SFC eliminates hydrodynamic dispersion but has the sampling advan-
tages of a column. The method is capable of measuring sorption or desorption with
“half-lives” as short as ⬃1 min or desorption with characteristic times of a few
tens of hours.
Uptake rates are measured by pumping influent of solute concentration Cin [M
L3] at flow rate u [L3 T1] through the suspension and measuring the effluent
concentration Ceff [M L3]. Release rates are measured likewise by pumping clean
liquid through the suspension of known initial sorbed and aqueous concentrations,
q0 [M M1] and C0 [M L3]. Provided adequate mixing occurs in the cell, the
mass balance equation is
∂q ∂C u
+ = (Cim − Ceff ) (68)
∂t ∂t VR
where q and C represent the instantaneous sorbed and aqueous concentrations in the
cell, is the bulk solids density [M L3], and is the volumetric liquid content [L3
L3]. Under well-mixed conditions C Ceff. The Ceff is made up of a component cs
representing uptake or release from the sorbent plus a component C* representing the
liquid phase of the chamber that would be observed in the case of no sorption:
Ceff = cs + C* (69)
⎛ −ut ⎞
C* = (C0 − Cin ) exp⎜ ⎟ + Cin (70)
⎝ VR ⎠
The value of C* is obtained by solving Eq. (68) for the case of no sorption. Veri-
fication of Eq. (70), which signifies thorough mixing conditions in the cell, can be
performed using a nonsorbing solute such as 3H2O. After subtracting C* from Ceff
(it can be seen that C* approaches a constant value at long times), Eq. (68) in terms
of cs becomes
∂q u ∂cs
= − cs − (71)
∂t VR ∂t
The first term on the right in Eq. (71) represents efflux from the chamber due to
the sorption process, whereas the second represents solution-phase chemical in-
side the cell due to the sorption process.
If an in-line detector is used and concentrations are measured at intervals of
t, then the finite difference approximation of Eq. (71), correct to terms of order
t2, is
q u( c j − 1 + 2 c j + c j + 1 ) ( c j + 1 − c j − 1 )
=− − (72)
t 4VR 2t
where the superscript j represents the jth measurement. If samples are collected,
reactor efflux is known exactly from c j and, therefore, the analogous approxima-
tion is
q uc j ( c j +1 − c j −1 )
=− − (73)
t 4VR 2t
The cumulative sorbed concentration is given by

q q0 t(q/t). (74)
For measurements with natural soils, the SCF technique suffers from many
problems. It is not very well suited for long-term uptake experiments because the
differences between Cin and Ceff quickly become too small. Some cells have built-
in stirring mechanisms: Magnetic stir bars grind particles and propeller seals have
a finite life because of continual abrasion from particles. Deitsch and Smith (1995)
simply placed the cell on a rotary shaker to effect mixing. There is always a trade-
off between loss of particles from the cell and clogging of the frits. Small particles
can be generated from large particles over time owing to particle–particle and par-
ticle–wall collisions.
D. ZERO-LENGTH COLUMNS
ZLCs (Eic and Ruthven, 1988; Ruthven and Eic, 1988) are columns that are
short enough (or the flow rate is fast enough) that dispersion is insignificant and
thus the response is sensitive only to sorption processes. The method is most suit-
able for desorption experiments. Basically, a short column of preequilibrated ma-
terial is swept with a stream of gas or liquid at a rate fast enough to maintain “zero”
concentration at the external particle surfaces, and desorbed chemical in the efflu-
ent is monitored with a sensitive detector or is trapped. This method is useful be-
cause it retains the advantages of columns in terms of relevance and ease of mea-
surement while eliminating the effects of longitudinal dispersion. Its advantage
over batch experiments is that it eliminates the need for agitation.
For a bed of spherical particles of radius a the total amount of diffusing sub-
stance leaving the column compared with the initial amount is given by Eq. (55)
(see Section V,A,1) provided the ratio of the rate parameter for external mass trans-
fer to the diffusivity is large. The observation that verifies this condition is that the
rate of efflux from the column is independent of flow velocity. The derivative of
Eq. (55) with respect to time is the desorption rate.
The ZLC technique has been employed in studies of TCE desorption from
columns of soil and silica sorbents (Farrell and Reinhard, 1994b; Grathwohl and
Reinhard, 1993).
VI. SORPTION KINETICS AND BIOAVAILABILITY
A. ASSIMILATION OF CHEMICALS IN SOIL SYSTEMS
It is well-known that bioavailability of chemicals (uptake, toxic effect, etc.) is

lower in a soil–water mixture compared to water alone. Bioavailability is reduced
for both thermodynamic and kinetic reasons: thermodynamic because a fraction
of the chemical is partitioned to the soil and is not available there; kinetic because
desorption can be rate limiting to uptake by the organism from the fluid phase. Al-
though the terms “available” and “sequestered” are often used in the literature in
regard to fractions of a sorbed chemical in soil, realistically there is a continu-
um from instantly available to completely unavailable. Furthermore, the term se-
questered has meaning only in the context of a given receptor, chemical, soil en-
vironment, mode, and duration of uptake.
Clearly, large multicell organisms assimilate chemicals only through the fluid
phase (liquid or vapor) and not directly from the particle surface or interior, al-
though they may be able to indirectly affect the flux of chemical from the particle.
For single-cell organisms the situation is less clear. Cells may attach to surfaces
by molecular forces or via extracellular exudates. Whether attached cells are able
to abstract sorbed organic molecules directly from the surface is inconclusive but
the preponderance of evidence is in the negative, at least for soil particles (Crock-
er et al., 1995; Shelton and Doherty, 1997). Experimental results are supported by
logic: (i) Most sorption sites lie within SOM interstices, which are physically in-
accessible to cells; (ii) most of the surface area of a particle is contained in meso-
pores and micropores (i.e., 50 nm), where even the smallest cells cannot fit; and
(iii) if we assume rapid local equilibrium sorption at the solution–solid interface
in the vicinity of the cell (Ke), the chemical potential, and therefore the activity, of
substrate is the same for dissolved and adsorbed forms. The following rate ex-
pressions apply,
Rate of uptake from solution kwa*fw (75)
Rate of uptake from surface ksa* (1 fw) (76)
where kw and ks are the rate constants for uptake from water and surface, a* is the
activity of the chemical in solution or on the surface (in equivalent units), and fw
is the fraction of cell surface area, , exposed to the water. A surface abstraction
mechanism can enhance bioavailability only if ks(1 fw) kw fw, which is doubt-
ful because molecules on the surface are likely to be less mobile than molecules
in solution.
An organism may affect the flux of chemical from soil particles indirectly in var-
ious ways. First, it can do so by steepening the concentration gradient across the
particle–fluid interface as a result of uptake from the fluid. This will accelerate
desorption and may explain why some bacteria seem able to access sequestered
fractions (Guerin and Boyd, 1992; Schwartz and Scow, 1999). Second, it can do
so by causing changes in soil properties through biological activity in ways that
affect Ke. Such change may result through direct action on the particle or through
effects on the surrounding medium. For example, dermal contact may involve
transfer of skin or hair oils to the particle that can facilitate uptake of hydrophobic
chemicals. Ingestion may expose particles abruptly to biosurfactants and radically
different pH regimes. Weston and Mayer (1998) found that stomach fluids increase
bioavailability of PAHs in soil. Solution pH affects soil minerology and SOM struc-
ture; acidification of a soil to below pH ⬃2 released sequestered fractions of halo-
genated hydrocarbons possibly by dissolving metal oxide cements (Pignatello,
1990b). Soil ingested by birds may be pulverized in the gizzard resulting in short-
er diffusion path lengths. Grinding in a ball mill has been shown to release resis-
tant fractions (Ball and Roberts, 1991b; Pignatello, 1990a; Steinberg et al., 1987).
Plant exudates may increase desorption by a surfactant effect or by a competitive
sorption effect; natural aromatic acids that are produced by living and decompos-
ing plants were shown to increase desorption of chlorinated aromatic hydrocarbons
and phenols by competitive displacement (Xing and Pignatello, 1998). Recent
studies show that competitive solutes increase sorption and desorption rates of the
principal solute (J. White and J. Pignatello, submitted for publication).
B. COUPLED SORPTION –BIODEGRADATION KINETIC MODELS
When biological uptake is relatively slow, or when the receptor moves rapidly
through the contaminated medium, the solution concentration is not altered ap-
preciably and bioavailability may be controlled simply by the existing solution
concentration. The equilibrium partition model being considered by the U.S. En-
vironmental Protection Agency for setting sediment quality criteria (Ankley et al.,
1996; Di Toro et al., 1991) is based on the assumption that bioavailability, or bio-
logical effect, can be predicted knowing the equivalent pore water concentration.
The pore water concentration is calculated from the total concentration present
in the solids (determined by exhaustive extraction) and the Koc determined ex-
perimentally or calculated from established Kow- or solubility-based LFERs
(Schwarzenbach et al., 1993). Although the database of Koc values and LFERs is
extensive, the values are primarily based on short equilibration times (48 h).
Their relevance to aged-contaminated systems, therefore, is highly questionable.
In many cases, the apparent Koc in historically contaminated samples has been as
much as two orders of magnitude greater than values obtained in freshly spiked
samples (Pignatello and Xing, 1996). Ronday (1997) found that, although the tox-
icity of pesticides to the springtail (Folsomia candida) correlated well with the
pore water concentration, the toxicity decreased over time and did not correlate
well with short-term Koc values.
When nonequilibrium conditions prevail during exposure it is necessary to
consider mass transfer rate laws describing the flux of chemical through the par-
Figure 16 The fraction or initial rate of phenanthrene desorbed (in the presence of Tenax infinite
sink) or mineralized to CO2 by two bacteria. The coincidence indicates that phenanthrene metabolism
is rate limited by desorption. The soils contained 1.4% OC (silt loam) and 44.5% OC (peat) (––) De-
sorption, (–䉭••) strain R biodeg, (– –••) strain P5-2 biodeg. (Reprinted with permission from Envi-
ronmental Toxicology and Chemistry, 1999. Correlation between the biological and physical availabil-
ities of phenanthrene in soils and soil humin in aging experiments, by J. C. White, M. Hunter, K. Nam,
J. J. Pignatello, and M. Alexander, 18, 1720–1727. Copyright Society of Environmental Toxicology
and Chemistry (SETAC), Pensacola, FL, 1999.)
ticle, across the particle–bulk fluid interface, and across the fluid–biomembrane
interface. An accurate bioavailability model will require linkage of uptake/depu-
ration kinetics with sorption/desorption kinetics. Some coupled models will be
discussed in this section. The discussion is restricted to the most widely re-
searched systems—those involving degradation of chemicals by microorgan-
isms. In the models we assume that the substrate is available only through the
aqueous phase.
There are many studies showing that biodegradation is rate limited by desorp-
tion of the substrate (White et al., 1999; Bosma et al., 1997; Rijnaarts et al., 1990).
For example, Fig. 16 shows the results of experiments (White et al., 1999) in which
phenanthrene was allowed to “age” in contact with soil for various times before
either adding bacterial degraders or carrying out desorption in the presence of the
infinite sink, Tenax. Normalized plots of initial desorption rate or initial bio-
degradation rate, each as a function of aging time, coincide. Likewise, normalized
plots of the amount desorbed or mineralized vs aging time coincide. This indicates
that the degraders metabolize phenanthrene molecules as they desorb.
Substrate is supplied to the solution by desorption and consumed via the solu-
tion phase by biodegradation:
∂C ∂q ⎛ ∂C ⎞
= − − ⎜ ⎟ (77)
∂t ∂t ⎝ ∂t ⎠ bd
Biodegradation of substrate when no other nutrient limitations exist is governed

by Monod kinetics:
⎛ ∂C ⎞ max C X (t )
− ⎜ ⎟ = ⋅ (78)
⎝ ∂t ⎠ bd Km + C YS
where C is the solution concentration experienced at cell surfaces [M L3], max

is the maximum growth rate [T1], Km [M L3] is the “half-saturation constant”
(the substrate concentration at which the rate is 50% of maximum), X the cell mass
concentration [M L3], and YS [M M1] is the specific bioconversion factor for
growth on the substrate (i.e., biomass produced per mass substrate consumed). The
values of max, Km, and Ys are obtained in separate soil-free growth experiments,
assuming the surface has no influence. When C is well below Km substrate uti-
lization is simplified to an expression that is first order in C:
⎛ ∂C ⎞ X (t )
− ⎜ ⎟ = max C (79)
⎝ ∂t ⎠ bd K m Ys
The cell mass concentration is a function of the rates of growth and decay, in-
cluding death and inactivation by the soil. Cells may grow on the chemical and on
utilizable natural organic matter (UOM). The general expression for the change in
cell mass is
∂X max CX ⎛ ∂[UOM] ⎞
= + YUOM ⎜ ⎟ − X (80)
∂t Km + C ⎝ ∂t ⎠
where YUOM [M M1] is the corresponding bioconversion factor for growth on

UOM, and [T1] is a first-order decay coefficient.
If degradation by indigenous organisms is being considered, natural growth and
decay may be assumed to have reached a steady state and the last two terms on the
right of Eq. (80) cancel out. If, however, degradation by cultured organisms is be-
ing considered, the last two terms—especially the decay term—may be signifi-
cant. Growth on UOM and decay processes in soils are complex and poorly un-
derstood. The values of YUOM, ![UOM]/!t, and are thus difficult to acquire,
especially since accurate assays for active cell population in the presence of soil
particles are lacking. [Somewhat better than order-of-magnitude estimates of ac-
tive degrader population may be made by a 14C-most-probable-number technique
if radiolabeled compound is available (Lehmicke et al., 1979).] If cells can be ac-

curately monitored, it may be possible to establish an empirical growth and decay
curve in the absence of substrate and input it into the model.
2. Biodegradation Coupled with First-Order-Type Sorption Models
In their study of 2,4-dichloroacetic acid (2,4-D) degradation by a 2,4-D de-

grading Alcaligenes species in unsaturated soils, Shelton and Doherty (1997) em-
ployed a four-compartment model: the biomass (X), the solution (C), sorbed avail-
able (A), and sorbed unavailable (U) compartments:
Mass transfer between the compartments obeyed the following Monod and sim-
ple first-order expressions:
dX max C
= X (81a)
dt Km + C
dC max C X
= − k1C + k−1SA − (81b)
dt K m + C YS
dA
= − k1C − ( k−1 + k2 ) SA + k−1SU (81c)
dt
dU
= k2 SA − k−1SU (81d)
dt
They assumed that the pesticide is the primary growth substrate and limiting nu-
trient, that there was no interference from indigenous organisms, and (apparently)
that there was no natural decay or growth. The sorption rate constants (k1, k1)
and (k2, k2) were obtained in independent experiments performed over 3- and 48-
h periods, respectively. Hence, the model is specific to the time frame of the ex-
periment.
3. Biodegradation Coupled with Linear Driving Force Sorption Model
Bosma et al. (1997) studied the biodegradation of -hexachlorocyclohexane

residues in field-contaminated soil. They assumed that the rate of desorption was
proportional to the concentration gradient of solute between liquids in distant

pores, where cells cannot enter, and the sites of bacterial colonies (0.8 to 3- m
macropores):
∂q
− = (Cd − Cv ) (82)
∂t
where Cd is the distal and Cc is the vicinal concentration with respect to the cell
surface, and [T1] is a desorption rate parameter. Under steady-state conditions
Eqs. (78) and (82) are equal, and through further manipulation it is possible to ob-
tain the so-called Best equation:
Cd + K m + vmax −1 ⎧⎪ ⎡ 4Cd vmax −1 ⎤

1/ 2 ⎫
⎪
v = vmax ⎨1 − ⎢1 − −1 2 ⎥ ⎬ (83)
2 vmax −1 ⎪⎩ ⎣ ( Cd + K m + v max ) ⎦ ⎪⎭
where v is the specific degradation velocity [M T1]; vmax ( maxX/YS ) is the

maximum degradation velocity [M T1], and the other variables are as defined
previously. The Best number (Bn),
1),
Bn /(vmaxKm (84)
is the index of mass transfer to biodegradation; the reaction is rate limited by
biodegradation when Bn 1 and rate limited by mass transfer when Bn 1. The
Bn for -hexachlorocyclohexane in soil slurry was 0.016 –0.03, indicating mass
transfer limitation.
4. Biodegradation Kinetics Coupled with Radial Diffusion Laws
These models (Rijnaarts et al., 1990; Scow and Alexander, 1992; Scow and Hut-
son, 1992) employ a simple radial pore diffusion law such as the one in Eq. (30)
in order to calculate an effective diffusivity using analytical (Crank, 1975) or nu-

merical solutions. It is normally assumed that the substrate concentration at the
cell surface is near zero.
Rijnaarts et al. (1990) used the measured mean particle diameter (122–182 m)
to obtain Deff, however, this value of Deff resulted in poor fits in the coupled mod-
el. Running the coupled model instead with a fitting parameter representing the
average “intraparticle diffusion distance” resulted in good fit when 14–18 m.
Rijnaarts et al. hypothesized that the bacteria were able to slightly penetrate the
particle. However, it is more likely that the length scale over which diffusion oc-
curs is simply smaller than the actual particle radius.
5. Biodegradation Coupled with Transport
With biodegradation the A–D equation for solute transport (Eq. 56; see Section
IV,A) becomes
∂C ∂q ∂2C ∂C ∂

∂t
+
∂t
= Dh 2 − v
∂x ∂x
−
∂t
(∑ i Ci ) , (85)
biodegradation term
where i refers to each solution compartment if more than one is applicable. Mod-
els published to date have assumed first-order biodegradation kinetics; that is, that
substrate concentration falls in the low-concentration region of the Monod curve
and the degrader population is at steady state.
The two-region (mobile–immobile) A–D model (Eq. 58) incorporating degra-
dation is thus
(m + fnKe Cim n −1 ) ∂C∂tm + (im + (1 − f )nKe Cim n −1 ) ∂C∂tim = m Dh

∂ 2 Cm
∂x 2
∂Cm
− m v − m m Cm − im im Cim , (86)
∂x
where the first-order biodegradation rate constants [T1] in the mobile ( m) and
immobile ( im) regions may be different due to nutrient availability, different pop-
ulations, or other factors.
Likewise, the two-site (equilibrium–kinetic) A–D model of Eq. (60) is given by
⎛ ⎞ ∂C ∂2C ∂C
1 + fnK e C n −1 + k−2 ( Ke C n − q ) = Dh 2 − v − C , (87)
⎝ ⎠ ∂t ∂x ∂x
where is the first-order biodegradation rate constant [T1]. An analytical solu-

tion to Eq. (86) or Eq. (87) is readily obtained in the linear sorption case (van
Genuchten and Wagenet, 1989).
Figure 17 Effect of dimensionless desorption rate parameter (*) and biodegradation rate para-
meter ( *) on the number of pore volumes needed to decrease the initial contaminant mass by a fac-
tor of 103 in an aquifer. Here, * L/v, * L/v,* Ke /, and D* Dh /vL. DII is the
Damkohler number II, the ratio of degradation rate to mass transfer rate. (Reprinted from Fry and Is-
tok, 1994, with permission from the American Geophysical Union.)
Researchers have used these models experimentally with some success (Angley
et al., 1992; Gamerdinger et al., 1990; Hu and Brusseau, 1998). Degradation de-
lays the breakthrough of the solute (Angley et al., 1992; Hu and Brusseau, 1998)
and, of course, decreases the amount recovered. Angley et al. studied alkylben-
zene transport in columns of nonsterile aquifer material, taking sorption to be lin-

ear. Predicted elution curves using the nonequilibrium two-site model were supe-
rior to the corresponding equilibrium single-site model ( f 1). The ’s were
highly dependent on flow velocity, however, increasing by a factor of three to eight
with increasing flow velocity from 5.76 to 65.8 pore volumes per day. Even at the
slowest flow velocity, the ’s were an order of magnitude greater than those in
nonagitated batch microcosm studies. This result underscored the “pseudo, or non-
constant nature of [ ]” and rendered extrapolation to the field “problematic” (Ang-
ley et al., 1992).
The relationship between desorption and bioavailability in an aquifer remedia-
tion scenario was examined theoretical by Fry and Istok (1994). They assumed
first-order biodegradation and the existence of a single sorption domain having a
linear isotherm and first-order desorption rate coefficient. Figure 17 shows the
number of pore volumes needed to decrease the initial contaminant mass in the
aquifer by three orders of magnitude as a function of the dimensionless desorption
coefficient, *, and the dimensionless biodegradation rate coefficient *. When
degradation is rate limiting (* is large relative to *), increasing the degradation
rate decreases the number of pore volumes needed to remediate the aquifer. How-
ever, when desorption is rate limiting (* is small relative to *), increasing the
degradation rate is predicted to be futile.
REFERENCES
Adamson, A. W., and Gast, A. P. (1997). “Physical Chemistry of Surfaces,” Sixth Edition/Ed. Wiley,
New York.
Alexander, M. (1995). How toxic are toxic chemicals in soil? Environ. Sci. Technol. 29, 2713 –2717.
Allen-King, R. M., Groenevelt, H., and Mackay, D. M. (1995). Analytical method for the sorption of
hydrophobic organic pollutants in clay-rich materials. Environ. Sci. Technol. 29, 148 –153.
Angley, J. T., Brusseau, M. L., Miller, W. L., and Delfino, J. J. (1992). Nonequilibrium sorption and
aerobic biodegradation of dissolved alkylbenzenes during transport in aquifer material: Column
experiments and evaluation of a coupled-process model. Environ. Sci. Technol. 26, 1404 –1410.
Ankley, G. T., Berry, W. J., Di Toro, D. M., Hansen, D. J., Hoke, R. A., Mount, D. R., Reiley, M. C.,
Swartz, R. C., and Zarba, C. S. (1996). Use of equilibrium partitioning to establish sediment qual-
ity criteria for nonionic chemicals: A reply to Iannuzzi et al. Environ. Toxicol. Chem. 15, 1019 –
1024.
Arocha, M. A., Jackman, A. P., and McCoy, B. J. (1996). Adsorption kinetics of toluene on soil ag-
glomerates: Soil as a bioporous sorbent. Environ. Sci. Technol. 30, 1500 –1507.
Ball, W. P., and Roberts, P. V. (1991a). Long-term sorption of halogenated organic chemicals by aquifer
material. 1. Equilibrium. Environ. Sci. Technol. 25, 1223 –1235.
Ball, W. P., and Roberts, P. V. (1991b). Long-term sorption of halogenated organic chemicals by aquifer
material. 2. Intraparticle diffusion. Environ. Sci. Technol. 25, 1237–1249.
Ball, W. P., Buehler, C., Harmon, T. C., Mackay, D. M., and Roberts, P. V. (1990). Characterization of
a sandy aquifer material at the grain scale. J. Contam. Hydrol. 5, 253 –295.
Ball, W. P., Xia, G., Durfee, D. P., Wilson, R. D., Brown, M. J., and Mackay, D. M. (1997). Hot-
methanol extraction for the analysis of volatile organic chemicals in subsurface core samples from
Dover Air Force Base, Delaware. Ground Water Monitoring Remediation 17, 104 –121.
Barrer, R. M. (1984). J. Membrane Sci. 14, 25 – 85.
Benzing, T. R., Rao, P. S. C., and Byd, S. A. (1996). Construction and performance of a gas purge sys-
tem. J. Environ. Quality 25, 1436 –1441.
Berens, A. (1989). Transport of organic vapors and liquids in poly(vinyl chloride). Makromol. Chem.
Macromol. Symp. 29, 95 –108.
Borglin, S., Wilke, A., Jepsen, R., and Lick, W. (1996). Parameters affecting the desorption of hydro-
phobic organic chemicals from suspended sediments. Environ. Toxicol. Chem. 15, 2254–2262.
Bosma, T. N. P., Middeldorp, P. J. M., Schraa, G., and Zehnder, A. J. B. (1997). Mass transfer limita-
tion of biotransformation: Quantifying bioavailability. Environ. Sci. Technol. 31, 248 –252.
Brusseau, M. L. (1992). Nonequilibrium transport of organic chemicals: The impact of pore-water ve-
locity. J. Contam. Hydrol. 9, 353 –368.
Brusseau, M. L. (1993). Using QSAR to evaluate phenomenlogical models for sorption of organic com-
pounds by soil. Environ. Toxicol. Chem. 12, 1835 –1846.
Brusseau, M. L., and Rao, P. S. C. (1989). Sorption nonideality during organic contaminant transport
in porous media. Crit. Rev. Environ. Control 19, 33 – 99.
Brusseau, M. L., Jessup, R. E., and Rao, P. S. C. (1990). Sorption kinetics of organic chemicals: Eval-
uation of gas-purge and miscible-displacement techniques. Environ. Sci. Technol. 24, 727–735.
Cameron, D. R., and Klute, A. (1977). Convective–dispersive solute transport with a combined equi-
librium and kinetic adsorption model. Water Resour. Res. 13, 183 –188.
Carroll, K. M., Harkness, M. R., Bracco, A. A., and R. R. B. (1994). Application of a permeant/poly-
mer diffusional model to the desorption of polychlorinated biphenyls from Hudson River sedi-
ments. Environ. Sci. Technol. 28, 253 –258.
Castellan, G. W. (1971). “Physical Chemistry,” 2nd ed. Addison-Wesley, Reading, MA.
Chang, M., Wu, S., and Chen, C. (1997). Diffusion of volatile organic compounds in pressed humic
acid disks. Environ. Sci. Technol. 31, 2307–2312.
Chen, N. Y. (1976). Hydrophobic properties of zeolites. J. Phys. Chem. 80, 60 – 64.
Chen, W., and Wagenet, R. J. (1995). Solute transport in porous media with sorption-site heterogene-
ity. Environ. Sci. Technol. 29, 2725 –2734.
Chiou, C. T. (1989). Theoretical considerations of the partition uptake of nonionic organic compounds
by soil organic matter. In “Reactions and Movement of Organic Chemicals in Soil” (B. L. Sawh-
ney and K. Brown, Eds.), pp. 1–29. Soil Science Society of America, Madison, WI. [Special pub-
lication]
Chiou, C. T., Rutherford, D. W., and Manes, M. (1993). Sorption of N2 and EGME vapors on some
soils, clays, and mineral oxides and determination of sample surface areas by use of sorption data.
Environ. Sci. Technol. 27, 1587–1594.
Connaughton, D. F., Stedinger, J. R., Lion, L. W., and Schuler, M. L. (1993). Description of time-vary-
ing desorption kinetics: Release of napthalene from contaminated soils. Environ. Sci. Technol. 27,
2397–2403.
Corley, T. L., Farrell, J., Hong, B., and Cponklin, M. H. (1996). VOC accumulation and pore filling in
unsaturated porous media. Environ. Sci. Technol. 30, 2884 –2891.
Cornelissen, G., van Noort, P. C. M., and Govers, H. A. J. (1997a). Desorption kinetics of chloroben-
zenes, polycyclic aromatic hydrocarbons, and polychlorinated biphenyls: Sediment extraction
with Tenax and effects of contact time and solute hydrophobicity. Environ. Toxicol. Chem. 16,
1351–1357.
Cornelissen, G., van Noort, P. C. M., Parsons, J. R., and Govers, H. A. J. (1997b). The temperature de-
pendence of slow adsorption and desorption kinetics of organic compounds in sediments. Envi-
ron. Sci. Technol. 31, 454 – 460.
Cornelissen, G., Rigterink, H., Ferdinandy, M. M. A., and Van Noort, P. C. M. (1998a). Rapidly de-
sorbing fractions of PAHs in contaminated sediments as a predictor of the extent of bioremedia-

tion. Environ. Sci. Technol. 32, 966 – 970.
Cornelissen, G., van Zuilen, H., and van Noort, P. C. M. (1999). Particle size dependence of slow de-
sorption of in situ PAHs from sediments. Chemosphere 37.
Crank, J. (1975). “The Mathematics of Diffusion,” 2nd ed. Clarendon, Oxford, UK.
Crocker, F. H., Guerin, W. F., and Boyd, S. A. (1995). Bioavailability of naphthalene sorbed to cation-
ic surfactant-modified smectite clay. Environ. Sci. Technol. 29, 2953 –2958.
Currie, J. A. (1960). Gaseous diffusion in porous media. Part 2. Dry granular materials. Br. J. Appl.
Phys. 11, 318–324.
Curthoys, G., Davydow, V. Y., Kiseliv, A. V., Kiselev, S. A., and Kuznetsov, B. V. (1974). Hydrogen
bonding in adsorption on silica. J. Colloid Interface Sci. 48, 58 –72.
Deitsch, J. J., and Smith, J. A. (1995). Effect of Triton X-100 on the rate of trichloroethene desorption
from soil to water. Environ. Sci. Technol. 29, 1069 –1080.
Deitsch, J. J., Smith, J. A., Arnold, M. B., and Bolus, J. (1998). Sorption and desorption rates of car-
bon tetrachloride and 1,2-dichlorobenzene to three organobentonites and a natural peat soil. En-
viron. Sci. Technol. 32, 3169 – 3177.
de Jonge, H., and Mittelmeijer-Hazeleger, M. C. (1996). Adsorption of CO2 and N2 on soil organic mat-
ter: Nature of porosity, surface area, and diffusion mechanisms. Environ. Sci. Technol. 30, 408–
413.
Di Toro, D. M., Zarba, C. S., Hansen, D. J., Berry, W. J., Swartz, R., Cowan, C. E., Pavlou, S. P., Allen,
H. E., and Paquin, P. R. (1991). Technical basis for establishing sediment quality criteria for non-
ionic organic chemicals using equilibrium partitioning. Environ. Toxicol. Chem. 10, 1541–1583.
Eic, M., and Ruthven, D. M. (1988). A new experimental technique for measurement of intracrystalline
diffusivity. Zeolites 8, 40 – 45.
Eick, M. J., Bar-Tal, A., Sparks, D. L., and Feigenbaum, S. (1990). Analyses of adsorption kinetics us-
ing a stirred-flow chamber: II. Potassium–calcium exchange on clay minerals. Soil Sci. Soc. Am.
J. 54, 1278–1282.
Emerson, M. T., Grunwald, E., Kaplan, M. L., and Kromhout, R. A. (1960). Proton transfer studies by
nuclear magnetic resonance. III. The mean life of the amine-water hydrogen bond in aqueous so-
lution. J. Am. Chem. Soc. 82, 6307– 6314.
Farrell, J., and Reinhard, M. (1994a). Desorption of halogenated organics from model solids, sedi-
ments, and soil under unsaturated conditions. 1. Isotherms. Environ. Sci. Technol. 28, 53 – 62.
Farrell, J., and Reinhard, M. (1994b). Desorption of halogenated organics from model solids, sedi-
ments, and soil under unsaturated conditions. 2. Kinetics. Environ. Sci. Technol. 28, 63 –72.
Fava, A., and Eyring, H. (1956). Equilibrium and kinetics of detergent adsorption—A generalized equi-
libration theory. Langmuir 60, 890 – 898.
Foster, R. (1969). “Organic Charge-Transfer Complexes.” Academic Press, London.
Fredrickson, G. H., and Helfand, E. (1985). Dual-mode transport of penetrants in glassy polymers.
Macromolecules 18, 2201–2207.
Frisch, H. L., and Stern, S. A. (1983). Diffusion of small molecules in polymers. CRC Crit. Rev. Solid
State Mater. Sci. 11, 123 –186.
Fry, V. A., and Istok, J. D. (1994). Effects of rate-limited desorption on the feasibility of in situ bio-
remediation. Water Resour. Res. 30, 2413 –2422.
Gamerdinger, A. P., Wagenet, R. J., and van Genuchten, M. T. (1990). Application of two-site/two re-
gion models for studying simultaneous nonequilibrium transport and degradation of pesticides.
Soil Sci. Soc. Am. J. 54, 957– 963.
Goring, D. A. I. (1989). The lignin paradigm. In “Lignin Properties and Materials. ACS Symposium
Series 397” (W. G. Glasser and S. Sarkanen, Eds.). American Chemical Society, Washington, DC.
Goss, K. (1992). Effects of temperature and relative humidity on the sorption of organic vapors on
quartz on sand. Environ. Sci. Technol. 26, 2287–2294.
Goss, K. (1994). Adsorption of organic vapors on polar mineral surfaces and on a bulk water surface:
Development of an empirical predictive model. Environ. Sci. Technol. 28, 640 – 645.
Goss, K.-U. (1997). Conceptual model for the adsorption of organic compounds from the gas phase to
liquid and solid surfaces. Environ. Sci. Technol. 31, 3600 – 3605.
Grathwohl, P., and Reinhard, M. (1993). Desorption of trichloroethylene in aquifer material: Rate lim-
itation at the grain scale. Environ. Sci. Technol. 27, 2360 –2366.
Greenland, D. J., and Hayes, M. H. B. (1981). Soil processes. In “The Chemistry of Soil Processes”
(D. J. Greenland and M. H. B. Hayes, Eds.), pp. 1– 35. Wiley, Chichester, UK.
Gregg, S. J., and Sing, K. S. W. (1982). “Adsorption, Surface Area and Porosity,” 2nd ed. Academic
Press, London.
Guerin, W. F., and Boyd, S. A. (1992). Differential bioavailability of soil-sorbed naphthalene to two
bacterial species. Appl. Environ. Microbiol. 58, 1142–1152.
Gustafson, D. I., and Holden, L. R. (1990). Nonlinear pesticide dissipation in soil: A new model based
on spatial variability. Environ. Sci. Technol. 24, 1032–1038.
Gustafsson, O., Haghseta, F., Chan, C., MacFarlane, J., and Gschwend, P. M. (1997). Quantification of
the dilute sedimentary soot phase: Implications for PAH speciation and bioavailability. Environ.
Sci. Technol. 31, 203 –209.
Haderlein, S. B., Weissmahr, K. W., and Schwarzenbach, R. P. (1996). Specific adsorption of ni-
troaromatic explosives and pesticides to clay minerals. Environ. Sci. Technol. 30, 612– 622.
Harmon, T. C., and Roberts, P. V. (1994). Comparison of intraparticle sorption and desorption rates for
a halogenated alkene in a sandy aquifer material. Environ. Sci. Technol. 28, 1650 –1660.
Hayes, M. H. B., and Himes, F. L. (1986). Nature and properties of humus–mineral complexes. In “In-
teraction of Soil Minerals with Natural Organics and Microbes” (P. M. Huang and M. Schnitzer,
Eds.), pp. 103–157. Soil Science Society of America, Madison, WI.
Hayes, M. H. B., MacCarthy, P., Malcolm, R. L., and Swift, R. S. (Eds.) (1989). “Humic Substances
II. In Search of Structure.” Wiley, Chichester, UK.
Hendricks, D. W., and Kuratti, L. G. (1982). Derivation of an empirical sorption rate equation by analy-
sis of experimental data. Water Res. 16, 829 – 837.
Holmén, B. A., and Gschwend, P. M. (1997). Estimating sorption rates of hydrophobic organic com-
pounds in iron oxide- and aluminosilicate clay-coated aquifer sands. Environ. Sci. Technol. 31,
105–113.
Horas, J. A., and Nieto, F. (1994). A generalization of dual mode transport theory for glassy polymers.
J. Polymer Sci. B Polymer Phys. 32, 1889 –1898.
Hu, M. Q., and Brusseau, M. L. (1998). Coupled effects of nonlinear, rate-limited sorption and
biodegradation on transport of 2,4-dichlorophenoxyacetic acid in soil. Environ. Toxicol. Chem.
17, 1673–1680.
Hu, Q., Wang, X., and Brusseau, M. L. (1995). Quantitative structure–activity relationships for eval-
uating the influence of sorbate structure on sorption of organic compounds by soil. Environ. Tox-
icol. Chem. 14, 1133 –1140.
Huang, L. Q., and Pignatello, J. J. (1990). Improved extraction of atrazine and metolachlor in field soil
samples. J. Assoc. Official Anal. Chemists 73, 443 – 446.
Huang, W., and Weber, W. J., Jr. (1997). A distributed reactivity model for sorption by soils and sedi-
ments. 10. Relationships between desorption, hysteresis, and the chemical characteristics of or-
ganic domains. Environ. Sci. Technol. 31, 2562–2569.
Huang, W., Schlautman, M. A., and Weber, W. J., Jr. (1996). A distributed reactivity model for sorp-
tion by soils and sediments. 5. The influence of near-surface characteristics in mineral domains.
Environ. Sci. Technol. 30, 2993 – 3000.
Huang, W., Young, T., Schlautman, M. A., Yu, H., and Weber, W. J., Jr. (1997). A distributed reactivi-
ty model for sorption by soils and sediments. 9. General isotherm nonlinearity and applicability
of the dual reactive domain model. Environ. Sci. Technol. 31, 1703 –1710.
Israelachvili, J. N. (1992). “Intermolecular and Surface Forces,” 2nd ed. Academic Press, London.
Jaynes, W. F., and Boyd, S. A. (1991). Hydrophobicity of siloxane surfaces in smectites as revealed by
aromatic hydrocarbon adsorption from water. Clays Clay Miner. 39, 428 – 436.
Kamiya, Y., Mizoguchi, K., Hirose, T., and Naito, Y. (1989). Sorption and dilation in poly(ethyl
methacrylate)–carbon dioxide system. J. Polymer Sci. B. 27, 879 – 892.
Kamiya, Y., Bourbon, D., Mizoguchi, K., and Naito, Y. (1992). Sorption, dilation, and isothermal glass
transition of poly(ethyl methacrylate)–organic gas systems. Polymer J. 24, 443 – 449.
Kan, A. T., Fu, G., Hunter, M. A., and Tomson, M. B. (1997). Irreversible adsorption of naphthalene
and tetrachlorobiphenyl to lula and surrogate sediments. Environ. Sci. Technol. 31, 2176 –2186.
Kan, A. T., Fu, G., Hunter, M., Chen, W., Ward, C. H., and Tomson, M. B. (1998). Irreversible sorp-
tion of neutral hydrocarbons to sediments: Experimental observations and model predictions. En-
viron. Sci. Technol. 32, 892– 902.
Kapoor, A., Yang, R. T., and Wong, C. (1989). Surface diffusion. Catal. Rev. Sci. Eng. 31, 129 –214.
Kärger, J., and Ruthven, D. M. (1992). “Diffusion in Zeolites and Other Microporous Solids.” Wiley,
New York.
Karickhoff, S. W., and Morris, K. R. (1985). Sorption dynamics of hydrophobic pollutants in sediment
suspensions. Environ. Toxicol. Chem. 4, 469 – 479.
Karickhoff, S. W., Brown, D. S., and Scott, T. A. (1979). Sorption of hydrophobic pollutants on natur-
al sediments. Water Res. 13, 241–248.
Kim, H., Annable, M. D., and Rao, P. S. C. (1998). Influence of air–water interfacial adsorption and
gas-phase partitioning on the transport of organic chemicals in unsaturated porous media. Envi-
ron. Sci. Technol. 32, 1253 –1259.
Kleineidam, S., Rügner, H., and Grathwohl, P. (1999). The impact of grain scale heterogeneity on slow
sorption kinetics. Environ. Toxicol. Chem. 18, 1673 –1678.
Kuhlbusch, T. A. J. (1998). Black carbon and the carbon cycle. Science 280, 1903 –1904.
Larson, J. W., and Wernett, P. (1988). Pore structure of Illinois No. 6 coal. Energy Fuels 2, 719 –720.
Lee, L. K. (1978). The kinetics of sorption in a biporous adsorbent particle. Am. Inst. Chem. Eng. J.
24, 531–533.
Lee, S. Y., Seader, J. D., Tsai, C. H., and Massoth, F. E. (1991). Restrictive diffusion under catalytic
hydroprocessing conditions. Ind. Eng. Chem. Res. 30, 29 – 38.
Lehmicke, L. G., Williams, R. T., and Crawford, R. L. (1979). 14-C Most-probable-number method
for enumeration of active heterotrophic microorganisms in natural waters. Appl. Environ. Micro-
biol. 38, 644–649.
Lick, W., and Rapaka, V. (1996). A quantitative analysis of the dynamics of the sorption of hydropho-
bic organic chemicals to suspended sediments. Environ. Toxicol. Chem. 15, 1038 –1048.
Lin, T., Little, J. C., and Nazaroff, W. W. (1994). Transport and sorption of volatile organic compounds
and water vapor within dry soil grains. Environ. Sci. Technol. 28, 322– 330.
Linz, D. G., and Nakles, D. V. (Eds.) (1997). “Environmentally Acceptable Endpoints in Soil: Risk-
Based Approach to Contaminated Site Management Based on Availability of Chemicals in Soil.”
American Academy of Environmental Engineers, Annapolis, MD.
Lorden, S. W., Chen, W., and Lion, L. W. (1998). Experiments and modeling of the transport of
trichloroethene vapor in unsaturated aquifer material. Environ. Sci. Technol. 32, 2009 –2017.
Lucht, L. M., Larson, J. M., and Peppas, N. A. (1987). Macromolecular structure of coals. 9. Molecu-
lar structure and glass transition temperature. Energy Fuels 1, 56–58.
Lyon, W. G. (1995). Swelling of peats in liquid methyl, tetramethylene and propyl sulfoxides and in
liquid propyl sulfone. Environ. Toxicol. Chem. 14, 229 –236.
Lyon, W. G., and Rhodes, D. E. (1993). Molecular size exclusion by soil organic materials estimated
from their swelling in organic solvents. Environ. Toxicol. Chem. 12, 1405 –1412.
Mader, B. T., Goss, K.-U., and Eisenreich, S. J. (1997). Sorption of nonionic, hydrophobic organic
chemicals to mineral surfaces. Environ. Sci. Technol. 31, 1079 –1086.
Maguire, R. J., Batchelor, S. P., and Sullivan, C. A. (1995). Reduction of the efficiency of extraction of
lipophilic chemicals from water by dissolved organic matter. Environ. Toxicol. Chem. 14, 389–393.
Mayer, L. (1999). Extent of coverage of mineral surfaces by organic matter in marine sediments.
Geochim. Cosmochim. Acta, 63, 207–215.
McBride, M. B. (1994). “Environmental Chemistry of Soils,” Oxford Univ. Press, New York.
McDonald, L. M., Jr., and Evangelou, V. P. (1997). Optimal solid-to-solution ratios for organic chem-
ical sorption experiments. Soil Sci. Soc. Am. J. 61, 1655 –1659.
McGroddy, S. E., Farrington, J. W., and Gschwend, P. M. (1996). Comparison of the in situ and de-
sorption sediment–water partitioning of polycyclic aromatic hydrocarbons and polychlorinated
biphenyls. Environ. Sci. Technol. 30, 172–177.
Miller, C. T., and Pedit, J. A. (1992). Use of a reactive surface-diffusion model to describe apparent
sorption–desorption hysteresis and abiotic degradation of lindane in a subsurface material. En-
viron. Sci. Technol. 26, 1417–1427.
Miller, D. M., Miller, W. P., and Sumner, M. E. (1989). A continuous-flow stirred reaction cell for study-
ing adsorption in suspensions. Soil Sci. Soc. Am. J. 53, 1407–1411.
Müller-Wegener, U. (1987). Electron donor acceptor complexes between organic nitrogen heterocy-
cles and humic acid. Sci. Total Environ. 62, 297– 304.
Nkedi-Kizza, P., Rao, P. S. C., Jessup, R. E., and Davidson, J. M. (1982). Ion exchange and diffusive
mass transfer during miscible displacement through an aggregated oxisol. Soil Sci. Soc. Am. J. 46,
471–476.
Nkedi-Kizza, P., Biggar, J. W., Selim, H. M., van Genuchten, M. T., Wierenga, P. J., Davidson, J. M.,
and Nielsen, D. R. (1984). Ion exchange and diffusive mass transfer during miscible displacement
through an aggregated oxisol. Soil Sci. Soc. Am. J. 46, 471– 476.
Nkedi-Kizza, P., Brusseau, M. L., Rao, P. S. C., and Hornsby, A. G. (1989). Nonequilibrium sorption
during displacement of hydrophobic organic chemicals and 45Ca through soil columns with aque-
ous and mixed solvents. Environ. Sci. Technol. 23, 814 – 820.
Novotny, F. J., Rice, J. A., and Well, D. A. (1995). Characterization of fulvic acid by laser-desorption
mass spectrometry. Environ. Sci. Technol. 29, 2464 –2466.
Nzengung, V. A., Nkedi-Kizza, P., Jessup, R. E., and Voudrias, E. A. (1997). Organic cosolvent effects on
sorption kinetics of hydrophobic chemicals by organoclays. Environ. Sci. Technol. 31, 1470–1475.
O’Dell, J. D., Wolt, J. D., and Jardine, P. M. (1992). Transport of imazethapyr in undisturbed soil
columns. Soil Sci. Soc. Am. J. 56, 1711–1715.
Ong, S., and Lion, L. (1991). Mechanisms for trichloroethylene vapor sorption onto soil minerals. J.
Environ. Quality 20, 180–188.
Papelis, C., Roberts, P. V., and Leckie, J. O. (1995). Modeling the rate of cadmium and selenite ad-
sorption on micro- and mesoporous transition aluminas. Environ. Sci. Technol. 29, 1099 –1108.
Parker, J. C., and van Genuchten, M. T. (1984). Determining transport parameters from laboratory and
field tracer experiments. bull. Virginia Agric. Exp. Station 84 – 3.
Paul, D. R., and Koros, W. J. (1976). Effect of partially immobilizing sorption on permeability and the
diffusion time lag. J. Polymer Sci. Polymer Phys. 14, 675 – 685.
Pedit, J. A., and Miller, C. T. (1994). Heterogeneous sorption processes in subsurface systems. 1. Mod-
el formulations and applications. Environ. Sci. Technol. 28, 2094 –2104.
Pedit, J. A., and Miller, C. T. (1995). Heterogeneous sorption processes in subsurfaces systems. 2. Dif-
fusion modeling approaches. Environ. Sci. Technol. 29, 1766 –1772.
Pellegrino, C., and Piccolo, A. (1999). Conformational arrangement of dissolved humic substances: In-
fluence of solution composition on association of humic molecules. Environ. Sci. Technol. 33,
1682–1690.
Petersen, L. W., Moldrup, P., El-Farhan, Y. H., Jacobsen, O. H., Yamaguchi, T., and Rolston, D. E.
(1995). The effect of moisture and soil texture on the adsorption of organic vapors. J. Environ.
Quality 24, 752–759.
Piatt, J. J., and Brusseau, M. L. (1998). Rate-limited sorption of hydrophobic organic compounds by
soils with well-characterized organic matter. Environ. Sci. Technol. 32, 1604 –1608.
Pignatello, J. J. (1989). Sorption dynamics of organic compounds in soils and sediments. In “Reactions
and Movement of Organic Chemicals in Soils,” pp. 45 – 80. Soil Science Society of America,
Madison, WI. [Special publication]
Pignatello, J. J. (1990a). Slowly reversible sorption of aliphatic halocarbons in soils. I. Formation of
residual fractions. Environ. Toxicol. Chem. 9, 1107–1115.
Pignatello, J. J. (1990b). Slowly reversible sorption of aliphatic halocarbons in soils. II. Mechanistic
aspects. Environ. Toxicol. Chem. 9, 1117–1126.
Pignatello, J. J. (1998). Soil organic matter as a nanoporous sorbent of organic pollutants. Adv. Colloid
Interface Sci. 7677, 445 – 467.
Pignatello, J. J. (1999). A revised physical concept of natural organic matter as a sorbent of organic
compounds. In “Kinetics and Mechanisms of Reactions at the Mineral/Water Interface” (D. L.
Sparks and T. J. Grundl, Eds.). American Chemical Society, Washington, DC.
Pignatello, J. J., and Xing, B. (1996). Mechanisms of slow sorption of organic chemicals to natural par-
ticles. Environ. Sci. Technol. 30, 1–11.
Pignatello, J. J., Ferrandino, F. J., and Huang, L. Q. (1993). Elution of aged and freshly added herbi-
cides from a soil. Environ. Sci. Technol. 27, 1663 –1671.
Preston, C. M., Schnitzer, M., and Ripmeester, J. A. (1989). A spectroscopic and chemical investiga-
tion on the de-ashing of a humin. Soil Sci. Soc. Am. J. 53, 1442–1447.
Rao, P. S. C., and Davidson, J. M. (1980). Estimation of pesticide retention and transformation para-
meters required in nonpoint source pollution models. In “Environmental Impact of Nonpoint
Source Pollution” (M. R. Overcash and J. M. Davidson, Eds.), pp. 23 – 67. Ann Arbor Science,
Ann Arbor, MI.
Rice, J. A., and MacCarthy, P. (1990). A model of humin. Environ. Sci. Technol. 24, 1875 –1877.
Rijnaarts, H. H. M., Bachmann, A., Jumelet, J. C., and Zehnder, A. J. B. (1990). Effect of desorption
and intraparticle mass transfer on the aerobic biomineralization of -hexachlorocyclohexane in a
contaminated calcareous soil. Environ. Sci. Technol. 24, 1349 –1354.
Rogers, C. E. (1965). Solubility and diffusivity. In “Physics and Chemistry of the Organic Solid State”
(D. Fox, M. M. Labes, and A. Weissberger, Eds.), Vol. 2, pp. 510 – 601. Interscience, New York.
Ronday, R., van Kammen-Polman, A. M. M., and Dekker, A. (1997). Persistence and toxicological ef-
fects of pesticides in topsoil: Use of the equilibrium partitioning theory. Environ. Toxicol. Chem.
16, 601–607.
Ruckenstein, E., Vaidyanathau, A. S., and Youngquist, G. R. (1971). Sorption by solids with bidisperse
pore structures. Chem. Eng. Sci. 26, 1305 –1318.
Ruthven, D. M. (1984). “Principles of Adsorption and Adsorption Processes.” Wiley, New York.
Ruthven, D. M., and Eic, M. (1988). Diffusion of linear paraffins and cyclohexane in NaX and 5A ze-
olite crystals. Zeolites 8, 472– 479.
Satterfield, C. N., Coltion, C. K., and Pitcher, W. H., Jr. (1973). Restricted diffusion in liquids within
fine pores. Am. Inst. Chem. Eng. J. 19, 628 – 635.
Sawhney, B. L., and Gent, M. P. N. (1990). Hydrophobicity of clay surfaces: Sorption of 1,2-dibro-
moethane and trichloroethene. Clays Clay Miner. 38, 14 –20.
Sawhney, B. L., Pignatello, J. J., and Steinberg, S. M. (1988). Determination of 1,2-dibromoethane
(EDB) in field soils: Implications for volatile organic compounds. J. Environ. Quality 17, 149–152.
Schlebaum, W., Badora, A., Schraa, G., and van Riemsdijk, W. H. (1998). Interactions between a hy-
drophobic organic chemical and natural organic matter: Equilibrium and kinetic studies. Environ.
Sci. Technol. 32, 2273 –2277.
Schulten, H.-R. (1995). The three-dimensional structure of humic substances and soil organic matter
studied by computational analytical chemistry. Fresenius J. Environ. Anal. Chem. 351, 62–73.
Schulten, H.-R. (1996). Three-dimensional, molecular structures of humic acids and their interactions
with water and dissolved contaminants. Int. J. Anal. Chem. 64, 147–162.
Schulten, H.-R., and Schnitzer, M. (1993). A state of the art structural concept for humic substances.
Naturwissenschaften 80, 29 – 30.
Schuster, P., Zundel, G., and Sandorfy, C. (1976). “The Hydrogen Bond.” North-Holland, Amsterdam.
Schwartz, E., and Scow, K. M. (1999). Using biodegradation kinetics to measure the availability of
aged phenanthrene to a bacteria inoculated into soil. Environ. Toxicol. Chem., in press.
Schwarzenbach, R. P., Gschwend, P. M., and Imboden, D. M. (1993). “Environmental Organic Chem-
istry.” Wiley, New York.
Scow, K. M., and Alexander, M. (1992). Effect of diffusion on the kinetics of biodegradation: Experi-
mental results with synthetic aggregates. Soil Sci. Soc. Am. J. 56, 128 –134.
Scow, K. M., and Hutson, J. (1992). Effect of diffusion and sorption on the kinetics of biodegradation:
Theoretical considerations. Soil Sci. Soc. Am. J. 56, 119 –127.
Seyfried, M. S., Sparks, D. L., Bar-Tal, A., and Feigenbaum, S. (1989). Kinetics of calcium–magne-
sium exchange on soil using a stirred-flow reaction chamber. Soil Sci. Soc. Am. J. 53, 406 – 410.
Shelton, D. R., and Doherty, M. A. (1997). A model describing pesticide bioavailability and biodegra-
dation in soil. Soil Sci. Soc. Am. J. 61, 1078 –1084.
Sparks, D. L. (1989). “Kinetics of Soil Chemical Processes.” Academic Press, San Diego.
Spurlock, F. C., Huang, K., and van Genuchten, M. T. (1995). Isotherm nonlinearity and nonequilibri-
um sorption effects on transport of fenuron and monuron in soil columns. Environ. Sci. Technol.
29, 1000–1007.
Steinberg, S. M., Pignatello, J. J., and Sawhney, B. L. (1987). Persistence of 1,2-dibromoethane in soils:
Entrapment in intraparticle micropores. Environ. Sci. Technol. 21, 1201–1208.
Swift, R. S. (1989). Molecular weight, shape, and size of humic substances by ultracentrifugation. In
“Humic Substances II. In Search of Structure” (M. H. B. Hayes, P. MacCarthy, R. L. Malcolm,
and R. S. Swift, Eds.). Wiley, Chichester, UK.
Ten Hulscher, T. E. M., Vrind, B. A., Van den Heuvel, J., Van der Velde, L. E., Van Noort, P. C. M.,
Beurskens, J. E. M., and Govers, H. A. J. (1999). Triphasic desorption of highly resistant
chlorobenzenes, polychlorinated biphenyls, and polycyclic aromatic hydrocarbons in field con-
taminated sediment. Environ. Sci. Technol. 33, 126–132.
Tognotti, L. M., Flytzani-Stephanopoulos, M., Sarofim, A. F., Kopsinis, H., and Stoukides, M. (1991).
Study of adsorption–desorption of contaminants on single soil particles using the electrodynam-
ic thermogravimetric analyzer. Environ. Sci. Technol. 25, 104 –109.
Travis, C. C., and Etnier, E. L. (1981). A survey of sorption relationships for reactive solutes in soil. J.
Environ. Quality 10, 8 –17.
Tye, R., Jepsen, R., and Lick, W. (1996). Effects of colloids, flocculation, particle size, and organic
matter on the adsorption of hexachlorobenzene to sediments. Environ. Toxicol. Chem. 15, 643 –
651.
Valocchi, A. J. (1985). Validity of the local equilibrium assumption for modeling sorbing solute trans-
port through homogenous soils. Water Resour. Res. 21, 808 – 820.
van Genuchten, M. T., and Wagenet, R. J. (1989). Two-site/two-region models for pesticide transport
and degradation: Theoretical development and analytical solutions. Soil Sci. Soc. Am. J. 53, 1303 –
1310.
van Genuchten, M. T., and Wierenga, P. J. (1976). Mass transfer studies in sorbing porous media I. An-
alytical solutions. Soil Sci. Soc. Am. J. 40, 473 – 480.
van Genuchten, M. T., and Wierenga, P. J. (1977). Mass transfer studies in sorbing porous media: II.
Experimental evaluation with tritium (3H2O). Soil Sci. Soc. Am. J. 41, 272–278.
van Genuchten, M. T., Wierenga, P. J., and O’Connor, G. A. (1977). Mass transfer studies in sorbing
porous media: III. Experimental evaluation with 2,4,5-T. Soil Sci. Soc. Am. J. 41, 278 –285.
van Oss, C. J., and Good, R. J. (1988). On the mechanism of hydrophobic interactions. J. Dispersion
Sci. Technol. 9, 355 – 362.
van Oss, C., Chaudhury, M. K., and Good, R. J. (1988). Interfacial Lifshitz–van der Waals and polar
interactions in macroscopic systems. Chem. Rev. 88, 927– 941.
Vieth, W. R. (1991). “Diffusion in and through Polymers.” Hanser, Munich.
Vieth, W. R., and Sladek, K. J. (1965). A model for diffusion in glassy polymer. J. Colloid Sci. 20,
1014–1033.
Wakao, N., and Smith, J. M. (1962). Diffusion in catalyst pellets. Chem. Eng. Sci. 17, 825 – 834.
Weber, W. J., Jr., and Huang, W. (1996). A distributed reactivity model for sorption by soils and sedi-
ments. 4. Intraparticle heterogeneity and phase-distribution relationships under nonequilibrium
conditions. Environ. Sci. Technol. 30, 881– 888.
Werth, C. J., and Reinhard, M. (1997). Effects of temperature on trichloroethylene desorption from sil-
ica gel and natural sediments. 2. Kinetics. Environ. Sci. Technol. 31, 697–703.
Weston, D. P., and Mayer, L. M. (1998). In vitro digestive fluid extraction as a measure of the bioavail-
ability of sediment-associated polycyclic aromatic hydrocarbons: Sources of variation and impli-
cations for partitioning models. Environ. Toxicol. Chem. 17, 820 – 829.
White, J., and Pignatello, J. (2000). Influence of bisolute competition on the kinetics of sorption and
desorption of polycyclic aromatic hydrocarbons in soil. Environ. Sci. Technol., in press.
White, J. C., Hunter, M., Nam, K., Pignatello, J. J., and Alexander, M. (1999). Correlation between the
biological and physical availabilities of phenanthrene in soils and soil humin in aging experiments.
Environ. Toxicol. Chem. 18, 1720 – 1727.
Wolf, D. C., Dao, T. H., Scott, H. D., and Lavy, T. L. (1989). Influence of sterilization methods on se-
lected soil microbiological, physical, and chemical properties. J. Environ. Quality 18, 39 – 44.
Wu, S., and Gschwend, P. M. (1986). Sorption kinetics of hydrophobic organic compounds to natural
sediments and soils. Environ. Sci. Technol. 20, 717–725.
Xia, G., and Ball, W. P. (1999). Adsorption-partitioning uptake of nine low-polarity organic chemicals
on a natural sorbent. Environ. Sci. Technol. 33, 262–269.
Xing, B., and Pignatello, J. J. (1996). Time-dependent isotherm shape of organic compounds in soil or-
ganic matter: Implications for sorption mechanism. Environ. Toxicol. Chem. 15, 1282–1288.
Xing, B., and Pignatello, J. J. (1997). Dual-mode sorption of low-polarity compounds in glassy
poly(vinyl chloride) and soil organic matter. Environ. Sci. Technol. 31, 792–799.
Xing, B., and Pignatello, J. J. (1998). Competitive sorption between 1,3-dichlorobenzene or 2,4-
dichlorophenol and natural aromatic acids in soil organic matter. Environ. Sci. Technol. 32, 614 –
619.
Xing, B., Gigliotti, B., and Pignatello, J. J. (1996). Competitive sorption between atrazine and other
organic compounds in soils and model sorbents. Environ. Sci. Technol. 30, 2432–2440.
Xue, S. K., and Selim, H. M. (1995). Modeling adsorption–desorption kinetics of alachlor in a typic
Fragiudalf. J. Environ. Quality 24, 896 – 903.
Yiacoumi, S., and Rao, A. V. (1996). Organic solute uptake from aqueous solutions by soil: A new dif-
fusion model. Water Resour. Res. 32, 431– 440.
Yiacoumi, S., and Tien, C. (1994). A model of organic solute uptake from aqueous solutions by soils.
Water Resour. Res. 30, 571– 580.
Yiacoumi, S., and Tien, C. (1995). Uptake of organic compounds from aqueous solutions by soils—A
comparison of two Laplace transform inversion techniques. Computers Chem. Eng. 19, 1041–
1050.
Young, D. F., and Ball, W. P. (1994). A priori stimulation of tetrachloroethene transport through aquifer
material using an intraparticle diffusion model. Environ. Prog. 13, 9 –20.
Young, T. M., and Weber, W. J., Jr. (1995). A distributed reactivity model for sorption by soils and
sediments. 3. Effects of diagenetic processes on sorption energetics. Environ. Sci. Technol. 29,
92–97.
Zhang, P.-C., and Sparks, D. L. (1993). Kinetics of phenol and aniline adsorption and desorption on an
organo-clay. Soil Sci. Soc. Am. J. 57, 340 – 345.
ENVIRONMENTAL INDICATORS
OF AGROECOSYSTEMS
O. H. Smith,1 G. W. Petersen, and B. A. Needelman

Department of Agronomy
Pennsylvania State University
University Park, Pennsylvania 16802
I.Introduction
II.Agroecosystems
III.Monitoring and Assessment Endpoints
IV. Environmental Indicators
A. Biological Indicators
B. Physical Indicators
C. Chemical Indicators
D. Landscape Indicators
V. Soil Organic Matter as a Candidate Environmental Indicator
A. The Functions of SOM in Soils
B. SOM and Atmospheric Carbon
C. Factors Controlling the SOM Content of Soils
D. Absolute and Relative Measures of SOM
E. The Measurement and Expression of SOM Quantities
VI. Indicator Ranking
VII. Conclusions and Recommendations
References
Conventional production agricultural practices are partly responsible for intensify-

ing the degradation of productive lands throughout the world. In monitoring the
impacts of these practices, a variety of biological, physical, chemical, landscape, and
economic measures are being used as indicators of environmental change. This
chapter is largely a review of both common and uncommonly used environmental
indicators of agricultural systems. Soil organic matter content is discussed in detail
as a candidate environmental indicator. A ranking scheme is proposed for the use
of multiple indicators in decision-making applications. © 2000 Academic Press.
1
Current address: VantagePoint Network LLC., 2057 Vermont Drive, Fort Collins, CO 80525.
75
0065-2113/00 $30.00
76 O. H. SMITH ET AL.
I. INTRODUCTION
Increased use of synthetic chemicals and intensification of management have

increased agricultural production but are often accompanied by deleterious envi-
ronmental effects. At a local level, these negative effects may include increases in
soil erosion and reductions in biodiversity. Regionally, incidents of groundwater
contamination and eutrophication of rivers and lakes are becoming more common,
and rates of pesticide resistance in insect populations and plant pathogens are in-
creasing (Matson et al., 1997). Regional challenges also include the loss of high-
value farmland for commercial, industrial, and residential land uses (U.S. Depart-
ment of Agriculture [USDA], 1997). Indicators of such changes require some a
priori knowledge of the scale of change in space and time and which indicators
are most appropriate for identifying critical impacts. The problem scope and its
spatial and temporal scales of resolution need to be defined, ideally prior to sam-
pling, and realistic scales of observation and management need to be selected
(Thomann, 1994). In general, less is known about indicators for terrestrial ecosys-
tems than for aquatic systems (U.S. Geological Survey Biological Resource Divi-
sion [USGSBRD], 1995). There are many interesting leads, but overall we are still
far from adequately characterizing major environmental indicators (EIs).
II. AGROECOSYSTEMS
An agroecosystem is an ecological system that is intensively managed for the

purposes of producing food, feed, and fiber. The functional components of an
agroecosystem relate to fluxes of energy, nutrients, matter, and biological species
(Altieri, 1995) within and across physical boundaries that are not spatially contin-
uous (Lanyon, 1995). Cropland, pasture, and range comprise more than 55% of
the total land area of the contiguous United States (USDA, 1997). Agroecosystems
are typically studied at field, whole farm, or regional scales (Food and Agriculture
Organization [FAO], 1994) and vary with physical, biological, socioeconomic,
and cultural influences, although methods are being developed in precision agri-
culture for study and management at the subfield scale. Major resource classes in
agroecosystems are (i) annually harvested herbaceous crops, (ii) perennial fruit
and nut crops, and (iii) pasture or range (U.S. Environmental Protection Agency/
Agroecosystems Resource Group [USEPA/ARG], 1993). In developed countries,
agroecosystems are generally managed for high production and are characterized
by intensive nutrient and pesticide inputs, fast growth and harvest cycles (hence,
disturbance dominated), and low crop and animal genetic diversity (Odum, 1984;
Matson et al., 1997). Agroecosystems usually contain parcels of unmanaged or
lightly managed areas, such as woodlots, fencerows, and riparian areas, that can
ENVIRONMENTAL INDICATORS OF AGROECOSYSTEMS 77
act both as refuges for beneficial predators of insect pests (Letourneau, 1997) and
as reservoirs of insect pests, weed seeds, and sources of plant pathogens. These ar-
eas harbor alternate hosts to fungal pathogens of crops such as cedar apple rust and
crown rust of oats (Schumann, 1991). Natural areas bordering agricultural fields
can act as similar refuges or reservoirs.
Recent analyses have estimated national and global economic benefits from
ecosystem services of soil formation, nitrogen fixation, decomposition, pest bio-
control, pollination, and others (Costanza and Daly, 1992; Pimentel et al., 1997).
Intensive agricultural management practices cause damage or loss of ecosystem
services, in the form of changes in nutrient cycling, primary productivity, species
diversity, species dominance, and population fluctuations (Odum, 1984; Rapport
et al., 1985; Duelli, 1997; Matson et al., 1997; Vitousek, 1997), in exchange for
economic productivity.
Humans play a major role in determining the nutrient, energy, water, and bio-
logical species inputs and outputs in agroecosystems. Therefore, to fully charac-
terize an agroecosystem or identify possible alternative management options for a
given agroecosystem, indicators of economic, cultural, and social forces that de-
termine human activities need to be included in an agroecosystem analysis (Aber
and Melillo, 1991; Altieri, 1995).
Agricultural practices that are ecologically sustainable may not be profitable,
thereby being economically unsustainable. Measuring crop productivity or animal
production alone also is not a sufficient indicator of agroecosystem status because
practices that achieve high yields may not be ecologically or socioeconomically
sustainable.
III. MONITORING AND ASSESSMENT ENDPOINTS
Previous and continuing federal efforts to monitor environmental conditions in

the United States include the National Agricultural Statistical Service (NASS;
USDA, 1997), USGS’s Biomonitoring of Environmental Status and Trends pro-
gram (National Research Council [NRC], 1995), and EPA’s Environmental Map-
ping and Assessment Program (EMAP; USEPA, 1997). The EMAP–ARG (USEPA-
ARG, 1993) is an interagency partnership between the USEPA and the USDA
Agricultural Research Service (ARS), the USDA NASS, and the USDA Natural
Resources Conservation Service.
The ARG produced a seminal paper (Meyer et al., 1992) which defined the fol-
lowing three assessment endpoints that summarize issues of concern to society and
can guide environmental scientists to appropriately interpret indicators:
• Sustainability: The capacity of agroecosystems to maintain commodity pro-
duction through time without threatening ecosystem structure and function.
• Contamination of natural resources: The degradation of the quality of air, soil,
water, or associated biota with by-products of agricultural practices, such as fer-

tilizers, pesticides, pathogens, and sediments.
• Quality of agricultural landscapes: The modification of land-use patterns that
comprise a matrix of landscape elements. Most significant is the ability of land-
scapes to support noncrop vegetation and wildlife populations.
The pressure–state–response framework (PSR; Pieri, 1995) is a different guide
used to conceptualize indicators of sustainability and land degradation that are de-
fined in groups as resource pressure indicators (pressure from human activities),
state indicators (resulting states of land quality and changes in their state, such as
from agricultural to urban land use), and response indicators (societal responses).
The PSR uses land quality indicators that differ by agroecological zones. Indica-
tors include intensity and diversity of land use, soil erosion, sediment loads in sur-
face water, soil fertility, price of farmlands, soil conservation practices, evenness
in income distribution, access to markets and services, and potential versus actu-
al farm yields (Pieri, 1995).
IV. ENVIRONMENTAL INDICATORS
Meyer et al. (1992) associated various indicators with assessment endpoints,

shown in Table I. All the candidate indicators discussed in this report are effec-
tively “state variables” that are integrative of environmental effects and therefore
cannot be considered statistically independent variables in subsequent analyses.
Of the 21 indicators identified, Meyer et al. (1992) selected the following 6 as
the most important, recommending that these should be evaluated by appropriate
research:
• Crop productivity: Crop yields are maintained by management inputs such as
seed, fertilizers, irrigation water, and pesticides. Crop productivity, therefore, is
a measure of efficiency, using crop-yield equivalents while accounting for the
inputs used to sustain them.
• Soil productivity: Soil productivity is characterized by the capacity of soil to
supply and maintain fertility, with soil fertility being an integrated measure of
nutrient-holding capacity, microbial activity to maintain nutrient delivery and
soil structure, the extent of contamination, and erosion rate.
• Quantity and quality of irrigation water: Impacts to irrigated agroecosystems
from water availability and quality as well as impacts of dissolved salts, sedi-
ments, fertilizers, and pesticides from agroecosystems on water quality.
• Abundance and diversity of beneficial insects: An indicator of the abundance
and diversity of insect pollinators as well as parasites and predators of insect
pests of crops.
Table I
Association between Agroecosystem Assessment Endpoints and Potential Indicatorsa
Contamination of Quality of agricultural

Indicator Sustainability natural resources landscape
Crop productivity X
Soil productivity X X
Nutrient-holding capacity X
Erosion X
Contaminants X X
Microbial component X X
Land use X X
Landscape descriptors X X
Wildlife populations X
Beneficial insect density X
Pest density X
Status of biomonitor species X
Irrigation water quantity X
Irrigation water quality X X
Agricultural chemical use X X
Non-point source loading X
Foliar symptoms X X
Livestock production X
Socioeconomic factors X
Genetic diversity X
a
From “Ecological Indicators” (1992), pp. 629– 658. Indicators of the ecological status of agro-
ecosystems, J. R. Meyer, C. L. Campbell, T. J. Moser, G. R. Hess, J. O. Rawlings, S. Peck, and W. W.
Heck; Table 35.2; with kind permission from Kluwer Academic Publishers.
• Agricultural chemical use: Crop productivity, non-point source pollution,

wildlife, beneficial species, and microbial processes are affected by amounts,
frequency, and types of agricultural chemicals used.
• Genetic diversity: Continued genetic improvement of crops and livestock re-
lies on the existence of wild-type species. The use of monocultures in agricul-
ture threatens the continuation of genetically diverse populations and invites po-
tential epidemics.
A. BIOLOGICAL INDICATORS
Measurement of species abundance, diversity, and extent are also useful as

agroecosystem EIs. Soil microbial status, invertebrate assemblages or invertebrate
exposures to toxins, pesticide resistance in insect pests, crop diversity, and water-
borne pathogens are discussed as important indicators.
1. Soil Microbial Status
Soils are possibly the most critical nonrenewable resource in terrestrial ecosys-
tems. Depletion of the soil resource can have catastrophic impacts on ecosystem
structure and function. Measurable effects may include reductions in net primary
productivity and species biodiversity following land-use conversion of forest
ecosystems to pasture or cropland. The disturbance effect may be attributable, in
part, to changes in soil quality and biota.
It has been suggested that, because of their rapid turnover times and genetic
information transfer, soil microbial communities are especially responsive to eco-
system disturbance (Edwards et al., 1996; Elmholt, 1996; Scow, 1997). The diffi-
culty in testing this hypothesis lies in problems of identifying and culturing soil
organisms without biasing the assessment toward those organisms that simply
grow well on the media chosen. Therefore, if soil microbial diversity and micro-
bially mediated processes are to be included as EIs, selection should emphasize
measurements of microbial diversity that minimizes bias and important specific
ecosystem processes, such as nitrification, that are mediated by a select group of
microorganisms and that are particularly sensitive to stress, such as nitrification or
CO2 fixation by algae.
Based on these considerations, two soil microbial indicators may be considered:
(i) some index of microbial community diversity. Several have been suggested
(Garland and Mills, 1991; Zak et al., 1994), but none is widely accepted so addi-
tional evaluations are required; and (ii) assessment of changes in ecosystem func-
tion by measuring nitrifying potential of soil (Schmidt and Belser, 1994) and pho-
tosynthetic activity by surface-dwelling cyanobacteria and algae. The former
approach is based on the fact that an important soil process whereby ammonium
is oxidized to nitrate is mediated by a relatively small group of bacteria that are
quite sensitive to environmental conditions. The latter approach relies on a process
that is sensitive to other types of ecosystem stress.
2. Earthworms
Earthworms are important decomposers that make plant residues and microbial
biomass available to other organisms. Earthworm activity contributes to soil struc-
ture and water infiltration (Altieri, 1995). Earthworms have also been used as in-
dicators of soil sustainability due to their sensitivity to pesticides (Paoletti et al.,
1991; Edwards et al., 1996; Bohlen et al., 1997).
3. Honeybees
Honeybees, which collect pollens from up to 7 km2, have been used to assess
regional pollutant levels (Bromenshenk, 1988). Recently, exposure to organophos-
phate and carbamate insecticides was determined using acetylcholinesterase ac-

tivity of honeybees (Stefanidou et al., 1996).
4. Other Biological Indicators
Other species used as EIs in agricultural systems include carabid beetles (Kromp,
1990), oribatid mites (Franchini and Rockett, 1996; Lebrun and van Straalen,
1995), lichens (Loppi, 1996), invertebrate assemblages (Paoletti et al., 1991; Sam-
ways et al., 1996), protozoa (Foissner, 1997), avian communities (Brooks et al.,
1998), and plants used to monitor pest populations (Berlinger et al., 1996). Further
development and standardization of biological indicators is greatly needed.
5. Pesticide Resistance in Insect Pests
Pesticide resistance is another useful metric of agroecosystem sustainability.

Pesticide resistance in arthropods, plant pathogens, and weed species is a major
obstacle to sustaining current levels of agricultural production (Georghiou, 1986;
Hoy, 1990). Ever-increasing resistance effectively forces farmers to make more
frequent pesticide applications and to use new pesticides, increasing chemical
loadings to fields, groundwater, and surface waters. Although crop rotations and
integrated pest management tactics are locall effective, current practices result in
the loss of susceptible genotypes on a regional basis. Because of these losses, re-
gional coordination of agricultural practices to slow resistance development is re-
quired. Changes in pesticide resistance can be interpreted by rightward shifts in
dose–mortality curves (ffrench-Constant and Roush, 1990). New transgenic crops
such as Bt corn (Roush, 1997) express insecticidal proteins in the crop to which
insect pests are expected to develop quick resistance (Alstad and Andow, 1995;
Roush, 1997; Tabashnik, 1997; Fox, 1998).
6. Crop Diversity
Crop diversity at the field, regional, and global scale is an important indicator
of agroecosystem sustainability and stability. At the field and regional scale, crop
diversity is an indicator of the vulnerability of a cropping system to pest outbreaks
(NRC, 1972). If an agroecosystem is dominated by a few crop species represent-
ing a few cultivars (genetic composition), a new pest could spread rapidly across
a large region. Such an event occurred in 1970 and 1971 when almost all of the
corn harvest in Illinois and Indiana was lost due to the fungus causing the south-
ern corn leaf blight (Helminthosporium maydis) and in the mid- and late 1980s
when an epidemic of blackleg (Leptosphaeria maculans) in rapeseed caused im-
mense yield losses ( Juska et al., 1997). Crop species diversity and the diversity of
genotypes of a given species at the global scale is an important indicator of agro-
ecosystem sustainability. Loss of crop and related wild species, as well as their nat-
ural habitats, indicates the loss of genetic resources important to the future of crop
breeding programs (Gliessman, 1998).
7. Fecal Pathogens
Surface water runoff from agricultural fields where livestock are present usual-
ly contains fecal pathogens. Fecal pathogens and parasitic protozoa can also be
used as biological indicators of water quality resulting from agricultural practices.
However, more research is needed to identify appropriate indicators of pathogens
and their biology in order to determine cost-effective water quality protection pro-
grams. For instance, fecal coliforms and fecal streptococci are not pathogens, but
they are indicators of water-borne pathogens that are difficult to monitor (Doran
et al., 1981). Confirmed human pathogens such as Cryptosporidium are a result of
runoff from dairy operations (Atwill, 1996). Particularly sensitive areas include
those underlain by limestone bedrock, where common sinkholes can allow surface
water to contaminate groundwater aquifers.
Challenges exist with the use of bioindicators: If a particular bioindicator
species exists in one area, it may be absent from another region. Related species
used as indicators may have different susceptibilities than those of the primary
species. Soil microorganisms, lichens, earthworms, beetles, bees, and birds all
have merits as indicators of agroenvironmental condition, each with varying
ranges of applicability. These ranges need to be experimentally determined.
Changes in pest resistance to pesticides will also be valuable indicators of im-
pacts due to the widespread adoption of transgenic crops such as Bt corn and Bt
cotton.
B. PHYSICAL INDICATORS
Physical EIs in agroecosystems were discussed in part in Table I in relation to

indicators of crop and soil productivity. Special mention, however, needs to be
made about soil degradation. Soil degradation, which reduces both profitability
and sustainability, is caused by water and wind erosion, excess salts, leaching of
nutrients, buildup of toxins, and changes in porosity, permeability, bulk density,
structural stability, and rates of organic matter decomposition (FAO, 1990). Soil
loss alone is reported to occur 13–40 times faster than soil renewal rates, mainly
from erosion on tilled and bare lands (Pimentel and Kounang, 1998). Soil degra-
dation is currently being assessed using remote sensing in conjunction with field
observations, soil databases, simulation models, and geographic information sys-
tems (GIS; FAO, 1990; Petersen et al., 1997; Nizeyimana and Petersen, 1997).
Landscape indicators, discussed later, can also be useful for coarse-scale assess-
ments of agroecosystem change.
An important database of landscape-scale natural resource trends is the Nation-

al Resource Inventory (NRI). The NRI is a comprehensive sampling of land cov-
er, land use, soil erosion, prime farmland, wetlands, and other characteristics on
nonfederal rural land in the United States (USDA, 1997). Every 5 years since 1982,
the USDA has published summaries of natural resource trends based on 800,000
sites in the NRI database. Several important discoveries were made using these
data: Although the rates of soil erosion and agricultural wetland loss have de-
creased, between 1982 and 1992, 6 million acres of prime farmland were converted
to nonagricultural uses. Future NRI surveys should continue to elucidate impor-
tant natural resource trends.
C. CHEMICAL INDICATORS
Chemical processes that occur in natural and managed ecosystems affect the
bioavailability of nutrients and toxins and can therefore impose significant con-
straints on ecosystem status. An essential or a toxic element is bioavailable if it is
in a chemical form that plants can absorb readily and if, once absorbed, it affects
the life cycle of the plant (Barber, 1984). Therefore, key to assessing bioavailabil-
ity is the knowledge of the form, speciation, or sequestration of a nutrient or po-
tentially toxic element in the ecosystem matrix. Soil chemical and biochemical
reactions such as complexation, sorption, ion exchange, mineral solubility, and
organic matter transformation set the initial constraints on the availability of es-
sential elements for plant uptake (Sposito, 1989). In addition to adequate supply,
plant growth is affected by the balance of nutrients in the soil. Nutrient imbalances
do not usually induce deficiency symptoms unless the supply of a given nutrient
reaches very low levels, but they do cause plants to grow more slowly (Clarkson
and Hanson, 1980).
Conversely, the balance of nutrient and toxic elements can induce significant
stress in ecosystems in which chemical species are taken up competitively from
the soil. In such cases, molar ratios can provide a valuable EI for stress “thresh-
olds” (Cronan and Grigal, 1995). For example, the calcium to aluminum (Ca/Al)
molar ratio of the soil solution has been used successfully to identify approximate
thresholds beyond which the risk of forest damage from Al stress and nutrient im-
balances increases. It has also been used as an indicator to assess forest ecosystem
changes over time in response to acidic deposition, forest harvesting, and other soil
acidifying processes (Cronan and Grigal, 1995).
Selective extraction procedures have been developed to quantify available nu-
trient species in soils (Van der Zee et al., 1987; Shuman et al., 1988; Bundy and
Meisinger, 1994). However, the problem with routine application of operational-
ly defined extractions to quantify chemical species is that the methods are not suf-
ficiently specific for target compounds and the actual species extracted are rarely
known with precision. Recent advances in molecular spectroscopy provide a set
of tools that may be used to assess chemical speciation in complex matrices such
as soils (Bertsch and Hunter, 1998). Spectroscopic approaches should be exploit-
ed in future work to corroborate the results of selective extractions and to assess
the viability of both approaches to evaluate element speciation and bioavailability.
Excess salinity, extremes in pH, and the status of nitrogen and phosphorous are
also important chemical indicators in agricultural systems. Crop harvest involve
the intentional removal of nutrients from agricultural lands. Unintentional nutri-
ent removal occurs through soil erosion, nutrient runoff or leaching, and volatiliza-
tion, depleting nutrient stocks that are necessary for sustained yields (Magdoff et
al., 1997). To offset nutrient depletion, synthetic fertilizers and/or manure are
added to the soil, which results in temporary excesses in N and P that are subject
to runoff and leaching, causing eutrophication of rivers and lakes. Nutrient flow
pattern, or nutrient dislocation, has demonstrable consequences at the field, farm,
region, and global scales (Magdoff et al., 1997).
D. LANDSCAPE INDICATORS
Landscape variables have been used extensively in efforts to link landscape pat-
terns with ecological processes (Turner, 1989; Forman, 1995; Gustafson and Gard-
ner, 1996; Ives et al., 1998). The theoretical underpinnings of pattern/process in-
teractions are related to issues of hierarchy and scale (Allen and Hoekstra, 1992;
Riera et al., 1998). Landscape descriptors, or metrics, reflect levels of landscape
disturbance, which have been used to test hypotheses relating to the effects of
landscape structure on such diverse entities as forest communities (Franklin and
Forman, 1987), beetles in rangelands (Wiens and Milne, 1989) and in croplands
(Colunga-Garcia et al., 1997), crop pest parasitoids (Marino and Landis, 1996),
bobcats (Litvaitis, 1993), prairie grasses (Wu and Levin, 1994), and avian guilds
(Herkert, 1994; Miller et al., 1997). Combinations of GIS and knowledge-based
systems have been employed to interpret relations within and among landscape
data themes for decision making on agricultural and natural resources (Coulson et
al., 1990; Petersen et al., 1995, 1997).
Spatial patterns of land cover types are characterized by a variety of landscape
metrics. Riitters et al. (1995) determined that 55 discrete metrices could be reduced
by factor analysis to the following metrics, exhibiting relative statistical indepen-
dence:
1. Average patch perimeter/area ratio (called fractal dimension)
2. Contagion
3. Average patch area normalized to the area of a square with the same perimeter
(linear vs planar patch shapes)
4. Patch perimeter/area scaling
5. Number of attribute classes
The USEPA (1994) ranked landscape metrics with regard to their suitability in
landscape characterization studies. The list, shown in Table II, shows that conta-
gion, fractal dimension, and dominance were ready for field test and implementa-
tion. Equations describing each metric are beyond the scope of this chapter; how-
ever, readers are referred to Gardner and O’Neill (1990) for further details.
Contagion is a measure of clumping, or aggregations of patches (Gardner and
O’Neill, 1990). Fractal dimension, more accurately described as average patch
perimeter/area ratio, is a measure of patch edge complexity (Wiens and Milne,
1989). Dominance is the proportion to which one or few land cover types occur
predominantly in the landscape (Gardner and O’Neill, 1990). Major improvements
to contagion and fractal dimension that are dramatically less sensitive to spatial
resolution and tesellation, termed the patch per unit area metric and square-pixel
metric, respectively, have been developed by Frohn (1997).
V. SOIL ORGANIC MATTER AS A CANDIDATE

ENVIRONMENTAL INDICATOR
No indicator can fully characterize agroecosystem status at multiple scales.

However, soil organic matter (SOM) content correlates with many aspects of agro-
ecosystem productivity, sustainability, and environmental integrity. Stevenson’s
(1994, p. 1) definition of SOM as “the whole of the organic material in soils, in-
Table II
Ranking of Landscape Stability and Resilience
Indicators by the EPA (1994)a
Indicator Status
Contagion C
Fractal dimension C
Dominance C
Lacunarity A
Diffusion rates A
Percolation backbone B
Percolation thresholds B
Miles of road B
Recovery time A
Landcover transition matrix A
aAbbreviations used: A, requiring further conceptual
development; B, requiring testing for feasilbility/sensi-

tivity; C, ready for field test and implementation.
cluding litter, light fraction, microbial biomass, water-soluble organics, and stabi-
lized organic matter (humus)” will be used in this text.
In general, SOM content is positively correlated with positive soil status
(Reeves, 1997; Bauer and Black, 1994). The methodology for SOM measurement
is established and in widespread use (Nelson and Sommers, 1996). With proper
techniques, the measurement of SOM is very precise. In comparison to many soil
measures, SOM content has small spatial and short-term temporal variability.
There is a large body of basic and applied research focused on SOM character and
the relationship between SOM and agroecosystem properties.
The SOM capacity of a soil is dependent on other soil properties such as soil
texture. Therefore, changes in SOM content must be interpreted in the context of
site-specific variables. SOM is not a comprehensive indicator. For example, SOM
content is not an accurate indicator of excess salinity and poor drainage. In addi-
tion, many larger scale parameters, such as land conversion and rural economic
prosperity, cannot be quantified with the SOM indicator.
A. THE FUNCTIONS OF SOM IN SOILS
SOM plays a mostly beneficial role in determining the biological, physical,

and chemical qualities of a soil (Stevenson, 1994, 1986). SOM is a nutrient and
energy source for soil organisms and a nutrient source for plants. It improves soil
structure, strengthens soil aggregates, increases water retention, chelates metals,
buffers the soil pH, interacts with xenobiotics, and retains cations and anions in
the soil system.
The positive relationship between SOM content and crop yield has been ob-
served in many agroecosystems (Reeves, 1997). Bauer and Black (1994) correlat-
ed an increase of 15.6 kg ha1 of wheat grain yield to a 1 Mg ha1 increase of
SOM in the northern Great Plains. The influences of increased SOM on soil
physical properties tends to reduce runoff and erodibility. This may be a result of
increased infiltration and percolation rates, water-holding capacity, aggregate
strength, and crusting resistance.
The relationships described previously between SOM and soil properties are not
necessarily causative. For example, SOM is correlated with the clay content of a
soil, making it difficult to separate the effects of SOM and clay on soil properties.
B. SOM AND ATMOSPHERIC CARBON
The carbon component of SOM is a dynamic pool in the global carbon cycle
and forms the largest terrestrial carbon pool (Paul and Clark, 1996). Kern and
Johnson (1993) sought to estimate the potential effect of conservation tillage on
U.S. terrestrial carbon pools. They reviewed 17 studies that compared SOM con-
tents in conventional tillage and no-tillage systems. They estimated that the in-
crease in SOM carbon that would result from the adoption of no-till agriculture on
76% of the planted cropland in the United States would be equivalent to 0.7–1.1%
of the total projected U.S. fossil fuel carbon emissions for the next 30 years. The
authors assumed that the effects of tillage were not dependent on climatic and soil
variables.
C. FACTORS CONTROLLING THE SOM CONTENT OF SOILS
1. The Factors of Soil Formation
The five factors of soil formation—climate, organisms, relief, parent material,

and time—largely determine the SOM content of soils ( Jenny, 1941; Stevenson,
1994). In general, the degree of impact of the factors of soil formation on SOM
decreases in the order climate vegetation topography parent material
age for loamy soils in the United States. Climate influences SOM content pri-
marily through temperature and precipitation. Temperature is negatively correlat-
ed to SOM content, probably due to increased microbial activity at higher tem-
peratures (Stevenson, 1986). In general, precipitation stimulates plant growth,
thereby increasing organic matter inputs into soils (Stevenson, 1994). Soil mois-
ture changes may decrease or increase SOM decomposition rates, depending on
the balance between water and oxygen availability. An example of the effect of
vegetation on SOM is the greater quantity of SOM found in grassland (Mollisols)
soils than in forest (Alfisols) soils. The mechanisms of this effect may include in-
creased biomass production, nitrification inhibition, decreased aeration, and a
more extensive rhizosphere in grasslands (Stevenson, 1986). Topography influ-
ences SOM contents through impacts on microclimate, drainage, and erosion.
Anaerobic conditions in poorly drained soils cause SOM accumulation due to
slow decomposition rates and incomplete catabolism. Erosion tends to remove
SOM-enriched soil (Rasmussen and Collins, 1991). The impact of parent mater-
ial is primarily through its influence on soil texture (Stevenson, 1986). The posi-
tive relationship between SOM content and fine soil texture is well established.
Soil age is most important in young soils, in which SOM accumulation rates typ-
ically exceed rates of decay.
2. Management
The tillage of natural lands leads to a loss of SOM (Jenny, 1941; Paustian et al.,
1997). This may be a result of erosion, increased oxygen availability, reduced bio-
mass production, and the disruption of aggregates. Erosion removes the surface
layer, which is frequently the soil with the greatest SOM concentration. In addi-
tion, erosion preferentially removes fine soil particles, which are SOM enriched.
Rasmussen and Collins (1991) argue that soil erosion is often overlooked in SOM
field experiments because soil erosion rates are rarely measured and field experi-
ments are usually located on fertile, level to gently sloping land.
In general, management practices that are considered beneficial to the health of
the agroecosystem, such as cover crops, conservation tillage, manure and residue
inputs, and erosion prevention practices, also increase SOM contents. Many agro-
nomic practices that increase yields will also increase biomass production, there-
by increasing organic matter inputs. However, there are many exceptions to these
generalizations.
The impacts of management practices on SOM are a result of a complex and in-
teracting set of factors. Most of the research concerning the impacts of manage-
ment practices on SOM contents has been conducted using long-term field plot
studies. Although much has been learned in these studies (Paustian et al., 1997),
additional research is needed to quantify the interactions between and among the
factors of soil formation and management practices. For example, Campbell et al.
(1999) investigated the impacts of texture in an 11-year comparison of no-till and
conventional tillage practices at three sites in the Brown soil zone in semiarid
southwestern Saskatchewan. They observed that the no-till systems contained
more SOM than conventionally tilled systems only in finer-textured soils.
D. ABSOLUTE AND RELATIVE MEASURES OF SOM
SOM content may be quantified in absolute or relative terms. Absolute measures

are simply the amount of SOM in a system. Relative measures estimate the change
in SOM content over time. The use of absolute measures would require the esti-
mation of SOM baseline conditions, most likely based on the factors of soil for-
mation and past management practices. Although this is an area of major research,
the relationship between the factors of soil formation, management practices, and
SOM contents is not sufficiently understood to make widespread, accurate base-
line estimates. In addition, the data set that would be required is not available. For
these reasons, relative measures of SOM content should be used. With sufficient
sampling intensity, a time step of 5 years should be appropriate for most applica-
tions. The impacts of management on SOM content are generally detectable after
several years but may take decades.
Relative measures need to be understood in the context of soil variability. For
example, a 10% increase in SOM content in a clay soil and in a sandy soil should
be interpreted differently. These interpretations are a function of the particular use
of the SOM measure. For example, for issues of atmospheric carbon, net changes
should be considered. The topic requires further research.
E. THE MEASUREMENT AND EXPRESSION OF SOM QUANTITIES
There are three standard methods to express SOM quantities: gravimetric, vol-
umetric, and equivalent mass. The gravimetric and volumetric methods may not
adequately account for soil depth and density (Hammer et al., 1995). The simplest
way to express the quantity of SOM in soil is as the mass of carbon or nitrogen per
mass of dry soil (mass basis). This method characterizes the average mass SOM
per mass soil in the sampling depth. Expression of SOM on a mass basis cannot
account for the total quantity of soil present (Doran and Parkin, 1994). For exam-
ple, this method may not accurately estimate SOM losses due to erosion.
In principle, the solution is to express SOM quantities as a mass carbon or ni-
trogen per unit area of soil—in effect integrating the mass per unit volume over
sufficient depth to include all the organic matter. However, it is generally imprac-
tical to sample to a depth that would include all the organic matter in the system.
A more practical solution is to express SOM quantities per unit volume of soil (vol-
umetric basis). This method is most effective when most of the SOM of the site is
contained in the sampling depth, which is often not the case. This method may
overestimate SOM in high bulk density soils. For example, consider a study that
compares a low and a high bulk density soil of otherwise identical composition.
Both soils will have identical gravimetric SOM contents, but the high bulk densi-
ty soil will have a greater volumetric SOM content. This is particularly problem-
atic when treatments, such as tillage, affect bulk density.
One solution to this problem is to express the quantity of SOM on an equiva-
lent mass basis (Ellert and Bettany, 1995). This approach adds a step to the volu-
metric method of calculation. Volumetric measurements are corrected by adjust-
ing the sampling depth to give an equivalent mass. This is done by sampling a
smaller depth in heavier soils or a greater depth in lighter soils. In the example of
a low and a high bulk density soil, a portion of the high bulk density soil would be
removed mathematically, restoring the equivalence in SOM content estimation.
Soil density, particularly that of the surface soil, is variable over short periods of
time. The equivalent mass basis is not sensitive to soil density changes.
Surface residues are rarely included in the quantification of SOM (Paustian et
al., 1997). This underestiimates SOM sequestration in high-residue systems such
as conservation tillage and forests. However, in soils with large SOM contents,
residues account for a small percentage of the total SOM storage of the system.
Carbon storage in residue is more important to consider in lower SOM content
soils.
Typically, SOM concentrations are stratified in soils with the greatest concen-
tration near the surface. The degree of stratification is dependent on soil proper-
ties and management. For example, tillage homogenizes SOM in the plow layer.
At the other extreme, forest soils contain a surface layer composed primarily of
organic materials. A vertically stratified sampling system should be used in order
to estimate SOM distribution. The vertical distribution of SOM is important in un-

derstanding SOM function and the causes of soil degradation. In general, sampling
depths should be based on soil horizons as opposed to standard depths. However,
standard depths may be used to divide soil horizons (e.g., 0–2 cm, 2–6 cm, and 6
cm to the bottom of the A horizon).
VI. INDICATOR RANKING
Although SOM as a candidate EI was discussed in particular detail, the caveats

for measuring, determining, and using indicator values apply to any indicator vari-
able that is examined. Furthermore, using a combination of indicators presents spe-
cial challenges in determining the weight and ranking assigned to each indicator.
Development of indicator rankings should include criteria of indicator appropri-
ateness based on biological relevance, repeatability of measure, integration of ef-
fects, existing databases, cost-effectiveness, and other criteria. When does a state
variable become an EI? In an interactive group decision-making process that in-
volves multiple and conflicting criteria, many experts, and many potential out-
comes, the Delphi method (Erffmeyer et al., 1986) can be useful in gaining con-
sensus. Characteristics of the Delphi technique include anonymity, controlled
feedback, and statistical group response (Khorramshahgol and Moustakis, 1988).
Once the evaluation criteria are weighted (Table III), each of the indicators can be
ranked (Table IV) according to the evaluation criteria by circulating brief, succes-
sive questionnaires among experts in the field, yielding a total numeric score for
each indicator. Table V contains the final indicator rankings, lumped by score
ranges into numeric categories, with 1 being the highest rank.
Applying EIs for regulatory or policy purposes necessitates the use of multiple
criteria decision-making tools such as the Delphi process.
Table III
Example Criteria, with Associated Example Weights, That May
Be Used to Rank Any Indicator
Criteria Average weight (sums to 100)
Biological relevance 35
Repeatability 20
Integrated measurement 20
Cost-effectiveness 10
Existing database 10
Infrastructural support 5
Etc. —
Table IV
Example Criteria for One Indicator and Its Associated Ranking and Weighted Rank Scores
Criteria relating to Indicator

crop productivity indicator Rank (0–9) Criteria weights weighted rank
Cost-effectiveness 3 10 30
Biological relevance 9 35 315
Repeatability 3 20 60
Integrated measurement 4 20 80
Existing database 4 10 40
Infrastructural support 2 5 10
Etc. — — —
Total for crop productivity 535
VII. CONCLUSIONS AND RECOMMENDATIONS
Concerns regarding issues of ecology and sustainability of agricultural systems

as well as detrimental environmental impacts from agriculture have resulted in the
development of indicators to estimate environmental trends and conditions. To ap-
propriately describe and evaluate the sustainability of an agroecosystem, a wide
Table V
Matrix of Candidate Agroecosystem Indicators Based
on Potential Rankings
Indicator Rank
Crop productivity 1
Soil productivity 1
Nutrient-holding capacity 3
Erosion 2
Contaminants 3
Microbial status 1
Irrigation water quantity and quality 2
Beneficial insect abundance and diversity 2
Agricultural chemical use 1
Genetic diversity 2
Status of biomonitor species 1
Landscape descriptors 3
Pest-resistance status 1
Socioeconomic factors 3
Etc. —
range of biological, physical, chemical, and economic indicators need to be as-

sessed. The following recommendations can be made with regard to the use of EIs:
• Determine which indicators are the most sensitive to change.
• Develop standards for validating bioindicators as EIs.
• Validate the use of landscape metrics with remote-sensing and GIS products for
monitoring landscape change.
• Use a multiscale approach to indicator validation, with studies targeted to local,
landscape, and regional levels.
• Provide funds for programs designed to improve terrestrial assessment methods
and long-term monitoring of soil biological processes, soil organic matter, crop
diversity, movement of pests and pathogens, ecological impacts of propagating
transgenic crops, changes in resistance to pesticides of insects, weeds, and plant
pathogens.
• Conduct field studies to explore linkages between ecological processes and spa-
tial/temporal patterns in agroecosystems.
• Institute policies to preserve crop and wild crop species diversity, which is crit-
ical for maintaining a genetic resource base for future plant breeding programs.
• Amend the USDA NASS questionnaire to include bioindicator information and
other data relevant to USEPA EMAP assessments.
ACKNOWLEDGMENTS
We thank John Chorover, Shelby Fleischer, Heather Karsten, and Andy Rogowski for their con-
structive criticisms and contributions.
REFERENCES
Aber, J. D., and Melillo, J. M. (1991). “Terrestrial Ecosystems.” Saunders, Philadelphia.

Allen, T. F. H., and Hoekstra, T. W. (1992). “Toward a Unified Ecology.” Columbia Univ. Press, New
York.
Alstad, D. N., and Andow, D. A. (1995). Managing the evolution of insect resistance to transgenic
plants. Science 268, 1894 –1896.
Altieri, M. A. (1995). “Agroecology: The Science of Sustainable Agriculture,” 2nd ed. Westview, Boul-
der, CO.
Atwill, E. R. (1996). Assessing the link between rangeland cattle and water-borne Cryptosporidium
parvum infection in humans. Rangelands 18(2), 48 – 51.
Barber, S. A. (1984). “Soil Nutrient Bioavailability.” Wiley, New York.
Bauer, A., and Black, A. L. (1994). Quantification of the effect of soil organic matter content on soil
productivity. Soil Sci. Soc. Am. J. 58, 185 –193.
Berlinger, M. J., Dijk, B. L., Dahan, R., and Lebiush-Mordechi, S. (1996). Indicator plants for moni-
toring pest population growth. Ann. Entomol. Soc. Am. 89(5), 611– 622.
Bertsch, P. M., and Hunter, D. B. (1998). Elucidating fundamental mechanisms in soil and environ-
mental chemistry: The role of advanced analytical and spectroscopic methods. In “The Future of
Soil Chemistry” (P. M. Huang, Ed.). SSSA, Madison, WI.
Bohlen, P. J., Parmelee, R. W., McCartney, D. A., and Edwards, C. A. (1997). Earthworm effects on
carbon and nitrogen dynamics of surface litter in corn agroecosystems. Ecol. Appl. 7(4), 1341–
1349.
Bromenshenk, J. J. (1988). Regional monitoring of pollutants with honeybees. In “Progress in Envi-
ronmental Specimen Banking” (S. A. Wise, R. Zeisler, and G. M. Goldstein, Eds.), Special Publ.
No. 740. U.S. Department of Commerce, National Bureau of Standards, Washington, DC.
Brooks, R. P., O’Connell, T. J., Wardrop, D. H., and Jackson, L. E. (1998). Toward a regional index of
biological integrity: The example of forested riparian ecosystems. Environ. Monitoring Assess-
ment 51(1), 131–143.
Bundy, L. G., and Meisinger, J. J. (1994). Nitrogen availability indices. P. 951– 984, In “Methods of
Soil Analysis, Part 2—Microbiological and Biochemical Properties” (R. W. Weaver et al., Eds.),
SSSA Book Series No. 5. SSSA, Madison, WI.
Campbell, C. A., Biederbeck, V. O., McConkey, B. G., Curtin, D., and Zentner, R. P. (1999). Soil qual-
ity: Effect of tillage and fallow frequency. Soil organic matter quality as influenced by tillage and
fallow frequency in a silt loam in Southern Saskatchewan. Soil Biol. Biochem. 31, 1–7.
Clarkson, D. T., and Hanson, J. B. (1980). The mineral nutrition of higher plants. Annu. Rev. Plant Phys.
31, 239–298.
Colunga-Garcia, M., Gage, S. H., and Landis, D. A. (1997). Response of an assemblage of Coccinel-
lidae (Coleoptera) to a diverse agricultural landscape. Environ. Entomol. 26, 797– 804.
Costanza, R., and Daly, H. (1992). Natural capital and sustainable development. Conservation Biol. 1,
37–45.
Coulson, R. N., Lovelady, C. N., Flamm, R. O., Spradling, S. L., and Saunders, M. C. (1990). Intelli-
gent geographic information systems for natural resource management. In “Quantitative Methods
in Landscape Ecology: The Analysis and Interpretation of Landscape Heterogeneity.” (M. G.
Turner and R. H. Gardner, Eds.), pp. 153 –172. Springer-Verlag, New York.
Cronan, C. S., and Grigal, D. F. (1995). Use of calcium/aluminum ratios as indicators of stress in for-
est ecosystems. J. Environ. Quality 24, 209 –226.
Doran, J. W., and Parkin, T. B. (1994). Defining and assessing soil quality. p. 3 –21. In “Defining Soil
Quality for a Sustainable Environment” ( J. W. Doran, D. C. Coleman, D. F. Bezdicek, and B. A.
Stewart, Eds.), SSSA Special Publ. No. 35. SSSA, Madison, WI.
Doran, J. W., Slepers, J. S., and Swanson, N. P. (1981). Chemical and bacteriological quality of pas-
ture runoff. J Soil Water Conservation 36(3), 166 –171.
Duelli, P. (1997). Biodiversity evaluation in agricultural landscapes: An approach at two different
scales. Agric. Ecosystems Environ. 62, 81– 91.
Edwards, C. A., Subler, S., Chen, S. K., and Bogomolov, D. M. (1996). Essential criteria for selecting
bioindicator species, processes, or systems to assess the environmental impact of chemicals on
soil ecosystems. In “Bioindicator Systems for Soil Pollution” (N. M. van Straalenand and D. A.
Krivolutskii, Eds.), pp. 67– 84. Kluwer, Dordrecht.
Ellert, B. H., and Bettany, J. R. (1995). Calculation of organic matter and nutrients stored in soils un-
der contrasting management regimes. Can. J. Soil Sci. 75, 529 – 538.
Elmholt, S. (1996). Microbial activity, fungal abundance, and distribution of penicillium and fusarium
as bioindicators of a temporal development of organically cultivated soils. Biol. Agric. Hort. 13(2),
123.
Erffmeyer, R. C., Erffmeyer, E. S., and Lane, I. M. (1986). The delphi technique: An empirical evalu-
ation of the optimal number of rounds. Groups Org. Stud. 11, 120 –128.
ffrench-Constant, R. H., and Roush, R. T. (1990). Resistance detection and documentation: The rela-
tive roles of pesticidal and biochemical assays. In “Pesticide Resistance in Arthropods” (R. T.
Roush and B. E. Tabashnik, Eds.). Chapman & Hall, New York.
Foissner, W. (1997). Protozoa as bioindicators in agroecosystems, with emphasis on farming practices,

biocides, and biodiversity. Agric. Ecosystems Environ. 62, 93 –103.
Food and Agriculture Organization (FAO) (1990). Use of high-resolution satellite data and geograph-
ic information system for soil erosion mapping, FAO RSC Series No. 56. FAO, Rome.
Food and Agriculture Organization (FAO) (1994). Agro-Ecological land resources assessment for agri-
cultural development planning: A case study of Kenya. Making land use choices for district plan-
ning, World Soil Resource Reports 71/9. FAO, Rome.
Forman, R. T. (1995). “Land Mosaics: The Ecology of Landscapes and Regions.” Cambridge Univ.
Press, Cambridge, UK.
Fox, T. (1998). EPA seeks refuge from Bt resistance. Nature Biotechnol. 15(5), 409.
Franchini, P., and Rockett, C. L. (1996). Oribatid mites as “indicator” species for estimating the envi-
ronmental impact of conventional and conservation tillage practices. Pedobiologia 40(3), 217–
225.
Franklin, J. F., and Forman, R. T. T. (1987). Creating landscape patterns by forest cutting: Ecological
consequences and principles. Landscape Ecol. 1, 5 –18.
Frohn, R. C. (1997). “Remote Sensing for Landscape Ecology: New Metric Indicators for Monitoring,
Modeling, and Assessment of Ecosystems.” CRC Press, Boca Raton, FL.
Gardner, R. H., and O’Neill, R. V. (1990). Pattern, process, and predictability: The use of neutral mod-
els for landscape analysis. In “Quantitative Methods in Landscape Ecology: The Analysis and In-
terpretation of Landscape Heterogeneity” (M. G. Turner and R. H. Gardner, Eds.), pp. 289– 307.
Springer-Verlag, New York.
Garland, J. I., and Mills, A. I. (1991). Classifications and characterization of heterotrophic microbial
communities on the basis of community level sole-carbon source utilization. Appl. Environ. Mi-
crobiol. 57, 2351–2359.
Georghiou, G. P. (1986). “Pesticide Resistance: Strategies and Tactics for Management.” National
Academy Press, Washington, DC.
Gliessman, S. R. (1998). “Agroecology. Ecological Processes in Sustainable Agriculture.” Ann Arbor
Press, Ann Arbor, MI.
Gustafson, E. J., and Gardner, R. H. (1996). The effect of landscape heterogeneity on the probability
of patch colonization. Ecology 77, 94 –107.
Hammer, R. D., Henderson, G. S., Udawatta, R., and Brandt, D. K. (1995). Soil organic carbon in the
Missouri Forest–Prairie Ecotone. In “Carbon Forms and Functions in Forest Soils” (W. W. McFee
and J. W. Kelly, Eds.), pp. 201–231. SSSA, Madison, WI.
Herkert, J. R. (1994). The effects of habitat fragmentation on midwestern grassland bird communities.
Ecol. Appl. 4(3), 461– 471.
Hoy, M. A. (1990). Pesticide resistance in arthropod natural enemies: Variability and selection re-
sponses. In “Pesticide Resistance in Arthropods” (R. T. Roush and B. E. Tabashnik, Eds.). Chap-
man & Hall, New York.
Ives, A. R., Turner, M. G., and Pearson, S. M. (1998). Local explanations of landscape patterns: Can
analytical approaches approximate simulation models of spatial processes? Ecosystems 1, 35 – 51.
Jenny, H. (1941). “Factors of Soil Formation.” McGraw-Hill, New York.
Juska, A., Busch, L., and Tanaka, K. (1997). The Blackleg epidemic in Canadian rapeseed as a “nor-
mal agricultural accident.” Ecol. Appl. 7(4), 1350 –1356.
Kern, J. S., and Johnson, M. G. (1993). Conservation tillage impacts on national soil and atmospheric
carbon levels. Soil Sci. Soc. Am. J. 57, 200 –210.
Khorramshahgol, R., and Moustakis, V. S. (1988). Delphic hierarchy process (DHP): A methodology
for priority setting derived from the Delphi method and analytic hierarchy process. Euro. J. Op-
erations Res. 37, 347– 354.
Kromp, B. (1990). Carabid beetles (Coleoptera, Carabidae) as bioindicators in biological and conven-
tional farming in Austrian potato fields. Biol. fertil. Soils 9(2), 182.
Lanyon, L. E. (1995). Does nitrogen cycle? Changes in the spatial dynamics of nitrogen with indus-
trial nitrogen fixation. J. Prod. Agric. 8, 70 –78.
Lebrun, P., and van Straalen, N. M. (1995). Oribatid mites: Prospects for the use in ecotoxicology. Exp.
Appl. Acarol. 19(7), 361– 379.
Letourneau, D. K. (1997). Plant–arthropod interaction in agroecosystems. In “Ecology in Agriculture”
(L. E. Jackson, Ed.), pp. 239 –290. Academic Press, San Diego.
Litvaitis, J. A. (1993). Response of early successional vertebrates to historic changes in land use. Con-
servation Biol. 7(4), 866 – 872.
Loppi, S. (1996). Lichens as bioindicators of geothermal air pollution in central Italy. Bryologist 99, 41.
Magdoff, F., Lanyon, L., and Liebhardt, B. (1997). Nutrient cycling, transformations, and flows: Im-
plications for a more sustainable agriculture. Adv. Agron. 60, 1–73.
Marino, P. C., and Landis, D. A. (1996). Effect of landscape structure on parasitoid diversity and par-
asitism in agroecosystems. Ecol. Appl. 6, 276 –284.
Matson, P. A., Parton, W. J., Power, A. G., and Swift, M. J. (1997). Agricultural intensification and
ecosystem properties. Science 277, 504 – 509.
Meyer, J. R., Campbell, C. L., Moser, T. J., Hess, G. R., Rawlings, J. O., Peck, S., and Heck, W. W.
(1992). Indicators of the ecological status of agroecosystems. In “Ecological Indicators” (D. H.
McKenzie, D. E. Hyatt, and V. J. McDonald, Eds.), pp. 629 – 658. Elsevier, Amsterdam.
Miller, J. N., Brooks, R. P., and Croonquist, M. J. (1997). Effects of landscape patterns on biotic com-
munities. Landscape Ecol. 12, 137–153.
National Research Council (1972). “Committee on Genetic Vulnerability of Major Crops. Genetic Vul-
nerability of Major Crops.” National Academy of Science, Washington, DC.
National Research Council (1989). “Alternative Agriculture.” National Academy Press, Washington,
DC.
National Research Council (1995, February). “A Review of the Biomonitorning of Environmental Sta-
tus and Trends (BEST) Program: The Draft Detailed Plan. Committee to Review the Department
of the Interior’s Biomonitoring of Environmental Status and Trends Program.” National Acade-
my Press, Washington, DC.
Nelson, D. W., and Sommers, L. E. (1996). Total carbon, organic carbon, and organic matter. In “Meth-
ods of Soil Analysis, Part 3. Chemical Methods” (D. L. Sparks et al., Eds.), SSSA Book Ser. No.
5, pp. 961–1010. ASA/SSSA, Madison, WI.
Nizeyimana, E., and Petersen, G. W. (1997). Remote sensing applications to soil degradation assess-
ments. In “Methods for Assessment of Soil Degradation” (R. Lal, W. H. Blum, C. Valentine, and
B. A. Stewart, Eds.), pp. 393– 403. CRC Press, Washington, DC.
Odum, E. P. (1984). Properties of agroecosystems. In “Agricultural Ecosystems: Unifying Concepts”
(R. Lowrance, B. R. Stinner, and G. J. House, Eds.), pp. 3–12. Wiley–Interscience, New York.
Paoletti, M. G., Favretto, M. R., Stinner, B. R., Purrington, F. F., and Bater, J. E. (1991). Invertebrates
as bioindicators of soil use. Agric. Ecosystems Environ. 34, 341– 362.
Paul, E. A., and Clark, F. E. (1996). “Soil Microbiology and Biochemistry,” 2nd ed. Academic Press,
San Diego.
Paustian, K., Collins, H. P., and Paul, E. A. (1997). Management controls on soil carbon. In “Soil Or-
ganic Matter in Temperate Agroecosystems” (E. A. Paul, E. T. Elliot, K. Paustian, and C. V. Cole,
Eds.), pp. 15–49. CRC Press, Boca Raton, FL.
Petersen, G. W., Bell, J. C., McSweeney, K., Nielsen, G. A., and Robert, P. C. (1995). Geographic in-
formation systems in agronomy. Adv. Agron. 55, 67–111.
Petersen, G. W., Nizeyimana, E., and Evans, B. M. (1997). Applications of geographic information sys-
tems in soil degradation assessments. In “Methods for Assessment of Soil Degradation” (R. Lal,
W. H. Blum, C. Valentine, and B. A. Stewart, Eds.), pp. 377– 391. CRC Press, Washington, DC.
Pieri, C., Dumanski, J., Hamblin, A., and Young, A. (1995). “Land Quality Indicators,” World Bank
Discussion Paper No. 315. World Bank, Washington, DC.
Pimentel, D., Wilson, C., McCullum, C., Huang, R., Dwen, P., Flack, F., Tran, Q., Saltman, T., and
Cliff, B. (1997). Economic and environmental benefits of biodiversity. BioScience 47, 747–757.
Pimentel, D. C., and Kounang, N. (1998). Ecology of soil erosion in ecosystems. Ecosystems 1, 416 –
426.
Rapport, D. J., Regier, H. A., and Hutchinson, T. C. (1985). Ecosystem behavior under stress. Am. Nat.
125, 617–640.
Rasmussen, P. E., and Collins, H. P. (1991). Long-term impacts of tillage, fertilizer, and crop residue
on soil organic matter in temperate semiarid regions. Adv. Agron. 45, 93 –134.
Reeves, D. W. (1997). The role of soil organic matter in maintaining soil quality in continuous crop-
ping systems. Soil Tillage Res. 43, 131–167.
Riera, J. L., Magnuson, J. J., Vande Castle, J. R., and MacKenzie, M. D. (1998). Analysis of large-scale
spatial heterogeneity in vegetation indices among North American landscapes. Ecosystems 1,
268–282.
Riitters, K. H., O’Neill, R. W., Hunsaker, C. T., Wickham, J. D., Yankee, D. H., Timmons, S. P., Jones,
K. B., and Jackson, B. L. (1995). A factor analysis of landscape pattern and structure metrics.
Landscape Ecol. 10, 23 – 40.
Roush, R. T. (1997). Bt-transgenic crops: Just another pretty insecticide or a chance for a new start in
resistance management? Pesticide Sci. 51, 328 – 334.
Samways, M. J., Caldwell, P. M., and Osborn, R. (1996). Ground-living invertebrate assemblages in
native, planted and invasive vegetation in South Africa. Agric. Ecosystems Environ. 59, 19 – 32.
Schmidt, E. L., and Belser, L. W. (1994). Autotrophic nitrifying bacteria. In “Methods of Soil Analy-
sis, Part 2—Microbiological and Biochemical Properties” (R. W. Weaver et al., Eds.), SSSA Book
Ser. No. 5, pp. 159 –178. SSSA, Madison, WI.
Schumann, G. L. (1991). “Plant Diseases and Their Social Impact.” American Phytopathological So-
ciety, St. Paul, MN.
Scow, K. M. (1997). Soil microbial communities and carbon flow in agroecosystems. In “Ecology in
Agriculture” (L. E. Jackson, Ed.), pp. 367– 403. Academic Press, New York.
Shuman, L. M., Rayner, P. L., Day, J. L., and Cordonnier, M. J. (1988). Comparison of four phosphorus
extraction methods on three acid southeastern soils. Commun. Soil Sci. Plant Anal. 19, 579–595.
Sposito, G. (1989). “The Chemistry of Soils.” Oxford Univ. Press, New York.
Stefanidou, M., Koutselinis, A., Pappas, F., and Methenitou, G. (1996). Bee head acetylcholinesterase
as an indicator of exposure to organophosphate and carbamate insecticides. Vet. Hum. Toxicol.
38(6), 420–422.
Stevenson, F. J. (1986). “Cycles of Soil: Carbon, Nitrogen, Phosphorous, Sulfur, Micronutrients.” Wi-
ley, New York.
Stevenson, F. J. (1994). “Humus Chemistry: Genesis, Composition, Reactions.” Wiley, New York.
Tabashnik, B. E. (1997). Seeking the root of insect resistance to transgenic plants. Proc. Natl. Acad.
Sci. 94, 3488–3490.
Thomann, R. V. (1994). The significance of resource scale in water quality and ecosystem modeling and
decision making. In “Towards a Sustainable Coastal Watershed: The Chesapeake Experiment,”
conference proceedings, June 1–3, pp. 20–27. Chesapeake Research Consortium, Edgewater, MD.
Turner, M. G. (1989). Landscape ecology: The effect of pattern on process. Annu. Rev. Ecol. System-
atics 20, 171–197.
U.S. Department of Agriculture (USDA) (1997). Agricultural resources and environmental indicators,
1996–97, Agricultural Handbook No. 712. USDA Economic Research Service, Natural Re-
sources and Environment Division, Washington, DC.
U.S. Environmental Protection Agency (1994). Landscape monitoring and assessment research plan,
EPA No. 620/R-94/009. Office of Research and Development, Washington, DC.
U.S. Environmental Protection Agency (1997). EMAP research strategy [Online]. http://www.epa.gov.
[1998, February 15]
U.S. Environmental Protection Agency–Agroecosystems Resource Group (1993). EMAP—Agro-

ecosystems program background and conceptual framework [Online]. http://www.darwin.arrc.
ncsu.edu, September 7. [1998, February 15]
U.S. Geological Survey/BRD (1995). Summary of a workshop for the selection of biomonitoring meth-
ods [Online]. http://orion.mesc.nbs.gov. [1998, February 15]
Van der Zee, S.E.A.T.M., Fokkink, L. G. J., and van Riemsdijk, W. H. (1987). A new technique for as-
sessment of reversibly adsorbed phosphate. Soil Sci. Soc. Am. J. 51, 599 – 604.
Vitousek, P. M., Mooney, H. A., Lubchenco, J., and Melillo, J. M. (1997). Human domination of earth’s
ecosystems. Science 277, 494 – 499.
Wiens, J. A., and Milne, B. T. (1989). Scaling of landscapes in landscape ecology, or, landscape ecol-
ogy from a beetle’s perspective. Landscape Ecol. 3(2), 87–96.
Wu, J., and Levin, S. A. (1994). A spatial patch dynamic modeling approach to pattern and process in
an annual grassland. Ecol. Monogr. 64(4), 447– 464.
Zak, J. C., Willig, M. R., Moorhead, D. L., and Wildman, H. G. (1994). Functional diversity of mi-
crobial communities: A quantitative approach. Soil Biol. Biochem. 26, 1101–1108.
GROWTH PROMOTION OF
PLANTS INOCULATED WITH
M. A. Whitelaw
School of Wine and Food Sciences
Charles Sturt University
Wagga Wagga, NSW 2678, Australia
I. Introduction
II. Soil Phosphorus
A. Phosphate Sorption
B. Precipitation of Phosphate Compounds
C. The Soil Phosphorus Cycle
D. The Relationship of Soil pH to Phosphate Solubility
E. Plant Uptake of Sparingly Soluble Phosphate
III. Phosphate-Solubilizing Soil Microorganisms
A. Solubilization of Phosphates by Free-Living Fungi
B. Precipitated Phosphate Agar Studies
IV. Liquid Medium Studies
A. Solubility of Phosphate Compounds
B. Production of Organic Acids in Liquid Medium Studies
C. Chelation of Cations by Organic Acids
D. Titratable Acidity
E. pH
F. N Source
G. Fluctuations in Soluble Phosphate Levels over Time
V. Plant Growth Promotion by Phosphate-Solubilizing Fungi
VI. Conclusion
References
Phosphorus is an important plant nutrient which is in short supply in many agri-

cultural soils. Because much of the soluble phosphate (P) applied to soils as fertil-
izer is “fixed” by the soil and rendered less available to plants, the long-term appli-
cation of P fertilizers has resulted in an accumulation of total soil P, most of which
is poorly soluble. Many soil fungi, predominantly of the genera Aspergillus and
Penicillium, have been shown to possess the ability to solubilize sparingly soluble
phosphates in vitro by secreting inorganic or organic acids. Growth promotion and
99
0065-2113/00 $30.
100 M. A. WHITELAW
increased uptake of P by plants inoculated with P-solubilizing fungi have also been
reported by many investigators. © 2000 Academic Press.
I. INTRODUCTION
Many soils, such as those found in much of Australia, are low in phosphate (P)
readily available for plant growth (i.e., “available P”). Historically, deficiency of
soil P has been one of the most important chemical factors restricting plant growth
in the acid soils of the southern wheat belt of New South Wales, Australia (Col-
well, 1963). Because low P status tends to limit the production of arable crops, the
application of soluble P fertilizers such as superphosphate has been widely used,
especially for cereal crops. However, much of the soluble P applied as fertilizer
may react with the soil and be “fixed” or converted into one of the many sparing-
ly soluble forms which are less available for uptake by plants than rapidly ex-
changeable forms of P (Stevenson, 1986). It has long been known that soil mi-
croorganisms play an important role in the cycling of many soil nutrients including
phosphate. Bacteria, yeasts, actinomycetes, mycorrhizal fungi, and free-living
fungi have been reported to cause increases in plant-available P in the soil. Many
in vitro studies have demonstrated the presence of soil microorganisms capable of
transforming soil P to forms available to plants. Many investigators have also re-
ported increases in yield and P uptake due to inoculation of plants by P-solubiliz-
ing fungi. The use of vesicular mycorrhizal (VAM) fungi to inoculate plants is lim-
ited by the inability of scientific researchers to grow the fungi in pure culture and
produce large amounts of inoculum (Kucey et al., 1989). This review attempts to
summarize information from microbial P solubilization in vitro studies and from
glasshouse and field studies involving free-living P-solubilizing fungi.
II. SOIL PHOSPHORUS
Much of the soluble P applied to soil as fertilizer may react with the soil and be
converted to one of many sparingly soluble forms of P. The major P fixation reac-
tions are shown in Fig. 1. Both biological and chemical processes are involved in
P fixation but the chemical processes of P sorption and precipitation are more im-
portant in the retention of fertilizer P (Stevenson, 1986). McLaughlin et al. (1988)
reported that wheat plants grown on a solonized brown soil (pH 8.3) used only 12%
of the fertilizer P applied with the crop. Of the remainder, 71% was either precipi-
tated by aluminium, ferric, or calcium compounds or adsorbed by hydrous oxides
of aluminium and ferric or by aluminosilicate (clay) materials. Four percent was
immobilized by soil microbes and 13% was incorporated into soil organic matter.
PHOSPHATE-SOLUBILIZING FUNGI 101
Figure 1 Phosphate fixation reactions in soil. [Adapted from Sauchelli (1951) in “Cycles of Soil,
Carbon, Nitrogen, Phosphorus, Sulfur, Micronutrients,” F. J. Stevenson. Copyright 1986 John Wiley &
Sons. Reprinted by permission of John Wiley & Sons.]
A. PHOSPHATE SORPTION
“P sorption” is a term used to describe two types of chemical reactions which

remove P from the soil solution: adsorption and absorption. Adsorption is the ini-
tial rapid process in which a layer of H 2PO 4 forms on the surface of solid alu-
minium or ferric hydrous oxides, whereas absorption is the slow diffusive pene-
tration into the bulk of the adsorbing solid which occurs after adsorption (Barrow,
1987). Because of similarities in sorption isotherms, several researchers have con-
cluded that the sorption mechanisms for aluminosilicates (clays) are similar to
those of the hydrous oxides (Sample et al., 1980). The adsorption of anions by the
edge faces of the clay, kaolinite, is believed to be similar to the adsorption by hy-
drous oxides (Hingston et al., 1972). Because montmorillonite and kaolinite clays
retain similar amounts of P, their P sorption mechanisms are also considered to be
similar (Wild, 1950).
Olsen and Watanabe (1957) found that acidic soils fixed twice the amount of
added P per unit surface area than neutral or calcareous soils. They also reported
that the P fixed was held with five times more bonding energy in acidic soils com-
pared to calcareous soils. Much of the P retained by acidic soils is specifically
102 M. A. WHITELAW
adsorbed by iron and aluminium hydrous oxides and aluminosilicate minerals.

Leaver and Russell (1957) demonstrated the importance of hydrated iron and alu-
minium oxides in P retention in soils. They found that treating soils with reagents
which formed insoluble or un-ionized compounds with Fe3+ and Al3+ reduced P
retention. Oxides and hydrous oxides of iron and aluminium exist as discrete com-
pounds in soil, coatings on other soil particles, and amorphous aluminium-hydroxy
compounds between layers of expandable aluminium silicates (Sample et al.,
1980). Soil is a dynamic system with surface-reactive amorphous aluminium and
iron hydrous oxides being continually added to the system through the weathering
process (Hsu, 1965). The surfaces of these oxides and hydrous oxides have a high
affinity for P and the adsorption reactions onto these are rapid (Van der Zee et al.,
1987).
The P adsorption capacity of neutral or high pH calcareous soils is influenced
mainly by the amount of exchangeable calcium ions and by the CaCO3 content
(Kamprath and Watson, 1980). However, when Holford and Mattingly (1975)
studied P adsorption by 24 calcareous soils, they found strongly bonded adsorp-
tion surfaces to be closely related to the content of iron hydrous oxides, indicating
that even in calcareous soils hydrous oxides are important in P adsorption.
B. PRECIPITATION OF PHOSPHATE COMPOUNDS
Because of the high surface area of most soils and the rapidity of adsorption re-
actions, it has been assumed that P fixation in soils is controlled mainly by ad-
sorption. However, precipitation reactions could be an important mechanism for
P fixation in some soils and Holford (1983) suggested that precipitation by alu-
minium was a major cause of P fixation in several acidic soils (pH 4.5). Pre-
cipitation was found to be greatest in a sandy soil which had the lowest quantity
of major P adsorbents such as hydrous oxides. In nonsandy soils, precipitation was
a relatively minor but still significant P-fixation mechanism.
The type of precipitate formed from dissolved P depends largely on the soil pH.
In acidic soils, P can be precipitated by the Fe3+ and Al3+ released when the alu-
minium and iron hydrous oxides are decomposed by acidic, highly concentrated
P solutions, whereas in alkaline soils dissolved P can be precipitated by Ca2+. The
initial products of the reaction of superphosphate in alkaline soils are mainly cal-
cium monohydrogen phosphate dihydrate (CaHPO4#2H2O) and calcium monohy-
drogen phosphate (CaHPO4), whereas the initial products in acidic soils are amor-
phous ferric phosphate and amorphous aluminium phosphate. In acidic soils,
aluminium phosphates are precipitated in preference to iron phosphates unless the
soils contain a large quantity of amorphous ferric hydroxide (Russell, 1980). Ini-
tially, the precipitates are amorphous and P is moderately available to plants (Ra-
jan, 1976). Hsu (1982) observed that the reaction products may remain in amor-
phous forms for at least 66 months. The precipitates are then thought to be slow-
ly converted to less soluble forms such as calcium orthophosphate [Ca3(PO4)2],
hydroxyapatite [(Ca5(PO4)3(OH)], and fluoroapatite [Ca10(PO4)6F2] in alkaline
soils or crystalline ferric phosphate [FePO4#2H2O(cr) or strengite] and crystalline
aluminium phosphate [AlPO4#2H2O(cr) or variscite] in acidic soils (Lindsay et al.,
1962; Rajan, 1976; Sample et al., 1980).
C. THE SOIL PHOSPHORUS CYCLE
The soil P cycle is a dynamic one involving soil, plants, and microorganisms.
In natural ecosystems there is virtually a closed cycle, with the P consumed by
plants being returned to the soil in plant or animal residues which are broken down
by soil microorganisms. However, soil under conventional cultivation contains
lower levels of organic matter and P is regularly removed in the harvest (Steven-
son, 1986). In order to produce high crop yields, farmers must apply larger quan-
tities of P fertilizer than is taken up by their crops. In many cases they may be ad-
vised to apply up to four times the P requirements of the crop to overcome the
problem of the conversion of P to unavailable forms (Goldstein, 1986). This situ-
ation inevitably leads to an accumulation of sparingly soluble inorganic P in the
soil, with the soil acting as a “sink” for P (Cosgrove, 1977).
P concentrations in soil solutions are small, ranging from 0.01 to 0.3 mg P
liter1 (Ozanne, 1980). Although the soil solution P is only a small proportion of
the total soil P content, this is where plants derive most of their immediate re-
quirements (Bolan, 1991). In the absence of significant levels of organic P, the so-
lution P is in equilibrium with a quantity of relatively labile inorganic P such that
in any one soil the ratio of labile inorganic P to solution P is constant (Stewart and
Sharpley, 1987). The soil solution is only adequate if the labile inorganic P is sol-
ubilized at least as quickly as the roots can extract if from the soil solution (Rus-
sell, 1980). When plants take up P from the soil solution, the depleted solution P
pool is immediately replenished from labile and moderately labile inorganic P
forms. If these pools are depleted, nonlabile P sources determine the soluble P con-
centration in soil (Stewart and Sharpley, 1987). The solution inorganic P equilib-
rium was described by Larsen (1967). When the soil solution is depleted, there is
rapid movement from labile soil P to soil solution P. However, when labile soil P
is depleted, the movement from nonlabile soil P to labile soil P is slow.
Soil solution P B labile soil P B nonlabile soil P

(solid phase) (solid phase)
The overall supply of P to plants is represented in Fig. 2 (Bolan, 1991). Labile

soil P, consisting mainly of P adsorbed by iron and aluminium oxides and alumi-
104 M. A. WHITELAW
Figure 2 Schematic representation of the supply of phosphate to plant roots in soil systems. (1)
Adsorption–desorption, (2) solid state diffusion, (3) precipitation–dissolution, (4) immobilization–
mineralization, (5) diffusion in solution, (6) movement into roots. [Reproduced from Plant Soil 134
(1991), 189–207. A critical review on the role of mycorrhizal fungi in the uptake of phosphorus by
plants. N. S. Bolan. © 1991, with kind permission from Kluwer Academic Publishers.]
nosilicate (clay) minerals, is able to replenish depleted P in the soil solution in a

short period of time. Nonlabile inorganic soil P (i.e., P from ferric, aluminium, and
calcium–phosphate precipitates and organic P and “occluded” P) is slower to re-
plenish soil solution P (Murrmann and Peech, 1969).
Soil is deficient in P when the rate of supply to plant roots is less than the rate
required for optimum growth. One can increase the rate of P supply by supplying
a low level of freshly applied fertilizer but with time the rate decreases again and
more P fertilizer is needed (Costin and Williams, 1983).
D. THE RELATIONSHIP OF SOIL PH

TO PHOSPHATE SOLUBILITY
Controversy exists regarding the effect of pH change on P solubility in soil. De-

pending on the soil and experimental conditions, sorption has been found to be de-
creased, increased, or not affected by increasing soil pH. This topic has been re-
viewed by White (1983), who points out that most studies of acidic soils in which
aluminosilicates (i.e., clays) are the dominant P adsorbents show that P solubility
either decreases or exhibits no significant change with increasing pH. The chemi-
cal basis for this common observation is thought to be the hydrolysis of exchange-
able aluminium at high pH and the subsequent reaction with soluble P in the soil
solution to form an insoluble aluminium–phosphate compound (Haynes, 1982).
Changes in P desorption with changes in soil pH have also been attributed to
other causes. Soil pH affects the charge on the P species in solution and on the sur-
faces of the adsorbing particles. As the soil pH rises the P adsorbing surfaces be-
come increasingly negatively charged. This makes the reaction with negatively
charged phosphate ions more difficult and therefore tends to decrease P adsorp-
tion as pH rises. However, raising the pH also changes the proportions of P species
in solution. In the normal range of soil pH, P is present in solution mainly as
H2PO 2 2
4 and HPO4 but the affiinity of reactive soil surfaces for HPO4 is much
2
greater than that for H2PO4 . As the concentration of HPO4 increases with pH,
this effect tends to offset the increase in the negative charge of the adsorbing par-
ticles and therefore increases P adsorption as pH rises (Barrow, 1990). Whether a
decrease in soil pH will increase or decrease P solubilization sometimes depends
on which of the effects is dominant.
The effects of pH on P sorption also differ depending on the amount of P which
had previously been added to the soil. Barrow (1984) found that increases in pH
up to 5.5 decreased sorption in unfertilized soils but increased sorption in fertil-
ized soils. However, in both fertilized and unfertilized soils, soil P desorption was
largest at low pH. Barrow offered two explanations for the increased P desorption
at low pH. First, acidity may have decomposed some of the P-retaining material
in soils. Second, desorption involved diffusion of P from the interior of the parti-
cles back to the surface and this diffusion probably becomes more rapid as the pH
decreases and the surface charge becomes less negative. Decreases in soil pH can
also enhance the release of P from precipitated calcium phosphates in neutral to
slightly acidic soils (White, 1983).
E. PLANT UPTAKE OF SPARINGLY SOLUBLE PHOSPHATE
There are three main mechanisms by which plants may be able to extract P from
normally insoluble sources in the soil. First, acid production (of plant or microbial
origin) may cause a lowering of rhizosphere pH. Second, organic acids capable of
chelating metal ions may compete with P for adsorption sites in the soil. Third, or-
ganic acids may form soluble complexes with metal ions associated with insolu-
ble P (Ca2+, Al3+, and Fe3+) thus releasing P (Kepert et al., 1979).
Hedley et al. (1982) demonstrated that the rhizosphere pH of rape (a plant which
is known to be efficient in absorbing P from P-deficient soils) grown in thin lay-
ers of 32P-labeled soil decreased from 6.1 to 4.1 in 28 days, leading to the disso-
lution of acid-soluble forms of inorganic P which were not exchangeable with 32P.
After 41 days, an extra 3.2 mmol P kg1 soil was released due to the decrease in
soil pH. The acidification was reported to be mainly due to H+ release from the
roots during periods when cation uptake by the plant exceeded that of anions and
not to organic acid production. Trolldenier (1992) also provided evidence for the
solubilization of calcium phosphates due to acidification of the rhizosphere by a
variety of plant species grown in agar. Hinsinger and Gilkes (1995) reported root-
induced dissolution of rock phosphate by two species of lupin and concluded that
106 M. A. WHITELAW
dissolution was probably due to proton excretion, as evidenced by a decrease in

the rhizosphere pH of about two pH units.
Plants which have the dense proliferation of rootlets arising from their lateral
roots known as “proteoid roots” are notable for their ability to grow on soil low in
available P (Grierson and Attiwill, 1989). Grierson and Attiwill reported that
Banksia proteoid roots were able to solubilize amorphous ferric phosphate. Grier-
son (1992) later demonstrated that large amounts of organic acids, of which 50%
was citrate, 18% malate, and 17% aconitic, were excreted by these proteoid roots.
Gardner et al. (1982) reported that both ferric and aluminium phosphates were dis-
solved in agar films around the proteoid roots of lupin plants and suggested that
chelation of the Fe3+ and Al3+ by a substance excreted by the roots was responsi-
ble. Gardner et al. (1983) later demonstrated that large amounts of H+ and organ-
ic acids including citric acid were excreted by these proteoid roots. They suggest-
ed that the excreted citrate reacts with ferric phosphate in the soil, allowing the
eventual release of Fe3+ from ferric phosphate and enabling the absorption of P by
the roots.
Armstrong et al. (1993) found that the availability to maize of various sparing-
ly soluble P sources compared to KH2PO4 (100%) was 53% for amorphous alu-
minium phosphate, 39% for amorphous ferric phosphate, 3% for AlPO4(cr), and
2% for FePO4(cr). These results demonstrate the uptake of sparingly soluble
sources of mineral P by some plants, although the mechanisms for solubilization
is not known. Sainz and Arines (1988), in a sequential P extraction study of three
acidic soils, reported that applied P was mainly retained in the inorganic P frac-
tions extracted with NH4F (i.e., P bound to Al3+) and NaOH [i.e., P bound to
Fe(III)] and that the P bound to Al3+ was the fraction mainly utilized by red clover
plants which were infected by native VAM fungi.
III. PHOSPHATE-SOLUBILIZING
SOIL MICROORGANISMS
The existence of soil microbes capable of transforming soil P to forms available

to the plant has been recorded by many investigators. Many bacterial, fungal,
yeast, and actinomycete species capable of solubilizing sparingly soluble P in pure
culture have been isolated from soil and rhizosphere samples (Kucey et al., 1989).
Gerretsen (1948) provided evidence that microorganisms in the rhizosphere can
dissolve sparingly soluble inorganic P. It has since been observed by many inves-
tigators that a high proportion of P-solubilizing microorganisms are concentrated
in the rhizosphere of plants (Sperber, 1957; Raghu and MacRae, 1966; Khan and
Bhatnagar, 1977; Whipps and Lynch, 1986; Martinez Cruz et al., 1990). Although
Katznelson and Bose (1959) did not find selective stimulation of P-solubilizing
microorganisms in the rhizosphere, they did find that rhizosphere bacteria had
greater metabolic activity and suggested that they might contribute significantly to
the phosphate economy of the plant.
Microorganisms in the rhizosphere obtain their nutrition from root exudates
(low-molecular-weight compounds such as sugars, organic acids, and amino acids
which leak from root cells), plant mucigel (an actively secreted substance), and
root lysates (compounds released by autolysis of root cortical cells, older epider-
mal cells, and root hairs) (Martin, 1977; Rovira et al., 1979). Whipps (1984)
showed that in wheat plants up to 33–40% of the total carbon fixed as photosyn-
thate could be excreted into the rhizosphere. Rhizosphere microorganisms are nor-
mal heterotrophs but they live in an environment with high levels of nutrients, such
as carbon and nitrogen, and tend to adapt rapidly to improvements of nutrient sup-
ply (Tinker, 1984).
A. SOLUBILIZATION OF PHOSPHATES BY FREE-LIVING FUNGI
Sperber (1958a) reported that many P-solubilizing bacteria lost the ability to sol-
ubilize P on serial subculturing. However, Kucey (1983) found that fungi, in con-
trast to bacteria, retained their P-solubilizing activity even after serial subcultur-
ing and could be kept actively solubilizing P for many years. P-solubilizing fungi
also showed greater P-solubilizing activity both on precipitated phosphate agar
and in liquid media than did bacteria in studies by Sperber (1958b), Kucey (1983),
Venkateswarlu et al. (1984), Darmwal et al. (1989), Chabot et al. (1993), and Na-
has (1996). Scanning electron microscopy by Chabot et al. (1993) showed that in
liquid culture the hyphae of fungi were attached to P mineral particles, whereas
bacteria were not. Fungi in soil are able to traverse distances more easily than bac-
teria and thus may be more important to P solubilization in soils (Kucey, 1983). P-
solubilizing fungi in the rhizosphere or bulk soil were often found to be predomi-
nantly Penicillium and Aspergillus spp. (Katznelson et al., 1962; Ramos et al.,
1968; Bardiya and Gaur, 1974; Khan and Bhatnagar, 1977; Banik and Dey, 1981a;
Kucey, 1983; Molla et al., 1984; Thomas et al., 1985; Darmwal et al., 1989; Mar-
tinez Cruz et al., 1990; Narsian et al., 1994).
B. PRECIPITATED PHOSPHATE AGAR STUDIES
Solubilization of various forms of precipitated calcium phosphate in unbuffered

solid agar medium plates has been used widely as the initial criterion for the iso-
lation of P-solubilizing microorganisms (Sackett et al., 1908; Pikovskaia, 1948;
Gerretsen, 1948; Sperber, 1957; Katznelson and Bose, 1959). Microorganisms
grown on precipitated clacium phosphate agar produce clear zones around their
108 M. A. WHITELAW
colonies if they are capable of solubilizing calcium phosphate minerals. Other

forms of sparingly soluble P, such as ferric or aluminium phosphate (Rose, 1957;
Sperber, 1957; Louw and Webley, 1959; Barthakur, 1978; Banik and Dey, 1983;
Martinez Cruz et al., 1990; Toro et al., 1996), have also been included in precipi-
tated phosphate agar for the isolation of P-solubilizing microorganisms. Rock
phosphate suspended in unbuffered solid agar plates has also been used for the iso-
lation of microorganisms capable of solubilizing rock phosphate (Ahmad and Jha,
1968; Bardiya and Gaur, 1974; Khan and Bhatnagar, 1977; Mba, 1996; Singh et
al., 1984; Toro et al., 1996).
The precipitated phosphate agar techniques are useful for isolating and select-
ing microorganisms for further investigations but have limited sensitivity. Factors
such as the rate of diffusion of excreted organic acids and the colony growth rate
affect the size of the clear zone. Cunningham and Kuiack (1992) reported that both
the nutrient composition (e.g., carbon source and nitrogen source) and the buffer-
ing capacity of the medium influenced the diameter of the zone of calcium phos-
phate solubilization. The day on which the clear zone is measured was also im-
portant: Some bacterial isolates were reported to show a clear zone early in growth
but later failed to keep the clear zone beyond their colony edge (Kucey, 1983).
However, Kucey (1983) found that fungal isolates which expressed P-solubilizing
ability early in colony growth continued to maintain a clear zone even when the
colonies were growing rapidly and that there was a significant correlation (r
0.70) between the ability of 23 fungal isolates to produce a clear zone on prec-
ipitated phosphate agar and their ability to solubilize rock phosphate in a liquid
medium.
Some investigators reported low correlation between the size of the clear zone
in precipitated phosphate agar studies and the more quantitative data from P sol-
ubilization in liquid media. For example, some isolates with little or no clear zone
on solid agar exhibited high efficiency for dissolving insoluble phosphates in a
liquid medium showing that the plate technique was insufficient to detect all P-
solubilizing microorganisms (Rose, 1957; Louw and Webley, 1959; Das, 1963;
Ahmad and Jha, 1968; Gupta et al., 1994). Conversely, some isolates showed
large clear zones on agar but low P solubilization in a liquid medium (Ahmad and
Jha, 1968). Gupta et al. (1994) modified the precipitated phosphate agar tech-
nique by including bromophenol blue, an acid indicator. Cunningham and Kuiack
(1992) made a similar modification to the precipitated phosphate agar technique
by including “alizarin red S” as the acid indicator. Gupta et al. (1994) recorded
that many fungi which were unable to produce a clear zone on calcium phosphate
agar produced yellow acidic zones on agar which contained bromophenol blue.
These fungi were able to solubilize Ca3(PO4)2 in a liquid medium with acid pro-
duction and the authors found a significant correlation between the yellow acidic
zone diameter on agar and the amount of P solubilized in the liquid medium
(r 0.69).
IV. LIQUID MEDIUM STUDIES
Many researchers have quantitatively investigated the ability of soil fungi to sol-
ubilize sparingly soluble inorganic phosphates in unbuffered pure liquid medium
cultures (Table I). A wide range of media have been used to allow the study of nu-
tritional effects, and levels of solubilized phosphate and metabolites such as or-
ganic acids have also been measured. Fungal strains differed widely in their abil-
ities to solubilize P.
P solubilization was often higher when the initial insoluble P levels were high,
as long as the volume of liquid medium was adequate for dissolution. The amounts
of insoluble phosphate P initially included in the medium varied from 87 mg P
liter1 (Illmer and Schinner, 1992) to 4460 mg P liter1 (Cunningham and Kuiack,
1992). Low levels of P solubilization by Aspergillus candidus and Aspergillus fu-
migatus were reported by Banik and Dey (1982) despite high initial insoluble P
levels. In this case, the dissolution of P may have been limited by the low volume
(15 ml) of liquid medium used (Table I).
The initial pH of the liquid medium probably affected the amount of P solubi-
lized in many liquid medium studies. Ahmad and Jha (1968) found that P solubi-
lization was highest when the initial pH was in the optimum pH range for the
growth of the isolates (e.g., a Penicillium isolate was most active when the initial
pH was 4.0). P solubilization was also affected by whether or not the medium was
shaken. Oxygen absorption rates in cultures are markedly increased by shaking
(Corman et al., 1957). Cunningham and Kuiack (1992) reported that the produc-
tion of citric acid by the P-solubilizing fungus Penicillium bilaii was promoted by
shaking, and Ahmad and Jha (1968) found that P solubilization by a Penicillium
isolate was greater when the culture was shaken (Table I).
The presence of alkaline substances in liquid media was detrimental to P solu-
bilization in some cases. Sperber (1958b) found that the presence of calcium car-
bonate greatly reduced solubilization of apatite by lactic acid, and Ahmad and Jha
(1968) found that the presence of calcium carbonate greatly reduced microbial sol-
ubilization of rock phosphate. Banik and Dey (1982) included calcium carbonate
in their culture medium for the microbial solubilization of aluminium and ferric
phosphates in an attempt to avoid the partial hydrolysis of AlPO4 and FePO4 dur-
ing autoclaving. Because the calcium carbonate would have neutralized any acid
produced, the solubilization of the aluminium and ferric phosphates was negligi-
ble (Table I).
A. SOLUBILITY OF PHOSPHATE COMPOUNDS
Another factor affecting the amount of P solubilized in liquid medium studies

is the solubility of the P mineral. The solubility of sparingly soluble P compounds
Table I
Phosphate Solubilization by Fungi
Phosphate Titratable acidity

solubilized Organic (mmol H N source and
(mg P liter1 acids liter1) and sugar concentration
Microorganism culture)a Phosphate source produced change in pH in medium (% w/v) Reference
Aspergillus niger (iso- 1000e (14 days) 2000 mg P liter1 Oxalate, citrate, ND 0.2% asparagine, 1% Rose (1957)
lated from New CaHPO4b aspartate (PC) glucose (temperature
Zealand, pH unknown) unknown, static)
1000e (14 days) 2000 mg P liter1 Oxalate, citrate ND
Ca3(PO4)2c (volume (PC)
unknown
100e (14 days) 2000 mg P liter1 Oxalate (PC) ND
FePO4d (volume
unknown)
Aspergillus terreus 1000e (14 days) 2000 mg P liter1 Oxalate and H2S ND
(isolated from soil) FePO4d production (PC)
Sclerotium rolfsii 1000e (14 days) 2000 mg P liter1 Oxalate (PC) ND
110
(isolated from soil) CaHPO4b (volume

unknown)
1000e (14 days) 2000 mg P liter1 Oxalate (PC) ND
Ca3(PO4)2c
400e (14 days) 2000 mg P liter1 Glyoxylate, citrate ND
FePO4d (vol- (PC)
ume unknown)
A. niger (isolated from 460e (max. at 460 mg P liter1 Citrate, glycollate, TA 55 0.05% yeast extract, Sperber (1958b)
soil or rhizoshere, 3 days) CaHPO4b in 50 ml succinate, gluconate 1% glucose (25°C,
Australia (PC) shaken)
Penicillium sp. (isolated 340e (max. at 460 mg P liter1 Lactate, glycollate (PC) TA 25
from soil or rhizo- 12 days) CaHPO4b in 50 ml
sphere)
Penicillium sp. (from 599 (time un- 1000 mg P liter1 ND ND 0.01% N (from NH4), Subba-Rao and
surface of legume known) Ca3(PO4)2 in 50 ml 0.05% yeast extract, Bajpai (1965)
root nodule, India) 1% glucose (tempera-
ture unknown)
Cephalosporium sp. 55 1000 mg P liter1 ND ND
Ca3(PO4)2 in 50 ml
Alternaria sp. 62 1000 mg P liter1 ND ND
Ca3(PO4)2 in 50 ml
Penicillium lilacinum 78 (21 days) 200 mg P liter1 ND pH 6.7 to 4.7 0.1% N (from NO3 ), Chhonkar and
(isolated from legume Ca3(PO4)2 in 50 ml 5% sucrose (28°C, Subba-Rao
root nodules, India) static) (1967)
A. niger (isolated from 50 (21 days) 200 mg P liter1 ND pH 6.7 to 1.8
legume root nodules, Ca3(PO4)2 in 50 ml
India)
Aspergillus sp. (isolated 35 (21 days) 200 mg P liter1 ND pH 6.7 to 5.5
from legume root Ca3(PO4)2 in 50 ml
nodules, India)
Aspergillus flavus (iso- 35 (21 days) 200 mg P liter1 ND pH 6.7 to 6.8
lated from legume Ca3(PO4)2 in 50 ml
root nodules, India)
A. terreus (isolated 18 (21 days) 200 mg P liter1 ND pH 6.7 to 8.6
from legume root Ca3(PO4)2 in 50 ml
nodules, India)
Penicillium sp. (Siwan) 260e (14 days) 1000 mg P liter1 ND ND 0.5% peptone, 1% su- Ahmad and Jha
(isolated from soil, (volume unknown) crose (room temper- (1968)
Bihar, India) ature, shaken)
Penicillium sp. (isolated 103 (21 days) 200 mg P liter1 ND pH 6.6 to 6.4 0.1% N (from NO3 ), Sethi and
from soil, Delhi and Ca3(PO4)2 in 50 ml 5% sucrose (28°C, Subba-Rao
Ludhiana, India) static) (1968)
Aspergillus sp. (isolated 99 (21 days) 200 mg P liter1 ND pH 6.6 to 4.0
111
from soil, Delhi and Ca3(PO4)2c in 50 ml

Ludhiana, India)
Fusarium sp. (isolated 85 (21 days) 200 mg P liter1 ND pH 6.6 to 5.1
from soil, Delhi and Ca3(PO4)2 in 50 ml
Ludhiana, India)
A. niger 175 (20 days) 1000 mg P liter1 ND pH 7.0 to 4.1 0.1% N (from NO3 ), Agnihotri (1970)
Ca3(PO4)2 in 50 ml 3% glucose (25°C (all pure
static) culture of spe-
cies commonly
found in forest
beds, Canada)
Aspergillus ustus 157 (20 days) 1000 mg P liter1 ND pH 7.0 to 6.0
Ca3(PO4)2 in 50 ml
Fusarium solani 159 (20 days) 1000 mg P liter1 ND pH 7.0 to 4.7
Ca3(PO4)2 in 50 ml
Mortierella nana 156 (20 days) 1000 mg P liter1 ND pH 7.0 to 5.1
Ca3(PO4)2 in 50 ml
S. rolfsii 139 (20 days) 1000 mg P liter1 ND pH 7.0 to 3.2
Ca3(PO4)2 in 50 ml
A. niger 162 (20 days) 200 mg P liter1 ND pH 7.0 to 5.6
fluoroapatite in 50 ml
continues
Table I—Continued

1
Cylindrocladium sp. 156 (20 days) 200 mg P liter ND pH 7.0 to 5.9
Penicillium sp. 131 (20 days) 200 mg P liter1 ND pH 7.0 to 5.9
Verticillium alboatrum 132 (20 days) 200 mg P liter1 ND pH 7.0 to 6.2
hydroxyapatite in
50 ml
hydroxyapatite in
112
50 ml
Aspergillus awamori 750e (7 days) 874 mg P liter1 ND pH 7.0 to 4.4 0.01% NH4, 0.05% Gaur (1972)
Ca3(PO4)2 in 100 ml yeast extract, 1% glu-
cose (temperature
unknown)
Penicillium digitatum 579e (7 days) 874 mg P liter1 ND pH 7.0 to 4.7
Ca3(PO4)2 in 100 ml
Aspergillus carbonarum 143e (7 days) 874 mg P liter1 ND pH 6.5 to 2.7 0.01% N (from NH4) Gaur et al. (1973)
(isolated from RP, India) Ca3(PO4)2 in 100 ml 0.05% yeast extract,
A. flavus (isolated from 89e (7 days) 874 mg P liter1 ND pH 6.5 to 6.1 1% glucose
RP, India) Ca3(PO4)2 in 100 ml (30°C, static)
Aspergillus wentii (isolated 82e (7 days) 874 mg P liter1 ND pH 6.5 to 6.4
from Delhi alluvial Ca3(PO4)2 in 100 ml
soil, India)
Aspergillus fumigatus f 57e (7 days) 874 mg P liter1 ND pH 6.5 to 7.1
(isolated from Delhi Ca3(PO4)2 in 100 ml
alluvial soil, India)
A. awamori 116e (21 days) 874 mg P liter1 ND pH 6.5 to 3.4 0.01% N (from NH4), Bardiya and Gaur
(isolated from rhizosphere Mussoorie RP in 0.05% yeast extract, (1974)
of soybean) 100 ml 1% glucose (30°C
static)
P. digitatum 77e (21 days) 874 mg P liter1 ND pH 6.5 to 4.0
(isolated from rhizo- Mussoorie RP in
sphere of soybean) 100 ml
Curvularia sp. (1F) 48e (21 days) 874 mg P liter1 ND pH 6.5 to 5.3
sphere of cowpea) 100 ml
Trichoderma sp. (26F) 45e (21 days) 874 mg P liter1 ND pH 6.5 to 5.4
sphere of black gram) 100 ml
A. niger (isolated from 298e (7 days) 1000 mg P liter1 ND pH 7.0 to 3.7 0.2% asparagine, 1% Khan and Bhat-
rhizosphere soil, Ca3(PO4)2 in glucose (28°C, nagar (1977)
Sindri, India. Soil 200 ml static)
available P: 6 ppm,
pH 6.4)
133e (7 days) 669 mg P liter1 ND pH 7.0 to 3.8
Jhabua RP in 20 ml
A. niger van Tiegh 46 (15 days) 100 mg P liter1 ND pH 4.8 to 3.2 0.01% N (from NH4), Barthakur (1978)
(isolated from rice Fe3(PO4)2 in 50 ml 1% glucose (33°C,
rhizosphere, Assam, shaken once every
India) 2 days)
113
67 (15 days) 100 mg P liter1 ND pH 5.2 to 2.8

FePO4 in 50 ml
A1PO4 in 50 ml
Ca3(PO4)2 in 50 ml
Trichoderma hamatum 33 (15 days) 100 mg P liter1 ND pH 6.0 to 3.6
A1PO4 in 50 ml
Ca3(PO4)2 in 50 ml
Mucor flavus 28 (15 days) 100 mg P liter1 ND pH 6.0 to 3.3
A1PO4 in 50 ml
Ca3(PO4)2 in 50 ml
F. solani 42 (15 days) 100 mg P liter1 ND pH 6.6 to 4.8
Ca3(PO4)2 in 50 ml
Fusarium oxysporum 20 (15 days) 100 mg P liter1 ND pH 5.2 to 3.0
FePO4 in 50 ml
F. oxysporum 39 (15 days) 100 mg P liter1 ND ph 6.6 to 5.6
Ca3(PO4)2 in 50 ml
Penicillium simplicissimum 37 (15 days) 100 mg P liter1 ND ph 6.6 to 3.4
Ca3(PO4)2 in 50 ml
continues
Table I—Continued

A. terreus 16 (15 days) 100 mg P liter1 ND pH 6.0 to 3.6

AlPO4 in 50 ml
39 (15 days) 100 mg P liter1 ND pH 6.6 to 5.3 0.01% N (from NH4), 1%
Ca3(PO4)2 in 50 ml glucose (33°C, shaken
15 (15 days) 100 mg P liter1 ND pH 5.2 to 3.0 once every 2 days)
FePO4 in 50 ml
A. awamori 642 (21 days) 874 mg P liter1 ND pH 7.0 to 0.01% N (from NH4), Arora and Gaur
Ca3(PO4)2 in 50 ml between 2.9 0.05% yeast extract, 1% (1979)
and 3.6 glucose (30°C, static)
625 (21 days) 874 mg P liter1 ND pH 7.0 to
hydroxyapatite in between 2.9
50 ml and 3.6
Mussoorie RP in between 2.9
114
50 ml and 3.6
A. niger 534 (21 days) 874 mg P liter1 ND pH 7.0 to
Ca3(PO4)2 in 50 ml between 3.3
and 3.8
50 ml and 3.8
50 ml and 3.8
P. digitatum 399 (21 days) 874 mg P liter1 ND pH 7.0 to
Ca3(PO4)2 in 50 ml between 3.3
and 3.8
50 ml and 3.8
50 ml and 3.8
A. niger 150 (max. at Ca3(PO4)2 (amount ND pH from 4.5 0.05% N (from NO3 ), 0.5% Ortuno et al.
(Spain) 7 days) unknown) in 250 ml to 3.2 yeast extract, 3% sucrose (1979)
A. niger van Tieghem 4 (10 days) Ca3(PO4)2 (amount Unidentified organic ND 0.01% N (from NH4), Banik and Dey
(isolated from soil, West unknown) in 15 ml acid (PC) 0.05% yeast extract, 1% (1981a)
Bengal, India. Soil sucrose (37°C, static)
available P: 7 ppm,
pH 5.4)
Aspergillus candidus 20 (10 days) 1000 mg P liter1 Oxalate (1.2 mM ), ND 0.01% N (from NH4), Banik and Dey
(isolated from alluvial Ca3(PO4)2 in 15 ml tartate (0.7 mM ) 0.05% yeast extract, 1% (1982)
soil, West Bengal, (PC) sucrose (37°C, static)
India. Soil available
P: 7 ppm, pH 7.4)
2 (10 days) 1000 mg P liter1 ND ND Above with 5 mM
AlPO4 in 15 ml CaCO3
FePO4 in 15 ml CaCO3
2 (10 days) 1000 mg P liter1 ND ND Above without 5 mM
Mussoorie RP CaCO3
in 15 ml
A. fumigatus f (isolated 19 (10 days) 1000 mg P liter1 Oxalate (4.5 mM ), ND Above with 5 mM
from alluvial soil, Ca3(PO4)2 in 15 ml tartate (1.2 mM ), CaCO3
West Bengal, India. citrate (1.1 mM )
115
Soil available (PC)

P: 7 ppm, pH 7.4) 1 (10 days) 1000 mg P liter1 ND ND Above with 5 mM
2 (10 days) 1000 mg P liter1 ND ND Above without 5 mM
Mussoorie RP CaCO3
in 15 ml
A. awamori (isolated 441 (15 days) 1000 mg P liter1 ND pH 5.2 to 2.7 0.1% N (from NO3 ), Singh et al.
from soil or compost, Ca3(PO4)2 in 100 ml 3% glucose (static) (1982)
India)
P. digitatum 395 (15 days) 1000 mg P liter1 ND pH 5.2 to 2.7
Ca3(PO4)2 in 100 ml
Paecilomyces fusisporus 388 (15 days) 1000 mg P liter1 ND pH 5.2 to 3.3

Ca3(PO4)2 in 100 ml
Papulaspora mytiline 284 (15 days) 1000 mg P liter1 ND pH 5.2 to 3.5
Ca3(PO4)2 in 100 ml
Masoniella grisea 191 (15 days) 1000 mg P liter1 ND pH 5.2 to 4.0
Ca3(PO4)2 in 100 ml
continues
Table I—Continued

1
Penicillium sp. (isolated 3 (10 days) 1000 mg P liter Oxalate (trace) (PC) 0.01% N (from NH4), Banik and Dey
from Fluvaquent soil, Ca3(PO4)2 in 15 ml 0.05% yeast extract, 1% (1983)
West Bengal, India; sucrose (37°C, static)
pH unknown)
Fungus “93” (isolated 125e (8 days) 412 mg P liter1 ND ND 0.005% N (from NO3 ) and Kucey (1983)
from soil, Southern Idaho RP in 50 ml 0.01% N (from NH4),
Alberta, Canada) 1% glucose (20°C, shaken)
Fungus “98” (isolated 121e (8 days) 412 mg P liter1 ND ND
from soil, Southern Idaho RP in 50 ml
Alberta, Canada)
Fungus “96” (isolated 110e (8 days) 412 mg P liter1 ND ND
from soil, Southern Idaho RP in 50 ml
116
Alberta, Canada)
Aspergillus sp. (isolated 524e (max. at Unknown amount of ND pH 7.4 to 6.5 0.05% yeast extract, 1% Molla et al.
from soil and rye grass 11 days) CaHPO4b in 100 ml sucrose (30°C, static) (1984)
or wheat rhizosphere,
Bangladesh)
Penicillium sp. 513e (max. at Unknown amount of ND pH 7.4 to 6.7
11 days) CaHPO4b in 100 ml
A. awamori 4 (14 days) 1000 mg P liter1 ND pH 8.0 to 6.9 0.01% N (from NH4), Singh et al.
Mussoorie RP 0.05% yeast extract, 1% (1984)
in 50 ml glucose (30°C, shaken
10 h day1)
A. niger (isolated from 666e (max. at 1000 mg P liter1 Lactate, glycollate, citrate, pH 6.5 to 2.3 0.01% N (from NH4), Venkateswarlu
soil, Western Rajasthan, 15 days) Ca3(PO4)2 in 50 ml succinate (PC) 0.05% yeast extract, 1% et al. (1984)
India; pH unknown) glucose (28°C, static)
Pencillium pinophillum 440e (max. at 1000 mg P liter1 Lactate, glycollate, citrate, pH 6.5 to 3.7
(isolated from soil, 9 days) Ca3(PO4)2 in 50 ml succinate (PC)
Western Rajasthan, India;
pH unknown)
A. niger (isolated from 375e (max. at 875 mg P liter1 Lactate, glycollate, citrate, ND As above except 2% glucose
soil, Western Rajasthan, 12 days) RP in 50 ml succinate (PC)
India; pH unknown)
Cylindrocarpon obtusisporum 209e (7 days) 994 mg P liter1 ND pH 6.5 to 6.0 0.01% N (from NH4), Surange (1985)
(isolated from lateritic Ca3(PO4)2 in 50 ml 0.05% yeast extract, 1%
forest soil, Maharashtra, glucose (27°C, shaken)
India; pH unknown)
Spegazzinia tessarthra 209e (7 days) 994 mg P liter1 ND pH 6.5 to 5.4
(isolated from lateritic Ca3(PO4)2 in 50 ml
forest soil, Maharashtra,
India; pH unknown)
Beltraniella humicola 189e (7 days) 994 mg P liter1 ND pH 6.5 to 6.0
(isolated from lateritic Ca3(PO4)2 in 50 ml
forest soil, Maharashtra,
India; pH unknown)
Scopulariopsis brumptii 156e (7 days) 994 mg P liter1 ND pH 6.5 to 5.4
Ca3(PO4)2 in 50 ml
Phoma exigua 139e (7 days) 994 mg P liter1 ND pH 6.5 to 5.1
Ca3(PO4)2 in 50 ml
Curvularia lunata 120e (7 days) 994 mg P liter1 ND pH 6.5 to 5.5
Ca3(PO4)2 in 50 ml
Myrothecium roridum 120e (7 days) 994 mg P liter1 ND pH 6.5 to 5.5
Ca3(PO4)2 in 50 ml
117
Humicola fuscoatra 117e (7 days) 994 mg P liter1 ND pH 6.5 to 4.7

Ca3(PO4)2 in 50 ml
Robillarda sessilis 117e (7 days) 994 mg P liter1 ND pH 6.5 to 5.0
Ca3(PO4)2 in 50 ml
Gliomastix murorum 107e (7 days) 994 mg P liter1 ND pH 6.5 to 5.0

Ca3(PO4)2 in 50 ml
Syncephalastrum racemosum 97e (7 days) 994 mg P liter1 ND pH 6.5 to 5.7
Ca3(PO4)2 in 50 ml
Aspergillus sp. (isolated 148e (15 days) 200 mg P liter1 ND pH 7.0 to 2.5; 0.01% N (from NH4), Thomas et al.
from soil, coconut planta- Ca3(PO4)2 in 50 ml TA 50 0.05% yeast extract, 1% (1985)
tions, Kerala, India. Soil glucose (30°C, static)
available P: 7–11 mg P
kg1, pH 5.5–6.3)
Penicillium sp. (24) (isolated 138e (15 days) 200 mg P liter1 ND pH 7.0 to 2.9;
from soil, coconut planta- Ca3(PO4)2 in 50 ml TA 46
tions, Kerala, India. Soil
available P: 7–11 mg P
kg1, pH 5.5–6.3)
Penicillium bilaii 298 (max. within 206 mg P liter1 ND pH 7.2 to 3.7 0.01% N (from NH4) and Asea et al.
12 days) Idaho RP in 100 ml 0.01% N (from NO3 ), 1% (1988)
glucose (24°C, shaken)
continues
Table I—Continued

Penicillium fuscum 203 (max. within 206 mg P liter1 ND pH 7.2 to 4.1 Asea et al.
12 days) Idaho RP in 100 ml (1988)
P. bilaii 46 (max. within 206 mg P liter1 ND pH 7.2 to 4.0 As above except 0.01% N
12 days) Idaho RP in 100 ml from NO3 only
P. fuscum (Sopp) Biorge 7 (max. within 206 mg P liter1 ND pH 7.2 to 6.2 As above except 0.01% N
Sensu. 12 days) Idaho RP in 100 ml from N03 only
A. niger van Tieghem 76 (11 days) 86 mg P liter1 ND pH 7.0 to 3.9; 0.02% N (from NH4) and Cerezine et al.
fluoroapatite in 30 ml TA 17 0.02% N (from NO3 ), 1% (1988)
glucose 13.3 mM sodium
citrate (30°C, static)
60 (11 days) 86 mg P liter1 ND pH 7.0 to 2.5; As above except 0.02% N
fluoroapatite in 30 ml TA 40 from NH4 only
8 (11 days) 86 mg P liter1 ND pH 7.0 to 5.3; As above except 0.01% N
fluoroapatite in 30 ml TA 7 from NO3
A. niger (isolated from 108 (max. at 437 mg P liter1 ND pH 4.9 to 3.2 0.01% N (from NH4), Darmwal et al.
compost material, 7 days) Ca3(PO4)2 (volume 0.05% yeast extract, 1% (1989)
118
Faizabad, India) unknown) glucose (28°C, static)

89 (max. at 437 mg P liter1 ND pH 4.8 to 3.8
10 days) Mussoorie RP
(volume unknown)
A. niger (isolated from soil. 455e (14 days) 500 mg P liter1 ND pH 7.0 to 2.4 0.01% N (from NH4), Martinez Cruz
Soil available P: 18–37 ppm, FePO4 in 50 ml 0.05% yeast extract, 1% et al. (1990)
pH 6.2 or 7.1; or from glucose (30°C, static)
sugarcane rhizosphere,
Havana, Cuba)
A. niger van Tieghem 401e (13 days) 1640 mg P liter1 Araxá ND pH 4.2 to 3.2; N from vinasse 4.5° Brix Nahas et al.
(isolated from soil) fluoroapatite RP in TA 59 (% glucose and fructose (1990)
30 ml mixture) (30°C, static)
Penicillium sp. (isolated 41e (max. at 204 mg P liter1 Oxalate (15.9 mmol liter1) pH 7.0 before 0.01% N (from NH4), 1% Parks et al.
from garden soil) 4.5 hours) hydroxyapatite in and itaconate (HPLC) inoculation to glucose (28°C, shaken) (1990)
100 ml spent culture approx. 2.5 after
incubation
without
hydroxyapatite;
to 5.4 after
equilibration with
hydroxyapatite
36e (max. at 180 mg P liter1 Oxalate (15.9 mmol liter1) pH 7.0 before
4.5 hours) Ca3(PO4)2 in 100 ml and itaconate (HPLC) inoculation to
spent culture 2.5 after
incubation
without
Ca3(PO4)2; to
5.4 after
equilibration
with Ca3(PO4)2
⬇8e (max. at ⬇40 mg P liter1 iron ore Oxalate and itaconate pH 7.0 before
10 min.) concentrate in 100 ml (HPLC) inoculation to
spent culture 2.8 after
incubation in
absence of iron
ore; to 5.7 after
equilibration
with iron ore
P. bilaii 1686 (1 h) 4560 mg P liter1 Nil citrate (TLC) pH 4.5 to 5.9 0.05% N (from NH4), 2% Cunningham and
CaHPO4 in 100 ml after equili- sucrose (20°C, shaken) Kuiack (1992)
“fresh” (72-h) culture bration with
119
CaHPO4
2412 (1 h) 4560 mg P liter1 Citrate and oxalate pH 5.0 to 5.6 0.03% N (from NO3 ), 2%
CaHPO4 in 100 ml (amount unknown) after equili- sucrose (20°C, shaken)
“spent” (72-h) culture (TLC) bration with
CaHPO4
Penicillium sp. (isolated 65 (max. at 44 mg P liter1 hydroxy- Gluconate (0.6 mM) and pH 7.0 to ⬇3.7 0.01% N (from NH4) and Illmer and
from forest soils, 8 days) apatite and 43 mg traces of lactate and 0.01% N (from NO3 ), 0.2% Schinner
Austria. Soil available P liter1 CaHPO4 citrate (isotachophoresis) glucose, 0.2% sucrose; (1992)
P: 0.9–1.66 ppm, pH (volume unknown) spent culture, diluted 1:2
4.2–4.6) (30°C, shaken)
Aspergillus sp. (KAR021) 162 (22 days) 1000 mg P liter1 ND pH 4.7 to 2.7 0.01% N (from NH4), Goenadi and
(isolated from soils, AlPO4 in 25 ml 0.05% yeast extract, 1% Saraswati
Indonesia) glucose (25°C, shaken) (1993)
Penicillium sp. (KAR023) 64 (22 days) 1000 mg P liter1 ND pH 4.7 to 2.7
(isolated from soils, AlPO4 in 25 ml
Indonesia)
A. awamori (isolated 154e (max. at 500 mg P liter1 ND pH 7.0 to 4.4 0.01% N (from NH4), Narsian et al.
from soil, Bhavnagar, 12 days) Ca3(PO4)2 in 100 ml 0.05% yeast extract, 1% (1993)
India) glucose (28°C, static)
continues
Table I—Continued

A. awamori e
218 (max. at 500 mg P liter1 ND pH 7.0 to 2.2 Narsian et al.
9 days) CaHPO4 in 100 ml (1993)
175e (max. at 500 mg P liter1 ND pH 7.0 to 3.4
15 days) Fe3(PO4)2 in 100 ml
15 days) AlPO4 in 100 ml
Aspergillus tamarii 222e (7 days) 1000 mg P liter1 Oxalate (PC) pH 5.9 to 4.8 0.01% N (from NH4), Gupta et al.
Ca3(PO4)2 in 50 ml 0.05% yeast extract, 1% (1994)
glucose (30°C, static)
Aspergillus clavatus 192e (7 days) 1000 mg P liter1 ND pH 5.9 to 3.6
Ca3(PO4)2 in 50 ml
Penicillium nigricans 191e (7 days) 1000 mg P liter1 ND pH 5.9 to 4.9
Ca3(PO4)2 in 50 ml
Aspergillus foetidus 205e (7 days) 1000 mg P liter1 Oxalate (PC) pH 5.9 to 5.0
120
Ca3(PO4)2 in 50 ml
A. awamori 190e (7 days) 1000 mg P liter1 Oxalate (PC) pH 5.9 to 4.8
Ca3(PO4)2 in 50 ml
Aspergillus aculeatus 144e (7 days) 1000 mg P liter1 ND pH 5.9 to 4.4
Ca3(PO4)2 in 50 ml
Aspergillus terricola 142e (7 days) 1000 mg P liter1 Oxalate (PC) pH 5.9 to 2.5
Ca3(PO4)2 in 50 ml
A. terreus 139e (7 days) 1000 mg P liter1 ND pH 5.9 to 4.3
Ca3(PO4)2 in 50 ml
A. fumigatus 135e (7 days) 1000 mg P liter1 ND pH 5.9 to 3.1
Ca3(PO4)2 in 50 ml
A. candidus 95e (7 days) 1000 mg P liter1 ND pH 5.9 to 4.2
Ca3(PO4)2 in 50 ml
Aspergillus amstelodemi 86e (7 days) 1000 mg P liter1 Oxalate (PC) pH 5.9 to 3.2
Ca3(PO4)2 in 50 ml
Aspergillus fischeri 72e (7 days) 1000 mg P liter1 ND pH 5.9 to 4.6
Ca3(PO4)2 in 50 ml
A. aculeatus (isolated 471e (max.at 500 mg P liter1 ND pH 7.0 to 3.6 0.01% N (from NH4), Narsian et al.
from rhizosphere of 4 days) Ca3(PO4)2 in 100 ml 0.05% yeast extract, 1% (1994)
Gram, Bhavnagar, India) e
273 (max. at 500 mg P liter1 ND pH 7.0 to 3.2 sucrose (30°C, static,
5 days) CaHPO4 in 100 ml shaken every 12 h)
A. aculeatus 159e (max. at 500 mg P liter1 ND pH 7.0 to 2.0
A. aculeatus 101e (max. at 219 mg P liter1 Sonari ND pH 7.0 to 3.0
15 days) RP (India) in 100 ml
A. niger (isolated from 449e (max. at 500 mg P liter1 ND pH 7.0 to 3.5
rhizosphere soil, 5 days) Ca3(PO4)2 in 100 ml
Bhavnagar, India)
12 days) CaHPO4 in 100 ml
64e (max. at 219 mg P liter1 Sonari ND pH 7.0 to 2.8
15 days) RP (India) in 100 ml
Aspergillus japonicus 180 (7 days) 437 mg P liter1 Parulia Oxalate (PC) pH 8.0 to 3.6 0.01% N (from NH4), Singal et al.
(isolated from low-grade RP in 50 ml 0.05% yeast extract, 1% (1994)
RP, India) glucose (37°C, static)
121
211 (7 days) 437 mg P liter1 Jhabua Oxalate (PC) pH 8.0 to 3.8

RP in 50 ml
Aspergillus foetidus 214 (7 days) 437 mg P liter1 Sagar Oxalate (PC) pH 8.0 to 3.3
(isolated from low-grade RP in 50 ml
RP, India)
Penicillium aurantiogriseum 4e (max. time 154 mg P liter1 Nil organic acid (HPLC) pH 5.9 to 6.0 0.02% N (from NH4), Illmer and
(isolated from forest soil, unknown) AlPO4(cr) (volume 0.8% glucose, 0.8% Schinner
Austria) unknown) sucrose (1995a)
A. niger 55e (unknown 154 mg P liter1 Citrate (18 M ), oxalate pH 7.0 to 2.4 As above except 0.01% N
incubation AlPO4(cr) (volume (5 M), gluconate from NH4
time) unknown) (0.5 M) (HPLC)
A. niger (isolated from 100e (max. at 16 154 mg P liter1 Citrate (18 M ), oxalate pH 6.9 to 1.8 0.01% N (from NH4) Illmer et al.
forest soil, Austria) days but still AlPO4(cr) (volume (5 M), gluconate (0.5 0.8% glucose, 0.8% (1995)
increasing) unknown) M ) (HPLC) sucrose
A. aculeatus (isolated from 470e (max. at 500 mg P liter1 ND pH 7.0 to 4.0 0.01% N (from NH4), Narsian et al.
rhizosphere of Gram 2 days) Ca3(PO4)2 in 50 ml 0.05% yeast extract, 1% (1995)
(Cicer ariatenum), India) glucose (28°C, static)
209e (2 days) 500 mg P liter1 ND pH 7.0 to 4.4 As above except 0.01 % N
Ca3(PO4)2 in 50 ml from NO3
continues
Table I—Continued

A niger (NB2) 292 (10 days) 384 mg P liter1 RP Citrate (22 mM ) TA 72 (10 0.05% N (from NO3 ), 0.5% Vassilev et al.
fluoroapatite from (colorimetric methods) days); pH 6.5– yeast extract, 3% sucrose, (1995)
Morocco in 50 ml 7.0 to 3.0 (20 10% sugar beet waste
days) (30°C, shaken for 5 min
day1)
A. niger (isolated from 219e (9 days) 540 mg P liter1 ND TA 10; pH 7.0 Unknown Nahas (1996)
soil, Sao Paulo State, CaHPO4 # 2H2O to 5.4
Brazil)
47e (9 days) 520 mg P liter1 Patos ND TA 2; pH 7.0
122
de Minas (Brazil) RP to 6.2

Penicillium solitum 302e (9 days) 540 mg P liter1 ND TA 13; pH 7.0
(isolated from soil, CaHPO4 # 2H2O to 5.1
Sao Paulo State, Brazil)
Eupenicillium shearii 285e (9 days) 540 mg P liter1 ND TA 15; pH 7.0
Aspergillus ochraceus 200e (9 days) 540 mg P liter1 ND TA 15; pH 7.0
Penicillium implicatum 20e (9 days) 540 mg P liter1 ND TA 22; pH 7.0
Penicillium minioluteum 290e (9 days) 540 mg P liter1 ND TA 16; pH 7.0
Penicillium viridicatum 160e (9 days) 540 mg P liter1 ND TA 13; pH 7.0
Penicillium purpurogenum 319e (9 days) 540 mg P liter1 ND TA 12; pH 7.0
Penicillium variabile P16 279 (5th batch, 1790 mg P liter1 Gluconate (86 mM; 5th ND 0.05% N (from NO3 ), 8% Vassilev et al.
(Spain) 2 days) sedimentary RP in batch) (HPLC and glucose. The fungus was (1996)
50 ml enzymatic methods) immobilized on 0.3 cm3
347 (3rd batch, polyurethane sponge
2 days) cubes (28°C, shaken)
123
Penicillium radicum (isolated 475 (max. at 1000 mg P liter1 Gluconate (17 mM) (HPLC) pH 6.4 to 3.4; 0.01% N (from NH4), 3% Whitelaw et al.
from wheat rhizosphere, 6 days) CaHPO4 in 50 ml TA 29 glucose (25°C, shaken) (1999)
Australia)
360 (max. at 1000 mg P liter1 Gluconate (20 mM) (HPLC) pH 6.1 to 4.1;
14 days) Ca3(PO4)2 in 50 ml TA 23
14 days) amorphous aluminium TA 24
phosphateg in 50 ml
213 (max. after 1000 mg P liter1 Gluconate (31 mM) (HPLC) pH 6.1 to 4.5; 0.01% N (from NO3 ), 3%
31 days) Ca3(PO4)2 in 50 ml TA 20 glucose (25°C, shaken)
5 days) CaHPO4 in 50 ml TA 12
Note. Abbreviations used: HPLC, high-performance liquid chromatography; ND, not determined; PC, paper chromatography; TLC, thin-layer chromatography; RP, rock phosphate.
a Time
at which determination was made is given. When solubilized P was monitored over a period of time, only the maximum (max.) solubilized P is given.
bPrecipitated
in situ by method of Gerretson (1948): K2HPO4 CaCl2 r CaHPO4 2KCl.
cPrecipitated in situ by method of Rose (1957): 2K HPO 3CaCl r Ca (PO ) 6KCl.
3 4 2 3 4 2
dPrecipitated in situ by method of Rose (1957): K HPO FeCl r FePO 2KCl HCl.
2 4 3 4
eNo indication of statistical significance was given by these authors.
f Human pathogen.
gPrepared by method of Deming and Cate (1963).

124 M. A. WHITELAW
in water varies widely. A brief comparison of solubility equilibrium constants

(logKsp) for a range of common crystalline P compounds in pure water at 25C is
shown in Table II. The approximate water solubilities (i.e., the maximum amount
which can be solubilized in water) are shown.
The pH of the solvent plays a major role in the dissolution of P compounds.
Poorly soluble P compounds, such as variscite and strengite, can only be effec-
tively dissolved in water if the pH is amended. In aqueous solution, AlPO4 has a
minimum solubility at a pH between 6.0 and 6.5 and FePO4 has a minimum solu-
bility at a pH between 5.0 and 6.0. As the pH decreases further, the solubility of
these P compounds tends to increase. Most of the calcium phosphates have mini-
mum solubilities above pH 8 in water and their solubility increases as the pH de-
creases below pH 8. The solubility of some common calcium, ferric, and alumini-
um phosphates is represented in Fig. 3 (Stumm and Morgan, 1995).
The solubility of P compounds also depends on the degree of crystallinity and
particle size. Amorphous compounds do not have a well-organized crystal struc-
ture and are almost always more soluble than the corresponding crystalline solid.
Table II
Solubility of Common Crystalline Phosphate Compunds
Solubility
(in water,
Phosphate Equilibrium reaction for the Reference for 25°C)a
compound dissolution of the phosphate compound Log Ksp log Ksp (g/100 g)
Ca(H2PO4)2 Ca(H2PO4)2(s) B Ca2 2H2PO

4
1.14 Snoeyink and 18
(monocalcium Jenkins
phosphate or (1980)
“triple super-
phosphate”)
CaHPO4 (calcium CaHPO4(s) B Ca2 HPO42 6.6 Stumm and 0.14
monohydrogen Morgan
phosphate) (1995)
-Ca3(PO4)2 -Ca3(PO4)2(s) B 3Ca2 2PO43 24.0 Snoeyink and 0.02
(-calcium Jenkins
orthophosphate (1980)
Ca5(PO4)3OH Ca5(PO4)3OH(s) B 5Ca2 3PO43 OH 55.9 Snoeyink and b
(hydroxyapatite) Jenkins
(1980)
AlPO4 # 2H2O AlPO4 # 2H2O(s) B Al3 PO43 2H2O 21.0 Stumm and b
(variscite) Morgan
(1995)
FePO4 # 2H2O FePO4 # 2H2O(s) B Fe3 PO43 2H2O 26.0 Stumm and b
(strengite) Morgan
(1995)
aAylward and Findlay (1994).

b“Practically insoluble.”
Figure 3 Solubility as a function of pH for selected discrete phosphate compounds in water.

(Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters,” 3rd ed., W. Stumm and J. J.
Morgan. Copyright 1995 John Wiley & Sons. Reprinted by permission of John Wiley & Sons, Inc.)
Amorphous solids are not stable and will transform slowly to the stable crystalline
solid phase (Snoeyink and Jenkins, 1980). The initial reaction product of the re-
action of superphosphate in acidic soils (amorphous aluminium phosphate) is more
soluble than crystalline AlPO4#2H2O (Lindsay, 1979), and this is probably due to
the lower particle size and thus higher surface area of the amorphous compound.
The surface area of amorphous aluminium phosphate is 10.5 m2 g1 (Juo and El-
lis, 1968), whereas the surface area of crystalline AlPO4#2H2O is 1.54 m2 g1
(Taylor and Gurney, 1964). The equilibrium constant (logKsp) for the dissolution
of AlPO4#2H2O,
AlPO4#2H2O B Al3+ H2PO4 20H
is higher for the more soluble amorphous aluminium phosphate (logKsp 28.1;
Veith and Sposito, 1977) than for crystalline AlPO4#2H2O (logKsp 30.5;
Bache, 1963). Many investigators have selected CaHPO4 or CaHPO4#2H2O as the
“insoluble” source of P in agar and liquid medium studies or in pot and field stud-
ies of plant growth promotion by P-solubilizing microorganisms. These calcium
phosphate minerals, however, are poor models for soil P because they are not in-
soluble. CaHPO4 is a moderately effective fertilizer in acidic soils in higher rain-
fall areas. Although it is far less soluble than Ca(H2PO4)2#H2O in water, it is sol-
uble enough in a soil environment to allow plant growth. The consequence of using
relatively soluble forms of P when investigating plant growth promotion by P-sol-
126 M. A. WHITELAW
ubilizing microorganisms is that the ability of the P-solubilizing microorganisms

to dissolve soil P is potentially underestimated.
B. PRODUCTION OF ORGANIC ACIDS

IN LIQUID MEDIUM STUDIES
Laboratory studies, reviewed by Stevenson (1967), Kucey et al. (1989), and Bar-
Yosef (1991), have shown that the microbial solubilization of soil phosphates in
liquid medium studies has often been due to the excretion of organic acids. In many
studies the presence of organic acids in liquid culture filtrates was determined by
paper chromatography or thin-layer chromatography, but more modern techniques
such as high-performance liquid chromatography, isotachophoresis, and enzymat-
ic methods have been used by others to allow more accurate identification of un-
known organic acids (Rose, 1957; Sperber, 1958b; Banik and Dey, 1982, 1983;
Venkateswarlu et al., 1984; Parks et al., 1990; Cunningham and Kuiack, 1992;
Illmer and Schinner, 1992, 1995a; Gupta et al., 1994; Singal et al., 1994; Illmer et
al., 1995; Vassilev et al., 1995, 1996; Whitelaw et al., 1999) (Table I).
Production of oxalic acid as a major organic acid was reported by many inves-
tigators from Aspergillus amstelodemi, A. awamori, A. candidus, A. foetidus, A. fu-
migatus, A. japonicus, A. niger, A. tamarii, A. terreus, A. terricola, an unknown
species of Penicillium, and Sclerotium rolfsii (Rose, 1957; Banik and Dey, 1982,
1983; Parks et al., 1990; Gupta et al., 1994; Singal et al., 1994; Illmer and Schin-
ner, 1995a). Cunningham and Kuiack (1992) also reported that oxalate was a ma-
jor organic acid anion produced by P. bilaii when the N source was nitrate,
although it was not detected when the N source was ammonium (Table I). The pro-
duction of citric acid as a major organic acid by A. niger was reported by
Rose (1957), Sperber (1958b), Venkateswarlu et al., (1984), Illmer and Schinner
(1995a), and Vassilev et al. (1996). Cunningham and Kuiack (1992) also reported
that citrate was a major organic acid anion produced by P. bilaii when the N source
was nitrate, although, as with oxalic acid, citric acid was not detected when the N
source was ammonium (Table I). The production of lactic acid has been reported
from an isolate of Penicillium and from A. niger and Penicillium pinophillum,
whereas tartaric acid production was reported from A. candidus and A. fumigatus
(Sperber, 1958b; Banik and Dey, 1982; Venkateswarlu et al., 1984).
Gluconate was reported to be the principal organic acid anion formed during P
solubilization by Penicillium aurantiogriseum (Illmer and Schinner, 1992), P.
variabile (Vassilev et al., 1996), and P. radicum (Whitelaw et al., 1999), with
smaller quantities being produced by A. niger (Sperber, 1958b; Illmer and Schin-
ner, 1995a) (Table I). Petruccioli et al. (1994) demonstrated the production of large
quantities of glucose oxidase and gluconic acid by P. variabile immobilized on
polyurethane sponge. Phosphate limitation has been reported to favor gluconic
acid synthesis in A. niger (Berry, 1975). De La Torre et al. (1993), in a study of

biochemical weathering of stone by P. frequentans, reported that gluconic acid pro-
duced by the fungus released a large amount of Ca2+ from limestone and appre-
ciable amounts of Fe3+, Ca2+, and Al3+ from sandstone and granite. Robert and
Razzaghe-Karimi (1975) and Eckhardt (1980) also observed that gluconic acid
was able to induce significant cation release from biotite.
C. CHELATION OF CATIONS BY ORGANIC ACIDS
The solubilization of soil P by plant organic acids and the mechanisms by which
fungal organic acids solubilize phosphates in liquid media are reported to be due
to either the lowering of pH or the chelation of the cation bound to P. Chelation of
cations has been shown to be an important mechanism for P solubilization when
the organic acid structure is favorable. Chelation involves the formation of two or
more coordinate bonds between a molecule (the “ligand”) and a metal ion, there-
by creating a ring structure complex. Chelation by an organic acid ligand occurs
through oxygen containing hydroxyl and carboxyl groups and takes place only
when either five-membered rings or the less stable six-membered rings can be
formed (Albert and Serjeant, 1984).
Many investigators have recorded that organic acids were able to solubilize
more P than was solubilized by inorganic acid at the same pH with the difference
presumed to be primarily due to chelation. Johnston (1959) found that HCl at pH
1.0 was able to solubilize less ferric and aluminum phosphate than many organic
acids, including citric, oxalic, malic, and lactic acids, at higher pH. Kim et al.
(1997) reported that HCl was able to solubilize less P from hydroxyapatite than
citric acid and oxalic acid at the same pH. Cunningham and Kuiack (1992) ob-
served that the presence of citrate at pH 4.5 increased the abiotic solubilization of
CaHPO4 by 837 mg P liter1 compared to inorganic acid alone at the same pH,
indicating that chelation by citric acid was probably involved. Artificial acidifi-
cation of culture media with HCl also solubilized less P from insoluble calcium
phosphates than that solubilized by unknown species of both Penicillium and
Pseudomonas (Illmer and Schinner, 1992) and Enterobacter agglomerans (Kim et
al., 1997), suggesting that chelation by organic acids produced by the microor-
ganisms may have been involved. Whitelaw et al. (1999) observed that either glu-
conic acid alone or P. radicum inoculation alone was able to solubilize more amor-
phous aluminium phosphate than HCl at the same pH. Johnston (1954) also
reported that at a similar pH, gluconic acid was able to solubilize more P from
Ca3(PO4)2 than could be solubilized by HCl alone, indicating that chelation was
probably occurring.
Many studies have investigated the ability of organic acid anions to release P
from soil or the P-adsorbing components of soil. Citrate was able to release P from
Table III
Stability Constants for Al-Organic Acid Complexes
Organic acid Structure Log KAl Reference

H
冟
H%C%COOH
冟
HOOC%C%OH
冟
H%C%COOH
冟
H
Citric Tricarboxylic acid, one -hydroxyl group 9.6 Bar-Yosef (1991)

HO O
C
冟
C
O OH
Oxalic Dicarboxylic acid, no hydroxyl groups 7.3 Bar-Yosef (1991)
H
冟
HOOC%C%OH
冟
HO%C%COOH
冟
H
Tartaric Dicarboxylic acid, two -hydroxyl group 5.3 Bar-Yosef (1991)
H
冟
HOOC%C%OH
冟
H%C%COOH
冟
H
Malic Dicarboxylic acid, one -hydroxyl group 5.4 Bolan et al. (1994)
OH
冟
H%C%COOH
冟
H%C%H
冟
H
Lactic Monocarboxylic acid, one -hydroxyl group 2.4 Bolan et al. (1994)
COOH
冟
H%C%OH
冟
HO%C%H
冟
H%C%OH
冟
H%C%OH
冟
CH2OH
Gluconic Monocarboxylic acid, one -hydroxyl group 2.0 Motekaitis and
HO O Martell (1984)
C
冟
H%C%H
冟
H
Acetic Monocarboxylic acid, no hydroxyl groups 1.5 Bar-Yosef (1991)

HO O
C
冟
H
Formic Monocarboxylic acid, no hydroxyl groups 1.4 Bar-Yosef (1991)
kaolin, goethite, ferric phosphate, and aluminum phosphate (Swenson et al.,

1949); citrate, tartrate, oxalate, and malate were able to desorb P from kaolinite
and gibbsite (Nagarajah et al., 1970); citrate and tartrate reduced P sorption by soil
(Earl et al., 1979); and citrate, oxalate, and malate were able to desorb P from
acidic soils over a range of pH (Lopez-Hernandez et al., 1979). Nagarajah et al.
(1970) suggested that the decrease in P adsorption caused by the organic acid an-
ions was due to competition with P for adsorption sites which would be determined
by the stability of the ferric or aluminium–organic anion complex. The ability to
desorb P generally decreased with a decrease in the stability constants for the
Fe(III)– or Al–organic acid complex (logKA1 or logKFe) in the following order:
citrate oxalate malonate tartrate acetate. Other studies reported that log-
KA1 was a good indicator of the ability of organic acids to release P from soils (Fox
et al., 1990) and rock phosphate (Kpomblekou-A and Tabatabai, 1994). Bolan et
al. (1994) reported that the addition of organic acids caused the dissolution of soil
components such as ferric and aluminium oxides and thereby decreased P adsorp-
tion by soil. They suggested that dissolution of ferric and aluminium oxides was
caused by the complexation of these metal ions with the organic acids. The de-
creases in P adsorption increased with an increase in logKA1. Organic acids also
increased the dissolution of rock phosphate, with P solubilization increasing with
increasing logKA1 values. (LogKA1 values for a selection of organic acids com-
monly produced by soil microorganisms are included in Table III).
The extent to which an organic acid is capable of chelating metal cations is
greatly influenced by its molecular structure. Struthers and Sieling (1950) demon-
strated that the ability of organic acids to prevent the precipitation of P by Fe3+ and
Al3+ (and by inference the ability to chelate Al3+ and Fe3+) increased progressively
as the number of hydroxyl groups was increased in an otherwise unchanged mol-
ecule. Johnston (1956) and Johnston and Miller (1959) described the importance
of hydroxyl and carboxyl structures to the P-solubilizing ability of carboxylic
acids. Ca2+ was chelated most strongly by tricarboxylic acids such as citric acid,
less strongly by dicarboxylic acids such as malic and tartaric acids, and to a small-
er extent by some monobasic acids such as lactic acid, which had hydroxyl groups
adjacent to carboxylic groups (i.e., -hydroxy acid structures). Among the dicar-
boxylic acids, those with hydroxyl groups formed the strongest complexes and -
hydroxy compounds were more effective than compounds which had hydroxyl
groups located on the second carbon atom from the carboxylic groups (i.e., -hy-
droxy compounds). The ability to solubilize calcium phosphates was attributed
both to the increased acid strength of -hydroxy acids in comparison to unsubsti-
tuted acids and to the enhancement of Ca2+ chelation by organic acids with such
structures. Bolan et al. (1994) also found that the ability of organic acids to pre-
vent P adsorption in soil decreased in the following order: tricarboxylic acid di-
carboxylic acid monocarboxylic acid.
Hue et al. (1986) found that the Al3+ detoxifying capacities of organic acids (and
130 M. A. WHITELAW
by inference Al3+ chelating ability) were correlated with the relative positions of
hydroxyl and carboxylic groups on their main carbon chain. Many effective chela-
tors of Al3+ had -hydroxy acid structures, which favored the formation of stable
five-bond ring structures with Al3+. Kpomblekou-A and Tabatabai (1994) also ob-
served that aliphatic acids with hydroxyl and carboxyl groups in positions suitable
for the formation of complexes with metal cations were more effective than other
aliphatic or aromatic acids in releasing P from rock phosphate.
On the other hand, Moghimi and Tate (1978) demonstrated that the ability of -
ketogluconic acid to solubilize hydroxyapatite was due to its ability to lower the
pH and not to the chelation of calcium ions. The stability constant for the Ca2+ –
-ketogluconic acid complex (i.e., logKCa ) was found to be negligible. -Ke-
togluconic acid is produced by many P-solubilizing bacteria and is one of the
strongest monobasic carboxylic acids, having a low acid dissociation constant
(pKa) of 2.66.
Illmer and Schinner (1995a) suggested that solubilization of AlPO4 by A. niger
and Penicillium simplicissimum in a liquid medium was probably caused by aci-
dolysis (i.e., the lowering of pH) followed by complex formation by citrate, where-
as calcium phosphates such as CaHPO4#2H2O and hydroxyapatite were solubi-
lized by acidolysis alone. This suggestion is partly supported by the fact that the
stability constant for the Al3+ –citrate complex (logKA1 9.6; Bar-Yosef, 1991)
is higher than the constant for the Ca2+ –citrate complex (logKCa 3.5; Martell
and Smith, 1977).
D. TITRATABLE ACIDITY
Titratable acidity (TA), determined by titrating the culture filtrate with NaOH
solutions using phenolphthalein indicator, was included in many P-solubilization
studies as a measure of the concentration of total dissociated plus undissociated
H+ (Table I). Peaks in TA and organic acid concentration often coincided with
peaks in soluble P, suggesting correlations between these variables. High correla-
tions between solubilized P and TA were reported by Thomas et al. (1985), Nahas
(1996), Vassilev et al. (1996), and Whitelaw et al. (1999), with r values of 0.57,
0.83, 0.81, and 0.91, respectively, indicating that acid production was the key P-
solubilization mechanism in many cases.
E. pH
In most cases, acidification (i.e., pH decrease) was a major P-solubilizing mech-

anism. High P solubilization was often associated with a low pH in the final cul-
ture solution and, conversely, low P solubilization was often associated with a high
pH of the culture solution pH (Table I). Some liquid medium studies compared the
changes in medium pH and P solubilization for many different soil fungi. Signif-
icant negative correlations for linear regression equations linking media pH and
solubilization of calcium phosphate minerals by a range of different soil microor-
ganisms were found by Venkateswarlu et al. (1984), Thomas et al. (1985), and
Illmer and Schinner (1992), with correlation coefficients (r) of 0.93, 0.75, and
0.49, respectively. A negative correlation between pH and P released by differ-
ent soil fungi was not found by some investigators (Sethi and Subba-Rao, 1968;
Gaur et al., 1973; Bardiya and Gaur, 1974; Kucey, 1983; Surange, 1985; Salih et
al., 1989; Nahas, 1996). Sometimes the culture filtrate pH was relatively high and
yet medium to high P solubilization occurred. Cylindrocarpon obtusisporum and
Beltraniella humicola solubilized 209 and 189 mg P liter1, respectively, after the
pH decreased from 6.5 to only 6.0 (Surange, 1985), and A. fumigatus increased the
culture filtrate pH from 6.5 to 7.1 while solubilizing 57 mg P liter1 (Gaur et al.,
1973). In the study by Bardiya and Gaur (1974), two fungal isolates caused the
culture pH to decrease to almost the same level (viz. A. awamori strain 18F and an
unidentified fungal isolate “21F” reached pH 3.4 and 3.3, respectively), whereas
the amounts of P solubilized from rock phosphate were quite different (viz. 116
and 48 mg P liter1, respectively), indicating that fungal P solubilization proba-
bly depends not only on the ultimate pH of the culture solution but also on other
factors such as the type of organic acids excreted (Table I).
Other liquid medium studies monitored changes in the pH and the amount of P
solubilized over time by a particular soil fungus. Ortuno et al. (1979) reported that
throughout the incubation period, the pH and the solubilization of calcium phos-
phate minerals by A. niger were negatively correlated (r 0.99, p 0.001).
Cerezine et al. (1988) also found a negative correlation between pH and the solu-
bilization of fluoroapatite by A. niger (r 0.45, p 0.01). Whitelaw et al.
(1999) found a positive correlation between the activity of H+ and P solubilized
from CaHPO4, Ca3(PO4)2, and amorphous aluminum phosphate. In contrast, Nar-
sian et al. (1995) reported that over an incubation period of 7 days, pH and solu-
bilization of Ca3(PO4)2 by A. aculeatus were not correlated. However, Ca3(PO4)2
solubilization by A. aculeatus reached a maximum, coinciding with a minimum in
pH, on the second day of incubation. After this time, the amount of P solubilized
decreased, fluctuating in a manner which did not appear to be related to pH. The
decrease in the soluble P concentration after the initial peak was probably due to
factors other than pH. The tendency for the relationship between pH and P solu-
bilization to become less significant after the initial peak in solubilized P was also
seen in P solubilization by P. aurantiogriseum (Illmer and Schinner, 1992), a Peni-
cillium isolate (Goenadi and Saraswati, 1993), and P. radicum (Whitelaw et al.,
1999).
132 M. A. WHITELAW
F. N SOURCE
The N source affected P solubilization in many liquid medium studies. Assim-

ilation of ammonium often favored higher acid production (i.e., lower pH and
higher TA) and greater solubilization of P in comparison to nitrate. Higher acid
production and P solubilization from ammonium compared to nitrate assimilation
has also been reported for the solubilization of fluoroapatite by A. niger (Cerezine
et al., 1988); rock phosphate by P. bilaii (Asea et al., 1988); Ca3(PO4)2 by A. ac-
uleatus (Narsian et al., 1995); and CaHPO4, Ca3(PO4)2, and colloidal aluminium
phosphate by P. radicum (Whitelaw et al., 1999) (Table I). Illmer et al. (1995) also
found that solubilization of AlPO4 by P. aurantiogriseum and P. simplicissimum
was more effective when small amounts of ammonium were included in the medi-
um compared to higher amounts of nitrate. However, Cunningham and Kuiack
(1992) reported greater P solubilization of CaHPO4 with nitrate than with ammo-
nium by a fresh P. bilaii culture (Table I).
Acid production in the form of H+ release in response to the assimilation of
cations such as ammonium is a well-known fungal phenomenon (Taha et al., 1969;
Banik and Dey, 1983; Kucey, 1983; Roos and Luckner, 1984; Asea et al., 1988).
The uptake of ammonium by fungi in a liquid medium commonly leads to a rapid
decrease in the pH of the medium (Cochrane, 1958). Roos and Luckner (1984)
found that an equal number of moles of H+ were produced by Penicillium cyclop-
ium per gram of mycelial growth from the same number of moles of ammonium
in a sucrose medium because of ammonium assimilation. In contrast, the initial
stages of nitrate assimilation produce an increase rather than a decrease in pH. The
reduction of nitrate to ammonia results in the internal release of OH or con-
sumption of H+ (Haynes, 1990). Nitrate-fed plant cells can cope with excess in-
ternal OH production by excreting OH or producing organic acid anions
(Haynes, 1990). The relatively high pH of culture filtrates in nitrate media found
by many investigators may be consistent with the excretion of either OH or or-
ganic acid anions in response to nitrate consumption.
G. FLUCTUATIONS IN SOLUBLE PHOSPHATE

LEVELS OVER TIME
In many liquid medium studies, soluble P concentrations, TA, and organic acid
concentrations increased and decreased several times during the incubation period.
Solubilization of P by A. niger (Cerezine et al., 1988; Nahas et al., 1990; Vassilev
et al., 1995), P. bilaii (Cunningham and Kuiack, 1992, P. aurantiogriseum (Illmer
and Schinner, 1992), A. aculeatus (Narsian et al., 1995), and P. radicum (Whitelaw
et al., 1999) followed a pattern whereby the amount of P solubilized reached a peak
after a number of days and then decreased. This indicates the importance of mon-
itoring microbial P solubilization at regular intervals over the entire incubation pe-
riod. Studies in which the soluble P concentration is determined at specific times
and not throughout the whole incubation period may not indicate the correct max-
imum amount of P solubilization. Illmer and Schinner (1992) suggested that the
rise and fall in P concentration observed when calcium phosphates were solu-
bilized by P. aurantiogriseum might be caused by the formation and secon-
dary solubilization of organic P complexes which might be used as energy or
nutrient sources are required by the fungus. A pattern of increases and decreases
in P concentration was also found by Illmer and Schinner (1995a) when a mixture
of CaHPO4#2H2O and hydroxyapatite was solubilized by a Pseudomonas sp. Be-
cause the decreases in soluble P occurred at the same time as the increases in the
mass of precipitated calcium phosphate, it was suggested that hydroxyapatite was
crystallized out of the solution once a threshold value in soluble P concentration
was reached. Parks et al. (1990) also reported that P solubilized from iron ore in-
creased to a maximum within 5 min after contact with spent medium containing
Penicillium sp. fungal metabolites but then decreased during the next 2 h. It was
suggested that this was due either to reprecipitation or to adsorption to ore parti-
cles or the vessel walls. Vassilev et al. (1995) suggested that the decrease in both
TA and P concentration observed during solubilization of rock phosphate by A.
niger on sugar beet waste was most likely a result of citric acid utilization by the
fungus under conditions of nutrient depletion which also caused sporulation of the
fungal culture. This explanation is in agreement with Cochrane (1958), who stat-
ed that with few exceptions acids which accumulate in fungal cultures are later uti-
lized for energy and growth.
V. PLANT GROWTH PROMOTION BY

The ability of P-solubilizing fungi to solubilize soil P compounds in vitro indi-

cates that inoculation by such fungi could enhance the utilization of sparingly sol-
uble soil phosphate by plants. Although it is unlikely that organic acids produced
in the rhizosphere would remain untouched for long enough to affect bulk P re-
lease from soil, it is possible that the short-term reduction of the rhizosphere pH
and the complexation of cations could produce an effective P-solubilizing mi-
croenvironment, resulting in increased P uptake by the plant roots (Kucey et al.,
1989). Much of the solubilized P would be quickly reprecipitated or readsorbed by
the soil particles, but plant roots might be able to take up some of the released P
before this occurs. P-solubilizing rhizosphere fungi would be supplied with nutri-
ents from the plant root (Martin, 1977; Rovira et al., 1979; Whipps, 1984; Rovi-
ra, 1991).
Table IV
Plant Growth Promotion and Enhancement of Soil P Availability by Phosphate-Solubilizing Fungi
Microorganism Soil type or growth mediuma Plant P source Yield increase (dry weight) (%) Increase in P (and N) uptake Reference
Aspergillus niger Hydroponics glasshouse experiment: American elm RP 600b,c (aboveground) 900%b,c (aboveground P) Rosendahl (1942)
nutrient solution, sand (Ulmus americana)
harvested at 3 months
Aspergillus awamori Greenhouse experiment: alluvial Wheat (Triticum Mussoorie 84 (grain) 83% (grain P) Gaur (1972)
soil (alkaline) from Delhi, India. aestivum Kalyan RP (219 kg
Soil available P: 9 mg kg1, pH sona) harvested at P ha1)
unknown maturity
Penicillium digitatum 80 (grain) 98% (grain P)
A. awamori (isolated Greenhouse experiment: red soil Groundnut (Arachis Nil P fertilizer ND 30%b (total plant P) Vidhyasekaran et al. (1973)
from soil, Tamil hypogaea L.) harvested
Nadu, India) at 30 days
Farmyard manure ND 383%b (total plant P)
(10 tons ha1)
A. awamori with Field trial: IARI, New Delhi, India Wheat (Sonalika and 44 kg P ha1 16 (straw) ND Gaur et al. (1980)
134
Pseudomonas striata HD2122) harvested Mussoorie RP 11 (grain)

at maturity and paddy straw
A. niger van Tieghen Flask culture experiment: typic No plant Nil P fertilizer ND 2 mg P kg1 (soil Banik and Dey (1981b)
(isolated from Ochragual soil, Kapgari, West available P)
Lateritic soil, India) Bengal, India. Soil available
P: 7 mg P kg1, pH 5.4
A. niger [isolated from Greenhouse experiment: sandy loam, Onion (Allium cepa var. Nil P fertilizer 24 (aboveground) 24% (shoot P) Manjunath et al. (1981)
rhizosphere of India. Soil available P: 3 mg chikkaballapur red.) NS (root) 51% more VAM spores
cowpea (Vignia P kg1, pH 5.6 62 (bulb fresh weight) in soil
unguiculata (L.)
alp., India]
A. niger with Bacillus 25 (aboveground) 25% (shoot P)
mobilis (N fixing 16 (roots) 44% (shoot N)
bacterium, isolated 88 (bulb fresh weight) 54% more VAM spores
from rhizosphere in soil
of cotton, India)
A. niger with Bacillus 57 (shoot) 48% (shoot P)
mobilis and Glomus 82 (roots) 71% (shoot N)
fasciculatus 155 (bulb fresh weight) 84% (root P)
125% more VAM sproes
in soil
Aspergillus fumigatusd Greenhouse experiment: partially No plant RP (30 kg P ha1) ND 3 kg P ha1 (soil Banik and Dey (1982)
[isolated from a sterilized Gangetic alluvial soil and farmyard available P)
Gangetic alluvial, (Fluvaquent) silty clay, West manure (40
(Fluvaquent) soil, Bengal, India. Soil available P: 7 kgN ha1)
West Bengal, India; mg P kg1, pH 7.4
pH 7.4]
Aspergillus candidus No plant RP (30 kg P ha1) ND NS (soil available P)
(isolated from a and farmyard
Gangetic alluvial manure (40
soil, West Bengal, kgN ha1)
India; pH 7.4)
A. awamori Field trial Soybean harvested at Nil P fertilizer NS (grain) 44% (grain P) Gaur (1985)
maturity
Penicillium bilaii Greenhouse experiment: brown Field beans (Phaseolus Idaho RP (45 mg 10.3 (aboveground) NS (aboveground P) Kucey (1987)
(isolated from soil, Chernozemic loamy sand from vulgaris) harvested P kg1)
Canada) Alberta, Canda. Soil sterilized and at maturity
allowed to reequilibrate for 6
weeks. Soil available P: 2 mg
kg1, pH 7.2
Field beans (P. vulgaris) Nil P fertilizer 53 (aboveground) 31% (aboveground P)
135
harvested at maturity
Wheat (Chester) harvested Idaho RP (45 mg 21 (aboveground) 9% decrease (aboveground P)
at maturity P kg1 soil)
Penicillium bilaii Wheat (Chester) harvested Nil P fertilizer 5 (aboveground) 3% (aboveground P)
at maturity
Field trial: orthic brown Chernozemic Wheat (Chester) harvested Straw and Idaho 25 (grain) 31% (aboveground P)
clay loam, Lethbridge, Canada. at maturity RP (20 kg 9 (straw)
Soil available P: 2 mg kg 1, P ha1)
pH unknown
Wheat (Chester) harvested Nil P fertilizer 27 (grain) 34% (aboveground P)
at maturity 17 (straw)
P. bilaii plus mixed Greenhouse experiment: brown Wheat (Chester) harvested Straw and Idaho 39 (aboveground) 46% (aboveground P)
culture of VAM Cherozemic loamy sand, Alberta, at maturity RP (45 mg
fungi Canada. Soil sterilized to kill native P kg1 soil)
VAM fungi. Soil available P: 2
mg kg1, pH 7.2
P. bilaii Greenhouse experiment: Orthic Wheat (Neepawa) Idaho RP (20 mg 28% (aboveground) 25% (aboveground P) Asea et al. (1988)
Brown Chernozem from Canada. harvested at the early P kg1 soil)
Soil available P: 3 mg kg1, heading stage
pH 8.0
continues
Table IV—Continued
P. bilaii Wheat (Neepawa) Nil P fertilizer 35% (aboveground) 27% (aboveground P)

harvested at the early
heading stage
Greenhouse experiment: dark brown Wheat (Chester) harvested Idaho RP (700 93 (aboveground) 47% (aboveground P) Kucey (1988)
Chernozemic clay loam soil, at 8 weeks mg P kg1 soil) 2 mg kg1 (soil
Lethbridge, Alberta, Canada. Soil available P)
available P: 4 mg kg1, pH 7.7 214% more P-solubilizing
fungi in the rhizosphere
Wheat (Chester) harvested Nil P fertilizer 86 (aboveground) 73% (aboveground P)
at 8 weeks 2 mg kg1 (soil available p)
283% more P-solubilizing
fungi in rhizosphere
Field trial: orthic dark brown Wheat (Chester) harvested RP (20 kg P ha1) 11 (grain) 8% (aboveground P)
Chernozemic clay loam soil, at maturity 4 (straw)
Lethbridge, Alberta, Canada.
136
Soil available P: 4 mg kg1,

pH 7.7
Wheat (Chester) harvested Nil P fertilizer 6 (straw) 15% (aboveground P)
at maturity 10 (grain)
P. bilaii Greenhouse experiment: brown Canola (Brassica napus Florida RP (20 NS (straw or pods) 36% (aboveground P) Kucey and Leggett (1989)
Chernozem soil (Loamy sand), L. “Westar”) harvested mg P kg1)
Canada. Soil available P: 2 mg at maturity
kg1, pH 7.2
P. bilaii Wheat (Chester) harvested MAP (20 mg NS (straw or pods) 19% (aboveground P)
at 8 weeks P kg1)
Wheat (Chester) harvested nil P fertilizer NS (straw or pods) 62% (aboveground P)
at 8 weeks
Penicillium sp. Greenhouse experiment: calcareous Sorghum (Sorghum bicolor RP from 11 (aboveground) 17% (aboveground P) Salih et al. (1989)
(isolated from soil (typic Torrifluvent) from Iraq. Moench) harvested at Akashatmine, 2 mg kg1 (soil
rhizosphere of Soil available P: 4 mg kg1, pH8.2 70 days Iraq (45 mg available P)
tomato, eggplant, P kg1)
or cucumber,
Baghdad, Iraq)
Wheat (Chester) harvested TSP (45 mg 16 (aboveground) 8% (aboveground P)
at 8 weeks P kg1) 9 mg kg1 (soil available P)
Aspergillus foetidus Greenhouse experiment: calcareous Sorghum (Moench) RP from 13 (aboveground) 19% (aboveground P)
(isolated from soil (typic Torrifluvent) from Iraq. harvested at 70 days Akashat mine, 1 mg kg1 (soil
rhizosphere of Soil available P: 4 mg kg1, pH 8.2 Iraq (45 mg available P)
tomato, eggplant, P kg1)
or cucumber,
Baghdad, Iraq)
Sorghum (Moench) TSP (45 mg 4 (aboveground) 4% (aboveground P)
harvested at 70 days P kg1) 5 mg kg1 (soil
available P)
A. awamori Field trial: alluvial loamy soil (Pura Rice (Saket-4) harvested RP (13 kg P ha1) 12 (grain) ND Tiwari et al. (1989)
Farm). Soil available P: 6 kg ha1, at maturity and TSP (13 kg
pH 7.5 P ha1)
Chickpea (T3) harvested RP (26 kg P ha1) 23 (rain) 110% more rhizobium
at maturity nodules
Wheat (HD-1553) RP (26 kg P ha1) NS (grain) ND
harvested at maturity
Field trial: moderate alkaline loamy Wheat (HD-1553) RP (26 kg P ha1) 14 (grain) ND
soil (Chakeri farm). Soil available harvested at maturity
P: 12 kg P ha1, pH 8.6
P. bilaii Pot trial: Shellbrook orthic gray-black Peas (Pisum sativum) Nil P fertilizer 22 (roots plus aboveground) NS (roots plus above- Downey and van Kessel
137
Chernozemic fine sandy loam from harvested at 43 days ground P) (1990)

Porcupine Plain, Saskatchewan,
Canada. Soil available P: 4 mg
kg1, pH 7.3
Peas harvested at 43 days TSP (44 kg NS (aboveground) NS (aboveground P)
P ha1)
P. bilaii and 17% decrease cf. Rhizobium
Rhizobium alone (roots plus above-
leguminosarum ground)
P. bilaii Field trials: 37 locations, 3 Canadian Wheat harvested at TSP (4.4 kg 2 (grain) ND Gleddie et al. (1991)
provinces, over 3 years. Low to maturity P ha1)
medium available P soils (0–22 kg
P ha1).
Wheat harvested at Nil P fertilizer 2 (grain) ND
maturity
P. bilaii Greenhouse experiment: brown Pea (Trapper) harvested Nil P fertilizer 48 (aboveground) 39% (aboveground P) Gleddie (1993)
Chernozemic soil (loamy sand) as first flower bud 55% (aboveground N)
from Saskatchewan, Canada. Soil opened at 41–51 days
available P: 13.5 mg kg1, pH 7.6
continues
P. bilaii Pea (Trapper) harvested TSP (10 mg 12 (aboveground) 18% (aboveground P)

as first flower bud P kg1) NS (aboveground N)
opened at 41–51 days
Field trials: four P-fertilizer responsive Pea (Trapper) harvested Nil P fertilizer NS (aboveground) NS (aboveground P)
locations, 1989 and 1991. Thin at 8 weeks after 7% (aboveground N)
black and black soil zones, Western emergence
Canada. Available soil P: “low to
medium levels”
Pea (Trapper) harvested TSP (4.4 kg 16 (aboveground) 18% (aboveground P)
at 8 weeks after P ha1) 19% (aboveground N)
emergence
Pea (Trapper) harvested TSP (8.7 kg NS (aboveground) 7% (aboveground)
at 8 weeks after P ha1) NS (aboveground N)
emergence
A. awamori Greenhouse experiment: Patharchatta Soybean (Glycine max, L. Nil P fertilizer 34 (straw) 58% (grain P) Singh and Singh (1993)
sandy loam (typic Hapludoll), merr.) (Bragg) harvested 44 (grain) 8 mg P kg1 (soil
India. Soil available P: 27 mg kg1, at maturity available P)
138
pH 6.2 40% (N uptake)

100% (nodule dry weight)
microbes in soil
Soybean (G. max, L. merr.) Mussoorie RP 10 (straw) 13% (grain P)
(Bragg) harvested at (86 mg P kg1) 5 (grain) 3 mg P kg1 (soil
maturity available P)
11% (N uptake)
NS (nodule dry weight)
microbes in soil
A. awamori and Nil P fertilizer 24 (grain) compared to 35% (grain P)
Bradyrhizobium sp. inoculation by Bradyrhizobium NS (soil available P)
sp. alone 21% (grain N)
NS (nodule dry weight)
A. awamori Field trial: alluvial loamy soil at Pura, Wheat (HD 1553) Nil P fertilizer NS (grain) ND Tiwari et al. (1993)
India. Soil available P: 6.3 kg ha1, harvested at maturity 12 (straw)
pH 7.5
Wheat (HD1553) Mussoorie RP NS (grain or straw) ND
harvested at maturity (26 kg P ha1)
Field trial: moderately alkaline soil Wheat (HD1553) Nil P fertilizer 8 (grain) ND
(loam), Chakeri, Kanpur (U.P.), harvested at maturity 7 (straw)
India. Soil available P: 12 kg ha1,
pH 8.6
Wheat (HD1553) Mussoorie RP NS (grain or straw) ND
harvested at maturity (26 kg P ha1)
Aspergillus sp. Greenhouse experiment: autoclaved No plant (15 days) Nil P fertilizer ND 14 mg P kg1 (soil Goenadi (1995)
(KAR0210) Sang Hyang Damar Ultisol soil. available P)
Soil available P: 0.5 mg P kg1
No plant (30 days) RP (300 mg ND 38 mg P kg1 (soil
P kg1) available P)
Aspergillus sp. Greenhouse experiment: Rajamandala No plant (14 days) RP (300 mg P ND 24 mg P kg1 (soil Goenadi et al. (1995)
(KAR0210) Ultisols soil. Soil available P: 37 kg kg1) available P)
ha1, pH 3.9
Penicillium glaucum Greenhouse experiment: red soil Sunflower (Helianthus SSP (82 mg P NS (plant height, 60 days) NS (shoot P at harvest) Gururaj and
(HE4) (isolated GKVK farm, Bangalore, India. annuus L.) (BSH-1) kg1 soil) NS (leaf area, 45 days) 47% more P-solubilizing Mallikarjunaiah (1995)
from sunflower 1% farmyard manure. Soil 35 (seed yield at harvest) fungi in the rhizosphere
rhizoshpere, available P: 7 mg kg1, pH
Bangalore, India) unknown
Penicillium Greenhouse experiment (no plant): No plant Nil P fertilizer ND 2 ng P ml1 soil solution Illmer and Schinner
aurantiogriseum soil from Austrian alps. Soil (soil available P) (No (1995b)
(isolated from available P: 0.6 or 0.7 g P dm1, significant difference when
forest soil, Austria) pH 4.2 or 4.6. Soil amended with soil not amended with
0.2% glucose, 0.2% sucrose, glucose, sucrose, or N)
139
0.002% N (from NH 4)
Paecilomyces Field trial: clayey soil, Junagadh, Groundnut (Arachis Nil P fertilizer NS (pod) 19% (P in kernel) Mehta et al. (1996)
fussiporus India. Soil available P: 31 kg P hypogaea) 24% (P in haulm)
ha1, pH 7.9 8% (nodule dry weight)
Aspergillus sp. Field trial: sandy loam Bangalore, Soybean (Hardee) RP (17.5 kg 7 (seed) NS (protein) Thimmegowda and
India. Soil available P: 14 kg harvested at maturity P ha1) 7 (oil) Devakumar (1996)
P ha1, pH 6.2
RP (35 kg P 5 (seed) NS (protein)
ha1) 8 (oil)
RP (17.5 kg 12 (seed) 7% (protein)
P ha1) and 12 (oil)
SSP (17.5 kg
P ha1)
A. awamori Greenhouse experiment: soil, India. Gram (JG 315) (legume) Nil P fertilizer 34 (aboveground) 36% (aboveground P) Vaishya et al. (1996)
Soil available P: 1.2 kg P ha1, harvested at flower- 206% (nodule dry weight)
pH 7.1 initiation stage
Gram (JG 315) (legume) SSP (20 kg 14 (aboveground) 24% (aboveground P)
harvested at flower- P ha1) 22% (nodule dry weight)
initiation stage
continues
A. awamori Gram (JG 315) (legume) SSP (40 kg 13 (aboveground) 52% (aboveground P)
initiation stage
Gram (JG 315) (legume) RP (17 kg NS (above ground) 38% (aboveground P)
initiation stage
Penicillium radicum Greenhouse experiment: red earth Wheat (Dollarbird) Nil P fertilizer 26 (grain) 23% (grain protein) Whitelaw et al. (1997)
(isolated from soil. Soil available P: 17 mg kg1 harvested at maturity 14 (aboveground) NS (grain P)
wheat rhizosphere, pH 4.6 at 20 weeks
Australia)
Wheat (Dollarbird) KH2PO4 (5 kg 10 (grain) NS (grain P)
harvested at maturity P ha1)
at 20 weeks
Wheat (Dollarbird) KH2PO4 (15 kg 15 (grain) 20% (grain P)
harvested at maturity P ha1) 15% (grain protein)
140
at 20 weeks
Field trial: red earth soil, Wagga Wheat (Dollarbird) Nil P fertilizer NS NS
Wagga, NSW, Australia. Soil harvested at maturity
available P: 16 mg kg1, pH 4.9 at 28 weeks
Wheat (Dollarbird) SSP (5 kg 18 (grain) 18% (grain protein)
at 28 weeks
Wheat (Dollarbird) SSP (15 kg 25 (grain) 33% (grain protein)
at 28 weeks
Wheat (Dollarbird) SSP (20 kg 16 (grain) NS
at 28 weeks
Note: Abbreviations used: MAP, monoammonium phosphate; ND, not determined; NS, not statistically different; RP, rock phosphate; SSP, single superphosphate [mixture of Ca(H2PO4)2 and CaSO4 produced by the
action of H2SO4 on RP]; TSP, triple superphosphate [Ca(H2PO4)2 “monocalcium phosphate” produced by action of H3PO4 on RP]; VAM, vesicular arbuscular mycorrhizal fungi.
a“Soil-available P” determined by extraction with NaHCO unless otherwise indicated.
3
bStatistical significance not given.
cCompared to sterile plants.
dHuman pathogen.
Growth promotion and increased P uptake by plants inoculated with P-solubi-

lizing fungi have been reported by many investigators (Table IV). Many of the
studies reported in Table IV have investigated the ability of P-solubilizing fungi
to promote P uptake and plant growth in soil under greenhouse conditions. Under
these conditions, rooting volumes are usually restricted so that if microbial P sol-
ubilization does take place, the plant response may be higher than that in field tri-
als (Kucey et al., 1989).
In an early sand and nutrient solution greenhouse study, inoculation of Ameri-
can elm with A. niger was able to increase the yield and P uptake by 600 and 900%,
respectively, but in this study inoculated plants were compared to plants in a ster-
ile medium (Rosendahl, 1942). Illmer and Schinner (1995b) point out that nearly
all rhizosphere microorganisms, not only those which solubilize P, increase the nu-
trient supply of plants. This means it is advisable to use nonsterile soil to ensure
that uninoculated control plants have adequate nutrition.
In soil greenhouse trials, yield and/or P uptake has been increased by inocula-
tion of wheat, onion, sorghum, soybean, and gram with P-solubilizing fungi (Gaur,
1972; Manjunath et al., 1981; Salih et al., 1989; Singh and Singh, 1993; Vaishya
et al., 1996; Whitelaw et al., 1999). Yield alone was also increased by inoculation
of sunflower and soybean (Gururaj and Mallikarjunaiah, 1995; Thimmegowda and
Devakumar, 1996). Penicillium bilaii (ATCC strain No. 20851) is commercially
available under the trade name Provide and has consistently increased grain yield
and P uptake by wheat grown on neutral or alkaline soils under greenhouse con-
ditions. Penicillium bilaii has also increased the yield and P uptake by field beans
and peas and the uptake of P by canola (Kucey, 1987, 1988; Asea et al., 1988;
Kucey and Leggett, 1989; Downey and Van Kessel, 1990; Gleddie, 1993) (Table
IV).
Growth promotion of plants by P-solubilizing fungi under field conditions has
also been reported by many investigators. P uptake was increased in groundnut in-
oculated with Paecilomyces fussiporus (Mehta et al., 1996), whereas both wheat
yields and P uptake were increased by inoculation with P. bilaii (Kucey, 1987,
1988). Wheat yield has been increased by inoculation with P. bilaii, A. awamori,
and P. radicum (Gleddie et al., 1991; Tiwari et al., 1989, 1993; Whitelaw et al.,
1999) and yields of pea, rice, chickpea, and soybean have been increased by in-
oculation with P-solubilizing fungi (Gleddie et al., 1991; Gleddie, 1993; Tiwari et
al., 1989; Thimmegowda and Devakumar, 1996) (Table IV).
Some studies of P-solubilizing fungi included the effect of P-solubilizing bac-
teria in a mixed inoculum. A mixed inoculum of A. awamori with Pseudomonas
striata increased the yield of wheat under field conditions (Gaur et al., 1980). VAM
fungi, which are known to enhance the ability of the host plant to absorb P, have
also been included in mixed inocula. The individual growth-promoting effects of
P-solubilizing fungi, bacteria, and VAM fungi have been reported to be additive.
An inoculum consisting of a mixture of A. niger, Bacillus mobilis, and the VAM
142 M. A. WHITELAW
fungus Glomus fasciculatus increased the P content of shoots and roots and the
bulb weight of onion grown on unsterilized soil to a greater extent than the indi-
vidual microbial components (Manjunath et al., 1981). Wheat plants inoculated
with both P. bilaii and VAM fungi and grown on soil which was sterilized to kill
native VAM fungi received greater benefit from rock phosphate addition than
plants receiving only one of the microorganisms (Kucey, 1987) (Table IV). Many
studies have been made on the effect of P-solubilizing fungi on legumes. Fungi
such as unknown species of Penicillium, Cephalosporium, and Alternaria and
Penicillium lilacinum, A. niger, A. flavus, and A. terreus were associated with
legume root nodules and were capable of solubilizing Ca3(PO4)2 (Subba-Rao and
Bajpai, 1965; Chhonkar and Subba-Rao, 1967) (Table I). Increased soil P avail-
ability is known to increase nodulation in legumes (Jones et al., 1977) and in some
studies inoculation of legumes with P-solubilizing fungi increased nodulation and
N uptake (Gleddie, 1993; Tiwari et al., 1989; Singh and Singh, 1993; Mehta et al.,
1996; Vaishya et al., 1996) (Table IV).
The effect of inoculation of unplanted soil on the NaHCO3-extractable P or “soil
available P” levels was investigated in several greenhouse studies. Increases in soil
available P in unplanted soil were reported by Banik and Dey (1981b) with A.
niger, Goenadi et al. (1995) with an unidentified Aspergillus sp., and Illmer and
Schinner (1995b) with P. aurantiogriseum. Soil available P was also determined
in some plant growth studies. Increases in the levels of plant available P found in
the soil after wheat, sorghum, and soybean experiments were reported by Kucey
(1988), Salih et al., (1989), and Singh and Singh (1993) (Table IV).
Inoculation of plants with P-solubilizing fungi was sometimes reported to en-
courage the proliferation of other P-solubilizing fungi in the rhizosphere. Kucey
(1988) found that after 8 weeks of growth, the total number of P-solubilizing fun-
gal isolates from the rhizosphere of wheat inoculated with P. bilaii was 283% high-
er than in uninoculated plants. Gururaj and Mallikarjunaiah (1995) also reported
an increase (47%) in the number of P-solubilizing fungi in the rhizosphere of sun-
flower inoculated with Penicillium glaucum at harvest.
The relative benefits of inoculation with P-solubilizing fungi have been ob-
served to decrease as the soil P availability increases. Downey and Van Kessel
(1990) and Gleddie (1993) found that the response to inoculation of pea with P.
bilaii under greenhouse conditions depended on whether fertilizer P had been
added to the soil. When soluble P fertilizer was added, the response to inoculation
was lower than when no P was added. Whitelaw et al. (1999) also found that the
response to inoculation of wheat with P. radicum under greenhouse conditions was
lower when fertilizer P was added. Gleddie et al. (1991) reported results of 55 sep-
arate field trials of wheat inoculated with P. bilaii and found that in soils which had
high available P levels, neither P fertilization nor inoculation induced a yield re-
sponse. Salih et al. (1989) reported higher responses in yield and P uptake to in-
oculation of sorghum with an unidentified Penicillium sp. or with A. foetidus in
soil treated with rock phosphate in comparison with soil treated with triple super-
phosphate. They surmised that this was probably due to the presence of adequate
quantities of readily available P in the soil treated with triple superphosphate,
whereas available P in the soil treated with rock phosphate appeared to be the lim-
iting factor for plant growth.
Many greenhouse or field experiments have investigated plant growth promo-
tion by P-solubilizing microorganisms using only a single level of P application.
Abbott and Robson (1984), while discussing VAM fungi which are capable of in-
creasing P availability to plants, state that there are advantages in studying a com-
plete P response curve (i.e., a full range of applied P with a low and at least two
high P application levels resulting in a plateau defining maximum yield response
to P). The study of the complete response curve is important to demonstrate that
the effect of the inoculum can be eliminated by applying P fertilizer, thus imply-
ing that the P fertilizer has replaced the advantage gained by the P-solubilizing
effect of the inoculum (Abbott and Robson, 1984). Where less data are available
(e.g., less than a full response curve), confirmation of a negative interaction be-
tween inoculation and the P rate is evidence for a P-solubilizing effect but it is not
possible to separate alleviation of P deficiency from a combination of alleviation
of P deficiency and other plant growth-promotion mechanisms.
Three classes of response to the inoculation and P application treatments need
to be considered when interpreting the data in the previously discussed experi-
ments. First, if the effect of inoculation on plant growth is caused by the allevia-
tion of P deficiency as a result of fungal-induced solubilization of soil inorganic
phosphates, then application of P fertilizer at a rate sufficient to eliminate the de-
ficiency should eliminate the response to inoculation (i.e., a negative interaction
between inoculation and P application rate would be expected). Second, if the ef-
fects of inoculation are other than the alleviation of P deficiency, then the response
should not be eliminated by P fertilizer. Therefore, a negative interaction between
inoculation and P application rate would not be expected and there would be a pos-
itive yield response to inoculation no matter how high the P application rate. Third,
if both P-solubilizing and other mechanisms are present, then a negative interac-
tion between inoculation and P application rate could be expected but a difference
in yield would still be expected at P rates high enough to eliminate P-deficiency
effects.
VI. CONCLUSION
Phosphorus is an important plant nutrient which is in short supply in many agri-

cultural soils. Because a large percentage of phosphatic fertilizer is fixed by soil
and thus rendered less available to plants, the long-term application of P fertiliz-
144 M. A. WHITELAW
ers has resulted in an accumulation of total soil P, most of which is poorly soluble.
Many soil fungi, predominantly of the genera Aspergillus and Penicillium, have
been shown to possess the ability to solubilize sparingly soluble phosphates in vit-
ro by secreting inorganic or organic acids. The microorganisms and soils reported
here are very diverse and it is difficult to generalize about the success of plant
growth promotion by P-solubilizing soil fungi. The relationship between soil pH
and phosphate solubility is not a simple one, and controversy exists over the effect
of increasing or decreasing pH on P solubility in soil. However, in unbuffered liq-
uid media studies mentioned previously, P solubilization appeared to often be as-
sociated with a lowering of pH. Some of the fungi tested appeared to solubilize P
in soils and in unbuffered liquid media. Field and greenhouse trials have demon-
strated that inoculation of plants with some P-solubilizing fungi increased the con-
centration of “available” P in the soil and enhanced the yield and P uptake by the
plant. Continued work in this area could yield plant fungal inoculants capable of
more consistent performance over a range of soil and climatic conditions.
ACKNOWLEDGMENTS
I thank Associate Professor Terence Harden of the School of Wine and Food Sciences, Charles Sturt
University, Wagga Wagga, and Dr. Mark Conyers of N.S.W. Agriculture, Wagga, Wagga, N.S.W. Aus-
tralia, for critically reviewing the manuscript.
REFERENCES
Abbott, L. K., and Robson, A. D. (1984). The effect of VA Mycorrhizae on plant growth. In “VA My-
corrhiza” (C. L. Powell and D. J. Bagyaraj, Eds.) pp. 113 –130. CRC Press, Boca Raton, FL.
Agnihotri, V. P. (1970). Solubilization of insoluble phosphates by some soil fungi isolated from nurs-
ery seed beds. Can. J. Microbiol. 16, 877– 880.
Ahmad, N., and Jha, K. K. (1968). Solubilization of rock phosphate by micro-organisms isolated from
Bihar soils. J. Gen. Appl. Microbiol. 14, 89 – 95.
Albert, A., and Serjeant, E. P. (1984). Chelation and the stability constants of metal complexes. In “The
Determination of Ionization Constants—A Laboratory Manual” (A. Albert and E. P. Serjeant,
Eds.), 3rd ed. Chapman & Hall, London.
Armstrong, R. D., Helyar, K. R., and Prangnell, R. (1993). Direct assessment of mineral phosphorus
availability to tropical crops using 32P labelled compounds. Plant Soil 50, 279 –287.
Arora, D., and Gaur, C. (1979). Microbial solubilization of different inorganic phosphates. Indian J.
Exp. Biol. 17, 1258 –1261.
Asea, P. E. A., Kucey, R. M. N., and Stewart, J. W. B. (1988). Inorganic phosphate solubilisation by 2
Penicillium species in solution culture and soil. Soil Biol. Biochem. 20, 459 – 464.
Aylward, G., and Findlay, T. (1994). “SI Chemical Data,” 3rd ed. Wiley, Brisbane.
Bache, B. W. (1963). Aluminium and iron phosphate studies relating to soils. I. Solution and hydroly-
sis of variscite and strengite. J. Soil Sci. 14, 113 –123.
Banik, S., and Dey, B. K. (1981a). Phosphate-solubilizing microorganisms of a lateritic soil. (I). Sol-
ubilization of inorganic phosphates and production of organic acids by microorganisms, isolated
in sucrose calcium phosphate agar plates. Zentralblatt Bakteriol. Parasitenkunde Infektion-

skarankheiten, Hygiene. 2. Naturwiss. Miikrobiol. Landwirtsch 136, 478 – 486.
Banik, S., and Dey, B. K. (1981b). Phosphate-solubilizing microorganisms of a lateritic soil. (II). Ef-
fect of some tricalcium phosphate-solubilizing microorganisms on available phosphorus content
of the soil. Zentralblatt Bakteriol. Parasitenkunde Infektionskarankheiten, Hygiene. 2. Naturwiss.
Miikrobiol. Landwirtsch 136, 487– 492.
Banik, S., and Dey, B. K. (1982). Available phosphate content of an alluvial soil as influenced by in-
oculation of some isolated phosphate-solubilizing micro-organisms. Plant Soil 69, 353 – 364.
Banik, S., and Dey, B. K. (1983). Alluvial soil microorganisms capable of utilizing insoluble alumini-
um phosphate as a source of phosphorus. Zentralblatt Mikrobiol. 138, 437– 442.
Bardiya, M. C., and Gaur, A. C. (1974). Isolation and screening of microorganisms dissolving low-grade
rock phosphate. Folia Microbiolo. (Praha. Acad. Scientiarium Bohemoslovaca) 19, 386–389.
Barrow, N. J. (1984). Modelling the effects of pH on phosphate sorption by soils. J. Soil Sci. 35, 283 –
297.
Barrow, N. J. (1987). “Reactions with Variable-Charge Soils.” Nijhoff, Dordrecht.
Barrow, N. J. (1990). Chemical processes and management systems. In “Phosphorus Requirements for
Sustainable Agriculture in Asia and Oceania,” pp. 199 –209. International Rice Research Institute,
Manilla.
Barthakur, H. P. (1978). Solubilization of relatively insoluble phosphate by some fungi isolated from
the rhizosphere of rice. Indian J. Agric. Sci. 48, 762–766.
Bar-Yosef, B. (1991). Root excretions and their environmental effects: Influence on availability of
phosphorus. In “Plant Roots: The Hidden Half ” (Y. Waisel, A. Eshel, and U. Kafkafi, Eds.), pp.
529–557. Dekker, New York.
Berry, D. R. (1975). The environmental control of the physiology of filamentous fungi. In “The Fila-
mentous Fungi, Vol. 1, Industrial Mycology” (J. E. Smith and D. R. Berry, Eds.), pp. 16–32. Ed-
ward Arnold, UK.
Bolan, N. S. (1991). A critical review on the role of mycorrhizal fungi in the uptake of phosphorus by
plants. Plant Soil 134, 189 –207.
Bolan, N. S., Naidu, R., Mahimairaja, S., and Baskaran, S. (1994). Influence of low-molecular-weight
organic acids on the solubilization of phosphates. Biol. Fertil. Soils 18, 311– 319.
Cerezine, P. C., Nahas, E., and Banzatto, D. A. (1988). Soluble phosphate accumulation by Aspergillus
niger from fluoroapatite. Appl. Microbiol. Biotechnol. 29, 501– 505.
Chabot, R., Cescas, M. P., and Antoun, H. (1993). Microbiological solubilization of inorganic P-frac-
tion normally encountered in soils. Phosphorus Sulfur Silicon 77, 329.
Chhonkar, P. K., and Subba-Rao, N. S. (1967). Phosphate solubilization by fungi associated with
legume root nodules. Can. J. Microbiol. 13, 749 –753.
Cochrane, V. W. (1958). “Physiology of Fungi.” Wiley, New York.
Colwell, J. D. (1963). The estimation of the P requirements of wheat in southern NSW by soil analy-
sis. Austr. J. Exp. Agric. Anim. Husbandry 3, 190 –197.
Corman, J., Tsuchiya, H. M., Koepsell, H. J., Benedict, R. G., Kelley, S. E., Feger, V. H., Dworschack,
R. G., and Jackson, R. W. (1957). Oxygen absorption rates in laboratory and pilot plant equip-
ment. Appl. Microbiol. 5, 313 – 318.
Cosgrove, D. J. (1977). Microbial transformations in the phosphorus cycle. In “Advances in Microbial
Ecology” (M. Alexander, Ed.). Plenum, New York.
Costin, A. B., and Williams, C. H. (1983). “Phosphorus in Australia.” Centre for Resource and Envi-
ronmental Studies, A.N.U., Canberra, Australia.
Cunningham, J. E., and Kuiack, C. (1992). Production of citric and oxalic acids and solubilization of
calcium phosphate by Penicillium bilaii. Appl. Environ. Microbiol. 58, 1451–1458.
Darmwal, N. S., Singh, R. B., and Rai, R. (1989). Isolation of phosphate solubilizers from different
sources. Curr. Sci. 58, 570 – 571.
146 M. A. WHITELAW
Das, A. C. (1963). Utilisation of insoluble phosphates by soil fungi. J. Indian Soc. Soil Sci. 11, 203 –
207.
De La Torre, M. A., Gomez-Alarcon, G., Vizcaino, C., and Garcia, M. T. (1993). Biochemical mecha-
nisms of stone alteration carried out by filamentous fungi living in monuments. Biogeochemistry
19, 129–147.
Deming, M. E., and Cate, W. E. (1963). Preparation of variscite. Soil Sci. 95, 206 –207.
Downey, J., and Van Kessel, C. (1990). Dual inoculation of Pisum sativum with Rhizobium legumi-
nosarum and Penicillium bilaji. Biol. Fertil. Soils 10, 194 –196.
Earl, K. D., Syers, J. K., and McLaughlin, J. R. (1979). Origin of the effects of citrate, tartrate and ac-
etate on phosphate sorption by soils and synthetic gels. Soil Sci. Soc. Am. J. 43, 674 – 678.
Eckhardt, F. E. W. (1980). Microbial degradation of silicates. Release of cations from alumino silicate
minerals by yeasts and filamentous fungi. In “Biodeterioration.” Proceedings of the Fourth Inter-
national Symposium, Berlin (T. A. Oxley, D. Allsopp, and G. Becker, Eds.), Pitman & Biodeteri-
oration Society, London.
Fox, T. R., Comerford, N. B., and McFee, W. W. (1990). Phosphorus and aluminium release from a
spodic horizon mediated by organic acids. Soil Sci. Am. J. 54, 1763 –1767.
Gardner, W. K., Parbery, D. G., and Barber, D. A. (1982). The acquisition of phosphorus by Lupinus
albus L. I. Some characteristics of the soil–root interface. Plant Soil 68, 19 – 32.
Gardner, W. K., Barber, D. A., and Parbery, D. G. (1983). The acquisition of phosphorus by Lupinus
albus L. III. The probable mechanism by which phosphorus movement in the soil–root interface
is enhanced. Plant Soil 70, 107–124.
Gaur, A. C. (1972). Role of phosphate solubilizing microorganism and organic matter in soil produc-
tivity. Symp. Soil Prod. Natl. Acad. Sci. 259 –268.
Gaur, A. C. (1985). Proc. Soil Biol. Symp. Hisar, p. 125; quoted in Kapoor, K. K., Mishra, M. M., and
Kukreja, K. (1989). Phosphate solubilisation by soil microorganisms. Indian J. Microbiol. 29,
119–127.
Gaur, A. C., Madan, M., and Ostwal, K. (1973). Solubilization of phosphatic compounds by native mi-
croflora of rock phosphates. Indian J. Exp. Biol. 11, 427– 429.
Gaur, A. C., Mathur, R. S., and Sadasivam, K. V. (1980). Effect of organic materials and phosphate-
dissolving culture on the yield of wheat and greengram. Indian J. Agron. 25, 501– 503.
Gerretsen, F. C. (1948). The influence of microorganisms on the phosphate intake by the plant. Plant
Soil 1, 51–81.
Gleddie, S. C. (1993). Response of pea and lentil to inoculation with the phosphate-solubilizing fun-
gus Penicillium bilaii (Provide). In “Proceedings of the Soils & Crops Workshop,” Saskatoon,
Saskatchewan, 25 –26, February, 1993, pp. 47– 52.
Gleddie, S. C., Hnatowich, G. L., and Polonenko, D. R. (1991). A summary of wheat response to Pro-
vide (Penicillium bilaji) in Western Canada. In “Proceedings of the Alberta Soil Science Work-
shop,” Lethbridge, Alberta, 19 –21 February, 1991, pp. 306 – 313.
Goenadi, D. H. (1995). Suitability of selected mixtures of clay minerals with humic substances as car-
rier of phosphate-solubilizing microbes. Menara Perkebunan 63, 102–113.
Goenadi, D. H., and Saraswati, R. (1993). Phosphate-solubilizing capabilities of selected phosphate-
solubilizing fungal isolates. Menara Perkebunan 61, 61– 66.
Goenadi, D. H., Saraswati, R., and Nganro, N. N. (1995). Nutrient-solubilizing and aggregate-stabi-
lizing microbes isolated from selected humic tropical soils. Menara Perkebunan 63, 60 – 66.
Goldstein, A. H. (1986). Bacterial solubilization of mineral phosphates: Historical perspective and fu-
ture prospects. Am. J. Alternative Agric. 1, 51– 57.
Grierson, P. F. (1992). Organic acids in the rhizosphere of Banksia integrifolia L.F. Plant Soil 144,
259–265.
Grierson, P. F., and Attiwill, P. M. (1989). Chemical characteristics of the proteoid root mat of Banksia
integrifolia L.F. Austr. J. Bot. 37, 137–143.
Gupta, R., Rekha, S., Aparna, S., and Kuhad, R. C. (1994). A modified plate assay for screening phos-
phate solubilizing microorganisms. J. Gen. Appl. Microbiol. 40, 255 –260.
Gururaj, R., and Mallikarjunaiah, R. R. (1995). Interactions among Azotobacter chroococcum, Peni-
cillium glaucum and Glomus fasciculatum and their effect on the growth and yield of sunflower
(Helianthus annuus L.). Helia 18, 73 – 84.
Haynes, R. J. (1982). Effects of liming on phosphate availability in acid soils. Plant Soil 68, 289 – 309.
Haynes, R. J. (1990). Active ion uptake and maintenance of cation–anion balance: A critical examina-
tion of their role in regulating rhizosphere pH. Plant Soil 126, 247–264.
Hedley, M. J., White, R. E., and Nye, P. H. (1982). Plant-induced changes in the rhizosphere of rape
(Brassica napus var. Emerald) seedlings. III. Changes in L value, soil phosphate fractions and
phosphatase activity. New Phytol. 91, 45 – 56.
Hingston, F. J., Posner, A. M., and Quirk, J. P. (1972). Anion adsorption by goethite and gibbsite. I. The
role of the proton in determining adsorption envelopes. J. Soil Sci. 23, 177–192.
Hinsinger, P., and Gilkes, R. J. (1995). Root-induced dissolution of phosphate rock in the rhizosphere
of lupins grown in alkaline soil. Austr. J. Soil Res. 33, 477– 489.
Holford, I. C. R. (1983). Differences in the efficiency of various soil phosphate tests for white clover
between very acidic and more alkaline soils. Austr. J. Soil Res. 21, 173 –182.
Holford, I. C. R., and Mattingly, C. E. G. (1975). Phosphate adsorption and plant availability of phos-
phate. Plant Soil 44, 377– 389.
Hsu, P. H. (1965). Fixation of phosphate by aluminium and iron in acidic soils. Soil Sci. 99, 398 – 402.
Hsu, P. H. (1982). Crystallization of iron(III) phosphate at room temperature. Soil Sci. Soc. Am. J. 46,
928–932.
Hue, N. V., Craddock, G. R., and Adams, F. (1986). Effect of organic acids on aluminium toxicity in
subsoils. Soil Sci. Soc. Am. J. 50, 28 – 34.
Illmer, P., and Schinner, F. (1992). Solubilization of inorganic phosphates by microorganisms isolated
from forest soils. Soil Biol. Biochem. 24, 389 – 395.
Illmer, P., and Schinner, F. (1995a). Solubilization of inorganic calcium phosphates—Solubilization
mechanisms. Soil Biol. Biochem. 27, 257–263.
Illmer, P., and Schinner, F. (1995b). Phosphate solubilizing microorganisms under nonsterile condi-
tions. Bodenkultur 46, 197–204.
Illmer, P., Barbato, A., and Schinner, F. (1995). Solubilization of hardly-soluble AlPO4 with P-solubi-
lizing microorganisms. Soil Biol. Biochem. 27, 265 –270.
Johnston, H. W. (1954). The solubilization of “insoluble” phosphates. III. A quantitative and compar-
ative study of the action of chosen aromatic acids on tricalcium phosphate. N. Z. J. Sci. Technol.
36, 281–284.
Johnston, H. W. (1956). Chelation between calcium and organic anions. N. Z. J. Sci. Technol. 37, 522–
537.
Johnston, H. W. (1959). The solubilization of “insoluble” phosphates. V. The action of some organic
acids on iron and aluminium phosphates. N. Z. J. Sci. 2, 215 –218.
Johnston, H. W., and Miller, R. B. (1959). The solubilization of “insoluble” phosphate. IV. The reac-
tion between organic acids and tricalcium phosphate. N. Z. J. Sci. 2, 109 –120.
Jones, G. D., Lutz, J. A., and Smith, T. J. (1977). Effects of phosphorus and potassium on soybean nod-
ules and seed yield. Agron. J. 69, 1003 –1006.
Juo, A. S. R., and Ellis, B. G. (1968). Chemical and physical properties of iron and aluminium phos-
phates and their relation to phosphorus availability. Soil Sci. Soc. Am. Proc. 32, 216 –221.
Kamprath, E. J., and Watson, M. E. (1980). Conventional soil and tissue tests for assessing the phos-
phorus status of soils. In “The Role of Phosphorus in Agriculture” (F. E. Khasawneh, E. C. Sam-
ple, and E. J. Kamprath, Eds.), pp. 433 – 469. ASA-CSSA-SSSA, Madison, WI.
Katznelson, H., and Bose, B. (1959). Metabolic activity and P-dissolving capability of bacterial iso-
lates from wheat roots, rhizosphere and non-rhizosphere soil. Can. J. Microbiol. 5, 79 – 85.
148 M. A. WHITELAW
Katznelson, H. E., Peterson, E. A., and Roualt, J. W. (1962). Phosphate dissolving microorganisms on
seed and in the roots zone of plants. Can. J. Bot. 40, 1181–1186.
Kepert, D. G., Robson, A. D., and Posner, A. M. (1979). The effect of organic root products on the
availability of phosphorus to plants. In “The Soil–Root Interface” ( J. L. Harley and R. Scott-Rus-
sell, Eds.), pp. 115 –124. Academic Press, London.
Khan, J. A., and Bhatnagar, R. M. (1977). Studies on solubilization of insoluble phosphates by mi-
croorganisms. Fertil. Technol. 14, 329 – 333.
Kim, K. Y., McDonald, G. A., and Jordan, D. (1997). Solubilization of hydroxyapatite by Enterobac-
ter agglomerans and cloned Escherichia coli in culture medium. Biol. Fertil. Soils 24, 347– 352.
Kpomblekou-A, K., and Tabatabai, M. A. (1994). Effect of organic acids on release of phosphorus from
phosphate rocks. Soil Sci. 158, 442– 452.
Kucey, R. M. N. (1983). Phosphate-solubilising bacteria and fungi in various cultivated and virgin Al-
berta soils. Can. J. Soil Sci. 63, 671– 678.
Kucey, R. M. N. (1987). Increased phosphorus uptake by wheat and field beans inoculated with a phos-
phorus-solubilising Penicillium bilaji strain and with vesicular–arbuscular mycorrhizal fungi.
Appl. Environ. Microbiol. 53, 2699 –2703.
Kucey, R. M. N. (1988). Effect of Penicillium bilaji on the solubility and uptake of P and micronutri-
ents from soil by wheat. Can. J. Soil Sci. 68, 261–270.
Kucey, R. M. N., and Leggett, M. E. (1989). Increased yields and phosphorus uptake by Westar canola
(brassica napus L.) inoculated by a phosphate-solubilizing isolate of Penicillium bilaji. Can. J.
Soil Sci. 69, 425– 432.
Kucey, R. M. N., Janzen, H. H., and Leggett, M. E. (1989). Microbially mediated increases in plant-
available phosphorus. Adv. Agron. 42, 199 –228.
Larsen, S. (1967). Soil phosphorus. Adv. Agron. 19, 151–210.
Leaver, J. P., and Russell, E. W. (1957). The reaction between phosphate and phosphate-fixing soils. J.
Soil Sci. 8, 113–126.
Lindsay, W. L. (1979). “Chemical Equilibria in Soils.” Wiley, New York.
Lindsay, W. L., Frazier, A. W., and Stephenson, H. F. (1962). Identification of reaction products from
phosphate fertilizers in soils. Soil Sci. Soc. Am. Proc. 26, 446 – 452.
Lopez-Hernandez, D., Flores, D., Siegert, G., and Rodriguez, J. V. (1979). The effect of some organic
anions on phosphate removal from acid and calcareous soils. Soil Sci. 128, 321– 326.
Louw, H. A., and Webley, D. M. (1959). A study of soil bacteria dissolving certain mineral phosphate
fertilizers and related compounds. J. Appl. Bacteriol. 22, 227–233.
Manjunath, A., Mohan, R., and Bagyaraj, D. J. (1981). Interaction between Beijerinckia mobilis, As-
pergillus niger and Glomus fasciculatus and their effects on growth of onion. New Phytol. 87,
723–727.
Martell, A. E., and Smith, R. M. (1977). “Critical Stability Constants. Vol. 3: Other Ligands.” Plenum,
New York.
Martin, J. K. (1977). Factors influencing the loss of organic carbon from wheat roots. Soil Biol.
Biochem. 9, 1–7.
Martinez Cruz, A., Aleman, I., Ferran, J., and Martinez, V. (1990). Phosphate dissolving microorgan-
isms in the Cuban red ferralitic and reddish brown fersialitic soils. Agrokemia,-es-Talajtah 39,
492–494.
Mba, C. C. (1996). Rock phosphate solubilizing Streptosporangium isolates from casts of tropical
earthworm. Resour. Conservation Recyclling 17, 211–217.
McLaughlin, M. J., Alston, A. M., and Martin, J. K. (1988). Phosphorus cycling in wheat-pasture ro-
tations. 1. The source of phosphorus taken up by wheat. Austr. J. Soil Res. 26, 323 – 331.
Mehta, A. C., Malavia, D. D., Kaneria, B. B., and Khanpara, V. D. (1996). Response of groundnut
(Arachis hypogaea) to farmyard manure, phosphorus and phosphate-solubilizing micro-organism.
Indian J. Agron. 41, 172–174.
Moghimi, A., and Tate, M. E. (1978). Does 2-ketogluconate chelate calcium in the pH range 2.4 to 6.4?
Soil Biol. Biochem. 10, 289 –292.
Molla, M. A. Z., Chowdhury, A. A., Islam, A., and Hoque, S. (1984). Microbial mineralization of or-
ganic phosphate in soil. Plant Soil 78, 393 – 399.
Motekaitis, R. J., and Martell, A. E. (1984). Complexes of aluminium (III) with hydroxy carboxylic
acids. Inorganic Chem. 23, 18 –23.
Murrmann, R. P., and Peech, M. (1969). Relative significance of labile and crystalline phosphates in
soils. Soil Sci. 107, 249 –255.
Nagarajah, S., Posner, A. M., and Quirk, J. P. (1970). Competitive adsorption of phosphate with
polygalacturonate and other organic anions on kaolinite and oxide surfaces. Nature 228,
83–85.
Nahas, E. (1996). Factors determining rock phosphate solubilization by microorganisms isolated from
soil. World J. Microbiol. Biotechnol. 12, 567– 572.
Nahas, E., Banzatto, D. A., and Assis, L. C. (1990). Fluoroapatite solubilization by Aspergillus niger
in vinasse medium. Soil Biol. Biochem. 22, 1097–1101.
Narsian, V., Thakkar, J., and Patel, H. H. (1993). Solubilization of natural rock phosphates and pure in-
soluble inorganic phosphates by Aspergillus awamori. Indian J. Exp. Biol. 31, 747–749.
Narsian, V., Thakkar, J., and Patel, H. H. (1994). Isolation and screening of phosphate solubilizing fun-
gi. Indian J. Microbiol. 34, 113 –118.
Narsian, V., Thakkar, J., and Patel, H. H. (1995). Mineral phosphate solubilization by Aspergillus ac-
uleatus. Indian J. Exp. Biol. 33, 91– 93.
Olsen, S. R., and Watanabe, F. S. (1957). A method to determine a phosphorus absorption maximum
of soils as measured by the Langmuir isotherm. Soil Sci. Soc. Am. J. 21, 144 –149.
Ortuno, A., Hernansaez, J., Noguera, J., Morales, V., and Armero, T. (1979). Phosphorus solubilization
by Aspergillus niger and Pseudomonas fluorescens. Anal. Edafol. Agrobiol. 38, 1311–1316.
Ozanne, P. G. (1980). Phosphate nutrition of plants—A general treatise. In “The Role of Phosphorus
in Agriculture” (F. E. Khasawneh, E. C. Sample, and E. J. Kamprath, Eds.), pp. 559 – 589. ASA-
CSSA-SSSA, Madison, WI.
Parks, E. J., Olson, G. J., Brinckman, F. E., and Baldi, F. (1990). Characterization by high performance
liquid chromatography (HPLC) of the solubilization of phosphorus in iron ore by a fungus. J. Ind.
Microbiol. 5, 183–190.
Petruccioli, M., Piccioni, P., Fenice, M., and Federici, F. (1994). Glucose oxidase, catalase and glu-
conic acid production by immobilized mycelium of Penicillium variabile P16. Biotechnol. Lett.
16, 939–942.
Pikovskaia, R. I. (1948). Mobilization of phosphorus in soil in connection with vital activity by some
microbial species. Mikrobiologia 17, 362– 370.
Raghu, K., and MacRae, I. C. (1966). Occurrence of phosphate-dissolving micro-organisms in the rhi-
zosphere of rice plants and in submerged soils. J. Appl. Bacteriol. 29, 582– 586.
Rajan, S. S. S. (1976). Phosphate reactions with soil constituents and prospects for manipulation. In
“Prospects for Improving Efficiency of Phosphorus Utilization” (G. J. Blair, Ed.), pp. 35 – 39.
Univ. of New England, Armadale, Australia.
Ramos, A., Callao, V., and de Carvalho, P. C. T. (1968). La solubilizacion de fosfatos por hongos del
suelo. Microbiol. Española 23, 1–15.
Robert, M., and Razzaghe-Karimi, M. (1975). Mise en évidence de deux d’évolution minralogique des
micas trioctaédriques en présence d’acides organiques hydrosolubles. C. R. Acad. Sci. Paris 280,
2175–2178.
Roos, W., and Luckner, M. (1984). Relationships between proton extrusion and fluxes of ammonium
ions and organic acids in Penicillium cyclopium. J. Gen. Microbiol. 130, 1007–1014.
Rose, R. E. (1957). Techniques for determining the effect of microorganisms on insoluble phosphates.
N. Z. J. Sci. Technol. 38, 773 –780.
150 M. A. WHITELAW
Rosendahl, R. O. (1942). The effect of mycorrhizal and nonmycorrhizal fungi on the availability of dif-
ficultly soluble potassium and phosphorus. Soil Sci. Soc. Am. Proc. 7, 477– 479.
Rovira, A. D. (1991). Rhizosphere research—85 years of progress and frustration. In “The Rhizo-
sphere and Plant Growth” (D. L. Keister and P. B. Cregan, Eds.), pp. 3– 13. Kluwer, Dordrecht.
Rovira, A. D., Foster, R. C., and Martin, J. K. (1979). Note on terminology: Organisms, nature and
nomenclature of the organic materials in the rhizosphere. In “The Soil–Root Interface” (J. L.
Harley and R. Scott-Russell, Eds.), pp. 1– 4. Academic Press, London.
Russell, E. W. (1980). “Soil Conditions & Plant Growth,” 10th ed. Longman, London.
Sackett, W. G., Patten, A. J., and Brown, C. W. (1908). The solvent action of soil bacteria upon the in-
soluble phosphates of raw bone meal and natural rock phosphate. Zentralblatt Bakteriol. 2, 688 –
703.
Sainz, M. J., and Arines, J. (1988). P absorbed from soil by mycorrhizal red clover plants as affected
by soluble P fertilization. Soil Biol. Biochem. 20, 61– 67.
Salih, H. M., Yahya, A. I., Abdul-Rahem, A. M., and Munam, B. H. (1989). Availability of phospho-
rous in a calcareous soil treated with rock phosphate or superphosphate as affected by phosphate-
dissolving fungi. Plant Soil 120, 181–185.
Sample, E. C., Soper, R. J., and Racz, G. J. (1980). Reactions of phosphate fertilizers in soils. In “The
Role of Phosphorus in Agriculture” (F. E. Khasawneh, E. C. Sample, and E. J. Kamprath, Eds.),
pp. 263–310. ASA-CSSA-SSSA, Madison, WI.
Sauchelli, V. (1951). “Manual on Phosphates in Agriculture.” Davidson Chemical, Baltimore.
Sethi, R. P., and Subba-Rao, N. S. (1968). Solubilization of tricalcium phosphate and calcium phytate
by soil fungi. J. Gen. Appl. Microbiol. 14, 329 – 331.
Singal, R., Gupta, R., and Saxena, R. K. (1994). Rock phosphate solubilization under alkaline condi-
tions by Aspergillus japonicus and A. foetidus. Folia Microbiol. 39, 33 – 36.
Singh, C. P., Mishra, M. M., and Kapoor, K. K. (1982). Solubilization of insoluble phosphate by
mesophilic fungi. Rev. Écol. Biol. Sol. 19, 17–25.
Singh, H. P., and Singh, T. A. (1993). The interaction of rockphosphate, Bradyrhizobium, vesicular–
arbuscular mycorrhizae and phosphate-solubilizing microbes on soybean grown in a sub-Hi-
malayan mollisol. Mycorrhiza 4, 37– 43.
Singh, H. P., Pareek, R. P., and Singh, T. A. (1984). Solubilization of rock phosphates by phosphate
solubilizers in broth. Curr. Sci. 53, 1212–1213.
Snoeyink, V. L., and Jenkins, D. (1980). “Water Chemistry.” Wiley, New York.
Sperber, J. I. (1957). Solution of mineral phosphates by soil bacteria. Nature 180, 994 – 995.
Sperber, J. I. (1958a). The incidence of apatite-solubilizing organisms in the rhizosphere and soil. Aus-
tr. J. Agric. Res. 9, 778 –787.
Sperber, J. I. (1958b). Solution of apatite by soil microorganisms producing organic acids. Austr. J.
Agric. Res. 9, 782–787.
Stevenson, F. J. (1967). Organic acids in soil. In “Soil Biochemistry” (A. D. McLaren and G. H. Pe-
terson, Eds.), pp. 130 –146. Dekker, New York.
Stevenson, F. J. (1986). “Cycles of Soil, Carbon, Nitrogen, Phosphorus, Sulfur, Micronutrients.” Wi-
ley, New York.
Stewart, J. W. B., and Sharpley, A. N. (1987). Controls on dynamics of soil and fertilizer phosphorus
and sulfur. In “Soil Fertility and Organic Matter as Critical Components of Production Systems.
“Soil Science Society of America, St. Paul, MN.
Struthers, P. H., and Sieling, D. H. (1950). Effect of organic anions on phosphate precipitation by iron
and aluminium as influenced by pH. Soil Sci. 69, 205 –213.
Stumm, W., and Morgan, J. J. (1995). “Aquatic Chemistry: Chemical Equilibria and Rates in Natural
Waters,” 3rd ed. Wiley, New York.
Subba-Rao, N. S., and Bajpai, P. D. (1965). Fungi on the surface of legume root nodules and phosphate
solubilization. Experientia 21, 386 – 387.
Surange, S. (1985). Comparative phosphate solubilizing capacity of some soil fungi. Curr. Sci. 54,
1134–1135.
Swenson, R. M., Cole, C. V., and Sieling, D. H. (1949). Fixation of phosphate by iron and aluminium
and replacement by organic and inorganic ions. Soil Sci. 67, 3 –22.
Taha, S. M., Mahmoud, S. A. Z., Halim El-Damaty, A., and Abd El-Hafez, A. M. (1969). Activity of
phosphate-dissolving bacteria in Egyptian soils. Plant Soil 31, 149 –160.
Taylor, A. W., and Gurney, E. L. (1964). Solubility of variscite. Soil Sci. 98, 9 –13.
Thimmegowda, S., and Devakumar, N. (1996). Effect of phosphorus management practices on protein
and oil yield of soybean. Indian Agriculturist 40, 61– 64.
Thomas, G. V., Shantaram, M. V., and Saraswathy, N. (1985). Occurrence and activity of phosphate-
solubilizing fungi from coconut plantation soils. Plant Soil 87, 357– 364.
Tinker, P. B. (1984). The role of microorganisms in mediating and facilitating the uptake of plant nu-
trients from soil. Plant Soil 76, 77– 91.
Tiwari, V. N., Lehri, L. K., and Pathak, A. N. (1989). Effect of inoculating crops with phospho-mi-
crobes. Exp. Agric. 25, 47– 50.
Tiwari, V. N., Pathak, A. N., and Lehri, L. K. (1993). Rock phosphate–superphosphate in wheat in re-
lation to inoculation with phosphate solubilizing organism and organic waste. Indian J. Agric. Res.
27, 137–145.
Toro, M., Azcon, R., and Herrera, R. (1996). Effects on yield and nutrition of mycorrhizal and nodu-
lated Pueraria phaseoloides exerted by P-solubilizing rhizobacteria. Biol. Fertil. Soils 21, 23 –29.
Trolldenier, G. (1992). Techniques for observing phosphorus mobilization in the rhizosphere. Biol. Fer-
til. Soils 14, 121–125.
Vaishya, U. K., Bapat, P. N., and Dubey, A. V. (1996). Phosphate sobulizing efficiency of microorgan-
isms on gram grown on Vertisol. J. Indian Soc. Soil Sci. 3, 524 – 526.
Van der Zee, S.E.A.T.M., Fokkink, L. G. J., and van Riemsdijk, W. H. (1987). A new technique for as-
sessment of reversibly adsorbed phosphate. Soil Sci. Soc. Am. J. 51, 599 – 604.
Vassilev, N., Baca, M. T., Vassileva, M., Franco, I., and Azcon, R. (1995). Rock phosphate solubiliza-
tion by Aspergillus niger grown on sugar-beet waste medium. Appl. Microbiol. Biotechnol. 44,
546–549.
Vassilev, N., Fenice, M., and Federici, F. (1996). Rock phosphate solubilization with gluconic acid pro-
duced by immobilized Penicillium variable P16. Biotechnol. Techniques 10, 585 – 588.
Veith, J. A., and Sposito, G. (1977). Reactions of aluminosilicates, aluminium hydrous oxides, and alu-
minium oxide with o-phosphate: The formation of x-ray amorphous analogs of variscite and mon-
tebrasite. Soil Sci. Soc. Am. J. 41, 870 – 876.
Venkateswarlu, B., Rao, A. V., and Raina, P. (1984). Evaluation of phosphorus solubilisation by mi-
croorganisms isolated from aridisols. J. Indian Soc. Soil Sci. 32, 273 –277.
Vidhyasekaran, P., Balaraman, K., Deiveegasundaram, M., and Viswanathan, G. (1973). Phosphate-
dissolving activity of Aspergillus awamori. Indian J. Microbiol. 13, 51– 53.
Whipps, J. M. (1984). Environmental factors affecting the loss of carbon from the roots of wheat and
barley seedlings. J. Exp. Bot. 35, 767–773.
Whipps, J. M., and Lynch, J. M. (1986). The influence of the rhizosphere on crop productivity. Adv.
Microbial. Ecol. 9, 187–244.
White, R. E. (1983). The enigma of pH–P solubility relationships in soil. In “Phosphorus: Indispens-
able Element. Proceedings of the 3rd International Congress on Phosphorus Compounds, Brus-
sels, 4–6 October, 1983,” pp. 53 – 64. Institut Mondial du Phosphate, Casablanca, Morocco.
Whitelaw, M. A., Harden, T. J., and Bender, G. L. (1997). Plant growth promotion of wheat inoculat-
ed with Penicillium radicum sp. nov. Austr. J. Soil Res. 35, 291– 300.
Whitelaw, M. A., Harden, T. J., and Helyar, K. R. (1999). Phosphate solubilisation in solution culture
by the soil fungus Penicillium radicum. Soil Biol. Biochem. 31, 655 – 665.
Wild, A. (1950). The retention of phosphate by soil. A review. J. Soil Sci. 1, 221–238.
HYDROLOGICAL FACTORS
FOR PHOSPHORUS TRANSFER
FROM AGRICULTURAL SOILS
P. M. Haygarth,1 A. L. Heathwaite,3 S. C. Jarvis,1 and T. R. Harrod2

1Institute of Grassland and Environmental Research and
2Soil Survey and Land Research Centre
Cranfield University
North Wyke, Okehampton
Devon EX20 2SB, United Kingdom
3Department of Geography
University of Sheffield
Sheffield S10 2TN, United Kingdom
I. Introduction
II. Temporal Variables
A. Effective Rainfall
B. Levels of Hydrological Activity
C. Timescales
III. Spatial Variables
A. Scale
B. Pathways
IV. Conclusions
References
Understanding and managing against phosphorus (P) transfer from agricultural

soils to receiving waters is a multidisciplinary task in which the role of hydrology is
particularly difficult to simplify. We review current knowledge to define the spatial
and temporal controls on P transfer from agricultural soils via the various hydro-
logical pathways. Rainfall intensity and duration and the interval between rainfall
events are key temporal variables which influence the discharge (and hence P load)
to receiving waters. In terms of understanding mechanisms, we postulate that lev-
els of hydrological activity may be nominally classified at two levels. Level 1 activ-
ity occurs during light or little rainfall for a high proportion of time; in contrast,
level 2 activity occurs less frequently but is more energetic and has a large capaci-
ty for P transfer over a small time period. The range of potential hydrological path-
ways of P transfer creates confusion because terminology varies. Often, process
terms such as leaching or generic terms such as drainage are confused with hydro-
logical pathways per se. Here we define the spatial variation in the hydrological
pathways responsible for P transfer at two scales—soil profile and slope/field—
which subsequently can help the understanding of P transfers into the wider catch-
153
0065-2113/00 $30.00
154 P. M. HAYGARTH ET AL.
ment, where problems occur. Our aim is to provide a simplified basis for classify-
ing the otherwise complex hydrochemical regimes which result in P transfer.
© 2000 Academic Press.
I. INTRODUCTION
The consequences of land use on environmental quality have an increasingly

high profile. From an agricultural standpoint, one area of critical current interest
is the transfer of pollutants from soil to water because soils play a pivotal role in
the protection of groundwaters and surface waters (National Rivers Authority,
1992). There is a need to identify and describe sources, media, and pathways of
pollutant transfer in order to aid the developing interactions of soil science with
hydrology (Boorman et al., 1995) and/or land use specialists with aquatic biolo-
gists. Such multidisciplinary links have the potential to create ambiguous, impre-
cise, and contradictory terminology or to omit key information. Definitions are
particularly important when describing environmental phosphorus (P) transfer
from agricultural soils to inland water bodies. Most studies on P transfer are from
an agronomic standpoint in which the role of hydrology is not fully considered
(Haygarth and Jarvis, 1999). We therefore deliberately take this opportunity to de-
scribe the hydrological factors in P transfer because, to date, there has been a ten-
dency for such factors to be neglected.
In many of the world’s agricultural soils, P accumulates in intensively man-
aged agricultural soils because of farm imports of fertilizer and livestock feeds
(Brouwer et al., 1995; Haygarth et al., 1998b). Consequently, agricultural soils
are now considered to be the main diffuse source of P reaching freshwaters (Foy
and Withers, 1995), in which concentrations as low as 35 –100 g total P liter1
may contribute to eutrophication (Organization for Economic Cooperation and
Development, 1982). Previous research on phosphorus transfer (PT) has em-
phasized measurable “soil factors” such as defining threshold P concentrations
using soil extractants (Heckrath et al., 1995; Sharpley et al., 1996), with rela-
tively few studies focusing on “hydrological factors” (Heathwaite, 1995). Ac-
cording to Heathwaite, soil characteristics and factors define the initial chemi-
cal form of P export, but the hydrological conditions determine whether or not
mobilization occurs and along which pathway. The scarcity of studies on hy-
drological interactions with PT is perhaps understandable because, unlike soil
properties, hydrological factors are less easy to measure, classify, and interpret
than many other soil properties. Hydrological conditions are also temporally and
spatially dynamic and have the added difficulties of variations with changes in
scale.
HYDROLOGICAL FACTORS FOR PHOSPHORUS TRANSFER 155
Conventionally, general-purpose soil classifications have been defined by pedo-

logical properties, although broad hydrological status might be implied from class
names, as in Avery’s (1980) “stagnogley.” Although pedological classifications
place high significance on expression or gleying (indicating duration of waterlog-
ging), it required the hydrology of soil types (HOST) classification of Boorman et
al. (1995) to provide a detailed description of the hydrologically driven system in
England and Wales. Inferences on overall hydrology, pathways, seasonal behav-
ior patterns, and responses to rainfall can be derived from HOST. Until the Host
classification was developed, only the very simple winter rain acceptance poten-
tial classification (Farquharson et al., 1978) and U.S. Department of Agriculture
(USDA, 1972) hydrologic soil grouping had, in a consistent way, addressed the
hydrology within the soil.
Studies of the fundamental role played by water flow characteristics on PT are
invaluable because hydrology provides both the energy and the carrier for the
transfer regime (Haygarth and Jarvis, 1999). Moreover, soils are hydrologically
diverse (Boorman et al., 1995), with very different pedological and hydrological
properties as well as varied responses to rainfall and land-use practices. Never-
theless, contrasting soils are often found in close proximity on hillslopes and even
within individual fields. Such differences are illustrated by the soils in Table I
which, along with the hydrologically similar soils they represent, are important in
various regions of the United Kingdom. The contrast is shown by the imperme-
able, clayey Hallsworth series soils (Avery, 1980), which are waterlogged from
October to May, whereas the permeable, loamy Denbigh series soils are water-
logged only in very wet weather. Thus, each soil series has different implications
for PT.
The aim of this review is to help to redress the current shortfalls in information
by providing a considered review of the hydrological issues and defining the role
of hydrology in context with other aspects of PT. The treatment of hydrological
factors in this review is provided in two sections dealing with temporal and spa-
tial variables.
II. TEMPORAL VARIABLES
A. EFFECTIVE RAINFALL
Precipitation, with its temporal variation, provides the energy source and the
physical carrier mechanism within the PT process (Haygarth and Jarvis, 1999).
Critical to PT is the relationship between rainfall input and runoff generation. In
catchment/watersheds in which the climate is characterized by low-intensity rain-
Table I
Hydrological Characteristics Describing the Contrast between the Denbigh and Hallsworth Soil Series
Soil classification Denbigh Hallsworth
Soil Survey and Land Research Centre (Avery, 1980) Typical brown earths Pelostagnogley soils
USDA (1972) Dystrochrepts Typic haplaquepts
Food and Agriculture Organization Dystric cambisols Dystric gleysols
Wetness (Hodgson, 1997)
Class I IV
Duration of waterlogging (days)
40-cm depth 0 days 180 days
70-cm depth 30 days 180 days
Saturated hydraulic conductivity (m/day) 1 (topsoil), 1 (subsoil) 0.8 (topsoil), 0.002 (subsoil)
Porosity
Retained water % (moisture content after drainage, i.e., at FC) 42 (topsoil), 32 (subsoil) 47 (topsoil), 49 (subsoil)
Air capacity % (drainable pores) 12 (topsoil), 21 (subsoil) 12 (topsoil), 0.1 (subsoil)
Workability/trafficability (Findlay et al., 1984)
Good machinery workdays relative to field capacity season 30 55
HOST (Boorman et al., 1995)
General description 1:No impermeable or gleyed layer 1: Gleyed layer within 40 cm
within 1 m
2: No significant aquifer or GW 2: Over slowly permeable substrate
3: Over impermeable or hard substrate 3: No significant aquifer
Hydrological pathways Vertical unsaturated flow; bypass flow Surface runoff likely; prolonged seasonal
in the substrate; some surface runoff saturated flow; short seasonal bypass
flow to the substrate
HOST class 17 24
SPR% (standard percentage runoff; from catchment scale 32 51
interpretation)
Base flow index (BFI% 1 total BF 0 no BF; from 0.6 0.31
catchment interpretation)
fall (15 mm h1), the current soil moisture state is more important in control-
ling the occurrence and magnitude of runoff than rainfall magnitude (Istok and
Boersma, 1986). This has implications for PT because most export occurs when
the land is at or close to field capacity (FC). For example, in the Slapton catch-
ment, Devon, England, 80% of PT occurred in winter (Heathwaite et al., 1989).
Snowmelt has been associated with high rates of PT from soil (Hawkins and
Scholefield, 1996; Timmons et al., 1977).
The concept of “field capacity” is useful in this context. Field capacity occurs
when a soil is thoroughly wetted and drainage starts. For example (see Table I), a
freely draining soil such as the Denbigh series would “drain” and reach FC in ap-
proximately 2 days from thorough wetting, whereas this takes weeks (Hodgson,
1997) in a poorly drained soil such as the Hallsworth series. The respective hy-
drological properties of these two soil types are summarized in Table I. Meteoro-
logical field capacity (MFC) occurs when received rainfall is greater than evapo-
transpiration, provided no soil moisture deficit exists. MFC is readily presented in
cartographic form (Jones and Thomasson, 1985). Excess precipitation, which oc-
curs when soils are at FC, can be called hydrologically significant precipitation.
This effective rainfall is an important driver for the PT process.
B. LEVELS OF HYDROLOGICAL ACTIVITY
Temporal variations in rainfall intensity, duration, and intervals between

storms (return period) affect the magnitude of discharge along various hydrolog-
ical pathways (Burgoa et al., 1993; DeWalle and Pionke, 1994; Evans, 1978;
Sharpley, 1980a,b; Thornes, 1979). Although flow rate is a continuum from low
to high discharge, different hydrological pathways may be triggered at different
rates of discharge or rainfall input, with different consequences for PT. Howev-
er, division into base-flow and storm-flow conditions is a helpful starting point
when considering discharges. Storm flows occur infrequently and result in over-
land and macropore flows, whereas base flows occur more frequently but may
only transfer water along predominately subsurface pathways. Pionke et al.
(1996) found that storm flow was important for P discharge from a 7.4-km2 agri-
cultural watershed. Dils and Heathwaite (1996) found that individual storm
events had different capacities to transport P; moreover, the forms of P varied be-
tween events. They suggested that the most important controls on P fractionation
were antecedent soil moisture (determining the likelihood of surface runoff ) and
the interval between storm events (determining the incidence of “old” and “new”
soil water and hence the amount of time for interaction between mobilized P and
soil water).
Base flow comprises non-storm-flow periods in which groundwater discharge,
including springs and near-stream seepage, may form the main component of flow
(Pionke and DeWalle, 1994). For base flow, subsurface hydrological pathways
such as throughflow are most important for PT. Despite the obvious importance of
storm events, the significance of subsurface pathways in transporting P during
base-flow or low-flow conditions should not be discounted and needs further re-
search in terms of PT.
A limitation of the storm-flow/base-flow concept is that it pertains solely to dis-
charge per se, whereas a more generalized temporal hydrological classification,
which accounts for precipitation (and thus rainfall erosivity, for example), may be
more useful. Haygarth and Jarvis (1997), Haygarth et al. (1998a), and Fraser et al.
(1998) identified two populations of data when studying PT, implicating an inter-
action of rainfall with flow. We therefore suggest that an effective means of clas-
sification must recognize at least two levels of hydrological activity. Level 1 ac-
tivity occurs during light or little rainfall for a high proportion of time (and will
incorporate base flows). In contrast, level 2 activities will be of low frequency but
high intensity, operating with greater energy than level 1 activity, and have a high
propensity for PT during a short period and resulting in storm flows. When level
2 occurs it will produce high-intensity rainfall having a greater erosivity and will
therefore result in larger particulate PTs than those of level 1 rain. In reality, a two-
tier classification may be too simplistic and a sliding scale of increasing activity
may be more appropriate: Nevertheless, this system provides useful initial con-
ceptual classification.
C. TIMESCALES
Shrinkage of soils such as the Hallsworth series (Table I), with coarsely struc-
tured, often clayey horizons (HOST classes 18 –25 of Boorman et al., 1995), dur-
ing summer soil moisture deficits opens vertical fissures. These permit vertical by-
pass flow particularly during autumnal rewetting (level 2; Fig. 1a). When rewetted,
the horizons revert to an impermeable state, dominated by saturated lateral flow
(level 1). The temporal extent of either state depends on the strength of the soil
moisture deficit and the duration of the field capacity period.
In Table II we present a nonexhaustive list of hydrological pathways which
we have attempted to classify in relation to timescales (more detailed discussion
of the relevance of this table is given in the following section, which defines scale).
At the slope/field scale, the time taken for overland flow to travel 100 m may be
of the order of minutes during intense rainfall (Horton, 1945), whereas water in
subsurface pathways to underlying aquifers may take months to years (thus giv-
ing rise to the terms new and old water; Bohlke and Dener, 1995). Again, many of
these variations can be defined by soil class, for example, as expressed in the
HOST classification of Boorman et al. (1995) and illustrated by data in Table I for
the Denbigh and Hallsworth series.
Figure 1 Soil profile scale pathways. The situation common (a) in late summer/autumn, when soils are dry and prone to bypass flow, and (b) in winter/spring,
when soils are saturated and prone to matrix flow.
Table II
Terminology Commonly Associated with Hydrochemical Transfer Pathways, Nominally Classified by Discipline, Time, and Spatial Scale
Term (sorted Nominal

alphabetically) Scale Generic Definition timescale
Arterial drainage Catchment Agronomic Major artificial drainage channel used to remove water after its Hours/days
discharge from field drains
Base flow Slope/field Hydrological Nominally the transfer of water underground in “background”
flow conditions, not pathway specific; also used to describe
a low magnitude of flow Hours/weeks/months
Bypass flow Soil Hydrological Implies a type of soil water movement—in the case of vertical Minutes/hours
movement along larger subsoil pathways, e.g., wormholes
and fissures, often occurring in unsaturated conditions
Darcian flow Soil Hydrological Not a pathway but describes discharge through the soil as related —
to the hydraulic gradient and hydraulic conductivity
Ditching Slope/field Agronomic Artificial (human-made) first-order drainage streams, often used Minutes/hours
160
on agricultural land to compliment arterial drainage

Interflow Slope/field Hydrological Lateral flows below the soil surface Minutes/hours
Land drainage Subcatchment Agricultural Water and solute ( entrained solids) export to catchment Minutes/hours
resulting from land drainage practices: anthropogenic
Leaching Soil Chemical Eluviation of chemicals vertically through the soil profile and Variable
vadose zone; despite misconceptions, this is a mechanism
and not a pathway
Leakage Slope/field Hydrological and General nonspecific term describing water and chemical Not applicable
chemical movement
Macropore flow Soil Hydrological As bypass flow; macropores are large enough (60 m) to Minutes/hours
allow gravitational drainage
Matrix flow Soil Hydrological Implies a type of soil water movement—in this case uniform Days
vertical movement downwards, common in very porous
media such as sandy textures; only occurs under saturated
conditions
Overland flow Slope/field Geomorphological/ Movement of water exclusively over the soil surface, down Minutes/hours
pedological slope, during heavy rain
Percolating water Soil Hydrological General nonspecific term describing water movement Not applicable
Pipe flow Slope/field Geomorphological Lateral subsurface preferential flow Minutes/hours
Piston flow Soil Hydrological As matrix flow Variable
Preferential flow Soil Hydrological As bypass flow Minutes/hours
Return flow Slope/field Hydrological Where a subsurface flow pathway emerges at the soil surface Minutes/days
River Subcatchment Hydrological Large-order drainage network Hours/days
Roadway Subcatchment Engineering Human road or path which can assist water transfer from slope/ Hours/days
field to catchment; little studied
Runoff Slope/field Hydrological General hydrological term describing the lateral movement of Minutes/hours
water off land above and below ground, causing a short-term
increase in flow at the catchment outlet; can refer to pathway
when qualified (e.g., surface runoff), but also has been used
to describe processes and water samples
Saturated (soil) flow Soil Hydrological As piston flow, but lateral and not vertical Days
Seepage Slope/field Hydrological General nonspecific term describing water movement; implies Not applicable
emergence at or near the ground surface
161
Soil solution Soil Chemical Nonspecific term describing water sampled from the soil Not applicable
environment by whatever means; not a pathway
Stream Subcatchment Hydrological and Small-order drainage network Hours/days
geomorphological
Subsurface flow Slope/field Geomorphological Lateral flows below the soil surface Minutes/hours
Surface runoff Slope/field Hydrological As overland flow Minutes/hours
Throughflow Soil and Hydrological As percolating water Not applicable
Slope/field
Unsaturated flow Slope/field Hydrological As preferential flow, but occurring laterally over capped, Minutes/hours
compacted, or slowly permeable horizons
Vertical saturated Soil Hydrological As piston flow Days
flow
Vertical unsaturated Soil Hydrological As bypass flow Minutes/hours
flow
III. SPATIAL VARIABLES
A. SCALE
Effects of space and scale are important when attempting to understand PT

along soil hydrological pathways (Beven et al., 1993; Konikow, 1991). Kirkby
(1988) demonstrated the importance of scale (drainage basin area) on hillslope
flow processes (Fig. 2). The concepts incorporated in Fig. 2 may be extended to
an examination of PT where lag times to peak concentration (Fig. 2a) are impor-
Figure 2 (a) Lag times and (b) peak runoff rates integrating spatial and temporal factors. NB Hor-
tonian infiltration excess overland flow. (Redrawn from “Hillslope Hydrology,” M. J. Kirkby. Copy-
right John Wiley & Sons Limited. Reproduced with permission.)
tant in determining exchange and uptake reactions with soil solution and conse-
quently the form of P transported. In contrast, peak runoff rates (Fig. 2b) also aid
our understanding of the potential P load reaching the drainage network. Thus, for
cultivated land, the higher peak runoff rates associated with Hortonian (infiltra-
tion-excess) overland flow (Fig. 2b) will have greater capacity to detach and trans-
port soil particles and sorb P. Although this pathway is restricted in catchment/wa-
tersheds in temperate climes such as that in the United Kingdom, where the soil
infiltration capacity is rarely exceeded (Kirkby, 1988), poor land management
(e.g., overgrazing) will exacerbate its incidence (Heathwaite, 1997).
Because soil hydrological and chemical processes operate at different scales, it
is essential that their spatial dimensions be identified. The challenge is to identify
mechanisms that connect soil profile and slope/field scale process studies with
larger scale catchment/watershed effects. Recent attempts to incorporate chemi-
cal–physical links initiating PT using a combination of runoff and erosion model-
ing (Gburek et al., 1996) have advanced knowledge of pathways of P delivery and
their interaction with soil profile characteristics. Similarly, knowledge of hydro-
logical pathways of water movement from land to stream has developed with in-
creasingly detailed means of field monitoring. Thus, simple models of pipe-flow
channeling of new water via the soil matrix to the stream have been replaced by
new concepts incorporating new/old water and bypass flow (Bohlke and Dener,
1995).
For the purpose of clarifying approaches to understanding the spatial controls
on PT, we have nominally selected two scales smaller than that of catchment/wa-
tershed. The size boundaries selected are arbitrary but provide a convenient means
for subdivision of landscape units. The smallest is the soil profile scale (centime-
ters to meters). The second is the slope/field scale; this has also been called the
hillslope or the subcatchment/watershed and will range from meters to hectares
and is generally restricted to catchment/watershed headwaters and/or first-order
streams (Haygarth and Jarvis, 1997; Heathwaite and Johnes, 1996). This scale is
likely to be relevant to individual soil mapping units (scales of 1:25,000 and
greater). For PT from agricultural land, this scale is perhaps the most critical be-
cause it is possible to integrate detailed soil process studies with characterization
of the patterns of PT at the slope/field scale (e.g., P fractionation in different hy-
drological pathways; Dils and Heathwaite, 1996; Haygarth et al., 1998a). The
largest scale is the catchment/watershed; although we will not focus on catchment/
watershed pathways, it is necessary to consider some definitions. Catchments (En-
glish term) or watersheds (American term) are variable in size and there are many
examples of catchment/watershed studies and classifications (Burt et al., 1996;
Gburek et al., 1996; Heathwaite and Johnes, 1996; Johnes et al., 1996; Kronvang,
1990; Molden and Cerney, 1994). Catchment/watersheds are appropriate for in-
terpretation and mapping of “reconaissance” maps (1: M in the United Kingdom,
as in appendix D of Boorman et al., 1995). Our scheme for subdividing the spa-
tial dimensions of PT is similar to that proposed by Kirkby (1988) in order to ex-

amine the time lag (in hours from precipitation input to stream output within a
catchment/watershed; see Fig. 2). Kirkby argues that “catchments may be thought
of as a sequence of moisture stores, some in series, some in parallel.” Temporal
rates of water movement (introduced in Section II) clearly interact with the spa-
tial scales and can give rise to the classification of (i) surface detention (0.1–1 h),
(ii) infiltration (1–20 h), (iii) unsaturated vertical percolation (1–50 h), (iv) satu-
rated downslope flow (1–12 h), (v) channel flow [depends on catchment/water-
shed area: 0.5 h (1 km2), 7 h (100 km2), 100 h (104 km2)]. Here, scales i–iv equate
with our soil profile scale, scale iv and part of scale v equate with our slope/field
scale, and v refers to the wider catchment/watershed. The store with the longest
residence time exerts the greatest control on water movement. This has implica-
tions for P form and the magnitude of P export because it indicates the amount of
time available for equilibration between rainfall and soil water and hence the po-
tential for P mobilization and transfer.
B. PATHWAYS
Generalized overviews and definitions of hydrological pathways have been pro-

vided by many authors (Anderson and Burt, 1990; Boorman et al., 1995; Cox et
al., 1997; Kirkby, 1978; Mangold and Tsang, 1991; McGechan and Wu, 1996;
Miyazaki et al., 1993; Youngs and Leeds-Harrison, 1990). When precipitation
reaches the soil, it is partitioned between overland flow and subsurface flow (Kirk-
by, 1988), each having varying potentials to entrain and retain P. Terms used to de-
scribe these two pathways (and their various forms) are sometimes confused in the
scientific literature and problems arise because
1. Hydrological pathway terms can become confused with process terms. The
terms “runoff” and “leaching” are good examples. Runoff has been used in the
context of a (rather vaguely defined) pathway (Boorman et al., 1995; Haygarth and
Jarvis, 1996, 1997), a water medium (Harms et al., 1974; Loehr, 1974), or, in oc-
casional circumstances, a process (Zobisch et al., 1994). Leaching has also been
used in an ambiguous way. It does not describe a pathway, although it has some-
times been used in this context to characterize an amalgam of all pathways of wa-
ter drainage through soil (Bromfield and Jones, 1972; Heckrath et al., 1995; Jor-
dan and Smith, 1985). Leaching is a process term and describes the elluviation of
solutes, such as P, down through soil and is common in porous soils (Wagenet,
1990; Weaver et al., 1988a,b).
2. Pathways tend to be classified according to research background and disci-
pline and may be defined in terms of geomorphological, hydrological, pedologi-
cal, chemical, or land-use criteria.
Table II provides a synthesis of common soil hydrological pathways and asso-

ciated terminology. We classified them both generically and by scale. A detailed
discussion is provided in the following subsections, which are separated into the
soil profile and slope/field scales.
1. Soil Profile Scale
Key hydrological pathways at the soil profile scale are illustrated in Fig. 1. Void
size (numbers and distribution), number of fissures (pores and macropores), and
the degree of structure and aggregation (Hodgson, 1997) all contribute to vari-
ability. The modes of hydrological transport at this scale are (i) matrix flow or sat-
urated/piston flow and (ii) preferential bypass or macropore flow. These do not de-
scribe pathways per se but rather a mode of flow. Saturated/piston flow refers to
the uniform movement of water through soil, which occurs after the soil pores have
become saturated. It is dependent on soil porosity and is common in sandy soils
(Tindall et al., 1986). For relatively homogeneous soils such as sands, simple mod-
els of water movement based on Darcian flow may be applied. Here, discharge
through the soil is related to the hydraulic gradient and hydraulic conductivity. The
latter is a function of the media (e.g., sand, silt, and clay) and the properties of the
fluid flowing through it. Darcian flow may explain the movement of a sharp wet-
ting front between wet and dry soil and models of soil water movement via this
pathway are relatively simple. During saturated flow, PT may be delayed through
interaction within soil peds en route. Preferential flow, alternatively termed macro-
pore or bypass flow (Thomas and Phillips, 1979), is the rapid and direct transfer
of water through soil fissures and macropores (root channels and animal burrows
coarser than 60 m; Hodgson, 1977); it may increase the infiltration capacity of
soils by one or two orders of magnitude (Burt et al., 1996).
Artificial drainage systems, both land drainage pipes and associated permeable
fill, and secondary treatments of mole drains and subsoiling are important routes
for preferential flow in impermeable agricultural soils (see Section II,B,2 for fur-
ther discussion on artificial drainage). Preferential flow is important in clay soils,
which are particularly vulnerable to shrinkage cracking, often after dry weather.
However, after a period of rewetting the cracks close. Macropores, such as earth-
worm burrows, function year-round. They are often very common in loamy and
sandy soil and can result in bypass flow, even in soils with abundant macropores.
Figure 1 provides an illustration of the seasonal variations which may lead to pref-
erential vs saturated/piston flow. Preferential flow can occur laterally (i.e., paral-
lel to the surface) or vertically (Beven and Germann, 1982). There has been con-
siderable research on preferential flow (Beven, 1991; Beven and Germann, 1982;
Hornberger et al., 1990; Phillips et al., 1995; Singh and Kanwar, 1991; Wopereis,
1994), although it remains difficult to model the spatial complexity of hillslope hy-
drology in terms of the potentially turbulent nature of flow where continuously
connected soil voids (macropores or pipes) bypass matrix flow. Three-dimension-

al macropore–matrix models are needed to further elucidate the hillslope hydro-
logical system (Holden et al., 1995).
To date, there has been little work on P movement via preferential flow, with
the exception of some initial studies and discussions (Beauchemin et al., 1998;
Dils and Heathwaite, 1996; Haygarth et al., 1998a; Heckrath et al., 1995; Stamm
et al., 1998). Preferential flow may be particularly effective in transporting P pre-
sent on the soil surface where it has accumulated after, for example, excreta, ma-
nure, or fertilizer deposition/application. Thus, this pathway may be important
for PT from grassland soils, in which such practices are common (Haygarth et
al., 1998a; Heathwaite, 1997), and for arable soils used for organic waste dis-
posal.
The soil hydrological pathways described previously for the soil profile scale
may be conveniently summarized by the HOST model (Boorman et al., 1995).
HOST assumes two extremes of soil water movement as either (i) predominantly
vertical or (ii) predominantly lateral and at or close to the surface. The classifica-
tion is based on the permeability of substrate and the depth of the water table. There
are three settings for the model in terms of relationships to groundwater or
aquifers; these are illustrated in Fig. 3. In HOST, soil hydrological pathways are
described with respect to
• Location in the soil, subsoil, and substrate
• Vertical or lateral direction of flow
• Saturated or unsaturated conditions
• Probability and duration of flow (local conditions)
• Existence of bypass flow
• Existence of leakage to substrate
Weaknesses in the HOST approach are that it does not consider capping/crusting
soils and the role of slopes. Although the international applicability of the HOST
model has yet to be tested, it shows great potential.
2. Slope/Field Scale
The pathways at this scale are illustrated in Fig. 4: This is the link between the
soil profile and the catchment/watershed, encompassing hydrological pathways
which may directly (e.g., by overland or surface flow) or indirectly (e.g., by inter-
flow or lateral subsurface flow) link land to ditch to stream. A significant propor-
tion of PT may be accounted for by erosion mechanisms and overland flow path-
ways at this scale (Sharpley and Smith, 1990); pathways which arise from artificial
land drainage may also be important. There are many studies which summarize
water flow on this scale (Atkinson, 1978; Emmett, 1978; Freeze, 1978; Hornberger
et al., 1991; Knapp, 1978; Trudgill and Coles, 1988; Whipkey and Kirby, 1978),
Figure 3 The hydrology of soil types (HOST) classification (redrawn from Boorman et al., 1995, Fig. 3.2, p. 27).
Figure 4 Slope/field scale pathways (redrawn from Haygarth and Jarvis, 1999).
and others illustrate techniques for sampling and studying transfers (Cuttle and
Mason, 1988; Hoover, 1987).
Although PT from agricultural catchment/watersheds is largely derived from
nonpoint sources, some parts of the catchment/watershed are more likely to con-
tribute water—and possibly P—flow than others (Gburek and Sharpley, 1998).
The link between the spatial variation in PT and hillslope hydrology lies in un-
derstanding the role of hillslope morphology in controlling water flow pathways.
There is often a close link between hillslope angle and soil (and therefore porosi-
ty/infiltration) patterns with, for example, permeable soils such as the Denbigh
series (Table I) occupying steeper slopes and impermeable soils such as the
Hallsworth series occupying gentler footslopes and interfulvial crests (see Fig. 44
in Findlay et al., 1984). Catchment/watershed morphology is a dominant control
on spatial patterns of hydrological response (Beven and Young, 1988), particular-
ly where surface or near-surface flows are the main mechanism by which water
reaches a stream. However, other sources of heterogeneity contributing to spatial
variation exist, such as local management and the effect of soil characteristics on
infiltration and thus overland flow.
Work on hillslopes (Anderson and Burt, 1990; Burt et al., 1996) demonstrated
the importance of topographical variation in hillslope form on the spatial variation
in runoff generation and hence solute transport. Valley side slopes were shown to
have high potential for solute export, whereas hillslope hollows form potential
point sources and may be important for transferring solutes, such as nitrate,
through the soil. Such hollows become important for PT where they coincide with
critical source areas (CSAs) of P (Pionke et al., 1997) usually as a result of land
use (Gburek et al., 1996). CSAs describe catchment/watershed areas where P
sources (e.g., fields with high manure applications and/or high soil P concentra-
tions) coincide with areas with hydrological connectivity to the drainage network
(e.g., variable source areas in near-stream zones). Hillslope hollows appear to be
activated only when flow converges into hollows and leads to accumulation of soil
moisture and activation of saturated overland flow; such activation is dependent
on rainfall duration, intensity, and magnitude as well as the interval between rain-
fall events.
For impermeable subsoils, runoff is the most commonly used term for describ-
ing water movement at the slope/field scale. As indicated previously, this is po-
tentially ambiguous: Runoff is often used as a general term describing the lateral
movement of water (and entrained solids) across or under the surface of a slope
and refers to the composite of all soil profile and slope/field pathways. The propen-
sity for runoff to affect PT is described by the effective depth of interaction (EDI)
(Ahuja, 1986; Ahuja et al., 1981; Sharpley, 1985). Experiments adding 32P to the
soil surface on simulated slopes showed that the top few centimeters of soil play
a critical role in the generation of P in runoff. This critical layer was termed the
“mixing zone” and under experimental conditions it was calculated that the EDI
for a 4% slope was 0.2 or 0.3 cm after application of 6.5 cm h1 rainfall for 30
min (Ahuja, 1986; Ahuja et al., 1981; Sharpley, 1985). Although the experiments
used extreme rainfall inputs, the research indicated that P residing in the surface 1
mm of soil was most vulnerable to export in runoff.
Overland flow, or surface runoff, describes transport which occurs exclusively
over the soil surface (Emmett, 1978). It is important in physically transporting P
attached to solids following soil erosion and also through incidental losses of fer-
tilizers and manures from the slope (Haygarth and Jarvis, 1999). It is also a path-
way with a strong affinity for PT because the surface soil has the greatest EDI
(Ahuja et al., 1981) and the highest concentrations of P (Haygarth et al., 1998a),
which can interact with mobile water. In the examples in Table I, the Hallsworth
series soils are at risk of overland flow during and after rain throughout the field
capacity season, whereas Denbigh series soils will be at risk only during heavy
rain or after severe surface damage by livestock or machinery. Overland flow oc-
curs through either “infiltration excess” (i.e., Hortonian) or “saturation excess”
mechanisms (Fig. 2). Infiltration excess overland flow was first described by Hor-
ton (1945) and occurs where rainfall intensity exceeds the infiltration capacity of
the soil (Dunne, 1983). The concept is applicable to arid/semiarid environments
in which it may occur over wide areas. In temperate zones, infiltration-excess over-
land flow is more appropriately described as partial-area runoff (Betson, 1964)
where only parts of a catchment/watershed contribute flow, for example, over-
grazed grassland (Heathwaite et al., 1990) and tracks or tractor wheelings (Heath-
waite et al., 1989)—physical effects resulting from land management. Discharge
per unit contour length (q) is given by
q (I f )a
where I is the rainfall intensity after interception, f is the infiltration rate, and a is
the area drained per unit contour length (Thornes, 1979). In Britain, for example,
rainfall intensities are generally low and the soil infiltration capacity is unlikely to
be exceeded (Kirkby, 1978); therefore, the runoff regime should be dominated by
saturation-excess and subsurface storm-flow mechanisms. There is evidence to
suggest, however, that land management practices are increasing the incidence of
infiltration-excess overland flow (Heathwaite et al., 1990). This has been record-
ed on a fieldwide basis at Slapton, Devon (Heathwaite and Burt, 1992), where
overgrazing of kale by sheep compacted the soil surface to the extent that the in-
filtration capacity was of the order 0.1 mm h1. In grassland, infiltration capaci-
ties ranging from 12 mm h1 for temporary grass to 5 mm h1 for lightly
grazed permanent grass and 1 mm h1 for heavily grazed permanent grass were
recorded. Snow and frozen ground impair infiltration and therefore increase rates
of overland flow with correspondingly high rates of PT (Hawkins and Scholefield,
1996). Other land management activities affecting the soil surface, and thus the
propensity for overland flow and PT, are poaching (also called pugging or tram-
pling by livestock), panning (lateral smearing and closing of pores due to the pas-
sage of agricultural machinery, often resulting in a subsurface effect), rutting/
wheeling (by agricultural vehicles), and capping/slaking/crusting (the degradation
of the structure of the soil surface), often by rain impact to impair infiltration and
general compaction.
Saturation-excess overland flow describes soil saturation via a high water table
or lateral water movement above an impeding horizon. It may incorporate return
flow (Table II and Fig. 4) and is a seasonally variable and potentially important
component of storm runoff, especially in catchment/watersheds with thick vege-
tation, thin soils, high water tables, and long, gentle concave slopes (Dunne, 1983).
Dunne and Black (1970) demonstrated that the major proportion of storm runoff
was produced as overland flow from a small proportion of the catchment/water-
shed. The authors identified partial areas contributing quick runoff; such areas
were controlled by rainfall intensity and expanded/contracted seasonally or dur-
ing an individual storm. Geology, topography, soils, and rainfall characteristics de-
fined their location. Under steady rainfall, lower rainfall intensities are required to
maintain saturation overland flow in comparison with infiltration-excess overland
flow. Whereas infiltration-excess overland flow may occur over wide areas such
as whole fields, the areas of a hillslope or subcatchment/watershed contributing
saturation-excess overland flow are both spatially and temporally variable. These
are called variable source areas (VSAs), which are zones of saturation within a
catchment/watershed or hillslope where saturation excess overland flow is likely
to occur: this is a hydrological term and does not relate to P per se. VSA should
not be confused with CSAs, which combine VSAs with the source of P and may
be a particularly important vehicle for PT (Gburek et al., 1996). Saturation-excess
overland flow is generally confined to areas subject to saturation, such as (i) areas
adjacent to perennial streams (base of hillslopes), (ii) areas of concave upward
slopes (slope profile concavities), (iii) hollows (areas of concave outward con-
tours), and (iv) areas with thin or impermeable soils.
In addition to the surface flow pathways described previously, subsurface mech-
anisms of PT are also important. Throughflow describes the transport of water be-
low the soil surface (Thornes, 1979), along or above an impermeable subsoil hori-
zon, such as clay or bedrock, often in a vector parallel or nearly parallel with the
slope surface (Whipkey and Kirby, 1978). Throughflow is common under many
soils in the United Kingdom and occurs either as saturated flow or as preferential
flow (at this scale, preferential flow may include pipe flow) (Table II and Fig. 4).
Return flow occurs where a subsurface flow pathway emerges at the soil surface
(Miller et al., 1977). Transmission of water through the soil produces different
rates of throughflow from different soil layers where, in general, saturated perme-
ability decreases with depth (Kirkby, 1978). Dunne (1983) suggests that through-
flow is an important component of storm runoff, in which soil conductivity is high.
It enters the drainage network (i) via groundwater, (ii) via lateral flow where soil
layers have vertical conductivity rainfall intensity, and (iii) where concave topo-
graphic contours create contributing areas (high water table and/or subsurface
impedance causes convergent flow). Zones generating subsurface flow are tem-
porally and spatially variable. The hydrological pathways are continuous and dy-
namic: A hillslope may generate only subsurface flow during a gentle rainstorm,
infiltration-excess overland flow during a deluge, or subsurface flow only during
a short rainstorm and saturation-excess overland flow during a long event (Dunne,
1983).
Finally, base flow, introduced earlier to contrast levels of hydrological activity
with storm flow, can also be used in a related manner as a spatial term describing
the deep percolation of water from the soil into groundwater and subsurface path-
ways. Deep percolation is probably only important where well-defined aquifers
are present. Thus, a large part of base flow results from percolation or throughflow
in the soil—a time lag of the order of months may describe water movement in
this instance from interfluve (catchment/watershed boundary) to stream.
Artificial land drainage provides an additional pathway of PT at the slope/field
scale. This involves placing outlets for water in the subsoil to supplement the nat-
ural downward flux of water (Armstrong et al., 1984; Armstrong and Garwood,
1991; Armstrong and Harris, 1996; Tyson et al., 1993; Youngs, 1983) and involves
both pipe drains and, in the United Kingdom, mole drains on heavy textured soils
(Armstrong and Harris, 1996). Under grazed pasture slopes, land drainage has
been shown to reduce PT (Haygarth et al., 1998a). Agricultural land drainage will
inevitably influence pathways of PT depending on (i) depth and spacing of pipe
drains, combined with the presence or absence of permeable fill; (ii) secondary
treatment such as mole drainage or subsoiling; (iii) ditching (human-made streams
to assist land drainage); and (iv) arterial drainage (a major artificial drainage chan-
nel used to remove water after its discharge from field drains). Other examples of
research examining PT and land drainage are given in Roberts et al. (1986) and
Sharpley and Syers (1979).
The previous discussion refers to situations at the slope/field scale in which the
subsurface soil and/or geology is impermeable. Where subsoils and bedrock are
permeable, most water (and associated P) will be transferred to groundwater and
aquifers (Denver, 1991), with negligible surface pathway activity. Where well-de-
fined aquifers are present at shallow depths, such as in The Netherlands (Chardon
and Oenema, 1995; Chardon et al., 1997) and on the east coast of the United States
(Delmarva Peninsula, Chesapeake Bay, and the Atlantic Coastal Plain in Mary-
land) (Bachman and Phillips, 1996; Bohlke and Dener, 1995; Phillips and Bach-
man, 1996), PT to groundwater becomes an important environmental problem, es-
pecially where it is associated with livestock intensification (Sims, 1996). The
behavior of P in deep groundwater has been a neglected area of study, in contrast
to N (Bachman and Phillips, 1996), and more studies of P in throughflow and per-
colation pathways are required.
IV. CONCLUSIONS
Water provides the energy and hydrological pathways offer the carrier route for
PT from land to stream. The spatial and temporal variabilities in water movement
along hydrological pathways are difficult to define and describe. However, the
temporal variation in precipitation amount and intensity is clearly important in
governing the magnitude of PT, whereas spatial variables control the pathway and
perhaps the form of P entering the drainage network. We believe that the defini-
tions of the spatial and temporal variability of hydrological pathways of PT at var-
ious scales given here will eliminate some of the misunderstandings that have oc-
curred in the past. If advances are to be made in quantifying and controlling PT
from diffuse agronomic sources, interpretations of PT processes must always be
qualified and defined in these spatial and hydrological terms.
ACKNOWLEDGMENTS
We are grateful to Rachael Dils, Andy Fraser, William Gburek, and Ben Turner for continuing in-
spiring “discussion” sessions. This work is funded by the Ministry of Agriculture, Fisheries and Food,
London. The Institute of Grassland and Environmental Research is supported by the Biotechnology
and Biological Sciences Research Council.
REFERENCES
Ahuja, L. R. (1986). Characterization and modelling of chemical transfer to runoff. Adv. Soil Sci. 4,
149–188.
Ahuja, L. R., Sharpley, A. N., Yamamoto, M., and Menzel, R. G. (1981). The depth of rainfall–runoff
interaction as determined by 32P. Water Resour. Res. 17, 969 – 974.
Anderson, M. G., and Burt, T. P. (1990). Process studies in hillslope hydrology: An overview. In
“Process Studies in Hillslope Hydrology” (M. G. Anderson and T. P. Burt, Eds.), pp. 1– 8. Wiley,
New York.
Armstrong, A. C., and Garwood, E. A. (1991). Hydrological consequences of artificial drainage of
grassland. Hydrol. Processes 5, 157–174.
Armstrong, A. C., and Harris, G. L. (1996). Movement of water and solutes from agricultural land: The
effects of artificial drainage. In “Advances in Hillslope Processes” (M. G. Anderson and S. M.
Brooks, Eds.), pp. 187–211. Wiley, Chichester, UK.
Armstrong, A. C., Atkinson, J. L., and Garwood, E. A. (1984). Grassland drainage economics experi-
ment, North Wyke, Devon, Report No. RD/FE/23 TFS 7442. Ministry of Agriculture, Fisheries
and Food, London.
Atkinson, T. C. (1978). Techniques for measuring subsurface flow on hillslopes. In “Hillslope Hy-
drology” (M. J. Kirkby, Ed.), pp. 73 –120. Wiley, New York.
Avery, B. W. (1980). “Soil Classification for England and Wales (Higher Categories).” Lawes Agri-
cultural Trust, Harpenden, UK.
Bachman, L. J., and Phillips, P. J. (1996). Hydrologic landscapes on the Delmarva Peninsula. Part 2:
Estimates of base-flow nitrogen load to Chesapeake Bay. Journal of the American Water Re-
sources Association 32, 779 –791.
Beauchemin, S., Simard, R. R., and Cluis, D. (1998). Forms and concentration of phosphorus in
drainage water of twenty-seven tile-drained soils. J. Environ. Qual. 27, 721–729.
Betson, R. (1964). What is watershed runoff? J. Geophys. Res. 69, 1541–1552.
Beven, K. (1991). Modeling preferential flow: An uncertain future? In “Proceedings ASAE Confer-
ence on Preferential Flow.” American Society of Agricultural Engineers, Chicago.
Beven, K., and Germann, P. (1982). Macropores and water flow in soils. Water Resour. Res. 18, 1311–
1325.
Beven, K. J., and Young, P. C. (1988). An aggregated mixing zone model of solute transport through
porous media. J. Contam. Hydrol. 3, 129 –143.
Beven, K. J., Henderson, D. E., and Reeves, A. D. (1993). Dispersion parameters for undisturbed par-
tially saturated soil. J. Hydrol. 143, 19 – 43.
Bohlke, J. K., and Dener, J. M. (1995). Combined use of groundwater dating, chemical, and isotopic
analyses to resolve the history and fate of nitrate contamination in two agricultural watersheds,
Atlantic coastal plain, Maryland. Water Resour. Res. 31, 2319 –2339.
Boorman, D. B., Hollis, J. M., and Lilly, A. (1995). “Hydrology of Soil Types: A Hydrologically-Based
Classification of the Soils of the United Kingdom.” Institute of Hydrology, Natural Environment
Research Council, Wallingford, UK.
Bromfield, S. M., and Jones, O. L. (1972). The initial leaching of hayed-off pasture plants in relation
to the recycling of phosphorus. Austr. J. Agric. Res. 23, 811– 824.
Brouwer, F. M., Godeschalk, F. E., Hellegers, P. J. G. J., and Kelholt, H. J. (1995). “Mineral Balances
at Farm Level in the European Union,” Agricultural Economics Research Institute (LEI-DLO),
Onderzoekverslag 137, The Hague.
Burgoa, B., Hubbard, R. K., Wauchope, R. D., and Davis-Carter, J. G. (1993). Simultaneous measure-
ment of runoff and leaching losses of bromide and phosphate using tilted beds and simulated rain-
fall. Community Soil Sci. Plant Anal. 24, 2689 –2699.
Burt, T. P., Heathwaite, A. L., and Johnes, P. J. (1996). Stream water quality and nutrient export in the
Slapton catchments. Field Studies 8, 613 – 627.
Chardon, W. J., and Oenema, O. (1995). “Leaching of Dissolved Organically Bound Phosphorus.”
DLO Research Institute for Agrobiology and Soil Fertility (AB-DLO), Wageningen, The Nether-
lands.
Chardon, W. J., Oenema, O., del Castilho, P., Vriesema, R., Japenga, J., and Blaauw, D. (1997). Or-
ganic phosphorus solutions and leachates from soils treated with animal slurries. J. Environ. Qual.
26, 372–378.
Cox, J. W., Kirkby, C. A., and Chittleborough, D. J. (1997). Pathways for losses of phosphorus from
rainfed pastures in South Australia. Conference Proceedings from XVIII International Grasslands
Congress, Winnipeg and Saskatoon, No. 2, pp. 35 – 36. (Full proceedings in preparation.)
Cuttle, S. P., and Mason, D. J. (1988). A flow-proportional water sampler for use in conjunction with
a V-notch in small catchment studies. Agric. Water Management 13, 93 – 99.
Denver, J. M. (1991). Groundwater sampling network to study agrochemical effects on water quality
in the unconfined aquifer, southeastern Delaware. In “Groundwater Residue Sampling Design”
(R. G. Nash and A. R. Leslie, Eds.), pp. 139 –149. American Chemical Society, Washington, DC.
DeWalle, D. R., and Pionke, H. B. (1994). Streamflow generation on a small agricultural catchment
during autumn recharge: II. Stormflow periods. J. Hydrol. 163, 23 – 42.
Dils, R. M., and Heathwaite, A. L. (1996). Phosphorus fractionation in hillslope hydrological pathways
contributing to agricultural runoff. In “Advances in Hillslope Processes” (M. Anderson and S.
Brooks, Eds.), pp. 229 –252. Wiley, Chichester, UK.
Dunne, T. (1983). The relation of field studies and modelling in the prediction of storm runoff. J. Hy-
drol. 65, 25–48.
Dunne, T., and Black, R. D. (1970). Partial area contributions to storm runoff in a small New England
watershed. Water Resour. Res. 6, 1296 –1311.
Emmett, W. W. (1978). Overland flow. In “Hillslope Hydrology” (M. J. Kirkby, Ed.), pp. 145–176. Wi-
ley, Chichester, UK.
Evans, R. (1978). Soil erosion in north Norfolk. J. Agric. Sci. Cambridge 90, 185 –192.
Farquharson, F. A. K., Mackney, D., Newson, M. D., and Thomasson, A. J. (1978). “Estimation of
Runoff Potential of River Catchments from Soil Surveys.” Lawes Agricultural Trust, Harpenden,
UK.
Findlay, D. C., Colborne, G. J. N., Cope, D. W., Harrod, T. R., Hogan, D. V., and Staines, S. J. (1984).
“Soils and Their Use in South West England.” Lawes Agricultural Trust, Harpenden, UK.
Foy, R. H., and Withers, P. J. A. (1995). The contribution of agricultural phosphorus to eutrophication.
Fertilizer Soc. Proc. 365, 1– 32.
Fraser, A. I., Harrod, T. R., and Haygarth, P. M. (1998). The effect of rainfall intensity and hydrologi-
cal pathways on phosphorus transfer from agricultural soils—Importance de l’intensité de la pluie
et des chemins hydrologiques sur le transfert du phosphore provenant des sols agricoles. In “Pro-
ceedings of the 16th World Congress of Soil Science.” ISSS, Montpellier. [CD-Rom Format]
Freeze, R. A. (1978). Mathematical models of hillslope hydrology. In “Hillslope Hydrology” (M. J.
Kirkby, Ed.), pp. 177–225. Wiley, Chichester, UK.
Gburek, W. J., and Sharpley, A. N. (1998). Hydrologic controls on phosphorus loss from upland agri-
cultural catchments. J. Environ. Qual. 27, 267–277.
Gburek, W. J., Sharpley, A. N., and Poinke, H. B. (1996). Identification of critical source areas for phos-
phorus export from agricultural catchments. In “Advances in Hillslope Processes” (M. G. Ander-
son and S. M. Brooks, Eds.), pp. 263 –282. Wiley, Chichester, UK.
Harms, L. L., Dornbush, J. N., and Anderson, J. R. (1974). Physical and chemical quality of agricul-
tural runoff. J. Water Pollution Control Federation 46, 2460 –2470.
Hawkins, J. M. B., and Scholefield, D. (1996). Molybdate-reactive phosphorus losses in surface and
drainage waters from permanent grassland. J. Environ. Qual. 25, 727–732.
Haygarth, P. M., and Jarvis, S. C. (1996). Pathways and forms of phosphorus losses from grazed grass-
land hillslopes. In “Advances in Hillslope Processes” (M. G. Anderson and S. M. Brooks, Eds.),
pp. 283–294. Wiley, Chichester, UK.
Haygarth, P. M., and Jarvis, S. C. (1997). Soil derived phosphorus in surface runoff from grazed grass-
land lysimeters. Water Res. 31, 140 –148.
Haygarth, P. M., and Jarvis, S. C. (1999). Transfer of phosphorus from agricultural soils. Adv. Agron.
66, 195–249.
Haygarth, P. M., Hepworth, L., and Jarvis, S. C. (1998a). Forms of phosphorus transfer in hydrologi-
cal pathways from soil under grazed grassland. Eur. J. Soil Sci. 49, 65 –72.
Haygarth, P. M., Jarvis, S. C., Chapman, P., and Smith, R. V. (1998b). Phosphorus budgets for two con-
trasting grassland farming systems in the UK. Soil Use Management 14, 160 –167.
Heathwaite, A. L. (1995). Sources of eutrophication: Hydrological pathways of catchment nutrient ex-
port. Int. Assoc. Hydrol. Sci. 230, 161–176.
Heathwaite, A. L. (1997). Sources and pathways of phosphorus loss from agriculture. In “Phosphorus
Loss to Water from Agriculture” (H. Tunney, O. T. Carton, P. C. Brookes, and A. E. Johnston,
Eds.), pp. 205–224. CAB. International, Cambridge, UK.
Heathwaite, A. L., and Burt, T. (1992). The evidence for past and present erosion in the Slapton Catch-
ment. In “Soil Erosion, Past and Present: Archeological and Geographical Perspectives” (M. Bell
and J. Boardman, Eds.), Vol. 22, pp. 89 –100. Oxbow Monograph, Oxford, UK.
Heathwaite, A. L., and Johnes, P. J. (1996). Contribution of nitrogen species and phosphorus fractions
to stream water quality in agricultural catchments. Hydrol. Processes 10, 971– 983.
Heathwaite, A. L., Burt, T. P., and Trudgill, S. T. (1989). Runoff, sediment and solute delivery in agri-
cultural drainage basins—A scale dependent approach. Int. Assoc. Hydrol. Sci. 182, 175 –191.
Heathwaite, A. L., Burt, T. P., and Trudgill, S. T. (1990). The effect of land use on nitrogen, phospho-
rus and suspended sediment delivery to streams in a small catchment in southwest England. In
“Soil Erosion on Agricultural Land” ( J. Boardman, L. D. L. Foster, and J. A. Dearing, Eds.), pp.
161–177. Wiley, Chichester, UK.
Heckrath, G., Brookes, P. C., Poulton, P. R., and Goulding, K. W. T. (1995). Phosphorus leaching from
soils containing different phosphorus concentrations in the Broadbalk Experiment. J. Environ.
Qual. 24, 904–910.
Hodgson, J. M. (1997). “Soil Survey Field Handbook—Describing and Sampling Soil Profiles,” 3rd
ed. Cranfield University, Silsoe, UK.
Holden, N. M., Scholefield, D., Williams, A. G., and Dowd, J. F. (1995). The simultaneous use of three
common hydrological tracers for fine spatial and temporal resolution investigation of preferential
flow in a large soil block. In “Tracer Technologies for Hydrological Systems,” pp. 77– 86. Inter-
national Association of Hydrological Sciences, Boulder, CO.
Hoover, J. R. (1987). Instrumentation system for determining watershed hydrologic characteristics.
Trans. ASAE 30, 1051–1056.
Hornberger, G. M., Beven, K. J., and Germann, P. F. (1990). Inferences about solute transport in macro-
porous forest soils from time series models. Geoderma 46, 249–262.
Hornberger, G. M., Germann, P. F., and Beven, K. J. (1991). Throughflow and solute transport in an
isolated sloping soil block in a forested catchment. J. Hydrol. 124, 81– 99.
Horton, R. E. (1945). Erosional development of streams and their drainage basins: Hydrophysical ap-
proach to quantitative morphology. Bull. Geol. Soc. Am. 56, 275 – 330.
Istok, J. D., and Boersma, L. (1986). Effect of antecedent rainfall on runoff during low-intensity rain-
fall. J. Hydrol. 88, 329 – 342.
Johnes, P. J., Moss, B., and Phillips, G. (1996). The determination of total nitrogen load and total phos-
phorus concentrations in freshwaters from land use, stock headage and population data: Testing
of a model for use in conservation and water quality management. Freshwater Biol. 36, 451–
473.
Jones, R. J. A., and Thomasson, A. J. (1985). “An Agroclimatic Databank for England and Wales.”
Lawes Agricultural Trust, Harpenden, UK.
Jordan, C., and Smith, R. V. (1985). Factors affecting leaching of nutrients from an intensively man-
aged grassland in County Antrim, Northern Ireland. J. Environ. Management 20, 1–15.
Kirkby, M. J. (1978). “Hillslope Hydrology.” Wiley, Chichester, UK.
Kirkby, M. (1988). Hillslope runoff processes and models. J. Hydrol. 100, 315 – 339.
Knapp, B. J. (1978). Infiltration and storage of soil water. In “Hillslope Hydrology” (M. J. Kirkby, Ed.),
pp. 43–72. Wiley, Chichester, UK.
Konikow, L. F. (1991). “Flow and Transport Problems in Groundwater Systems.” Miljoministeriet
Miljostyrelsen, Elinsore, Denmark.
Kronvang, B. (1990). Sediment associated phosphorus transport from two intensively farmed catch-
ment areas. In “Soil Erosion on Agricultural Land” ( J. Boardman, I. D. L. Foster, and J. A. Dear-
ing, Eds.), pp. 131– 330. Wiley, Chichester, UK.
Loehr, R. C. (1974). Characteristics and comparative magnitude of non point sources. J. WPFC 46,
1849–1872.
Mangold, D. C., and Tsang, C.-F. (1991). A summary of subsurface hydrological and hyrochemical
models. Rev. Geophys. 29, 51–79.
McGechan, M. B., and Wu, L. (1996). Modelling pollutant transport processes following the land
spreading of farm wastes. In “AgEng.” Universidad Politecnica de Madrid, Madrid.
Miller, W. W., Guitjens, J. C., and Mahannah, C. N. (1977). Quality of irrigation water and surface re-
turn flows from selected agricultural lands in Nevada during the 1974 irrigation season. J. Envi-
ron. Qual. 6, 193.
Miyazaki, T., Hasegawa, S., and Kasubuchi, T. (1993). “Water Flow in Soils.” Dekker, New York.
Molden, B., and Cerney, J. (1994). Small catchment research. In “Biogeochemistry of Small Catch-
ments” (B. Molden and J. Cerney, Eds.), pp. 1–29. Wiley, New York.
National Rivers Authority (1992). “Policy and Practice for the Protection of Groundwater.” Stanley
Hunt, Rushden, UK.
Organisation for Economic Cooperation and Development (OECD) (1982). “Eutrophication of Waters,
Monitoring, Assessment and Control.” OECD, Paris.
Phillips, P. J., and Bachman, L. J. (1996). Hydrologic landscapes on the Delmarva Peninsular. Part 1:
Drainage basin type and base flow chemistry. J. Am. Water Resour. Assoc. 32, 767–778.
Phillips, R. E., Quisenberry, V. L., and Zeleznik, J. M. (1995). Water solute movement in an undis-
turbed, macroporous column: Extraction pressure effects. Soil Sci. 59, 707–712.
Pionke, H. B., and DeWalle, D. R. (1994). Streamflow generation on a small agricultural catchment
during autumn recharge: I. Nonstormflow periods. J. Hydrol. 163, 1–22.
Pionke, H. B., Gburek, W. J., Sharpley, A. N., and Schnabel, R. R. (1996). Flow and nutrient export
patterns for an agricultural hill–land watershed. Water Resour. Res. 32, 1795 –1804.
Pionke, H. B., Gburek, W. J., Sharpley, A. N., and Zollweg, J. A. (1997). Hydrologic and chemical con-
trols on phosphorus losses from catchments. In “Phosphorus Loss to Water from Agriculture” (H.
Tunney, O. T. Carton, P. C. Brookes, and A. E. Johnston, Eds.), pp. 225–242. CAB International,
Cambridge, UK.
Roberts, G., Hudson, J. A., and Blackie, J. R. (1986). Effect of upland pasture improvement on nutri-
ent release in flows from a “natural” lysimeter and a field drain. Agric. Water Management 11,
231–245.
Sharpley, A. N. (1980a). The effect of storm interval on the transport of soluble phosphorus in runoff.
J. Environ. Qual. 9, 575 – 578.
Sharpley, A. N. (1980b). The enrichment of soil phosphorus in runoff sediments. J. Environ. Qual. 9,
521–526.
Sharpley, A. N. (1985). Depth of surface soil runoff as affected by rainfall, soil slope and management.
Soil Sci. Soc. Am. J. 49, 1010 –1015.
Sharpley, A. N., and Smith, S. J. (1990). Phosphorus transport in agricultural runoff: The role of soil
erosion. In “Soil Erosion on Agricultural Land” ( J. Boardman, L. D. L. Foster, and J. A. Dearing,
Eds.), pp. 351–366. Wiley, Chichester, UK.
Sharpley, A. N., and Syers, J. K. (1979). Loss of nitrogen and phosphorus in tile drainage as influenced
by urea application and grazing animals. N. Z. J. Agric. Res. 22, 127–131.
Sharpley, A. N., Daniel, T. C., Sims, J. T., and Pote, D. H. (1996). Determining environmentally sound
soil phosphorus levels. J. Soil Water Conserv. 51, 160 –166.
Sims, J. T. (1996). “The Phosphorus Index: A Phosphorus Management Strategy for Delawares Agri-
cultural Soils.” Univ. of Delaware, Newark.
Singh, P., and Kanwar, R. S. (1991). Preferential solute transport through macropores in large undis-
turbed saturated soil columns. J. Environ. Qual. 20, 295 – 300.
Stamm, C., Fluhler, H., and Gachter, R. (1998). Preferential transport of phosphorus in drained grass-
land soils. J. Environ. Qual. 27, 515 – 522.
Thomas, G. W., and Phillips, R. E. (1979). Consequences of water movement in macropores. J. Envi-
ron. Qual. 8, 149–152.
Thornes, J. B. (1979). Fluvial processes. In “Process in Geomorphology” (C. Embleton and J. Thornes,
Eds.), pp. 213–271. Edward Arnold, London.
Timmons, D. R., Verry, E. S., Burwell, R. E., and Holt, R. F. (1977). Nutrient transport in surface runoff
and interflow from an aspen–birch forest. J. Environ. Qual. 6, 188 –192.
Tindall, T. A., Morrill, L. G., and Henderson, L. J. (1986). The influence of soil structure on movement
rate of salts through soils. Soil Sci. 142, 108 –113.
Trudgill, S. T., and Coles, N. (1988). Application of simple soil–water models to the transfer of sur-
face-applied solutes to streams. J. Contam. Hydrol. 3, 367– 380.
Tyson, K. C., Hawkins, J. M. B., and Stone, A. C. (1993). “Final Report on the AFRC–ADAS Drainage
Experiment 1982–1993.” Institute of Grassland and Environmental Research, Okehampton, UK.
U.S. Department of Agriculture (1972). “National Engineering Handbook, Section 4, Hydrology.” U.S.
Government Printing Office, Washington, DC.
Wagenet, R. J. (1990). Quantitative prediction of the leaching of organic and inorganic solutes in soil.
Philos. Trans. R. Soc. London 329, 321– 330.
Weaver, D. M., Ritchie, G. S. P., and Anderson, G. C. (1988a). Phosphorus leaching in sandy soils. II.
Laboratory studies of the long-term effects of the phosphorus source. Austr. J. Soil Res. 26, 191–
200.
Weaver, D. M., Ritchie, G. S. P., Anderson, G. C., and Deeley, D. M. (1988b). Phosphorus leaching in
sandy soils. I. Short term effects of fertiliser applications and environmental conditions. Austr. J.
Soil Res. 26, 177–190.
Whipkey, R. Z., and Kirkby, M. J. (1978). Flow within the soil. In “Hillslope Hydrology” (M. J. Kirk-
by, Ed.), pp. 121–144. Wiley, Chichester, UK.
Wopereis, M. C. S. (1994). Reducing bypass flow through a dry, cracked and previously puddled rice
soil. Soil Tillage Res. 29, 1–11.
Youngs, E. G. (1983). The contribution of physics to land drainage. J. Soil Sci. 34, 1–21.
Youngs, E. G., and Leeds-Harrison, P. B. (1990). Aspects of transport processes in aggregrated soils.
J. Soil Sci. 41, 665 – 675.
Zobisch, M. A., Richter, C., Heiligtag, B., and Schlott, R. (1994). Nutrient losses from cropland in the
central highlands of Kenya due to surface runoff and soil erosion. Soil Tillage Res. 33, 109 –116.
CASSAVA, Manihot esculenta Crantz,
GENETIC RESOURCES: THEIR
COLLECTION, EVALUATION,
AND MANIPULATION
Nagib M. A. Nassar
Departamento de Genética e Morfologia
Universidade de Brasília
Brasília 70919, Brazil
I. Wild Taxa of Cassava Manihot Species

A. Taxonomy
B. Determination of Wild Manihot Species Localities with Emphasis on
Probable Origin
C. Relationships between Manihot Species
D. Genetic Variation of Wild Manihot Species
II. Broadening the Genetic Base of Cassava, M. esculenta Crantz, and
Development of Interspecific Hybridization
A. Production of Cassava Interspecific Hybrids
B. Development of Cassava Interspecific Hybrids
C. Development of Cassava Interspecific Hybrids for Savanna (Cerrado)
Conditions
D. Overcoming Crossing Barriers between Cassava, M. esculenta Crantz, and
a Wild Relative, M. pohlii Warwa
III. Development and Selection for Apomixis in Cassava, M. esculenta Crantz
A. Genetic Study of Apomixis in Cassava
B. Molecular and Embryonic Evidence of Apomixis in Cassava
IV. Production of Polyploid Types
A. Prospects of Polyploidizing Cassava, M. esculenta Crantz, by Unreduced
Microspores
B. Induction of a Productive Aneuploid in Cassava M. esculenta Crantz
C. Production of Triploid Cassava, M. esculenta Crantz, by Hybrid Diploid
Gametes
V. Protein Contents in Cassava Cultivars and Its Hybrid with Wild Manihot
Species
References
179
0065-2113/00 $30.00
180 NAGIB M. A. NASSAR
Wild species of Manihot are progenitors of cassava. They constitute valuable ge-
netic reservoirs with genes of new characters. Screening these species showed some
of them to have a notably high percentage of protein combined with a low per-
centage of hydrocyanic acid. Study of natural habitats revealed toxicity and adapta-
tion to cool temperature. Hybridizations between different wild Manihot species
and cassava have been carried out and hybrids were obtained, some of which showed
high root productivity and resistance to stem borers. Polyploid types were produced
by manipulation of 2n gametes. Apomixis was discovered in the wild and transferred
successfully to the cultivate. © 2000 Academic Press.
I. WILD TAXA OF CASSAVA Manihot SPECIES
A. TAXONOMY
Cassava Manihot esculenta Crantz does not grow wild. However, it is known that
about 98 species belong to the genus Manihot (Rogers and Appan, 1973). These
species include subshrubs, shrubs, and trees. A majority of them produce latex and
contain cyanogenic glucoside. Roots of wild species, contrary to the cultigen, are fi-
brous and slender, but some species exhibit a limited number of tuberous roots. The
root surface is smooth or rought. The subepidermis varies in color from red or yel-
low to white, and the cortexes of tuberous rooted species are white, cream, or yellow.
The stem varies in height from almost acaulescent in subshrubs to about 20 m
in tree species. Frequently, the stems of shrubs native to the Brazilian savanna die
back to the crown in the dry season. The color of the stem varies from gray or
brown to red. The stem normally branches dichotomously or trichotomously. The
branching point exhibits a terminal inflorescence. In wild species frequently the
young stem has a varying degree of pubescence. This characteristic is rarely en-
countered in the cultigen (Grattapaglia et al., 1986).
Leaves vary from subsessile to long petiolated. All species with the exception
of three have palmately lobed leaves. Infloresence is terminal and monoecious with
the exception of the acaulescent species native to central Brazil. Flowers have a
single perianth composed of five petals; their length ranges from 0.5 to 2.0 cm.
Buds of staminate flowers are ovoid or spheric, whereas pistillate buds are conic.
The fruit is capsule with three locules. The caruncle of the seed varies in size. The
chromosome number in all species investigated is 2n 36 (Nassar, 1978a).
All species of the genus Manihot are native to the New World; in Brazil and
Mexico they form distinct centers of diversity (Nassar, 1978b). They normally
grow sporadically in their habitat, rarely becoming dominant in the vegetation.
Due to the monoecious or dioecious structure of the inflorescence, wild Manihot
CASSAVA, M. esculenta Crantz, GENETIC RESOURCES 181
species are typically alogamous plants. However, in cultivated cassava a shift to-
ward autogamous plants has occurred. Nassar and O’Hair (1985) attributed this to
the monoclonal system of cultivation and domestication history. Observations of
frequent hybridization of the wild species and the cultigen and between wild
species suggest weak barriers in the genus (Nassar, 1980a). This is probably due
to the polyploid origin on the genus level.
B. DETERMINATION OF WILD Manihot SPECIES LOCALITIES

WITH EMPHASIS ON PROBABLE ORIGIN
From May to July 1975, I collected seeds of the wild species of Manihot in
northeastern Brazil in three states: Pernambuco, Ceará, and Bahia. the geograph-
ical distribution of the Manihot species was studied by Rogers and Appan (1973).
Manihot specimens collected by the expedition of Reading University and de-
posited at IPA herbarium, Recife, were also examined. Table I lists the wild species
of Manihot that were collected from different localities of northeastern Brazil. It
is apparent that western Pernambuco and central Bahia had the greatest variabili-
ty in Manihot. It should be noted that certain species reported by the Reading Uni-
versity expedition occur in some localities from which they no longer can be col-
lected (e.g., specimens of M. glaziovii collected about 12 km west of Ibimirim,
PE). Unfortunately, it was found that vegetation in that place had been cleared and
the land cultivated by mamona. Unlike most Manihot species, M. glaziovii grows
in large numbers and not as sporadic plants. Extinction of some wild Manihot
species from their natural habitats may be due to another factor: The majority of
these species are poisonous to grazing animals because of the presence of HCN.
Table I
Wild Species of Manihot Collected from Different Localities
in Northeastern Brazil
Species Locality
M. caerulescens Pohl Aparipina, PE

M. heptaphylla Ule Seabra, BA
M. cichotoma Ule Jequie, BA
M. catingae Ule Itaberaba, BA
M. brachyandra Pax et Hoffmann Petrolina, PE
M. maracasensis Ule Itambé, BA
M. epruinosa Pax et Hoffmann Bentecoste, Fortaleza, CE
M. glaziovii Mueller Arcoverde, Ouricure, Serratatlada, PE
M. jacobinensis Mueller Virtoria da Conquista, BA
M. quinquefolia Pohl Senhor do Bonfim, Juazeiro, BA
They are known among people of northeastern Brazil as “maniçoba,” i.e., the poi-
sonous cassava. Therefore, many plants are exterminated by farmers.
By studying the geographic distribution of Manihot species provided by Rogers
and Appan (1973) combined with localities determined on my trip, it became pos-
sible to create a map of concentration of wild species. It shows that in central Brazil
(southern Goias and eastern Minas Gerais) there are approxiimately 38 wild
species of the 98 recognized. Thus, this region includes a large number of wild
Manihot species and represents the highest diversity of these species. In this re-
gion the following species occur:
M. acuminatissima Mueller
M. sparsifolia Pohl
M. pruinosa Pohl
M. alutacea Rogers et Appan
M. divergens Pohl
M. cecropiaefolia Pohl
M. triphylia Pohl
M. pentaphylla Pohl
M. anomala Pohl
M. procumbens Mueller
M. crotalariaeformis Pohl
M. Pusilla Pohl
M. logepetiolata Pohl
M. tomentosa Pohl
M. purpureo-costata Pohl
M. attenuata Mueller
M. orbicularis Pohl
M. tripartita (Sprengel) Mueller
M. pilosa Pohl
M. sagittato-partita Pohl
M. falcata Rogers et Appan
M. quinqueloba Pohl
M. violacea Pohl
M. irwinii Rogers et Appan
M. mossamedensis Taubet
M. fruticulosa (Pax) Rogers et Appan
M. gracilis Pohl
M. warmingii Mueller
M. reptans Pax
M. stipularis Pax
M. oligantha Pax
M. nana Mueller
M. stricta Baillon
M. salicifolia Pohl
M. weddelliana Baillon
M. peltata Pohl
M. janiphoides Mueller
M. handroana N. D. Cruz
The second largest center of diversity is southwestern Mexico, which includes
M. pringlei Watson
M. aesculifolia Pohl
M. oaxaca Rogers et Appan
M. rhomboidea Mueller
M. walkarae Croizat
M. divisiae Croizat
M. michaelis McVaugh
M. websterae Rogers et Appan
M. auriculata Mcvaugh
M. rubricaulis I. M. Hohnson
M. chlorosticta Standley et Goldman
M. subspicata Rogers et Appan
M. caudata Greenman
M. angustiloba (Torrey) Mueller
M. tomatophylla Standley
M. foetida Pohl
The third largest center of diversity is northeastern Brazil, which includes
M. zenhtneri Ule
M. surinamensis Rogers et Appan
M. quinquefolia Pohl
M. pseudoglaziovii Pax et Hoffmann
M. maracasensis Ule
M. quinquepartita Huber
M. caerulescens Pohl
M. marajoara Chermont de Miranda
M. tristis Mueller
M. glaziovii Mueller
M. epruinosa Paz et Hoffmann
M. brachyandra Pax et Hoffmann
M. dichotoma Ule
M. leptophylla Pax
M. reniformis Pohl
M. heptaphylla Ule
Finally, the fourth largest center of diversity is western South Mato Grosso and
Bolivia. It includes the following species:
M. guaranitica Choda et Hassier
M. pruinosa Pohl
M. jacobinsis Mueller
M. condesata Rogers et Appan
M. xavantinensis Rogers et Appan
M. flemingiana Rogers et Appan
Vavilov (1951) showed that variation in cultivated plants is confined to rela-
tively few restricted areas or centers. In 1920, he set up 6 main geographic centers
for cultivated plants and in 1935 increased the number of centers to about 10. He
assumed that the center of diversity for cassava was in the Brazilian–Bolivian cen-
ter. Vavilov proposed that centers of diversity are places of origin of cultivated
plants. Since this exposition of the center of diversity in the 1920s, much more in-
formation has been gathered and it has become clear that not all centers of diver-
sity represent centers of origin. Harlan (1961) showed that more than one center
of diversity may be formed for a given crop through introgression. This phenom-
enon explained why in many cases there are centers of diversity for a given crop
far from areas of much diversity of wild relatives. Since Harlan proposed this the-
ory (giving a convincing example of the evolved species of Helianthus) much ev-
idence has supported it. Dobzhansky (1973) stated many conspicuous cases, such
as the formation of species of Iris, Eucalyptus, Liatris, Penstemon, and Trago-
pogon. Thus, this phenomenon serves as a model for what apparently occurred dur-
ing the formation of these four centers of diversity of Manihot. Assuming that cas-
sava was domesticated for the first time in one place and then carried by Indians
through immigrations, there could then result an extensive hybridization between
the cultivated species and local wild ones, giving rise to numerous new species
through introgression.
Cassava does not grow wild. The large variation of cassava cultivars due to
maintaining them by vegetative reproduction over hundreds of years makes it dif-
ficult to designate definite characteristics for M. esculenta. Thus, it is believed that
this species did not arise by natural selection. Hybrids between some wild species
may have been domesticated and maintained afterwards through vegetative re-
production. Surely if these cultivars were left to sexual reproduction and subject-
ed to natural selection, this would have led to different populations with specific
gene pools depending mainly on local environments. Our assumption is that do-
mestication included some natural hybrids and that the selected plants were main-
tained by vegetative reproduction for hundreds of years. This assumption is sup-
ported by the fact that many experimental crosses and observations led to frequent
hybridity of cultivars of M. esculenta and local wild species (Abraham, 1975;
Bolhuis, 1953; Cruz, 1968; Jennings, 1959; Lanjouw, 1939; Magoon et al., 1966;
Nichols, 1947). In this genus, systems of genetic and cytologic barriers are not well
established. Additional support may come from Schmidt’s (1951) statement about
the very rapid response of selection in different wild species to increase starch con-
tent in tubers and tuber formation through only a few generations. It seems that
many different wild species have the potential to increase tuber formation and
starch content. I observed two tree species of Manihot (M. epuinosa and M.
brachyandra) frequently grown in dooryards at Goiania with considerable tuber
production. These two species are natives of Bahia. It seems that they were car-
ried by people of this state. Immigration was common during the past 30 years due
to the rapid development of Goias. The assumption that domestication included
hybrids and did not include a certain wild species has been discussed by Rogers
(1963) using the expression “species complexity.”
The place of domestication still needs much discussion. I prefer to use “place
of domestication” and not “center of origin” because it is obvious that this crop
was not brought to existence as a wild species by means of natural selection. Study-
ing the history of ethnological groups in Brazil and their immigrations throws light
on the subject. It is reported that the “Aruak,” who lived in north Amazonia more
than 1000 years ago (Schmidt, 1951), knew cassava and practiced a developed
agriculture. Aruak in the Indian language means “people who eat tubers.” Nu-
merous reports indicate that they cultivated cassava many centuries before the ar-
rival of Columbus. The Aruak were obliged to immigrate in the eleventh century
to Central America, crossing the Caribbean and establishing themselves in the
West Indies. Many reasons were given to explain their immigration, but the most
likely are that they were escaping from enemies or possibly looking for a place
where man does not die. However, the most important reason given was that they
were searching for a better soil to cultivate cassava. However, this immigration co-
incides with the formation of a center of diversity of Mexico. Cassava carried by
the Aruak to Mexico would be expected to hybridize with local wild species, thus
creating a center of diversity. The fact that the Aruak continued on to Planalto Bo-
liviano and to central Brazil is in agreement with the existence of the two centers
of diversity in these regions. The northeastern Brazilian center of diversity is be-
lieved to be the result of immigration of the Tupi-Guarani group.
We must still determine which of these four centers constitutes the primary cen-
ter of diversity of Manihot. In other words, Manihot as a “biological group” must
have passed their differentiation in a certain region from which species spread to
other regions. Central Brazil, with its enormous number of species of Manihot, is
the primary center. Indeed, this region is an ancient area long available for growth
of the angiosperms. Considering Stebbins (1950) explanations of Vavilov’s inter-
pretation of diversity patterns may be useful here: Vavilov’s concept is an elabo-
ration of Willy’s age-and-area hypothesis (i.e., the longer a given biological enti-
ty occupies an area, the more variability of Manihot species), and this area might
constitute its primary center of diversity. This assumption is supported by the fact
that species which exhibit the most primitive characteristics are restricted to this
region: M. stipularis Pax, M. pusilla Pohl, M. longipetiolata Pohl, with their di-
oecious inflorescences, and M. stricta Baillon, M. purpureo-costata Pohl, and M.
salicifolia Pohl, with their nonlobed and sessile leaves.
C. RELATIONSHIPS BETWEEN Manihot SPECIES
According to Rogers and Appan (1973), 98 Manihot species have been recog-
nized. Only 1 species, Manihotoides pauciflora, is known in the closest related
genus, Manihotoides. Several of its attributes are not found in any Manihot
species, including uniflorous inflorescences, which is a primitive characteristic
compared with the multiflowered inflorescence in Manihot, and leaves born at the
apex of short, condensed stems arising from branchlets. Rogers and Appan classi-
fied Manihot species into 19 sections, varying from trees in the section Glazio-
viannae to subshrubs, nearly acaulescent, in the section Stipularis. The species in
this latter section are also characterized by being more dioecious than monoecious,
a condition reversed in all other Manihot species. Other sections, such as Tripar-
titae and Graciles, are perennial subshrubs with large woody roots; their stems fre-
quently die back to the root crown in response to dry periods or fires.
All Manihot species are native to tropical regions of the New World, particu-
larly Brazil and Mexico. Nassar (1978b) defined four centers of diversity for these
species: Mexico and northeast, central, and southwest Brazil. Microcenters of di-
versity of these species exist within central Brazil, where large numbers of species
are concentrated in small areas (50 km in diameter) (Nassar, 1978b,c,d,e,f,
1979a,b, 1980a,b, 1982, 1984, 1985, 1986). Nassar attributes the formation of
these microcenters to the frequent hybridization between species and the hetero-
genic topography of their habitats, which help isolate fragmented gene pools that
lead to speciation. Tree-like species, such as M. glaziovii and M. pseudoglaziovii,
are found in northeastern Brazil, whereas short species and subshrubs are found in
central Brazil.
Natural hybridization occurs between wild Manihot species and between these
and cassava (Nassar, 1984, 1989). Barriers within the genus appear to be weak due
to recent evolution of the group. All wild Manihot species examined cytogeneti-
cally have a chromosome number of 2n 36 (Nassar, 1978a) (Table II). Despite
this high chromosome number, Manihot species behave meiotically as diploids.
Therefore, they are believed to be allopolyploids and this seems to have antici-
pated the emergence of the whole group and is responsible for their rapid specia-
tion and their weak interspecific barriers, leading to interspecific hybridization. An
extremely heterozygous gene pool is thus created, followed by differentiation; this
begins a sequence of hybridization followed by speciation. Nassar (1980a) re-
ported frequent hybridization between M. reptans Pax and m. alutacea Rogers et
Table II
Chromosome Number in Wild Manihot Species
Species Growth habitat n 2n
M. handroana Shrub — 36
M. jolyana Shrub — 36
M. tripartita Shrub — 36
Shrub 18 —
M. tweedieana Shrub — 36
M. humilis Subshrub — 36
M. pedicellaris Shrub — 36
M. gracilis Subshrub — 36
Subshrub 18 —
M. dichotoma Tree — 36
M. giaziovii Tree 18 —
Tree — 36
M. anomala Shrub 18 —
M. zehntneria Shrub 18 —
M. olighanta Subshrub 18 —
M. nana Subshrub 18 —
M. tomentosa Subshrub 18 —
Appan in sympatric natural habitats in which their population boundaries over-

lap. Morphological marker gene leaf color and bract size were used to identify
this interspecific hybridization. The range of M. reptans has expanded during the
past 100 years (Nassar, 1984) and this is attributed to the continuing gene intro-
gression of Manihot species. Introgression of M. reptans with germplasm from
other species allowed its ecotypes to penetrate and colonize areas where M. rep-
tans (pure) had previously been unable to do so. This phenomenon was also not-
ed in other species such as M. cearulescens (Nassar, 1980a). From a plant breed-
ing viewpoint, the value of these hybrids lies in their ability to cross with the
cultigen.
Marker genes lobe shape, the presence of stem nodes, flower disc color, fruit
color, and fruit shape were discovered in controlled crosses between cassava and
wild Manihot species as well as in natural hybrids between cassava and different
species. These genes were used by Nassar to identify hybridization. Interspecific
hybrids of cassava with M. glaziovii, M. pseudoglaziovii, M. aesculifolia, M. pi-
losa, M. corymbilora, M. dichotoma, M. pohlii, M. neusana, and M. anomala were
obtained by Nassar through controlled crosses, although their frequency was low.
The meiotic behavior of several hybrids (cassava with M. neusana and cassava
with M. pseudoglaziovii) was studied by Nassar (1992), and results indicated low
hybrid fertility between these species and cassava.
Grattapaglia et al. (1986) conducted a biosystematic analysis of wild Manihot
species based on soluble seed protein pattern. Nineteen species were analyzed
electrophoretically (Table III). A similarity matrix between species, which con-
sidered differences in band density and number, was established (Table IV). Sev-
eral species were found to be highly similar, for example, M. fruticolosa and M.
pentaphylla and M. pilosa and M. corymbiflora. These results correlate well with
the taximetric analysis made by Rogers and Appan (1973). Manihot pilosa and M.
corymbiflora are the most similar species to cassava. Profile analysis confirmed
the introgression between M. cearulescens and cassava. The electrophoresis of
0.1% of SDS was conducted according to Laemilli (1970). The concentrator gel
containing 5.5% of acrylamide Tris–HCl (pH 6.8) was prepared and fixed dur-
ing 12 h in 5% trichloroacetic acid. The bands were colored by Coomassie bril-
liant blue (0.65%). Every species had its profile revealed in four different gels. The
approximate molecular weights (AMWs) were determined according to Webber
and Osborn (1969).
Table III
Wild Manihot Species and Their Identification Number in the Germplasm Bank
at the Universidade de Brasília
No.
herbarium
Species Section Habitat No. collection
M. esculenta Crantz (var. EAB) I. (A) Manihot Brasília (DF) 01 01

M. esculenta Crantz (var. RB) (B) Manihot Brasília (DF) 02 01/a
M. zehntnieri Ule II. (C) Heterophyllae Goiânia (GO) 173 02
M. grahami Hooker (D) Heterophyllae Maringá (PR) 375 03
M. pilosa Pohl (E) Heterophyllae São Miguel de Antes (MG) 601 04
M. corymbioflora Pax (F) Heterophyllae São Miguel de Antes (MG) 605 05
M. pohlii Wawra (G) Heterophyllae Lençóis (BA) 139 06
M. glaziovii Muell III. (H) Glaziovinae Pentocoste (CE) 221 08
M. pseudoglaziovii Pax et Hoffmann (I) Glaziovinae Remigio (PB) 545 09
M. epruinosa Pax et Hoffmann (J) Glaziovinae Serra Talhada (PE) 554 10
M. brachyandra Pax et Hoffmann (K) Glaziovinae Currais Novos (RN) 524 11
M. reptans Pax IV. (L) Crotalariaeformes Corumbá (GO) 602 13
M. alutacea Rogers et Appan V. (M) Quinquelobae Goiás Velho (GO) 115 07
M. fruticulosa Rogers et Appan VI. (N) Graciles Alexânia (GO) 162 10,938
M. pentaphylla Pohl (O) Graciles Goiás Velho (GO) 103 11,755
M. stipularis Pax VII. (P) Stipulares Alexânia (GO) 184 14
M. salicifolia Pohl VIII. (Q) Brevipetolatae Xavantina (MT) 195 —
M. caerulescens subsp. Caerulescens IX. (R) Caerulescentes Picos (PI) 258 15
M. caerulescens (no classification) (S) Caerulescentes Morro do Chapéu (BA) 567 16
M. caerulescens (no classification) (T) Caerulescentes Jequié (BA) 269 17
M. leptophylla Pax X. (U) Peuvianae Barra do Corda (MA) 517 12
M. neuzana Nassar —. (V) — Maringá (PR) 360 18
Table IV
Classification of Bands in Wild Manihot Species According to Approximate Molecular Weight (AMW)a
Section
I II III IV V VI VII VIII IX X —

No. AMW
(kDaltons) A B C D E F G H I J K L M N O P Q R S T U V
81–75 — — — — — — — — — — — — — — — — — — — — — —
75–66 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 1 3 2 2
66–62 1 1 1 1 1 1 1 — 1 1 1 1 1 1 1 — 1 — — 1 — 1
62–50 3 4 4 3 3 4 4 3 4 4 3 3 3 3 3 1 5 1 1 4 3 3
50–37.5 5 5 6 6 6 6 6 5 6 6 6 4 5 3 4 4 5 4 4 5 3 4
37.5–33 1 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2
33–30 1 1 2 2 1 1 — 2 2 — 2 3 1 2 2 3 1 1 1 2 1 2
30–27 2 2 2 2 2 2 2 1 1 1 1 2 3 1 1 3 1 1 1 1 1 1
27–25 1 1 — — — — — — — — — — — — — — — — — 1 — —
25–24 — — 1 — — — — — — — 1 — — — — — — — — — — 1
24–21 1 1 — — 1 — — 1 1 — 1 1 1 1 1 1 1 — — — — 1
21–20 — — 2 2 — — 1 1 — — — 1 1 — 1 — — — — 1 — —
20–18 1 1 1 2 — — 1 — — 1 — — 1 — — — 1 — — 1 — —
18–13 3 1 2 3 3 3 3 3 2 3 3 3 3 3 2 2 2 2 3 2 3 3
No. bands 21 20 24 24 20 20 21 19 20 19 20 21 20 17 18 18 20 11 12 22 15 20
No. reference 15 15 15 14 15 15 14 15 15 15 15 15 15 14 14 14 14 15 15 15 15 14
bands
No. total 36 35 39 38 35 35 35 34 35 34 36 36 35 31 32 32 34 26 27 37 30 34
bands
aFor identification species see Table II.

To proceed with the analysis of profiles, 15 bands were selected as references;

their aspects were evaluated in all four profiles obtained for every species (Table
V). They were classified into four categories of intensity: A, absent; B, slightly
visible; C, visible; D, dense band; and E, very dense band. To compare quanti-
tatively the protein patterns, the total values of every band were calculated and
expressed in numbers. The protein profiles varied in band intensity, and 15 bands
were selected as references. The variability of wild Manihot species in mor-
phology, growth habit, and geographic distribution was reflected in the elec-
trophoretic profiles as differences in number and intensity of visible bands
(Table VI).
The two studied varieties of cassava showed a 78% similarity with different
species of the section Glaziovinae. Manihot glaziovii Mueller and M. pseudo-
glaziovii Pax et Hoff showed a high index of similarity based on electrophoresis
analysis. A high similarity was also found in species of the section Gracilis in
which its level reached 78%. The same high similarity was found in species of the
section Heterophyllae. Within this section, the species M. pilosa and M. corymbi-
flora showed the highest similarity to the cultivated M. esculenta Crantz, coinsid-
ereding at the same time with morphological affinities between them and cassava.
They are probably part of the complex from which the cultigen had originated
(Nassar, 1978b). The high similarity between species in various sections indicates
their recent speciation and accords with the taxonomic classification. Genetically,
they probably share the same gene pool.
D. GENETIC VARIATION OF WILD Manihot SPECIES
Through the program at the Universidade de Brasília, wild Manihot species

were collected from South and Central America, evaluated, reproduced, and
maintained in a living collection (Nassar, 1986). The description and identifica-
tion of the wild Manihot species was made according to Rogers and Appand
(1973) and Nassar (1978b). Natural habitats were described, and 30 accessions
of each species were examined for the following characteristics: (i) tuber for-
mation and their protein and hydrocyanic acid (HCN) content—tubers were tak-
en 1 year after planting and analyzed according to the Association of Official
Analytic Chemists (AOAC) methods (1970); (ii) adaptation to various soil
types—chemical analysis of soil was carried out according to Black et al.
(1965); (iii) data regarding adaptation to various habitats were extracted from
records of federal meteorological stations; and (iv) seeds, cuttings, or whole
plants of the collected species were planted in a living collection at the Univer-
sidade de Brasília. The collected wild Manihot species were screened rapidly
for tuber formation and growth habit. Results of this screening are given in
Table VII.
Table V
Distribution of Reference Bands According to Density in Studied Wild Manihot Species Profilesa
Section

Reference band
No.AMW A B C D E F G H I J K L M N O P Q R S T U V
01/81 B B B B B B B B B B B B B B B B B B B B B B
02/75 B B B B B B B B B B B B B B B B B B B B B B
03/66 C C C C C C C C C C C C C C C C C C C C C C
04/62 C C C C C C C C C C C C C C C C C C C C C C
05/50 C D C C D C C B C C C C B C B B C B B D C C
06/37.5 C D C C C C C C C C C C B C B B C B B D C C
07/33 E E B B D B C D D C C B B D D D D B B C D D
08/30 C B B C C C C C C E D C B C C B C C C C C C
09/27 C C C C C C C B B D D C B B B A C C C C C C
10/25 B B B B B B C B B B B C B B C B C D D C B C
11/24 B B B C B B B B D B B D B B B B C C C C D C
12/21 D D B B C B C C C C C C B E E E C B B C B C
13/20 D D B A C C A C C C C C B E E E B B C C B A
14/18 D D D C C D C D D D C C C B B B D B B D D B
15/13 D D D C C C C D C D C C C A A C A C C C C C
Note. Abbreviations used: A, absent band; B, slightly visible band; C, visible band; D, dense band; E, very dense band.
aSee Table III for species identification.
Table VI
Matrix of Similarity between Studied Manihot Speciesa
Section

Section
species A B C D E F G H I J K L M N O P Q R S T U V
A — 78 54 45 67 64 58 66 64 58 58 58 50 45 43 43 54 30 32 54 47 50
B — 49 38 68 68 68 61 60 56 54 50 52 42 41 44 52 28 31 53 43 50
C — 62 51 65 48 51 49 51 54 54 59 31 30 32 45 33 33 50 44 40
D — 47 53 65 40 45 40 50 54 47 30 29 30 39 32 33 50 39 59
E — 75 61 70 75 63 74 66 62 46 44 45 60 36 39 66 53 62
F — 58 67 71 67 70 70 71 42 40 41 58 41 44 58 56 56
G — 51 54 51 65 65 52 38 39 38 50 33 35 59 41 78
H — 74 71 60 55 61 49 50 48 56 36 39 50 54 52
I — 59 64 70 45 45 43 43 62 36 38 58 63 60
J — 74 55 52 43 41 42 52 34 36 50 49 48
K — 71 59 41 39 38 52 35 37 64 47 54
L — 59 38 39 46 56 41 45 70 51 59
M — 40 43 50 50 45 44 36 47 50
N — 78 55 50 32 34 37 43 42
O — — 51 33 36 38 37 43
P 36 32 35 37 38 41
Q — 35 35 51 49 56
R — 88 38 49 36
S — 39 50 38
T — 43 58
U — 45
aSee Table III for species identification.

Table VII
Tuber Formation, Growth Habit, and Other Characteristics of the Wild Manihot Species
Tuber Growth Other

Species formation habit characteristics
M. oligantha Pax subsp. Subshrub Abundant cylindric tubers

M. tripartita Mueller Arg. Subshrub Abundant spheric tubers
M. anomala Pohl Tall shrub Abundant spheric tubers with strong
HCN smell
M. zehntneri Ule Tall shrub Abundant conic tubers
M. gracilis Pohl Subshrub Rare tubers; grows throughout cen-
tral Brazil
M. paviaefolia Pohl Subshrub Rare tubers; occurs in poor sandy
soil
M. pruinosa Pohl Subshrub Grows in limestone soil
M. falcata Rogers et Appan Subshrub Grows on slopes and well-drained
soil
M. reptans Pax Shrub Adapted to large range of soil; ex-
hibits different leaf shapes; hy-
bridizes easily with other Manihot
species occurring in its natural
habitat
M. alutacea Rogers et Appan Subshrub Grows on rocky slopes at ⬃ 1200-m
altitude
M. pentaphylla Pohl Subshrub Grows in limestone soil
M. caerulescens Pohl Tall shrub Adapted to dry areas of northeastern
Brazil
M. procumbens Mueller Arg. Subshrub Grow in poor soil, with high con-
centration of aluminum
M. stipularis Pax Subshrub Adapted to high altitude, ⬃1450 m
1. Tuber Formation Pattern and Protein Content
Among the wild species collected from Goias state, Brazil, four species formed
abundant tubers (Nassar, 1978d) and these were screened for tuber formation pat-
tern and fiber and protein content (Nassar, 1978e): M. oligantha Pax emend. Nas-
sar subsp. nestili, collected from Cristalina; M. tripartita Mueller, collected from
Serra Dourada, municipal Goiania; M. zehntneri Ule, collected from Goianesia;
and M. anomala Pohl, collected from Goiania–Inhumas road. These species dif-
fered greatly in tuber formation pattern and tuber composition. Manihot oligantha
subsp. nestili forms abundant cylindrical tubers superficially (10–30 cm below the
soil surface) which are dark brown with a rough surface. Manihot tripartita forms
spherical tubers deep in the ground (50 cm) which are bright brown and smooth
with a creamy cortex. Manihot anomala forms superficial tubers at a depth of about
20–30 cm that are oval shaped with a rough surface and light brown color with a
creamy cortex. Manihot zehntneri forms cylindrical to oval tubers at a depth of
about 50–70 cm which are dark brown with a rough surface and a white cortex
(Nassar, 1978f). Protein and fiber contents are shown in Table VIII. The compo-
sition of cassava as reported in the literature is variable. This variation results from
the fact that bitter cultivars differ from sweet ones not only in the amount of HCN
they contain but also in the proportions of nutrients. According to Bolhuis (1953),
cultivars with roots containing 50 mg of HCN per kilogram are considered
sweet. However, many reports state that crude protein content ranges from 2.2 mg/
kg in sweet to 2.7 mg/kg in bitter cultivars and fiber ranges from 3.1 to 10.3%
(Anonymous, 1968). Notably high percentages of protein occur in wild species in
comparison to cultivated cassava. Some reports have referred to a protein per-
centage in some cassava cultivars as high as 6 or 7% (Anonymous, 1968). Since
this estimation of protein was based on total nitrogen, it must be viewed with cau-
tion because it is not certain whether the breakdown products of cyanogenic glu-
cosides enhance the total nitrogen content. Narty (1969) showed that the hy-
drolytic products of glucosides are incorporated into amino acids for protein
synthesis in cassava. Therefore, it is not unlikely that the reported cultivars with
high nitrogen content were simply bitter cultivars with a high glucoside content.
A wild species attracting attention is M. oligantha subsp. nestili due to its high
protein content combined with a very low level of HCN (Nassar, 1978d). The au-
thor observed cows and horses eating the vegetative parts and tubers of this species
when grazing in its natural habitat. In the literature, two other wild Manihot species
have been reported to have high protein content: M. melanobasis (Jennings, 1959)
and M. saxicola (Lanjouw, 1939). However, because there is no reference to their
HCN content, it is not possible to determine to what extent crude protein estimates
were affected by hydrolytic products of glucosides. It seems logical that there are
wild cassava with a high protein content since selection for cultivation is aimed at
increased tuber size and decreased fiber content and not concerned with protein
Table VIII
Average Protein and Fiber Content of Wild Manihot Species on a Percentage
Dry Matter Basisa
Species Crude protein (%) Crude fiber (%)
M. oligantha Pax subsp. nesteli 7.10 & 0.58 26.67 & 4.86
M. tripartita Mueller Arg. 6.88 & 1.48 33.48 & 6.36
M. anomala Pohl 3.74 & 0.63 23.44 & 4.82
M. zehntneri Ule 3.06 & 0.82 21.52 & 4.82
aFour replicates of 20 tubers of each species were analyzed.

content. This could have led to the discarding of protein-producing genes from the
cultivated varieties.
2. Hydrocyanic Acid Content
Tubers of M. tripartita, M. anomala, M. zehntneri, and M. oligantha subsp.

nestili, which form abundant tubers, and M. gracilis, which grows widely in cen-
tral Brazil, were analyzed for HCN content (Table IX). HCN in fresh unpeeled tu-
bers was within a range of 238 mg/kg in M. tripartita to 62 mg/kg in M. oligan-
tha subsp. nesteli. On a dry matter basis, the results were similar, except that M.
anomala had the highest HCN content. HCN content in cassava tubers is reported
to vary between cultivars.
The analysis of about 100 cultivars for HCN content by Raymond et al. (1941)
indicated an average of 158 mg/kg fresh whole tuber with a maximum value of
438 mg/kg. Little information is available on HCN content in roots of wild Mani-
hot species. Bolhuis (1953) reported 430 mg/kg HCN in fresh roots of M. saxico-
la, but this is probably the only wild Manihot species in which HCN content has
been estimated. He considered that the high HCN content represented an obstacle
to the use of this species in breeding cassava, despite its high protein content. Bol-
huis stated that the minimum lethal dose of HCN for a human being is 50 –60 mg.
However, chronic toxicity due to the continuous intake of small amounts of HCN
is considered more important than acute toxicity because of its association with
many diseases (Oke, 1973).
The occurrence of species with low HCN levels is a valuable discovery. The
species M. oligantha Pax emend. Nassar subsp. nestili, with its notably low HCN
content, can be considered a useful parent.
Table IX
Hydrocyanic Acid Content of Unpeeled Tubers of Wild Manihot Species
HCN content HCN content

in fresh root on dry matter
Species (mg/kg) basis (mg/kg)
M. tripartita Mueller Arg. 281.1a 357.2b

M. anomala Pohl 192.2a 1026.3a
M. zehntneri Ule 125.8b 504.2b
M. gracilis Pohl 97.2c 291.2c
M. oligantha Pax emend. 62.3d 183.2d
Nassar subsp. nestili
Note. Means followed by the same letter are not significantly different using Duncan’s multiple
range test (p 0.5)
3. Growth Habitat and Natural Habitats
Results of screening wild Manihot species for growth habitat and natural habi-
tat are presented in Table VII. Of particular interest is M. paviaefolia, which forms
tubers, has limited vegetative growth, and is adapted to very poor soil. As a par-
ent in cassava breeding programs, it offers potential for additional adaptation to
poor soil conditions. Manihot reptans readily forms hybrids with other species in
its natural habitat, producing intermediate forms. Collections made by Ule in 1892
were restricted to the northern border of Minas Gerais, near Goias (Rogers and Ap-
pan, 1973), but Rogers and Appan found it widespread over most of Goias. In the
past 80 years this species may have expanded its geographical distribution and eco-
logical range through genetic variation and interspecific hybridization. In our sam-
ples of M. reptans, leaf shape was found to vary widely, reflecting the extent of hy-
bridization with other Manihot species. For example, M. reptans from Goias Velho
was distinguished by bright red leaf veins, a characteristic of the native M. alu-
tacea. Manihot reptans was identified by its characteristic growth habit, flower,
and inflorescence morphology. Donations of genes from different species adapted
to different environments could allow this species to expand rapidly over the whole
state of Goias. Harlan (1961) used an example of Helianthus annuus (the annual
sunflower) which has acquired a vast gene pool due to the formation of hybrids
with at least six other Helianthus species.
Manihot pruinosa forms tubers with about 3.8% protein by dry weight com-
pared to cassava with 2%. As seen in Table X, M. pruinosa and M. pentaphylla
may represent a source of adaptation to limestone soils. The adaptation of M. alu-
tacea to high altitude makes it a good candidate for breeding programs concerned
with producing cultivars tolerant to low temperature. Manihot falcata may provide
the potential for breeding cultivars with limited vegetative growth which are adapt-
ed to well-drained soil.
Table X
Analysis of Soil from Natural Habitats of Six Manihot Species
Ca2 Mg2 P K Al3

Species pH (meq/kg) (ppm) (ppm) (meq/kg)
M. paviaefolia 4.9 2 — 16 4
M. pruinosa and
M. pentaphylla 5.5 191 1 136 —
M. caerulescens 5.2 1 1 9 2
M. procumbens 4.9 2 0 18 5
M. stipularis 5.0 3 1 28 6
4. Adaptation to Climatic Conditions
Study of annual rainfall, evaporation, and temperature ranges in natural habi-

tats of wild cassavas showed a species of particular interest: M. caerulescens Pohl
collected from Araripina, state of Pernambuco, and from Posse, state of Goias
(shrubs 1–3 m tall, with a deeply extended root to 2 m underground). They rarely
form tubers: Tubers were intermittent, at depths exceeding 50 cm; the external col-
or was brown, the surface was smooth, and the cortex was white. Protein content
was 3.9% on a dry matter basis and HCN content was 125 mg/kg unpeeled fresh
root. Chemical analysis of soil showed it to be very poor (Table X). A fascinating
aspect of the ecology of M. caerulescens is its habitats in the western part of Per-
nambuco and South Caera, which are among the most semiarid of the world’s trop-
ics. The mean average rainfall in this region is about 500 mm, with a high evapo-
ration capacity and high temperature (Table XI). This unfavorable climate,
coupled with poor soil, suggests that this species is capable of affording a poten-
tial source of resistance to drought. It seems likely that adaptation of M. caeru-
lescens to this arid region depends on its deeply growing root system. However,
this species has some characteristics that distinguish it from other Manihot species.
For example, it has very large ribbed fruit,—four to six times the normal size of
Manihot fruit. I was informed by local inhabitants that seeds are eaten in times of
Table XI
Mean Monthly Precipitation, Evaporation, and Temperature in Natural Habitats
of M. caerulecens
Precipitation Evaporation Temperature

(mm) potential (mm) (°C)
Month Picos Posse Picos Posse Picos Posse
January 98.8 286.2 149.1 129.0 26.3 22.5

February 168.3 89.4 138.4 129.3 25.9 22.2
March 130.3 68.5 124.8 264.9 25.5 25.3
April 31.4 117.5 120.1 167.9 25.4 22.5
May 12.6 16.4 122.4 224.3 24.5 22.5
June 4.1 17.6 121.9 230.9 24.3 22.0
July 1.1 0.7 121.2 364.1 24.1 21.9
August 1.3 20.2 137.7 352.3 26.4 22.8
September 3.2 26.3 147.0 269.1 27.8 24.5
October 17.9 195.2 160.6 202.4 26.1 24.1
November 32.2 204.3 150.3 135.6 26.0 23.9
December 61.7 344.1 145.8 128.0 26.2 24.9
Total 556.7 1386.4
famine. It has a large range extending from northeastern to central Brazil. Some
biotypes of this species have apparently spread through this area. They tolerate a
wide range of environmental conditions from severe drought in the regions of
Araripina, Picos, and Crato in Pernambuco, Piaui, and Ceara states, respectively,
to a considerable amount of moisture at Posse in Goias state (Table X).
Manihot stipularis Pax collected from Brasília, a very short subshrub approxi-
mately 20 cm tall, does not form tubers, has a woody root, and grows on rocks
banks. Soil analysis indicates a poor soil (Table X). This species is characterized
by dioecious flowers. This characteristic, together with its very short height, dis-
tinguishes it from other Manihot species. Manihot stipularis was collected from
an altitude of about 1450 m in one of the highest regions of Brazil. This species
may offer genes for adaptations to coolness. Manihot procumbens Mueller Arg.
collected from Columba is a procumbent subshrub (⬃ 40 cm high) with a large
woody root and yellow latex and grows in very poor soil. It shows potential for
tolerance to soil aluminum toxicity and the absence of major elements. The fact
that wild Manihot species hybridize easily because of their weak interspecific bar-
riers (Nassar, 1984) must have contributed to the vast variation within this group
and the evolution of the large number of species in the genus Manihot.
II. BROADENING THE GENETIC BASE OF CASSAVA,

M. esculenta Crantz, AND DEVELOPMENT
OF INTERSPECIFIC HYBRIDIZATION
A. PRODUCTION OF CASSAVA INTERSPECIFIC HYBRIDS
Cassava cultivars are deficient in many economic characteristics such as resis-

tance to insects, diseases, and drought and have low protein content (Nassar and
Dorea, 1982; Nassar and Grattapaglia, 1986). This can be attributed to the mode
of evolution in the species and modifications of the allogamy system of the plant
(Nassar and O’Hair, 1985). Lost genes can be restored to the gene pool of the culti-
gen by interspecific hybridization with wild relatives which possess these genes
(Nassar et al., 1986). Wild species of cultivated crops have frequently been used
as an important source of genetic diversity and have been employed effectively in
a variety of breeding programs (Stalker, 1980). Controlled introgression of genes
could alleviate stress problems in cassava given the availability of wild relatives
which exhibit diversity in adaptations and attributes. There are interspecific barri-
ers to hybridization (Nassar et al., 1986), but these can be broken in different ways.
Nassar, (1989) reported production of interspecific hybrids of two Manihot
species (M. neusana Nassar and M. anomala Pax) with cassava through controlled
crosses by vector insects. Two wild Manihot species, M. anomala and M. neusana,
maintained in the living collection at the Experimental Biology Station, Universi-

dade de Brasília, were used for this experiment. In October 1982 the species were
planted in three rows alternated with cassava. In June 1983, 200 seeds were col-
lected from each species and replanted in October 1984 for identification of possi-
ble natural hybridization. The following marker genes were used to identify inter-
specific hybrids: variegated color of fruit dominant to smooth, red color of flower
disk dominant to yellow, setaceous bracteole dominant to foliaceous, and noded
stem dominant to smooth. Observations of growth habit, height, stem texture, and
tuber formation were also recorded. In addition to open pollination for the previ-
ously mentioned species, 400 manual crosses with pollen of cassava cultivar Cate-
lo were realized. Of 200 seeds of M. neusana, only 43 seedlings emerged, of which
2 hybrids were identified. Interspecific hybrids were identified by dominant mark-
ers from cassava; noded stem, setaceous bracteoles, ribbed fruit, and tuberculated
root (Table XII). Other characters provided indirect evidence of hybridization. The
200 seeds collected from M. anomala gave rise to 112 seedling. Of these, 3
seedlings showed characteristics of interspecific hybridization. Only 1 seedling sur-
vived to maturity. This hybrid plant exhibited dominant phenotypes from cassava,
namely, ribbed fruit, red color of the flower disk, noded stem, and tuberous root
(Table XII). These results show that glabrous stem, setaceous–foliaceous bracte-
oles, red-creamy color of flower disks, variegated-green color of fruit, and ribbed–
nonribbed fruit are simple marker genes that can be used to recognize interspecif-
ic hybridization. It is evident that interspecific barriers between Manihot species
can be broken by the use of an abundant diversity of pollinator gametes transmit-
ted by insect vectors. Interspecific crosses were difficult to fertilize manually in the
Table XII
Comparison of Morphological Characteristics of M. neusana, Cassava, and Their Hybrid
Characteristic M. neusana Cassava Hybrid
Growth habit Procumbent shrub (1.5–2 m) Erect shrub (1.5–2 m) Erect shrub (1.5–2 m)
Young stem texture Hairy Glabrous Hairy
Bracteoles Foliaceous Setaceous Setaceous
Fruits Globose, without Ovoid, ribbed, green Ovoid, ribbed,
ribs, variegated variegated
Tuber formation None Forms tubers Forms tubers
Growth habit Erect shrub (2–2.5 m) Erect shrub (1.5–2 m) Erect shrub (1.5–2 m)
Young stem texture Hairy Glabrous Hairy
Bracteoles Semifoliaceous Setaceous Setaceous
Flower disk color Creamy Red Red
Leaf form Anomala Lobed; 5 lobes Anomala
Fruit Globose, without ribs Ovoid, ribbed Ovoid, ribbed
Tuber formation Scarely forms tubers Forms tubers Forms tubers
present and in previous crosses (Nassar, 1980a). The evidence suggests that barri-
ers between cassava and other Manihot species are weak and recently evolved. It
seems they have arisen not as a primary isolating event but rather secondarily after
geographic isolation. Nassar (1978b) postulated that cassava is an interspecific hy-
brid that appeared by domestication approximately 2000 years ago or less.
B. DEVELOPMENT OF CASSAVA INTERSPECIFIC HYBRIDS
The wild Manihot species of M. neusana Nassar was hybridized with the cassa-
va clone Catelo through controlled hybridization with the help of pollinating insects
(Nassar, 1989). An interspecific hybrid that combined marker genes of both parents
was obtained. The marker genes were ribbed fruit, acquired from cassava, and var-
iegated fruit color from M. neusana. This hybrid (HN) was backcrossed with cassa-
va and used as a pollinator in the first trial and as a fruit carrier in the second trial.
Seeds were obtained from both crosses, but only one plant could be raised from each;
HO1 was the result of the interspecific hybrid (HN) as maternal plant (seed carrier),
and HO4 resulted from crosses in which the interspecific hybrid (HN) was used as
pollinator. The three hybrid plants (HN, H1, and H4) were studied cytogenetically
for both meiotic and mitotic behavior. For the study of meiosis, inflorescences were
fixed in a mixture of three parts absolute alcohol and one part glacial acetic acid and
kept in a refrigerator for 24 h. The anthers were smeared with acetic carmine. Chro-
mosome configurations in the metaphase, chromosome distribution in anaphase I,
and tetrad formation were also studied. Pollen viability had been determined by us-
ing acetocarmine and iodine stain (Nassar, 1978a). For the mitotic study, root tips
were left in 0.2% colchicine for 2 h and then fixed in acetic alcohol for 24 h. Before
smearing with acetocarmine, they were treated with 1 N HCl for 10 min.
1. Meiotic Behavior of the F1 Hybrids (HN)
One hundred pollen mother cells (PMCs) were studied in metaphase I of the in-
terspecific hybrid M. neusana with cassava; 30 PMCs in metaphase II and 1000
tetrads of the same material were also investigated. The study of metaphase I
showed different chromosome configurations, as shown in Table XIII. The aver-
age bivalent frequency in all cells of metaphase I was 16.13 per cell. The high fre-
quency of univalents was attributed to lack of synapses between chromosomes or
failure of the two species to remain associated. Virtually the only other report on
this subject is that of Magoon et al. (1970), in which chromosome pairing in the
interspecific hybrid M. glaziovii (rubber tree) and cassava was studied, and a reg-
ular synapsis led the authors to conclude that there is a strong relationship between
this species and cassava. Nassar et al. (1986) suggested that the material used by
Magoon et al. was not a pure M. glaziovii but rather a natural interspecific hybrid
between this species and cassava. If this is true, the supposed interspecific hybrid
Table XIII
Frequency of Chromosome Configuration of Metaphase I in Interspecific
Manihot Hybrids and Their Parents
Mean
PCMs (No.) Trivalents Bivalents Univalents
M. neusana 20 — 18.00 —
Cassava 20 — 18.00 —
GN 100 — 17.00 1.58
HO1 30 1.86 16.13 0.13
HO4 100 1.63 12.41 8.84
would be a backcrossed progeny. The study of anaphase I showed that of 40 PMCs

studied, 38 cells exhibited laggards, which were attributed to the occurrence of
univalents resulting from nonhomologous chromosomes.
Anaphase II showed meiotic restitution. Of 33 PMCs studied in this phase, 5
cells exhibited a second meiotic restitution (SMR), forming 36 chromosomes on
each pole. Apparently this phenomenon is a consequence of meiotic disturbance
in the hybrid. An example of this disturbance was the breakdown of anaphase I.
This was probably due to disharmony between the two different genomes. Nassar
(1991) documented this phenomenon in the interspecific hybrids of cassava with
M. pseudoglaziovii. The presence of such restitution was confirmed in the follow-
ing tetrad stage, in which the formation of both dyads and tetrads was observed.
In various crops, interspecific hybridization has led to the disturbance of mei-
otic division, with consequent meiotic restitution, e.g., in Trifolium pratense by
Parrot and Smith (1984) and in Medicago spp. by Vorsa and Bingham (1979). In
manioc species, Hahn et al. (1990) reported 2n pollen formation in wild species in
addition to certain clones of cassava. The detection of this phenomenon enabled
these researchers to isolate triploid and tetraploid types from progeny that came
from crosses of cassava with certain wild Manihot species, namely, M. glaziovii
and M. epruinosa. These types proved much more productive than commercial
clones used in Nigeria. Nassar (1991) manipulated the meiotic restitution occur-
ring in interspecific hybrids of M. pseudoglaziovii with cassava to produce triploid
types that showed very good productivity under semiarid conditions. The discov-
ery of the frequent occurrence of this phenomenon in interspecific hybrids of cas-
sava offers an effective tool for the production of polyploid types by sexual means
instead of the traditional method of colchicine applied to vegetative parts, which
normally induces unstable, chimeral types (Abraham et al., 1964). An additional
advantage is that production of triploid types may lead to production of trisomics
among their progeny. If genes which control productivity in cassava are polygenes
with addictive model action, as is the case for many crops, certain trissomics of
this crop may be more productive than their diploid ancestors. In general, the pro-
duction of polyploidy via sexual means is advantageous from both genetic and
evolution standpoints because it offers a vigorous heterotic effect and releases use-
ful genetic variability adaptations.
For the study of the tetrad stage, 1065 PMCs were investigated. Of these, 62 cells
formed dyads and 62 cells formed micronuclei. The presence of dyads and tryads at
this stage confirms that observed earlier in the anaphase: the occurrence of first- and
second-divisioin meiotic restitution (FMR and SMR). Both types are capable of pro-
ducing 2n gametes. However, the FMR is more valuable than the SMR since it pre-
serves the major part of its heterosis and epistatic interaction (Mendiburu and Pelo-
quin, 1977; Parrot and Smith, 1983; Vorsa and Bingham, 1979).
2. Cytogenetic Behavior of the Backcrossed Generation (HO1)
Table XIV shows the frequency of chromosome configurations at metaphase I.

Bivalents, trivalents and univalents were present, with a mean of 16.1 bivalents,
1.86 trivalents, and 0.13 univalents. The presence to trivalents indicated aneu-
ploidy in this hybrid. This was confirmed by mitotic counting, which showed 2n
38, i.e., 2n 2. In anaphase I, the presence of laggards with different frequen-
cies was recorded. The study of 900 tetrads showed 644 normal ones, 218 mi-
cronuclei, 12 dyads, and 26 tryads. Analysis of pollen viability revealed that only
35.8% were viable (Table XV).
3. Cytogenetic Behavior of the Backcrossed Generation (HO4)
One hundred PMCs at metaphase I were studied. Again, bivalents, trivalents,

and univalents with averages of 12.4, 1.6, and 8.8, respectively, were observed.
The total chromosome count for the different configurations was 38. This showed
a constitution of 2n 1 1, which was confirmed by root tip mitotic counting.
In anaphase I, of 32 PMCs studied, 31 proved to have laggards. In anaphase II, 35
PMCs were examined; of these, 7 cells appeared as restitution nuclei, and this was
later confirmed in the tetrad stage. In the tetrad stage, 1196 sporocytes were ob-
Table XIV
Diploid Pollen in Interspecific Manihot Hybrids and Their Parents
Diploid pollen
Pollen grains
examined (No.) No. %
M. neusana 818 3 0.36

Cassava 1162 3 0.26
HN 1235 20 1.62
HO1 1128 8 0.71
HO4 1007 6 0.60
Table XV
Viability of Pollen of the Interspecific Manihot Hybrids and Their Parents
Viable pollen
Pollen Nonviable
analyzed (No.) No. % pollen
M. neusana 1001 818 81.72 183

HN 1235 1162 94.09 73
HO1 1830 655 35.80 1175
HO4 1542 273 17.70 1269
served. Of these, 326 were normal, 826 contained micronuclei, 25 were tryads, and
19 were dyads. The study of pollen viability showed a very low viable grain per-
centage of 17.7% (Table XV).
4. Cytogenetic Behavior of the Parents
The cassava clone EB 01 showed a regular meiotic division in all of the 20

PMCs studied, with 18 bivalents formed. Of the 950 tetrads that were examined,
942 were normal, 5 contained micronuclei, and 3 were tryads (Table XVI). The
study in which M. neusana was used as a parent showed it to have a regular meio-
sis with 18 bivalent formations. Of 1011 tetrads studied, 1003 were normal, where-
as 6 had micronuclei and 2 were dyads. The pollen viability was 81.72% (Table
XV). Mitotic counting of root tips showed 2n 36. This was the first report of a
chromosome number for this new species described by Nassar (1985).
5. Evolutionary and Plant Breeding Significance
The fertility of the hybrid HO1 indicates the possibility of further manipulation
of this hybrid through backcrosses to transfer useful genes of M. neusana to cas-
sava. The backcrossed generations produced were aneuploid 2n 1 1 in both
Table XVI
Analysis of Tetrads of Interspecific Manihot Hybrids and Their Parents
Tetrads PMCs Tryads Dyads
Total Normal % No. % No. % No. %
M. neusana 1011 1003 99.22 6 0.59 2 0.19 — —

Cassava 9502 942 99.15 5 0.53 3 0.31 — —
HN 1065 694 65.15 262 24.60 62 5.82 47 4.41
HO1 900 644 71.55 218 24.22 26 2.88 12 1.33
cases studied (HO1 and HO4). In the case of hybrid HO4, the plant was completely
sterile, having a chromosomal constitution of 2n 1 1. Obviously, this hybrid
resulted from fertilization of a pollen gamete n 1 1 of the parent hybrid (HN),
with a cassava ovule of “n.” On the other hand, when the interspecific hybrid HN
was used as a maternal plant, a fertile progeny was obtained. When it was used as
a pollen parent in the backcross with cassava, it resulted in the production of a ster-
ile progeny (HO4). This was probably due to the elimination of fertile embryo
genotypes in the progeny because of incompatibility or disharmony between them
and the endosperm.
The partial fertility of the backcrossed generation (HO1) shows that the species
M. neusana may be classified within the secondary gene pool of cassava accord-
ing to the concept of Harlan and de Wet (1971). Other Manihot species that may
be included in this category are M. melanobasis (Jennings, 1959), M. glaziovii
(Magoon, 1970), M. reptans, M. zenhtneri, M. anomala, M. oligantha, M. pohlii
(Nassar et al., 1986), and M. dichotoma, M. epruinosa, and M. leptophylla (Hahn
et al., 1990). It was concluded that the cassava hybrid with M. neusana showed ir-
regular meiotic behavior in the lack of complete chromosome pairing, formation
of univalents in metaphase I, chromosome retardation in anaphase I, micronuclei
in the tetrad, and meiotic restitution. When backcrossed to cassava, the interspe-
cific hybrid produced two aneuploids 2n 1 1. These showed irregular meio-
sis, partial chromosome pairing, and the presence of meiotic restitution. The two
backcrossed F2 hybrids differed in regard to pollen-viable genetic constitutions.
The meiotic restitution continued to occur in F2 hybrids, which showed that it must
be correlated with interspecific meiotic irregularity.
C. DEVELOPMENT OF CASSAVA INTERSPECIFIC HYBRIDS

FOR SAVANNA (CERRADO) CONDITIONS
Interspecific hybrids of cassava with M. glaziovii and M. pseudoglaziovii were

produced by Nassar in 1991 (Nassar et al., 1996a). They were propagated by cut-
tings and planted in November 1992 alternately with clone Sonora. This clone was
chosen because of its high resistance to bacteriose. Seeds were collected from cas-
sava and planted in November 1993. Out of 182 seeds whole plants. In March
1994, these plants were reproduced by cuttings; six of each were planted for eval-
uation of productivity and survival during the drought season of June to October
(5 months). In November 1994, plants which survived were evaluated for root for-
mation. Ten clones were selected and given to the Semi Arid Centre at Pernam-
buco for propagation and cultivation under semiarid conditions of northeastern
Brazil. The selected clones were characterized morphologically according to
Rogers and Appan (1973) and Nassar and Grattapaglia (1986). This characteriza-
tion was aimed at detecting the association of different morphological characters
with tolerance to drought. Since polyploidy offers another alternative for this tol-
erance, the previously mentioned interspecific hybrids and several others available
in our living collection were investigated cytologically to detect formation of di-
ads and triads as evidence of unreduced 2n microspore production (Vasquez and
Nassar, 1994). This will enable the selection of possible progenitors of polyploid
types.
For cytogenetic study, 10 interspecific hybrids and/or their progenies were used:
F1 M. glaziovii
cassava (3 genotypes), F2 M. epruinosa
cassava, F2 M. anom-
ala
cassava (3 genotypes), F3 M. pseudoglaciovvi
cassava (2 genotypes), and
F4 M. pseudoglaciovvi
cassava.
Chromosome association in meiotic metaphase I and the occurrence of dyads,
tryads, and pollen viability were studied. For the meiotic study, inflorescence was
fixed in a 3:1 mixture of absolute alcohol and glacial acetic acid and kept at 5C
for 24 h. The anthers were smeared in acetocarmine stain. Chromosome configu-
rations at metaphase I and sporad formation were studied. For the pollen viability
study, one or two flowers per plant were selected, and their pollen was crushed
from anthers in acetocarmine preparations and scanned for viability. Pollen counts
and percentage of stained normal pollen were calculated.
Of the 182 seeds planted, only 35 germinated, established, and developed to the
flowering phase. This is due to dormancy of wild seed retained in F2 progeny.
These 35 plants were multiplied by cuttings to evaluate their performance for root
production. Only those plants which gave more than 2-kg roots with a medium 2.3
kg per plant by 8 months were selected for multiplication for further cultivation.
Morphological characterization showed that certain characters were associated
with tolerance to drought. As seen in Tables XVII and XVIII, all selected clones
have a notable brown, thick, and rough superficial epiderm. It seems that the
brown-colored thick epiderm is associated with tolerance to drought because of its
isolative nature, which impedes evaporation. All the wild species investigated by
the author have fibrous roots with brown external color and their epidermic layer
is thick. It is believed, therefore, that this character is inherited from the wild.
Graner (1942) reported that this character behaves dominant to white. Anatomi-
cally, the distinct portion of the enlarged root is composed of three sections. First,
a layer referred to as the phelloderm which is generally composed of the previ-
ously mentioned epidermis, a subepidermis, and a thicker inner layer. The phello-
derm is thick and easily separated from the next inner layer. Second, a layer of
parenchymatous cells that constitutes the bulk of the root and is the carbohydrate
storage region. Third, a portion called the cortex of flesh at the center of the root
is a well-defined central vascular core. As noted previously, the outer epidermis is
so fine that it is difficult to measure, but it is possible to evaluate its thickness us-
ing the naked eye. It is about 0.5 mm in the thickest types.
The second interesting case in selected clones is the prominence of leaf scars on
stems. All selected clones have a prominent enlarged leaf scar. This character
Table XVII
Characters Used in Clone Characterization
Root characters
1. Surface of root
A. Smooth
B. Rough
2. Thickness of epiderm layer
A. Thick (>. 0.3 mm)
B. Thin (<. 0.3 mm)
3. External color of root
A. Light brown-yellow
B. Brown, dark brown, reddish brown
C. Light brown, tan, light tan
D. Pinkish brown, pinkish tan
E. Pinkish white, light pink, pink
4. Root flesh color
A. White to cream
B. White to cream with pink
C. Cream-yellow to yellow
D. Cream-yellow to yellow with pink
5. Color of stem
A. Silver
B. Silver-brown
C. Brown
D. Yellow
6. Nature of scars on stem
A. Smooth
B. Slightly raised
C. Moderately raised
D. Very large
7. Branching of plant
A. One branch at top or no branches
B. One or two branches but not one branch at top
C. More than two branches
Leaf characters
8. Leaf lobe shape
A. Obovate
B. Linear
9. Sinuosity of leaf lobes
A. Pandurate
B. Some sinuosity
C. Simple (not sinuous)
D. Logical (linear)
10. Petiole color
A. Red
B. Greenish red
C. Reddish green
D. Green
11. Color of young foliage
A. Reddish blue
B. Bluish green
C. Green
Table XVIII
Morphological Characteristics of Selected Clones
Characteristic 1 2 3 4 5 6 8 9 10 11
1. Surface of root B B B B B B B B B B
2. Thickness of epiderm layer A A A A A A A A A A
3. External color of root B B B B B B B B B B
4. Root flesh color A B B C A B A A A A
5. Color of stem B B B B B B B B B B
6. Nature of scars on stem C C D C C C C C C D
8. Leaf lobe shape B A A A B A A B A A
9. Sinuosity of leaf lobes A A A D A A C B A D
10. Petiole color B D A A A C D B B B
11. Color of young foliage C C A C A A C B C B
Note. Numbers and letters correlate with those used in Table XVII.
which seems to be well associated with enlarged root formation in the hybrid prog-
eny, apparently came from cassava. All wild species studied by the author have a
smooth stem without any leaf scar. All selected clones gave deep fibrous roots in
addition to enlarged ones. It seems that this character is inherited from the wild.
This appears to be a mechanism of cassavas to tolerate drought since they are ca-
pable of capturing water from long distance. Both wild species and their interspe-
cific hybrids produced long, deep roots from the fourth month onwards, reaching
4 or 5 m when the plants were 1 year old.
Another mechanism of tolerance to drought is the thick epiderm layer, proba-
bly because of its structure, it impedes evaporation. The dieback of the vegetative
parts to the crown in dry season was the third common character shown by all the
selected clones. Presumably this habit helps plants to reduce respiration and con-
sumption of carbohydrate deposits. From this study, it is obvious that breeders can
make use of morphological characterization as a criterion to detect the association
of morphological characters and drought tolerance and consequently selection of
genotypes complying with this objective.
D. OVERCOMING CROSSING BARRIERS BETWEEN

CASSAVA, M. esculenta Crantz,
AND A WILD RELATIVE, M. pohlii WARWA
Manihot pohlii has a high resistance to stem borers, Coelosternus spp., and bac-
teria Xanthomonas Manihotis. Hybrids of M. pohlii and other species with cassa-
va can be obtained but only at a very low frequency due to interspecific crossing
barriers (Nassar et al., 1986). However, one technique for overcoming such barri-
ers involves the use of pollen mixes which combine the pollen intended for syn-
gamy with “mentor” pollen of the maternal species. This technique has been re-
ported for many plant taxa, including Populus (Knox et al., 1972), Malus (Day-
ton, 1974), Cosmos (Howlett et al., 1975), and Petunia (Sastri and Shivanna,
1976). The role of the mentor pollen is to facilitate fertilization by foreign pollen.
Apparently, the mentor pollen supplies proteinaceous substances which permit the
foreign incompatible pollen grains to germinate (Knox et al., 1972). A treatment
destroys the generative function of the pollen grain without affecting germination
and growth of the pollen tube and consequently does not affect stimulation. The
stimulating effect of destroyed pollen is due to protein recognition substances.
They are liberated by pollen grains while germinating and have enzymatic and
antigenic properties. These substances are localized in the internal layer of the
pollen surface and are correlated with pollen germination and growth of its tube
on the stigma surface. They have been localized by cytochemical means in the cel-
lulose intine (Knox et al., 1972). Many studies using this technique have been
successful in overcoming an incompatibility barrier (Brewer and Henstra, 1974;
Williams and Church, 1975). Subsequent work indicated the need to refine the
preparation of the mentor pollen by using freeze–thaw cycles or methanol treat-
ment (Knox et al., 1972). Nassar et al. (1996b) successfully used freezed mentor
pollen in hybridizing cassava with M. pohlii. In addition to the mentor effect, oth-
er techniques seem to have potential for overcoming interspecific barriers. One of
these may be the use of a bridge species. Nassar et al., inspired by the capacity of
M. neusana Nassar to cross easily with all Manihot species growing in the vicin-
ity, used it as a bridge species. Very little information is available regarding the
bridge technique. Probably the only case is that of Dionne (1963), who used
Solanum acaule Bitter as a bridge between Solanum tuserosum L. and Solanum
bulbocostanum Dunal.
A natural hybrid between M. pohlii and M. neusana Nassar was used by Nassar
et al. (1996b) as a bridge species. The hybrid was identified by fruit marker genes,
which produce variegation of fruit color in M. neusana and a straight white line in
M. pohlii fruit. Crosses of this hybrid (named HNP) and cassava (clone EB 05)
were carried out from January 1993 to January 1994. Flowers were taped shut for
2 days until they had been pollinated manually. Pollination of both the hybrid HNP
and cassava was done with pollen mixes of cassava and M. pohlii. Manihot pohlii
pollen used as mentor pollen was sucessively frozen for 5 min at 4C and thawed
for 30 min for a period of 105 min. The purpose of this treatment was to kill the
mentor pollen and increase the chance of obtaining interspecific hybrid seed. To
verify the presence of any autoincompatibility that would interfere in the con-
trolled crosses in M. polii and cassava, a controlled autopollination was undertak-
en. Crosses between cassava and M. pohlii (POH) were carried out using mentor
pollen in one trial and untreated pollen in the second trial. Seeds were collected
from both crosses and planted during the next growing season.
The pollination of M. pohlii by untreated cassava pollen did not produce any
fruit set (Table XIX), whereas crosses with mixed pollen of cassava and mentor
resulted in the production of 21 seeds, which is 4.9% of the possible maximum
(assuming that every fruit has three ovules). Using cassava as a maternal parent,
pollination by untreated pollen M. pohlii did not result in any seed. This clearly
demonstrates the effect of mentor pollen in the crosses of cassava with M. pohlii.
It likewise means that the freeze–thawing treatment administered to M. pohlii
pollen, although it did kill the pollen, did not affect its stimulatory function so that
all the seed produced by the mentor effect had embryos and endosperm. This re-
sult indicates that only the stigmatic barrier functions in preventing crossing in
these species.
Postfertilization mechanisms fail to prevent crossing. Only one plant ger-
miseeds to severe dormancy. This plant bore fruits carrying the marker genes of
both M. pohlii and cassava: a straight white line from M. pohlii and winged fruit
from cassava. The mentor effect has also been successfully used in Populus and
Cosmos. In these genera, interspecific incompatibilities have been overcome by
using compatible but dead pollen (Knox et al., 1972). These studies suggest that
this phenomenon is due to the release of proteinaceous recognitioin factors from
the wall of the killed compatible pollen, masking the rejection reaction of the re-
cipient stigma. Nassar et al.’s report (1996b) represents the first case of obtaining
hybrid seed of M. pohlii and cassava and its further reproduction. Despite the use-
ful characters of M. pohlii, no successful breeding program has been carried out
due to a lack of hybrids between this species and cassava.
The use of M. neusana as a bridge species through the hybrid M. neusana–M.
pohlii has improved seed set. When used as a maternal parent pollinated by cas-
sava, it gave seed in 3.5% of cases, whereas the reciprocal crosses had greatly im-
proved seed set, yielding 25.9% (Table XX). The combined treatment of both
species (bridge and mentor) produced 3.5% seed production. The success of M.
neusana as a bridge species between M. pohlii and cassava may occur because the
two genomes of M. neusana carry different genetic mechanisms of cross incom-
patibility. This hypothesis is confirmed by observations on crossing behavior of
this species in the living collection of wild Manihot species. Manihot neusana has
Table XIX
Crosses Attempted between M. pohlii and M. esculenta (Cassava) and Numbers of Fruit
and Seed Produced
Treatmenta Flowers pollinated No. Fruit No. Seed
1. Cassava
M. pohlii 145 0 0
2. M. pohlii
cassava 142 10 21
a1, Without mentor; 2, with mentor.

Table XX
Crosses Attempted between Cassava and M. neusana Hybrids and Numbers
of Fruit and Seed Produced
Treatment Flowers pollinated No. Fruit No. Seed
HNP
cassava 161 14 17
Cassava
HNP 90 14 70
HNP
cassava mentor 208 198 22
hybridized naturally with several Manihot species grown in the vicinity, e.g., M.
pseudoglaziovii Pax et Hoffmann, M. glaziovii Mueller Arg., M. tripartita Mueller
Arg., M. caerulescens Pohl, M. salicifolia Pohl, M. pohlii, and M. esculenta (Nas-
sar, 1989). Its hybrids with the previously mentioned species were easily identi-
fied by the marker dominant gene of variegated fruit which came from M. neu-
sana. Manihot neusana is a newly emerging species, described and identified by
Nassar (1985).
III. DEVELOPMENT AND SELECTION FOR APOMIXIS

IN CASSAVA, M. esculenta Crantz
A. GENETIC STUDY OF APOMIXIS IN CASSAVA
Cassava is a labor-intensive crop. It is propagated vegetatively by stem cuttings

which are often responsible for the spread of diseases and pests in the tropics. Nas-
sar and O’Hair (1985) hypothesized that the use of seed to replace stem cuttings
for cassava production would eliminate these problems and potentially reduce pro-
duction costs. One limiting factor, though, is the lack of rapid and uniform seed
germination. Another difficulty is genetic segregation and lack of true breeding
lines. Apomixis, which is synonymous with agamospermy or asexual seed forma-
tion (Asker, 1979), may provide a solution to these problems. It can be used as a
means for fixing heterozygosity responsible for vigor, high productivity, and wide
adaptation of cassava clones to different conditions of soil and climate (Nassar,
1992).
Nassar began a project in 1981 with the objectives of selecting for easy germi-
nating seed of cassava clones (Nassar and O’Hair, 1985) and of examining for
apomictic occurrence within the selected populations. Three clones (011, 040, and
060) were selected from an easy germinating population developed by Nassar. The
offsprings of these clones were studied through several generations for transfer-
ence of combinations of the following characters: abundant fruit formation in
plants which have sterile pollen, uniform progenies identical to the maternal types,
and the occurrence of multiple seedlings per seed as described by Asker (1979);
and segregation of several distinct characters of root [external color and surface
morphology (smooth or rough)], stem [color and surface morphology (smooth or
rough)], stem [size of scars and storey length (distance between two ramifica-
tions)], flower (petal color), and fruit (color) as defined by Rogers and Fleming
(1973) and Nassar and Grattapaglia (1986). The presence of apomixis was con-
firmed experimentally by controlled crosses made between clones 011, 040, and
060, which carry the recessive gene for green, young shoots, and clone 051 used
as a pollinator. Clone 051 carries the dominant, homozygous gene for dark red,
young shoots which serves as a marker gene. An F2 clone, 031, was included in
the crosses because it behaved as if it were apomictic. Cytogenetical studies were
conducted with mature plants of these selected clones to examine their meiotic be-
havior as follows: Flower buds were fixed in acetic acid: ethanol 3:1 (vol/vol) for
24 h, preserved in 70% ethanol, smeared and stained with 1% acetocarmin, and
then examined microscopically. Fifteen pollen mother cells of each clone were ex-
amined at metaphase I for chromosome association (Nassar, 1978a). Pollen via-
bility was estimated by staining with 1% acetocarmin. Round, regular-shaped, and
regular-sized pollen were considered viable.
Crosses of clone 021, an interspecific hybrid of cassava (M. esculenta
M. di-
chotoma used as a pollinator), and cultivated clone Branca Santa Caterina were
carried out to transfer apomictic genes from the wild species M. dichotoma. This
species, maintained in the living collection by the author, has been observed to
have apomictic behavior. One hundred F2 seeds resulted in 72 seedlings from
which the 3 most vigorous were selected and allowed to open pollinate and bear
fruit. One hundred seeds from each plant were collected and planted in rows, and
the raised offspring were examined for apomixis using the criterion for transfer-
ence of maternal combination of characters as described previously. This enabled
selection of clone 031 which was included in the controlled crosses. There were
variations in transference of maternal type to the progeny showing different de-
grees of apomixis.
Meiotic study (Table XVIII) revealed that clones 011, 031, and 040 were aneu-
ploids having 37 chromosomes compared with 36 chromosomes (18 bivalents) in
normal meiosis of cassava clone 060 and cultivar Branca Santa Caterina. The chro-
mosome associations in clones 011, 031, and 040 consisted of 17 bivalents and 1
trivalent (Table XXI). Pollen viability from offsprings of the easy germinating
clones was low (17.8–26.0%) compared with that of the normal clone 060 (89.9%)
and cultivar Branca Santa Caterina (86.5%). Gustafsson (1946) reported a corre-
lation between apomixis and polyploidy, specifically euploidy. Brown and Emery
(1958) found that apomixis in the Gramineae subfamily Ponicoideae was restrict-
ed to euploid species. Hybridization between cassava and its wild relatives was re-
ported to occur frequently in nature (Nassar, 1992). Wild Manihot species, con-
Table XXI
Pollen Viability, Chromosome Number, and Predominant Chromosome Association
of an Easy-Germinating F2 Hybrid and Cultivar Branca Santa Caterina
Easy-germinating F2
Branca Santa
Clone 011 040 060 031 Caterina
Pollen viability 20.4 17.8 89.9 26.0 86.5

Chromosome No. 37 37 36 37 36
Chromosome association 17 1 17 1 18 17 1 18
sidered the source of many useful characters (Nassar, 1978b), seemed to have con-
ferred apomixis to cultivared cassava. Specifically, the interspecific hybridization
of Brazilian cassava indigenous clones (Nassar, 1978b, 1992) may be responsible
since their crossings resulted in partial or complete sterility as well as formation
of aneuploid forms. Although obligate apomixis is clearly most favorable for use
in cassava breeding, a low percentage of seeds produced by gamete formation and
fertilization have been known to occur in different taxa of different crops which
were considered obligate in the past (Asker, 1979). Asker questions if a 100% ob-
ligate exists.
B. MOLECULAR AND EMBRYONIC EVIDENCE

OF APOMIXIS IN CASSAVA
To confirm the presence of apomixis and study its anatomic nature in cassava,
Nassar et al. (1997, 1998a) selected two putative apomictic cassava clones. The
first was identified by number 031 based on vigor in an F2 population resulting
from a cross between an interspecific hybrid (M. dichotoma
M. esculenta) and
cultivated clone Branca Santa Catarina. Clone 200 is an F1 individual from hy-
bridization between cultivated cassava and M. glaziovii. Progenies from both
clones were obtained by open pollination. Thirty-seven progeny plants of each
clone were used in the anatomic and molecular studies.
1. Embryo Sac Analysis
Morphostructural development of embryo sacs was studied histologically. Un-

pollinated pistillate buds at presumably 1 day preanthesis and pollinated pistils at
postanthesis were sampled and immediately fixed in Farmer’s fixative (1:3
glacial acetic acid:95% ethanol) in the field between 7:30 a.m. and 12:00 p.m.
Fixed pistils were dissected under a dissection microscope (magnification
40,
transmitted light). Dissected ovules were dehydrated in ethanol series and cleared
overnight in the benzyl-benzoate-four-and-a-half fluid (lactic acid:chloral hy-
drate:phenol:clove oil:xylene:benzyl benzoate 2:2:2:2:1:1, w/w) devised by
Herr (1982) and treated in a modified Herr’s fluid as previously reported by Og-
buria and Adachi (1994). Transparent ovules were then observed using an Olym-
pus BX50 microscope equipped with Nomarski’s differential interference contrast
(DIC) optics and a 100-W high-pressure mercury lamp with appropriate filters for
optimal viewing and photography. Both megasporangia and megagametophyte
components were photographed and printed by a Sony color video printer Mavi-
graph UP-1200. One hundred twenty-seven ovules from clone 031 and 134 ovules
of clone 200 were used in the embryo sac analysis.
2. DNA Extraction and RAPD Assay
Total genomic DNA was isolated from 200 mg of fresh leaf tissue ground in liq-
uid nitrogen using the CTAB protocol of Doyle and Doyle (1987), modified by the
addition of 1% PVP and 1% 2-mercaptoethanol to the isolation buffer. DNA con-
centration was estimated by gel electrophoresis comparing the fluorescence in-
tensities of the ethidium bromide-stained samples to those of lambda DNA stan-
dards. For the RAPD assay, working stocks of genomic DNA were diluted in water
at a concentration of 2.5 ng/ l. Arbitrary 10-base primers (kits OP-A–OP-Z) were
obtained from Operon Technologies, Inc. (Alameda, CA). Amplification reactions
(13 l) were carried out according to William et al. (1990) with the following mod-
ifications: 0.4 mM 10-base primer, 10 g/ul nonacetylated bovine serum albumin
(New England Biolabs), 5–10 ng genomic DNA, and 1 unit of Taq DNA poly-
merase.
Amplifications were performed in 96-well microwell plates using a MJ Re-
search PT-100 thermal controller. RAPD products were analyzed by electrophore-
sis in 1.5 or 2.0% agarose gels containing 0.2 mg/ml ethidium bromide. Gel im-
ages were captured and digitalized with an Eagle-Eye II system (Stratagene). Gel
scoring was performed directly from the gel images on a computer screen and im-
ages were stored electronically on laser CD. A set of 24 arbitrary primers selected
earlier for high multiplex content and discrimination power in cassava (Gratta-
paglia et al., 1996) were used. The presence and absence of RAPD fragments was
scored by visual inspection of the gel images. Informative RAPD markers were
identified as described previously (Grattapaglia and Sederoff, 1994). Two repli-
cate RAPD analysis experiments, including DNA extractions, RAPD assays, and
marker scoring, were carried out with the set of select primers on the putatively
apomictic individuals to confirm the patterns of bands observed. The two mater-
nal parents and 67 offspring individuals were genotyped with 24 selected arbitrary
10-base primers. Each selected primer amplified an average of 8.25 clearly inter-
pretable RAPD fragments with a range of 5 to 14 fragments. A total of 198 clear-
ly interpretable and reproducible RAPD markers were surveyed in this study. This
number of markers was considered to provide a representative genome coverage
for the objective of this study. On average, 81% of the scored RAPD bands were
monomorphic between maternal parent and progeny set in the two families. The
remaining bands were polymorphic and thus useful for testing the hypothesis pro-
posed herein.
Progenies of clones 031 and 200 displayed a high uniformity of DNA finger-
prints. However, it was possible to find markers that readily showed that individ-
uals were not derived from apomixis, except for one individual in each progeny.
In the progeny of clone 031, individual 4 showed a pattern of RAPD bands iden-
tical to that observed in individual 5 in the maternal one for all primers examined.
The same pattern was observed in individual 5 in the maternal one for all primers
examined. The same pattern was also observed in individual 5 in the progeny of
clone 200. Although it seems very unlikely that the maternal parent and a zygotic
progeny individual could have an identical combination of more than 100 RAPD
fragments, it could be argued that it is possible. As described in a previous report
(Grattapaglia et al., 1996), to exclude this possibility the statistical procedure de-
scribed by Novy et al. (1994) was used to show that the complete similarity be-
tween the two samples in not an artifact resulting from a limited number of RAPD
markers surveyed and rather has a biological basis. Given the number of markers
surveyed, the probability that complete uniformity in RAPD markers between the
maternal parent and the respective progeny individuals occurred due to change was
equal or 103 in both clones. Therefore, putative apomixis was detected in prog-
enies of both clones (031 and 200) at a rate of 3.13 and 2.70, respectively. These
results clearly indicate that the type of apomixis detected in this study is faculta-
tive and occurs at very low frequency in cassava.
A total of 261 ovules were analyzed histologically. Both clones studied indicat-
ed the presence of aposporic sacs inside the sexual embryo sacs. The presence of
two embryo sacs in an ovule is an indicator of the apospory nature of apomixis. It
is supposed that one of them is derived from somatic cells at its location in the
ovule, whereas the second embryo sac is derived from a normal megaspore moth-
er cell. At a certain stage, before the complete maturation of the sexual embryo
sac, it may abort and be replaced by the developing apospporous sac; otherwise,
it continues to develop, giving rise to two embryos sacs and two embryos in the
ovule. This abnormality was verified in 2.36 and 1.49% of the ovules of clones 031
and 200, respectively. Similar results using the same histological clarification tech-
nique were reported in Cenchrus ciliaris (Young et al., 1979). Using a DIC mi-
croscope, it was possible to see in the cleared pistils details of the cassava embryo
sac. The normal sacs showed an egg, two polar nuclei, and three antipodals. Syn-
ergids were occasionally seen, and the egg was often inconspicuous. The antipo-
dals were distinguished by a swollen, tear-drop shape, dense cytoplasm, chalazal
position, and the absence of a wall separating them from the cavity of the sac. The
aposporous sacs lacked antipodals and had only one nuclei per sac. Sometimes
there was a single polar nucleus and an egg. These results strongly suggest that the
mechanism responsible for apomixis in cassava is apospory (the development of
aposporic embryo sacs). Apospory in the angiosperms is the most common mech-
anism responsible for apomixis. In this type of apomixis the apospory embryo sac
originates from one or more somatic cells of the ovule followed by its enlargement
and vaculation (Asker, 1980). This type of apomixis explains the multiple plants
per seed found in clone 031 by Nassar (1994). The presence of apomixis in clones
031 and 200 shows the potential of the utilization of wild species as a source of
genes for apomixis in cassava since clone 031 represents generation F2 of a cross
between cassava and M. dichotoma and clone 200 is an F1 of cassava and M.
glaziovii. Sources of genes controlling apomixis in wild species that are relatives
of corn, sugar beet, wheat, and several forage grasses have already been reported
(Asker, 1979). Nassar et al. (1998) further validated the occurrence of apomixis in
cassava by presenting a combination of molecular and embryonic evidence from
two putative apomictic clones.
IV. PRODUCTION OF POLYPLOID TYPES
A. PROSPECTS OF POLYPLOIDIZING CASSAVA, M. esculenta

Crantz, BY UNREDUCED MICROSPORES
One of the most promising approaches to improve cassava is to produce triploid

types which may offer resistance to drought and insects, combined with high pro-
ductivity (Hahn et al., 1990; Nassar, 1992). The triploid tree produced by Nassar
(1992) showed a doubling of root production under semiarid conditions. Polyploid
types were obtained by Nassar and were found to be produced by unreduced ga-
mete fertilization (Nassar, 1992). It is believed that if parents with a high frequency
of unreduced gametes could be selected, this would enhance the possibility of ob-
taining polyploid types in their progeny.
1. Unreduced Microspores in Cassava, M. esculenta Crantz, Clones
The formation of unreduced microspores (gametes with somatic chromosome

number) appears to be a common phenomenon in angiosperms (Harlan and de Wet,
1975; de West, 1980). They most likely have a major role in the evolution and ori-
gin of polyploids. From a plant breeding viewpoint, these gametes may lead to the
development of highly productive tetraploids and triploids by sexual means for pre-
serving their heterozygosity (Mendiburu and Peloquin, 1977). Probably the most
convincing cytological evidence of their occurrence in higher plants has come from
microspores. Prakken and Swaminathan (1952) observed that diads (with two re-
duced microspores) and tetrads (with four reduced microspores) occur together in
the same plants of several Solanum species. This was confirmed later by several
workers using different crop plants (Roads and Dempsey, 1966). Nassar (1992) re-
ported for the first time in cassava interspecific hybrids as a consequence of meiot-
ic irregularity. Now, there is agreement that diads form due to spindle abnormali-
ties which may be visible at meiotic metaphases I and II (Nassar, 1996).
Cassava clones grown in the germplasm bank of the Centro Agronomico Trop-
ical de Investigacion y Ensenanza were used to prove the previously mentioned
hypothesis (Vasquez and Nassar, 1994). Floral buds were collected and fixed for
24 h in 3:1 ethanol:acetic acid mixture, transferred to 70% ethanol, and stored at
5C. The anthers were squashed in a drop of 1% acetocarmine. Metaphase I was
used for chromosome counting and study of chromosome associations. The tetrad
stage was observed to determine the frequency of diads and triads. About 300
tetrads were counted in each clone. Photomicrographs were taken from temporary
preparations using the Zeiss standard research microscope.
Of the nine cassava clones studied, eight showed regular metaphase with com-
plete pairing and formation of 18 bivalents. The ninth clone (No. 6477), known as
Chioriqui, showed a sectorial chimera in the inflorescence. One sector of inflores-
cence developed flowers with normal metaphase I and complete pairing of chro-
mosomes. No laggards at anaphase I and no micronuclei at the tetrad stage were
observed. The other sector of the inflorescence had flowers with extreme abnor-
mal metaphase I in all the flowers investigated. This was accompanied by the pres-
ence of empty anthers. The abnormality at metaphase I was due to lack of chro-
mosome pairing asynapsis. In the 20 metaphases examined the chromosome
association was 11 or 12 bivalents instead of 18 normal bivalents. The remaining
chromosomes formed only univalents. This resulted in the formation of micronu-
clei at the tetrad stage. There were 1–15 micronuclei per tetrad (Table XXII). Ap-
parently, this chimeral sector was due to a gene mutation in the second layer (LII)
of the apex of the growing shoot. Since only a lateral sector exhibits this abnor-
mality, the chimera must be sectorial rather than mericlinal or periclinal. This is
the first report of a mutation which affects chromosome pairing in cassava. Since
cassava reproduces vegetatively, it is likely that this mutation has been preserved
for a long time in this indigenous Costa Rican clone. To trace unreduced mi-
crospore formation, diads and triads were checked among 300 tetrads of each
clone. The most interesting case was found in clone No. 11,965 which formed
11 diads per 300 sporocytes studied, indicating first meiotic restitution (Table
XXIII). No other clone showed diad formation. Regarding triad formation, four
clones, namely, 9959 (Mangi), 6429, Vegna Mochera, and 6399, formed triads in
the range of 1–1.3%. The clone 11,965, the only diad producer among the clones
studied, also produced triads with a frequency of 1%.
As can be seen from Table XXIII there is variation for 2n gamete production
Table XXII
Chromosome Association and Diad and Triad Frequencies in Investigated Germplasm
Diads Triads
Chromosome
Cultivar association Frequency 2% Frequency 2%
9,959 (Mangi) 18II — — 4 1.33

6,417 (White) 18II — — — —
10,861 (Num-4-RB) 18II — — — —
6,429 (Negra Muchera) 18II — — 4 —
11,965 (Sim Nombre) 18II 11 3.7 3 1
6,379 (Amarilla-1) 18II — — 3 1
6,473 (Vagana 4208) 18II — — — —
6,477 (Chiriqui)
Chimera a 18II
Chimera b 11II 14I
among the cassava clones. Several workers reported that in species with a tenden-
cy to form unreduced microspores, the frequency of such gametes may vary from
one line to another (Mok and peloquin, 1975). In the meantime, stable high 2n ga-
mete production has been observed in 2n gamete producer lines (Ballington and
Galleta, 1976). Thus, the variation in the frequency of 2n gametes may be attrib-
uted mainly to their genotypic differences (Veilleux and Lauer, 1981). It seems that
unreduced microspore formation is gene controlled and not due to the disturbance
of chromosome asynapsis. However, it is agreed that this is due to the occurrence
of nonfunctional spindle, resulting in all the metaphase chromosomes remaining in
the center instead of separating to poles (Ramana, 1974; Vorsa and Bingham, 1979).
2. Unreduced Microspores in Interspecific Hybrids
The causes of unreduced microspores in plants are variable, ranging from sim-
ple recessive genes (Mok and Peloquin, 1975) to disturbed spindle function in in-
Table XXIII
Frequency of Diads and Five Kinds of Tetrads in the Sectorial Chimera of Clone 6477
(300 Tetrads Studied)
Normal Tetrads with different Nos. of

Parameter Diad tetrads micronuclei
Frequency 1 18 26 41 26 188
% 0.3 6 8.7 13.9 8.9 62.6
terspecific hybrids. Nassar et al. (1995) suggested that the disturbance of meiotic
division in interspecific hybrids may lead to a higher frequency of aneuploid ga-
metes, consequently making it possible to select polyploids among their progeny.
Brazilian Manihot species are believed to be progenitors of cassava, and Brazil
contains its centers of diversity (Nassar, 1978b). Interspecific hybridizations have
been carried out systematically by Nassar since 1980 and many interspecific hy-
brids have been produced (Nassar, 1989). Some of these, which represent differ-
ent interspecific hybrids and cover a vast array of variation, were used by Nassar
in the current investigation. Ten interspecific hybrids and/or their progenies were
used in this investigation: F1 M. glaziovii
cassava (3 genotypes), F2 M. eprui-
inosa
cassava, F2 M. anomala
cassava (3 genotypes), F3 M. pseudoglaziovii
cassava (two genotypes), and F4 M. pseudoglaziovii

cassava. These hybrids
and progenies were chosen because they provide large genetic variation. Chro-
mosome associations in meiotic metaphase I and the occurrence of dyads and tri-
ads and also pollen viability were studied. For the meiotic study, inflorescences
were fixed in a 3:1 mixture of absolute alcohol and glacial acetic acid and kept at
5C for 24 h. The anthers were smeared in an acetocarmine stain. Chromosome
configurations at metaphase I, together with spared formation, were studied. For
the pollen viability study, one to three flowers per plant were selected, and their
pollen was crushed in acetocarmine and scanned. Pollen counts and the percent-
age of stained normal pollen were calculated.
Chromosome associations and their frequency in meiotic metaphase I of PMCs
of all interspecific hybrids have shown regular pairing of 18 bivalents except the
F2 progeny of M. glaziovii
cassava (first genotype), which revealed the pres-
ence of two univalents in 5% of the cells examined. This genotype also showed a
high frequency (3.7%) of dyad formation and low pollen viability (42%). The
study of PMCs in the tetrad phase revealed the formation of abnormal tetrads hav-
ing one to three micronuclei in all hybrids used in the experiment, except F1 M.
glaziovii
cassava (second genotype). The abnormal tetrads ranged from 0.8%
in f1 M. glaziovii
cassava (first genotype) to 94% in the F1 hybrid M. epruinosa
cassava (Table XXIV).

The high frequency of unreduced microspores in the F2 progeny of M. glaziovii
cassava will facilitate selection of this genotype and its progeny as a possible
progenitor of polyploids in the future. Apparently, the dyad formation is due to
meiotic restitution. The low pollen viability may be due to univalent formation and
the consequent irregular chromosome distribution leading to unbalanced gametes.
Cassava is a natural allopolyploid judging by its chromosome number and the
complete pairing of its meiotic metaphase chromosomes (Nassar, 1978b; Vásquez
and Nassar, 1994). If unreduced gametes were responsible for its natural poly-
ploidization in the past, it is possible that one can detect them among wild rela-
tives and their hybrids with cassava. The presence of unreduced microspores in the
hybrid progeny confirms this hypothesis.
Table XXIV
Frequency and Percentage of Diads and Triads in Different Sporads
Pollen
Diad Triad Tetrad Abnormal tetrad viability
Microspores Pollen
Manihot hybrids examined n (%) n (%) n (%) n (%) examined n (%)
M. glaziovii
cassava 1217 10 (0.82) 8 (0.66) 1158 (95.15) 41 (3.37) 2979 1197 (40.18)
F1 M. glaziovii
cassava
cassava 1168 43 (3.70) 8 (0.70) 1107 (94.80) 10 (0.80) 1713 723 (42.21)
(1st genotype)
F1 M. glaziovii
cassava
cassava 1044 0 (0.00) 4 (0.40) 1040 (99.60) 0 (0.00) 1463 1123 (76.76)
(2nd genotype)
M. epriminoso
cassava 1252 0 (0.00) 0 (0.00) 1252 (100.00) 0 (0.00) 603 567 (94.03)
F2 M. anomala
cassava (1st genotype) 1374 1 (0.07) 0 (0.00) 1326 (96.51) 47 (3.42) 836 183 (21.89)
F2 M. anomala
cassava (2nd genotype) 1145 1 (0.00) 0 (0.00) 1117 (95.56) 28 (2.44) 1297 441 (34.00)
F2 M. anomala
cassava (3rd genotype) 1136 1 (0.09) 0 (0.00) 1106 (97.36) 29 (2.55) 801 452 (56.43)
F3 M. pseudoglaziovii
cassava 1416 0 (0.00) 0 (0.00) 1134 (95.62) 62 (4.38) 1427 873 (61.18)
F4 M. pseudoglaziovii
cassava 1210 0 (0.00) 0 (0.00) 1207 (98.80) 3 (0.20) 1130 1034 (94.50)
F1 M. dichotoma
cassava 1138 1 (0.09) 0 (0.00) 1125 (98.80) 12 (1.05) 1704 862 (50.59)
(1st genotype)
F1 M. dichotoma
cassava 491 2 (0.40) 3 (0.60) 476 (98.86) 10 (2.04) 1273 299 (23.49)
(2nd genotype)
Vásquez and Nassar (1994) reported a high frequency of unreduced microspores

among cassava clones. It is believed, therefore, that this character is heritable and
genetically controlled. It was probably acquired by cassava from its wild ances-
tors (Nassar et al., 1995), representing an evolutionary remnant in cassava. The
significance of this phenomenon is that it provides direct evidence on poly-
ploidization from lower ploidy levels through the mechanism of unreduced ga-
metes and not through other types of somatic doubling (Harlan and de Wet, 1975;
de Wet, 1980). According to the genes of wild ancestors they have probably had
the opportunity to combine and produce larger rooted cassava (Nassar, 1992; Nas-
sar et al., 1996a). It seems that this character is correlated with meiotic irregular-
ities provided there is some univalent formation in the genotype producing unre-
duced microspores.
It is worth mentioning that this unreduced microspore-producing genotype is
a progeny of a natural hybrid of M. glaziovii with cassava. This natural hybrid
has been maintained by farmers for hundreds of years through vegetative repro-
duction. The unreproduced microspore gene has probably arisen through recur-
rent mutation and been maintained through vegetative reproduction. In other
germplasm that reproduces sexually, it would be eleminated by natural selection
due to the abortion of gametes which carry it.
B. INDUCTION OF A PRODUCTIVE ANEUPLOID

IN CASSAVA, M. esculenta Crantz
Within the wild Manihot species which have been evaluated and hybridized
with cassava by Nassar to transfer their useful genes (Nassar, 1978a, 1980, 1986),
two interspecific hybrids between cassava and wild Manihot species have been ob-
tained and/or maintained in the Nassar program at the Universidade de Brasília.
One of the two interspecific hybrids was between cassava and M. pseudoglaziovii
Pax (Nassar, 1982); the second hybrid was between cassava and M. neusana Nas-
sar (Nassar, 1989). These hybrids were cloned by vegetative reproduction and seed
produced by open pollination. Ten seeds were collected from each hybrid for ger-
mination studies. Four seeds from progenies of cassava with M. pseudoglaziovii
germinated, of which one plant was selected for vigor. Only two seeds of proge-
nies of the hybrid of cassava with M. neusana germinated, of which one plant was
selected for vigor.
These three plants were cloned by vegetative propagation of cuttings. When the
plants flowered, the three clones and their parents were studied meiotically for
chromosome association in metaphase I, segregation in anaphase I, and pollen vi-
ability. Mitotic analysis was also performed on root tips taken from cuttings. Af-
ter 9 months of growth, these clones were examined for root productivity. For the
study of meiosis, buds were fixed in a 3:1 mixture of ethyl alcohol and glacial
acetic acid for 24 h, transferred to 70% alcohol, and then smear stained with 1%
acetic carmin. For mitotic counting of chromosomes, root tips were treated with
0.2% colchicine for 2 h, washed in distilled water, transferred to the previously
mentioned fixative fluid for 24 h, treated with 1 N HCl for 10 min, and smear
stained with 1% acetic carmin.
Meiotic examination in metaphase I of the clone, progeny of the interspecific
hybrid of cassava with M. pseudoglaziovii (HPS), showed a mean of 16–19 biva-
lents, 154 trivalents, and 1.00 univalent for the 15 sporocytes examined. These
results suggested an aneuploid having a 2n 2 chromosome constitution. The
parental hybrid showed 2n 36, with a mean association of 17.42 bivalents and
1.58 univalents. Estimation of pollen viability revealed a viability percentage of
54% in 2917 aneuploid grains and 40% in 1350 of the parental hybrid. The hybrid
of cassava and M. neusana (HO1) had a mean of 1.86 trivalents, 16.13 bivalents,
and 0.13 univalents among 30 PCMs examined. The chromosome constitution was
2n 38. This was confirmed by chromosome counting at mitosis of root tip cells.
The second clone of this cross (HO4) showed a mean chromosome association of
1.63 trivalents, 12.41 bivalents, and 8.84 univalents in metaphase I. This clone also
had 2n 38. The parent had a mean of 17.42 bivalents and 1.58 univalents, which
means 2n 36. Pollen viability was found to be 35.8% in HO1, 17.7% in HO4,
and 36.8% in their parent.
The progeny of the interspecific hybrid of cassava with M. pseudoglaziovii had
a very large starched root with a mean of 6.1 kg per plant at 10 months for the 15
plants evaluated, compared to 3.7 kg for the cultivated cassava clones. The other
two aneuploids had fibrous roots.
These aneuploids apparently resulted from disturbance of meiotic division of
their interspecific hybrid parents. This disturbance was interpreted by Nassar et al.
(1995) as an inhibition of spindle function and chromosome asynapsis. Both have
led to laggard formation and unbalanced gametes. Thus, the origin of these aneu-
ploids can be attributed to the fertilization of an n 1 egg by an n 1 male ga-
mete based on the fact that gametes with this structure are more viable than ga-
metes with n 2. The other possibility is the fertilization of an n 2 egg with an
external n male gamete because the plants were open pollinated. It was not possi-
ble to determine whether the male gamete came from selfing or from another pol-
linator plant. In various crops, interspecific hybridization has led to the disturbance
of meiotic division. Similar findings were reported in Trifolium pratense by Par-
rot and Smith (1984) and in Medicago spp. by Vorsa and Bingham (1979). No
aneuploids, however, were obtained in these crops.
The two additioinal chromosomes in clone HPS resulted in the production of a
large tuber root that is superior to normal root production in cultivated cassava
clones. In all likelihood this size increase in due to additive polygenes. These poly-
genes may be distributed on more than one chromosome.
Nassar (1978b) postulated the origin of cassava as a product of hybridization
between two wild species, both having fibrous nonstarchy root as a result of an ac-
cumulation of additive polygenes derived from ancestral wild parents.
C. PRODUCTION OF TRIPLOID CASSAVA, M. esculenta

Crantz, BY HYBRID DIPLOID GAMETES
Through the Nassar program of cassava germplasm collection and utilization

in Brazil, a natural hybrid of M. pseudoglaziovii and cassava was collected in
1978 from Remigo county, Paraíba state, and grown in the living collection at
the Biological Experimental Station, Universidade de Brasília (Nassar, 1982).
This natural hybrid was identified by the marker genes of winged fruit which
came from cassava and peltated, pendurated leaf, two characteristic genes of M.
pseudoglaziovii overlaps cassava plantations. The hybrid was multiplied vege-
tatively and cloned 20 times to individuals maintained in the living collection
and used in this experiment. This clone was left for open pollination “entre si.”
Seeds collected were planted in 1980 and the progeny of 22 plants were pro-
duced. Evaluation of the progeny for root production was carried out in the fol-
lowing years.
The natural hybrid and its selected progeny were studied meiotically. The an-
thers were fixed for 24 h in a 3:1 solution of ethanol and acetic acid, transferred to
70% ethanol, and stored at 5C. The anthers were squashed in 1% acetocarmine.
Pollen analysis was conducted by collecting pollen from undehisced natural an-
thers of a flower onto a microscope slide and staining with 1% acetocarmine.
Pollen grain diameter was assessed at 450x using an eyepiece micrometer. A min-
imum of 1000 stained grains per plant were counted. Pollen stainability was used
as a criterion for viability. Well-filled, darkly stained pollen grains were consid-
ered fertile and partially filled, and unstained ones were considered sterile. The fre-
quency of mature 2n pollen grains was based on counts of 1000 pollen grains
stained with 1% acetocarmine. Classification of n vs 2n pollen was made by visu-
al size discrimination of stained pollen, given that the latter should have twice the
nuclear volume of the former.
To obtain data on ploidy status, mitosis in root tips was examined and chromo-
some number was counted to confirm the plant’s chromosome constitution.
Screening the hybrid cloned plants for tuber formation showed them to have com-
plete fibrous roots but they extended deep in the ground, almost 5 or 6 m in length.
The presumed progenitor, M. pseudoglaziovii grows in one of the driest pockets
of Paraíba state, Remigio county, where the wild species and the hybrid were col-
lected. Apparently, the wild species had acquired this deep-rooted character as a
mechanism for removal of subterranean water. This character may be useful in sub-
sequent improved generations for production of drought-tolerant cassava culti-
vars. Two years after planting, the hybrid reached 3 or 4 m in height. Its leaves
were peltate and frequently pendulous. These characters are typical of M. pseudo-
glaziovii and one of the striking features that distinguish this species from the oth-
er 98 species recognized by Rogers and Appan (1973). The hybrid carried few fruit
(4 or 5 fruits compared to 60–70 fruits in the pure M. pseudoglaziovii of the same
age). The fruit is winged, a marker gene characterizing cassava fruit (Nassar,
1989).
Cytological analysis of PMCs revealed a regular occurrence of 18 bivalents at
metaphase I. No multivalents were observed in 50 metaphase I cells studied. How-
ever, disturbed anaphases I was noted in 6 of the 30 anaphase I cells studied. Dyads
and triads were noted totalling 14, compared to 86 normal tetrads. Pollen viabili-
ty as measured by stainability with acetocarmine was as low as 41%. Of these
viable pollen, 21% were small. The high percentage of sterility despite regular
pairing may be attributed to the maintenance of the hybrid through vegetative re-
production by farmers in Paraiba state. This mode of reproduction may have led
to the accumulation of sterility mutations which have not been eliminated by nat-
ural selection through the sexual cycle (Nassar, 1978c). Of the 20 plants cloned
from the hybrid, 82 fruits were collected. They gave 48 seeds with an average of
0.59 seeds per fruit. The average number of fruits collected per plant is 4.1 com-
pared to 60 –70 fruits in the wild species plant. This low fertility is expected in
view of the high sterility of pollen grains.
By planting the seeds, 25 seedlings were obtained but all of them, with the ex-
ception of four plants that survived, were chlorotic or deformed and died. The
plants that survived were identified in the living collection by numbers 120 –123.
Of the four plants, three gave fibrous roots, but the fourth (No. 121) was very vig-
orous, resistant to stem borers, and produced large tuberated roots by the end of
the year. When multiplied vegetatively and compared with other cassava clones it
showed superiority in tuber formation by the age of 12 months. This superiority
reached approximately 28% more tubers than the Sonora clone, one of the best
commercial cultivars. Its production was 2.13 kg/plant compared with 1.67 for
Sonora. An evaluation of this selection under arid conditions was carried out dur-
ing the years 1985 –1989 by the National Center for Tropical Semi-arid Region
and confirmed its superiority (Table XXV). The meiotic study of this plant showed
formation of trivalents, bivalents, and univalents in metaphase I with different fre-
quencies (Table XXVI). The meiotic division triploid resulted from the fertiliza-
tion of a 2n gamete with n gamete in the progenitor hybrid. Viability of pollen of
this triploid is as low as 19%. Seeds rarely formed. However, five seeds were col-
lected and have been reproduced in our program.
The 2n pollen formation which led to production of this triploid is reported in
cassava by Nassar for the first time. The cytological mechanism of its formation
is probably inhibition of spindle at metaphases I and II judging from the triad
formation. In the literature, several mechanisms were reported. Clement and
Stanford (1961) and Tyagi (1988) attributed it to abnormal cytokinesis, whereas
Table XXV
Average Productivity per Plant of the Triploid (Selection 121) and Other Cassava Clones under
Savanna and Semiarid Conditions
Clone No. plants Productivity Region, location of trial, and harvest age (kg/plant)
Triploid 121 30 2.13 Central Brazil, conducted at the experimental sta-

Sonora 30 1.67 tion, Univ. of Brasília; harvest age 12 months
Triploid 121 29 5.9 Under semiarid conditions; trial conducted by the
Nagib 01/84 27 1.8 CPATS, Petrolina, northeastern Brazil; harvest
Nagib 02/84 29 3.7 age 18 months
Nagib 03/84 28 1.1
Nagib 04/84 22 4.1
Cigana 28 2.9
Mendiburu and Peloquin (1977) proposed parallel spindle as an additional mech-

anism. The 2n pollen mutations were described in several plant species: Zea
mays (Roads and Dempsey, 1966), Solanum (Hogland, 1970), Medicago sativa
(Vorsa and Bingham, 1979), and Lolium (Sala et al., 1989). The vigor of Nassar-
selected triploid as seen in its productivity both under central Brazil conditions
and in the semiarid tropics adds evidence of the usefulness of this 2n gamete phe-
nomenon as a powerful mechanism for transferring variability and fitness to
polyploid offspring. If a partially fertile type of triploid could be produced, it will
serve in establishing a founder population of a new chromosome race. Its prog-
eny may rehybridize with new polyploids and diverse genotypes, producing ad-
ditive heterotic potentialities. Since trivalent occurrence in this triploid is pre-
dominantly demonstrating gene exchange between wild and cultivated genomes,
it is likely that this will generate more variability in the progeny. The wild par-
ent and its interspecific hybrid progeny are highly tolerant to drought as indicat-
ed by their deep roots. They normally survive the frequent years of drought in
their natural habitat. This triploid is a potential progenitor of cultivars adapted to
these conditions.
Table XXVI
Chromosome Associations at Metaphase I in the Triploid and Its Parent
Average chromosome association
Type No. PMCs examined III II I
Natural hybrid 50 — 18 —
Triploid 30 10.2 9.2 3.6
V. PROTEIN CONTENTS IN CASSAVA CULTIVARS

AND ITS HYBRID WITH WILD Manihot SPECIES
Tubers of 16 cassava clones, 8 months old and maintained in the germplasm col-
lection at the Experimental Station, University of Brasília, were analyzed chemi-
cally for protein content. Total nitrogen and dry matter basis were determined by
the 1970 AOAC procedures (Nassar and Dorea, 1982). Percent protein was ob-
tained by multiplying percent N by the factor 6.25. For every clone two samples
were analyzed, one from a small tuber (50 g) and the second from older tubers (200
g). Tubers were peeled, and protein was estimated in the coth peel and pulp. Tu-
bers of hybrid between cassava clone Catelo and M. oligantha were analyzed in
the same way. Seed of cassava and wild Manihot species maintained in the living
collection at the University of Basília were analyzed for protein content by the
same procedures.
The concentration of protein in tubers of cassava clones and tubers of the hy-
brid is presented in Table XXVII. It can be seen that protein percent (N%
6.25
protein percent) ranged from 0.7 to 1.2% for larger tubers (200 g), whereas
the range was 0.9 to 1.4% for tubers 50 g. In the same clone, protein content
of the pulp was higher in small tubers than in larger ones. These data agree with
those found by Akinrele (1964), who reported 0.7% for protein content in peeled
tuber on a dry matter basis, and Chada (1961), who found 1.2%. However, these
researchers did not pay attention to the effect of tuber size on protein content.
Jennings (1959) stated that protein in cassava root tends to be concentrated in the
outer zone of the root. This may explain why the small tubers have higher pro-
tein content since they have a larger proportion of the outer zone than do the old-
er and larger ones. The very little variation of protein content among the 16
clones collected throughout Brazil shows that selection for this characteristic in
cassava clones will not bring any notable improvement. From Table XXVII it can
also be seen that protein is higher in peel than in pulp in all the examined sam-
ples and in all clones. This may be explained by the previously mentioned state-
ment of Jennings that protein in cassava roots tends to be concentrated in the out-
er layers.
The analysis of hybrid tubers showed a notable increase in the hybrid of cas-
sava with M. oligantha at 4.6%. In an earlier paper (Nassar and Costa, 1978),
the protein content in M. oligantha was reported to be 7.1% on a dry matter ba-
sis. Moreover, crosses of this species with cassava were highly fertile (Nassar,
1980). Nassar and Fitchner (1978) also showed a low HCN content in the wild
species M. oligantha. This may exclude any possibility that high protein content
in the species is due to HCN nitrogen. Table XXVIII shows the results of seed
analysis of some wild Manihot species maintained in our living collection. The
highest protein content is that of M. Brachyandra followed by M. alutacea.
Table XXVII
Protein Content of Tubers of Cassava Clones
Clone Approximate size (g) Protein in peel (%) Protein in pulp (%)
CBM 0206 200 2.13 0.90

50 2.09 1.22
EAB 348 200 1.41 0.85
50 1.69 1.04
BGM 188 200 — —
50 1.68 1.45
CPM 0231 200 — —
50 1.56 1.26
CPM 2002 200 — —
50 2.08 0.99
CPM 0232 200 2.00 1.02
50 1.82 1.15
BGM 808 200 1.63 0.93
50 — —
CPM 0225 200 1.38 0.89
50 1.25 0.95
BGM 204 200 1.24 1.06
50 — —
CPM 1805 200 1.14 0.72
50 1.37 1.00
EAB 1156 200 1.58 0.84
50 1.28 1.16
EAB 484 200 1.96 1.07
50 — —
BGM 048 200 1.41 0.82
50 1.11 1.17
BGM 020 200 1.80 0.98
50 1.53 1.23
CPM 1060 200 — —
50 1.58 1.19
EAB 675 200 1.36 0.70
50 1.51 0.93
Hybrid 200 6.63 4.56
50 8.06 4.56
Manihot brachyandra is native to western Pernambuco and northern Bahia, two

of the driest areas in Brazil. Nassar (1980) reported that seed of wild cassava is
eaten by the population of these regions particularly in times of famine. Jones
(1959) reported that cassava seed is eaten in several parts of west and central
Africa. Thus, the discovery of the high protein content in native cassava hybrids
may open a new door to better protein-balanced food for people of the tropical
world.
Table XXVIII
Protein Content in Wild Manihot Species Seed
on Dry Matter Basis
Species Protein%
M. glaziovii 30.09
M. caerulescens 27.91
M. brachyandra 35.35
M. pseudoglaziovii 31.15
M. alutacea 37.33
M. zenhtneri 28.99
M. dichotoma 29.24
M. reptans 33.25
M. esculenta 26.81
ACKNOWLEDGMENT
The living collection of wild Manihot species was established at the Universidade de Brasília with
the help of the International Development Research Center Ottawa, Ontario, Canada, for which the au-
thor is grateful.
REFERENCES
Abraham, A. (1975). Breeding of tuber crops in India. J. Genet. Plant Breeding 17, 212–217.
Abraham, A., Panicher, P. K., and Mathew, P. M. (1964). Polyploids in relation to breeding in tuber
crops. J. Indian Bot. Soc. 43, 278 –283.
Akinrele, I. A., Collins, C., Cook, A. S., Hogate, R. A., Junaid, Y., and Baumer, G. (1962). Pilot plant
(1 ton a day) results of 3 month trial run, Research Reprints, Vol. 13, pp. 1– 30. Federal Institute
of Industrial Research, Nigeria.
Anonymous (1968). Tabla de composicion de pastes y otros alimentos de Centro America, Publication
No. E-440. Instituto Nacional de Centro America e Panama.
Asker, S. (1979). Progress in apomixis research. Hereditas 91, 231–240.
Asker, S. (1980). Gametophytic apomixis: Elements and genetic regulation. Hereditas 93, 277–293.
Bolhuis, G. G. (1953). A survey of some attempts to breed cassava varieties with a high content of pro-
teins in the roots. Euphytica 20, 107–112.
Brewer, J. L., and Henstra, S. (1974). Pollen of Pyrethrum. Fine structure recognition. Euphytica 23,
567–663.
Brown, W. V., and Emery, W. H. P. (1958). Apomixis in the Gramineae: Panicoideae. Am. J. Bot. 45,
253–263.
Chada, Y. R. (1961). Sources of starch in Commonwealth territories. III. Cassava. Trop. Sci. 3, 101–113.
Clement, W. M., Jr., and Stanford, E. M. (1961). A mechanism for the production of tetraploid and per-
toploid progeny from diploid and tetraploid cross alfafa. Crop Sci. 1, 11–14.
Cruz, N. D. (1968). Citologia no gênero Manihot Adans. 1 Determinação do número de cromossomo
em algumas espécies. anal. Acad. Bras. Ci. 40, 81– 95.
de Wet, J. M. J. (1980). Origin of polyploids. In “Polyploidy: Biological Relevance” (W. H. Lewis,

Ed.). Plenum, New York.
Dione, L. A. (1963). Studies on the use of Solanum acaule as a bridge between Solanum tuberosum
and species in the series Bulbocastana, cardiophylla and Pinnatisecta. Euphytica 12, 263 –269.
Dobzhansky, T. (1973). Genética do processo evolutivo. In “Tradução de Celso Abbade Mourão.”
Poligono, São Paulo, Brazil.
Doyle, J. J., and Doyle, J. L. (1987). Isolation of plant DNA from fresh tissues. Focus 12, 13 –15.
Food and Agriculture Organization (FAO) (1994). “Production Yearbook.” FAO, Rome.
Graner, E. A. (1942). Genet. Manihot Braganita 2, 13 –19.
Grattapaglia, D., and Sederoff, R. R. (1994). Genetic linkage maps of Eucalyptus grandis and E. uro-
plylla using a pseudo-testcross strategy and RAPD markers. Genetics 137, 1121–1137.
Grattapaglia, D. E., Nassar, N. M. A., and Dianese, J. (1986). Biossistemática de espécies brasileiras
do Gênero Manihot baseada em padrões de proteína da semente. Ciencia Cultura (J. Brazilian
Assoc. Advancement Sci) 19(3), 294 – 300.
Grattapaglia, D., Costa e Silva, C., and Nassar, N. M. A. (1996). Strict maternal inheritance of RAPD fin-
gerprints confirm apomixis in cassava (Manihot esculenta Crantz). Can. J. Plant Sci. 76, 379–382.
Gustafsson, A. (1946). Apomixis in higher plants. Part 1. The mechanisms of apomixis. Lunds Univ.
Arsskr. Adv. 42, 1– 66.
Hahn, S. K., Bai, K. V., and Asiedu, R. (1990). Tetraploids, triploids and 2n pollen from diploid inter-
specific crosses with cassava. Theor. Appl. Genet. 79, 433 – 439.
Harlan, J. (1961). Geographic origin of plants useful to agriculture. Am. Assoc. Adv. Sci. 66, 3 – 9.
Harlan, J. R., and de Wet, J. M. J. (1975). On O wings and a prayer; The origins of polyploids. Bot.
Rev. 41, 361–390.
Hogland, M. (1970). Meiosis in Solomum phureja. Jereditas 66, 183 –188.
Jennings, D. L. (1959). Manihot melanobasis Muell. Arg.—A useful parent for cassava breeding. Eu-
phytica 8, 157–162.
Knox, R. B., Willing, R. R., and Prior, L. D. (1972). Interspecific hybridization in populus using recog-
nitioin pollen. Silvae Genet. 21, 263 –298.
Laemilli, U. K. (1970). Cleavage of structural proteins during assembly of the head of bacteriophage
T4. Nature 227, 680 – 685.
Lanjouw, J. (1939). Two interesting species of Manihot L. from Surinam. recueil Travaux Botaniques
Neerlandais 36, 542– 549.
Magoon, M. L., Jos, J. S., and Appan, S. G. (1966). Cytomorphology of interspecific hybrid between
cassava and ceara rubber. Chromosome Information Service (Japan), pp. 8 –10.
Mendiburu, A. O., and Peloquin, S. J. (1977). The significance of 2n gametes in potato breeding. The-
or. Appl. Genet. 49, 53 – 61.
Mok, D. W. K., and Peloquin, S. S. (1975). The inheritance of three mechanisms of diploandroid (2n
pollen) formation in diploid potatoes. Heredity 35, 295 – 302.
Narty, F. (1969). Studies on cassava, Manihot utilissima Pohl. 1. Cyanogenesis: The biosynthesis of
linamarin and lutaustalin in etioled seedlings. Phytochemistry 7, 1307–1312.
Nassar, N. M. A. (1978a). Genetic resources of cassava: Chromosome behavior in some Manihot
species. Indian J. Genet. Plant Breeding 38, 135 –137.
Nassar, N. M. A. (1978b). Conservation of the genetic resources of cassava (Manihot esculenta); De-
termination of wild species localities with emphasis on probable origin. Econ. Bot. 32, 311– 320.
Nassar, N. M. A. (1978c). Microcenters of wild cassava Manihot spp. diversity in central Brazil. Tur-
rialba 28(4), 345 – 347.
Nassar, N. M. A. (1978d). Hydrocyanic acid content in some wild Manihot (cassava) species. Can. J.
Plant Sci. 58, 577– 578.
Nassar, N. M. A. (1978e). Some further species of Manihot with potential value to cassava breeding.
Can. J. Plant Sci. 58, 915 – 916.
Nassar, N. M. A. (1978f ). Wild Manihot species of central Brazil for cassava breeding. Can. J. Plant
Sci. 58, 257–261.
Nassar, N. M. A. (1979a). A study of the collection and maintenance of the germoplasm of wild cas-
savas Manihot spp. Turrialba 29(3), 221.
Nassar, N. M. A. (1979b). Three brazilian Manihot species with tolerance to stress conditions. Can. J.
Plant Sci. 59, 533– 555.
Nassar, N. M. A. (1980a). Attempts to hybridize wild Manihot species with cassava. Econ. Bot. 34,
13–15.
Nassar, N. M. A. (1980b). The need for germplasm conservation in wild cassava. Indian J. Genet. Plant
Breeding 39(3), 465 – 470.
Nassar, N. M. A. (1982). Collecting wild cassava in Brazil. Indian J. Genet. Plant Breeding 42, 405–411.
Nassar, N. M. A. (1984). Natural hybrids between Manihot reptans Pax and M. alutaced Rogers & Ap-
pan. Can. J. Plant Sci. 64, 423 – 425.
Nassar, N. M. A. (1985). Manihot neusana Nassar: A new species native to Parana, Brazil. Can. J. Plant
Sci. 65, 1097–1100.
Nassar, N. M. A. (1986). Genetic variation of wild Manihot species native to Brazil and its potential
for cassava improvement. Field Crop Res. 13, 177–184.
Nassar, N. M. A. (1989). Broadening the genetic base of cassava, Manihot esculenta Crantz, by inter-
specific hybridization. Can. J. Plant Sci. 69, 1071–1073.
Nassar, N. M. A. (1991). Production of triploid cassava, Manihot esculenta Crantz, by hybrid diploid
gametes. Field Crop Res. 13, 173 –182.
Nassar, N. M. A. (1992). Cassava in South America: A plant breeder view. Ciëncia Cultura (J. Brazil-
ian Assoc. Advancement Sci.) 44, 25 –27.
Nassar, N. M. A. (1994a). Selection and development of apomitic cassava clone. Ciëncia Cultura (J.
Brazilian Assoc. Advancement Sci. 41(6), 168 –171.
Nassar, N. M. A. (1994b). Relationship between Manihot species: A review of the proceedings of the
2nd Internacional Scientific Meeting of the Cassava Biotechnology Network. Bogor, 90 –100.
Nassar, N. M. A. (1995). Development of cassava interspecific hybrids for savanna (cerrado) condi-
tions. J. Root Crops 22, 9 –17.
Nassar, N. M. A., and Costa, C. P. (1978). Tuber formation and protein content in some wild cassava
(mandioca) species native to central Brazil. Experiëncia 33, 1304 –1306.
Nassar, N. M. A., and Dorea, G. (1982). Protein contents of cassava cultivars and its hybrid with Mani-
hot species. Turrialba 32(4), 429 – 432.
Nassar, N. M. A., and Grattapaglia, D. (1986). Variabilidade de clones de mandioca em relação ä fer-
tilidade e aspectos morfológicos. Turrialba 36(4), 555 – 559.
Nassar, N. M. A., and O’Hair, S. K. (1985). Variation among clones in relation to seed germination. In-
dian J. Genet. 45(2), 429 – 432.
Nassar, N. M. A., Silva, J. R., and Vieira, C. (1986). Hibridação insterespecifica entre mandioca e es-
pécies silvestres do Manihot. Ciëncia Cultura ( J. Brazilian Assoc. Advancement Sci.) 33(6),
1050–1055.
Nassar, N. M. A., Nassar, H. N. M., Vieira, C., and Saraiva, S. L. (1995). Cytogenetic behavior of in-
terspecific hybrids of cassava and M. neusansa Nassar. Can. J. Plant Sci. 75, 675 – 678.
Nassar, N. M. A., Nassar, H. N. M., Carvalho, C. G., and Vieira, C. (1996a). Induction of a productive
aneuploid in cassava, Manihot esculenta Crantz. Brazilian J. Genet. 19, 123 –125.
Nassar, N. M. A., Carvalho, C. G., and Vieira, C. (1996b). Overcoming barriers between cassava Mani-
hot esculenta Crantz and wild relative M. pohili Warwa. Brazilian J. Genet. 19, 617– 620.
Nassar, N. M. A., Vieira, M. A., Vieira, C., and Grattapaglia, D. (1998a). Molecular and embionic ev-
idence of apomix in cassava interspecific hybrids (Manihot spp). Can. J. Plant Sci. 78, 348 – 352.
Nassar, N. M. A., Vieira, M. A., Vieira, C., and Grattapaglia, D. (1998b). Molecular and embryonic ev-
idence of apomix in cassava (Manihot esculenta Crantz). Euphytica 102, 9 –13.
Nichols, R. F. W. (1947). Breeding cassava for resistance. East Afr. Agr. J. 12(3), 184 –194.
Novy, R. G., Kobak, C., Gofreda, J., and Vorsa, N. (1994). RAPDs identify varietal misclassification
and regional divergence in cranberry (Vaccinium macrocarpon (Ait.) Pursh. theor. Appl. Genet.
88, 1004–2020.
Oke, O. L. (1973). The mode of cyanide detoxification. In “Chronic Cassava Toxicity. Proceeding of
Inderdisciplinary Workshop, London, 29– 30 January 1973” (B. Nestel and R. MacIntyte, Eds),
pp. 97–104. IDRC-OLOe, Ottawa, Ontario, Canada.
Parrot, W. A., and Smith, R. R. (1984). Production of 2n pollen in red clover. Crop. Sci. 24, 467– 472.
Prakken, R., and Swaminathan, M. S. (1952). Cytological behaviour of some interspecific hybrids in
the genus Solanum. Genetic 54, 505 – 522.
Raymond, W. D., Jojo, W., and Zicodemus, Z. (1941). The nutritive value of some Tanganyka foods:
Cassava. East Afr. Agric. J. 6, 154 –159.
Roads, M. M., and Dempsey, E. (1966). Induction of chromosome doubling at meiosis by elongated
gene in maize. Genetics 54, 505 – 522.
Rogers, D. J. (1963). Studies on Manihot esculenta Crantz and related species. Bull. Torrey Bot. Cl. 99,
43–54.
Rogers, D. J., and Appan, C. (1973). Manihot, Manihotoides, Euphorbiaceae. In “Flora Neotropica.”
Hafner, New York.
Rogers, D. J., and Fleming, H. S. (1973). A monograph of Manihot esculenta with an explanation of
the taximetrics method used. Econ. Bot. 27, 1–113.
Sastri, D. C., and Shivanna, K. R. (1976). Attempts to overcome interspecific incompatibility in
Sesanum by using recognition pollen. Ann. Bot. 40, 891– 893.
Schmidt, C. B. (1951). A mandioca, contribuição para o conhecimento de sua origem. Boletim Agric.
1, 1.
Stebbins, G. L. (1950). “Variation and Evolution in Plants.” Columbia Press, New York.
Vavilov, N. I. (1951). Phytogeographic basis of plant breeding. The origin, variation, immunity and
breeding of cultivated plants. Chronica Bot. 13, 1– 366.
Vásquez, N., and Nassar, N. M. A. (1994). Unreduced microspores in cassava, Manihot esculenta
Crantz clones. Indian J. Genet. Plant Breeding 54, 436 – 441.
Vorsa, N., and Bingham, E. T. (1979). Cytology of pollen formation in diploid alfalfa, Medicago sati-
va. Can. J. Genet. Cytol. 21, 526 – 530.
Webber, K., and Osborn, M. J. (1969). The reliability of molecular weight determination by sulfate–
polyacrilamide gel electrophoresis. J. Biol. Chem. 244, 4406 – 4412.
Williams, J. G. K., Kubelik, A. R., Livak, K. J., Rafalski, J. A., and Tingey, S. V. (1990). DNA poly-
morphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Res. 18,
6531–6535.
Young, B. A., Sherwood, R. T., and Bashaw, E. C. (1979). Cleared-pistil and thick-sectioning tech-
niques for detecting aposporous apomixis in grasses. Can. J. Bot. 57, 1668 –1672.
Index
A Apospory
in cassava, 214 –215
Absorption Aquifer remediation
defined, 3–4 modeling, 65
of phosphate in soil, 101 Arrhenius equation, 32
Acidolysis, 130 Aruak Indians, 185
Adsorption Aspergillus, see Fungi, phosphate-solubilizing
defined, 3
on mineral surfaces, 7 B
of phosphate in soil, 101, 102
Bacteria
pH and, 104–105
phosphate solubilizing, 141–142
Advection-dispersion equation, 50
Base flow
Agricultural chemicals, see Herbicides; Pesti-
described, 160t, 172
cides
phosphorus transfer and, 157–158
Agricultural management
Batch solution-solid techniques, 45 – 49
adverse environmental effects, 76, 77
Bioavailability
effects on soil organic matter, 87– 88
coupled sorption-degradation kinetic models,
Agroecosystem Resources Group, 77
58 – 65
Agroecosystems
overview of, 56 – 58
environmental degradation and, 76
of soil chemicals, 83
environmental indicators
Biodegradation
biological, 79–82
sorption-degradation kinetic models, 58 – 65
chemical, 83–84
Biporous diffusion model, 36 – 37
of greatest significance, 78 –79
“Black carbon”
landscape, 83, 84–85
sorption by, 13
physical, 78, 82–83
Bolivia
ranking of, 90, 91t
wild species of Manihot in, 184
recommendations for using, 92
Brazil
soil organic matter, 85 – 90
domestication of cassava in, 185
environmental monitoring and, 77–78
wild species of Manihot in, 181t, 182–184
overview of, 76–77
Bridge species, 208, 209 –210
Aluminum
Bypass flow, 160t, 165
chelated by organic acids, 129 –130
in soil phosphorus dynamics, 102, 103 C
Aluminum phosphate
solubility in liquid media, 125 Calcium
Ammonium in soil phosphorus dynamics, 102, 103
effects on phosphate solubilization by fungi, Calcium phosphate
132 solubility in liquid media, 125
Aneuploidy Carbon, atmospheric
in cassava, 220–222 soil organic matter and, 86– 87
Apomixis Carboxylic acids
in cassava, 210–215 chelation of phosphate metal cations, 129
231
232 INDEX
Cassava, see also Manihot Crop diversity

apomixis in as environmental indicator, 81– 82
as apospory, 214 –215 Crop yield
embryo sac analysis of, 212–213, 214 –215 as environmental indicator, 78
genetic study of, 210 –212 soil organic matter and, 86
molecular analysis of, 213 –215
genetic origins of, 184 –185, 221–222 D
interspecific hybrids, 184 –185, 187
aneuploidy in, 221–222 Darcian flow, 160t, 165
apomixis and, 211–212 Darken equation, 31– 32
drought tolerance and, 204 –207 Delphi technique, 90
with M. anomala, 198 –200 Denbigh soil
with M. neusana, 198 –204 hydrological characteristics, 155, 156t
overcoming crossing barriers to M. pohlii, Desorption, see also Sorption
207–210 defined, 3
polyploidy and, 217–220 nonequilibrium
protein content, 225 –226 mechanisms in, 17–27
unreduced microspores in, 217–220, 223 – uptake and release profiles, 16 –17
224 rate of, 14
place of domestication, 185 zero-length columns, 56
polyploidy in, 218 Diffusion
advantages of, 215 defined, 17–18
aneuploids, 220–222 intraorganic matter diffusion, 20–23
through interspecific hybridization, 201– soil pore diffusion, 18 –20
202, 217–220 Diffusion equations, 31– 32
triploids, 222–224 Diffusion models
unreduced microspores and, 215 –220, combined organic matter/pore diffusion mod-
223–224 el, 39 – 41
protein content, 194, 196, 225, 226t coupled with biodegradation kinetics, 62– 63
relationship to Manihot species, 190 for fixed-pore systems, 32– 37
sectorial chimera in, 216 general considerations in, 30 – 32
Catchment/watershed hydrological pathways, for multiple particle sizes, 41– 43
163, 164, 166, 168 –172 for organic matter, 37– 39
Charge-transfer interactions, 6 for soil columns, 52– 54
Chelation Drought tolerance
by organic acids, 127–130 cassava interspecific hybrids and, 204–207
Chemisorption, 5 Dual resistance diffusion model, 36 – 37
Chimeras
in cassava, 216 E
Citric acid
excreted by phosphate-solubilizing fungi, 126 Earthworms
Clays as environmental indicators, 80
hindered diffusion and, 20 Electrodynamic thermogravimetric analyzer, 49
Coal Environmental indicators
sorption by, 13 agricultural chemical use, 79
Condensation biological, 79
defined, 4 crop diversity, 81– 82
Continuous-flow stirred tank reactor, 54 – 56 crop productivity, 78
Coordination complexes earthworms, 80
chemisorption and, 5 fecal pathogens, 82
INDEX 233
genetic diversity, 79 promotion of plant growth by, 133, 134 –140t,

honeybees, 80–81 141–143
insects, 78, 81
pesticide resistance, 81 G
soil microbial status, 80
landscape, 83, 84–85 Genetic diversity
physical, 78, 82–83 as environmental indicator, 79
ranking of, 90, 91t Gluconate
recommendations for using, 92 excreted by phosphate-solubilizing fungi,
soil chemicals, 83–84 126 –127
soil organic matter, 85 – 90 Graciles, 186
Environmental monitoring Groundwater
assessment endpoints, 77–78 phosphorus transfer and, 172
federal agencies in, 77
pressure-state-response framework, 78 H
Environmental Protection Agency
ranking of landscape metrics, 85 Hallsworth soil
Equilibrium partition bioavailability model, 58 hydrological characteristics, 155, 156t
Erosion Herbicides
effects on soil organic matter, 87– 88 environmental health and, 79
Honeybees
F as environmental indicators, 80 – 81
HOST classification, see Hydrology of soil
Fava-Eyring model, 30 types classification
Fecal pathogens Humic acids, 8
as environmental indicators, 82 Humic substances, 8
Ficks’s laws, 31–32 Humin, 8
Field capacity, 157 Hydrocyanic acid
Film resistance sorption model, 30 in Manihot tubers, 194, 195
First-order solute transport model Hydrogen bonding
with biodegradation term, 61 in organic molecules, 5
with sorption kinetic term, 52 Hydrological pathways
Freundlich equation, 34 phosphorus transfer and
Fulvic acids, 8 overview of, 154 –155
Fungi scale effects, 162–164
vesicular mycorrhizal, 100, 141–142 at the slope/field scale, 166, 168 –172
Fungi, phosphate-solubilizing, 100, 144 at the soil profile scale, 165 –166
agar studies, 107 temporal aspects of, 158, 159f
liquid medium studies terminology and, 164
chelation of cations by organic acids, 127– timescales in, 158 –161
130 types of, 160 –161t
nitrogen source effects, 132 variable source areas, 171
pH effects, 109, 124, 130–131 Hydrology of soil types (HOST) classification,
production of organic acids, 126 –127 155, 166, 167f
results from, 110–123t Hydrophobic effect, 13 –14
solubility of phosphate compounds, 109, Hydroxylated mineral surfaces, 6
124–126 Hysteresis
time-related soluble phosphate fluctuations, isotherm, 24 –26
132–133 kinetic, 26 –27
titratable acidity and, 130 overview of, 24
234 INDEX
I drought tolerance and, 204 –207

overcoming crossing barriers, 207–210
Infiltration-excess overland flow, 170 –171 polyploidy and, 217–220
Insects production of hybrids, 198 –200
as environmental indicators, 78 protein content, 225 –226
Intraorganic matter diffusion unreduced microspores in, 217–220, 223 –
overview of, 20–22 224
structure activity relationships in, 22–23 interspecific hybridization in, 186 –187, 196
Ion-exchange forces, 6 natural habitats, 193t, 196
Iron relationships between species, 186 –190,
in soil phosphorus dynamics, 102, 103 191t, 192t
Isotherm hysteresis, 24 –26 species in Brazil, 181t, 182–184
taxonomy of, 180 –181
K tuber formation patterns, 193 –194
tuber hydrocyanic acid content, 194, 195
Kerogen tuber protein content, 194 –195, 196, 225 –
sorption by, 13 226, 227t
-Ketogluconic acid, 130 M. anomala, 198 –200
Kinetic hysteresis, 26–27 M. brachyandra, 225, 226
M. caerulescens, 197–198
L M. dichotoma, 211
M. falcata, 196
Land drainage M. glaziovii, 181, 200, 204, 218, 220
phosphorus transfer and, 172 M. oligantha, 225
Landscape M. oligantha subsp. nestili, 194, 195
in environmental assessments, 83, 84 – 85 M. paviaefolia, 196
National Resource Inventory and, 83 M. pohlii, 207–210
Langmuir kinetic model, 27–28 M. procumbens, 198
Leaching M. pruinosa, 196
described, 160t M. pseudoglaziovii, 201, 204, 220, 221, 222–
use of term, 164 223
Linear driving force sorption models, 29 – 30 M. reptans, 186 –187, 196
with biodegradation term, 61– 62 M. saxicola, 195
M. stipularis, 198
M Manihot esculenta, see Cassava
Manihot neusana
Macropore flow, 160t, 165 as bridge species in hybridization, 208, 209 –
Macropores 210
modeling diffusion in, 33 – 35 hybridization with cassava, 198 –200, 220
size of, 18 cytogenetic behavior of backcrossed gener-
Manihot, see also Cassava ation, 202–203
adaptation to climatic conditions, 197–198 cytogenetic behavior of parents, 203
centers of diversity, 184, 185 –186 evolutionary and breeding significance of,
chromosome number, 186, 187t 203 –204
genetic variability, 190, 193 –198 meiotic behavior of F1 hybrids, 200–202
growth habit, 193t, 196 Manihotoides pouciflora, 186
hybridization with cassava, 184 –185, 187 Meiotic restitution
aneuploidy in, 221–222 in cassava interspecific hybrids, 201, 202
apomixis and, 211–212 Mentor pollen, 208
characterization of hybrids, 200 –204 Mercuric chloride, 47
INDEX 235
Mesopores Pesticide resistance

modeling diffusion in, 33 – 35 as environmental indicator, 81
size of, 18 Pesticides
Meteorological field capacity, 157 environmental health and, 79
Mexico pH, see also Soil pH
wild species of Manihot in, 183 effects on phosphate solubility in liquid me-
Micropores dia, 109, 124, 130 –131
modeling diffusion in, 35 Phenathrene, 17
size of, 18 Phosphate, see also Fungi, phosphate-solubiliz-
Microspores, unreduced ing; Phosphorus transfer; Soil phosphorus
in cassava, 216–217 in liquid medium studies
interspecific hybrids, 217–220, 223 –224 chelation of metal ions by organic acids,
overview of, 215–216 127–130
Mineral surfaces pH effects and, 109, 124, 130–131
sorption and, 6–7 solubility of, factors affecting, 109, 124 –
types of, 6 126
time-based fluctuations in, 132–133
N titratable acidity and, 130
precipitation in soils, 102–103
National Resource Inventory, 83 sorption in soils, 101–102
Nitrate Phosphate fertilizers
effects on phosphate solubilization by fungi, soil accumulation of phosphate, 100, 103, 154
132 Phosphorus transfer
Nitrogen effective rainfall and, 155, 157, 158
effects on phosphate solubilization by fungi, hydrological pathways
132 scale effects, 162–164
Nonaqueous phase liquids at the slope/field scale, 166, 168 –172
sorption by, 13 at the soil profile scale, 165 –166
Nonlinear driving force sorption models, 29 – 30 temporal aspects of, 158, 159f
levels of hydrological activity and, 157–158
O
overview of, 154 –155, 173
Organic acids Physisorption
excreted by phosphate-solubilizing fungi, on mineral surfaces, 7
126–127 overview of, 5 – 6
chelation of phosphate metal cations by, rates of, 14 –15
127–130 Piston flow, 161t, 165
excreted by roots, 106 Plants
stability constants for, 128t uptake of soil phosphate, 105 –106
Organoclays Pollen
sorption in, 20 mentor, 208
Overland flow Pollutant transfer, 154, see also Phosphorus
described, 160t, 164 transfer
phosphorus transfer and, 170 –171 Polycyclic aromatic hydrocarbons
Oxalic acid sorption by soot, 13
excreted by phosphate-solubilizing fungi, 126 Polyploidy, in cassava, 218
advantages of, 215
P aneuploids, 220 –222
through interspecific hybridization, 201–202,
Partial-area runoff, 170 217–220
Penicillium, see Fungi, phosphate-solubilizing triploids, 222–224
236 INDEX
Polyploidy (continued) typing by hydrological characteristics, 155,

unreduced microspores and, 215 –220, 223 – 156t
224 Soil aggregates, 33
Precipitation Soil chemicals
phosphorus transfer and, 155, 157, 158 bioavailability of, 83
Preferential flow, 161t, 165 –166 as environmental indicators, 83 – 84
Pressure-state-response framework, 78 Soil columns
Probability density functions, 43 – 44 advection-dispersion equation, 50
Propylene oxide, 47 overview of, 49 – 51
transport models with sorption kinetic terms,
R 51– 54
zero-length, 56
Radial diffusion laws Soil degradation
coupled with biodegradation kinetics, 62– 63 as environmental indicator, 82
Rainfall Soil microorganisms
phosphorus transfer and, 155, 157, 158 as environmental indicators, 80
Return flow, 161t, 171 phosphate solubilizing, 106 –108 (see also
Rhizosphere Fungi, phosphate–solubilizing)
solubilization of phosphates in, 105 –106 soil phosphorus dynamics and, 100
Roots Soil nutrients
of drought tolerant cassava hybrids, 205 –207 as environmental indicators, 83 – 84
Runoff Soil organic matter
described, 161t atmospheric carbon and, 86 – 87
phosphorus transfer and, 169 –170 components of, 8
use of term, 164 defined, 85 – 86
diffusion models for, 37– 41
S as environmental indicator, 85 – 90
functions of, 86
Saturated flow, 161t, 165 hindered diffusion in, 20 –22, 23
Saturation-excess overland flow, 171 hysteresis and, 25 –26
Self-diffusion, 31–32 measures of, 88
Shale models of sorption in, 8 –12
sorption by, 13 quantities of, measuring and expressing, 89 –
Siloxane mineral surfaces, 6 90
Slope/field hydrological pathways, 162, 164, rubber-glassy polymer concept of, 9 –12
166, 168–172 in soil aggregates, 33
Sodium azide, 47 soil content of, factors controlling, 87– 88
Soil sorption rates, 15
accumulation of phosphate in, 100, 103, 154 Soil particles
bioavailability of chemicals in, 56 – 58 description of, 32– 33
soil organic matter in diffusion models and
factors controlling content of, 87– 88 for multiple particle sizes, 41– 43
functions of, 86 for particles with fixed pore sizes, 32– 37
sorption in Soil pH
heterogeneous soils, 23 –24 phosphate solubility and, 104 –105
mineral surfaces, 6 –7 Soil phosphorus, see also Fungi, phosphate-sol-
other carbonaceous material, 13 ubilizing
soil organic matter, 8 –12 cycle, 103 –104
types of hydrological pathways through, 158 – deficiency in, 104
161 overview of, 100
INDEX 237
phosphate sorption, 101–102 Sorption, nonequilibrium

plant uptake of, 105–106 retardation mechanisms
precipitation of phosphate compounds, 102–103 hysteresis, 24 –27
soil pH and, 104–105 intraorganic matter diffusion, 20–23
Soil pores overview of, 17–18
condensation and, 4 pore diffusion, 18–20
diffusion models soil heterogeneity and, 23 –24
for biporous particles, 36 – 37 uptake and release profiles, 16 –17
combined pore diffusion/organic matter Sorption kinetics
model, 39–41 bioavailability of soil chemicals and, 56 –
for macropores, 33– 35 58
for mesopores, 33– 35 models
for micropores, 35 based on bond energetics, 27–29
diffusion through, 18–20 coupled sorption-degradation models, 58 –
size classes, 18 65
water concentration and bioavailability rela- diffusion models, 30 – 43, 62– 63
tionship, 58 linear driving force models, 29 – 30, 61–
Soil productivity 62
as environmental indicator, 78 solute transport models, 50 – 54, 63 – 65
Soil profile hydrological pathways, 165 –166 stochastic models, 43 – 45
Soil solution rates of elementary processes in, 14 –15
phosphorus dynamics and, 103 –104 significance of, 2
Solute transport models Static batch reactor, 49
advection-dispersion equation, 50 Stirred-flow cells, 54 – 56
with biodegradation term, 63 – 65 Stochastic sorption models, 43 – 45
with sorption kinetic term Storm flow
diffusion model, 52– 54 phosphorus transfer and, 157, 158
first-order model, 52 Storm runoff, 171
two-region model, 51– 52 Subsurface flow, 161t, 164
Soot Successive-dilution technique, 48
sorption by, 13 Surface runoff
Sorbate, 2 described, 161t
Sorbent, 2 phosphorus transfer and, 170 –171
Sorption, see also Sorption, nonequilibrium;
Sorption kinetics T
categories of, 3–5
defined, 3 Throughflow
experimental methods described, 161t
batch techniques, 45 – 49 phosphorus transfer and, 171–172
column techniques, 49 – 54 Tillage
stirred-flow cell techniques, 54 – 56 effects on soil organic matter, 87, 88
time frames, 45 Titratable acidity, 130
zero-length columns, 56 Tortuosity
factors affecting, 2 in diffusion through pores, 18
on mineral surfaces, 6 –7 Transgenic crops
of phosphate in soil, 101–102 pest resistance and, 81
pH and, 104–105 Transport diffusion, 31– 32
steps in, 3 Tripartitae, 186
thermodynamics of, 13 –14 Triploidy
types of molecular interactions in, 5 – 6 in cassava, 222–224
238 INDEX
Tupi-Guarani Indians, 185 W

Two-region solute transport model, 51– 52
Water films
V sorption and, 5
Van Der Waals forces Z

in physisorption, 5
Vesicular mycorrhizal fungi, 100, 141–142 Zero-length columns, 56

Advances in Agronomy v.69

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

Advances in Agronomy v.69

Uploaded by

Copyright:

Available Formats

Agronomy

Martin Alexander Ronald Phillips

Kenneth J. Frey Larry P. Wilding

Prepared in cooperation with the

THE MEASUREMENT AND INTERPRETATION OF SORPTION

ENVIRONMENTAL INDICATORS OF AGROECOSYSTEMS

GROWTH PROMOTION OF PLANTS INOCULATED WITH

V. Plant Growth Promotion by Phosphate-Solubilizing Fungi . . . . . . . . 133

HYDROLOGICAL FACTORS FOR PHOSPHORUS TRANSFER

CASSAVA, Manihot esculenta Crantz, GENETIC

Sorption is fundamental to the fate of organic chemicals in soil environments.

II. THE NATURE OF ELEMENTARY SORPTION

Figure 1 Different types of sorption available to organic molecules. A, Adsorption; B, absorption

• Association with water ﬁlms: Depending on the relative humidity, unsaturated

A. INTERMOLECULAR INTERACTIONS AVAILABLE

Organic compounds can undergo chemisorption, physisorption, or ion pair for-

Charge-transfer interactions (often referred to as donor–acceptor interactions)

B. PROPERTIES OF SOIL COMPONENTS

Two principal types of surface exist on natural minerals:

2. Soil Organic Matter

It is well established that sorption of hydrophobic compounds out of aqueous

where KD is the (linear) dissolution domain coefﬁcient made up of inseparable

3. Carbonaceous Materials Other Than SOM

C. THERMODYNAMIC DRIVING FORCE FOR SORPTION

D. RATES OF ELEMENTARY PROCESSES

Figure 5 Energy diagram for a physisorbing molecule approaching the surface.

III. SLOW SORPTION AND DESORPTION

A. UPTAKE AND RELEASE PROFILES

Evidence for slow (“nonequilibrium”) sorption phenomena has come from

1. Desorption is highly temperature dependent, being signiﬁcantly enhanced by

tration (more correctly, a gradient in chemical potential) so as to achieve maxi-

(1986) appear to be among the ﬁrst to employ it to describe intraparticle diffusion

2. Intraorganic Matter Diffusion

turn is proportional to the thermodynamic free energy of sorption, Gsorp. The

3. Effect of Soil Heterogeneity on Sorption Kinetics

Hysteresis refers to the apparent asymmetry (nonsingularity) of the sorption/

to rate-limited diffusion can lead to an underestimation of equilibrium sorbed con-

made between sorbate-induced changes in the conformation of humic molecules

er for desorption than sorption. “Thermodynamic” and “kinetic” hysteresis may

IV. SORPTION KINETIC MODELS

A. MODELS BASED ON BOND ENERGETICS

= ka′ p( ST − So ) − kd′ So (8)

by replacing p with aqueous concentration C [M L3]; replacing So with sorbed

The equilibrium expressions are as follows.

B. DRIVING FORCE MODELS

2. Diffusion in Fixed-Pore Systems

a. Macropores and Mesopores

sp [M L3] is the local sorbed concentration (i.e., c cp  (1  )sp, where 

∂cp ∂sp ⎡ ∂ 2 cp 2 ∂cp ⎤

Pore diffusivity taking into account surface diffusion can be expressed as

To simplify modeling it has been customary to (i) neglect surface diffusion;

aggregate (Ra) made up of individual microporous grains (r ). This model com-

where C(t) [ML3] is the external solution concentration.

3. Diffusion in Organic Matter

Rubbery SOM, in which diffusant molecules are considered to be dissolved, is

Therefore, we can eliminate sL in Eq (38). After derivatizing we ﬁnd that

4. Combined Pore Diffusion/Organic Matter Diffusion Model

The governing equation is

∂cp ∂N ⎛ ∂cp2 2 ∂cp ⎞

5. Mixtures of Particle Sizes

respectively. Multiple-particle class models better represented long-term sorption

where the arithmetic mean of k is k̄ exp(  2 /2) and the variance of k is V

In a review such as this, it is worth discussing the methodology used in sorp-

In solution–solid systems, agitation (stirring or shaking) is usually required to

by replacing p with aqueous concentration C [M L3]; replacing So with sorbed

sp [M L3] is the local sorbed concentration (i.e., c cp (1 )sp, where

where the arithmetic mean of k is k̄ exp( 2 /2) and the variance of k is V

where L a/D and the n values are the roots of

(m + fnKcCim n −1 ) ∂C∂tm + (im + (1 − f )nKc Cim n −1 ) ∂C∂tim =

(im + (1 − f )Kc nCim n −1 ) ∂C∂tim = (Cm − Cim ) (59)

Rate of uptake from surface ksa* (1 fw) (76)

where C is the solution concentration experienced at cell surfaces [M L3], max

where YUOM [M M1] is the corresponding bioconversion factor for growth on

Bosma et al. (1997) studied the biodegradation of -hexachlorocyclohexane

Cd + K m + vmax −1 ⎧⎪ ⎡ 4Cd vmax −1 ⎤

where v is the speciﬁc degradation velocity [M T1]; vmax ( maxX/YS ) is the

(m + fnKe Cim n −1 ) ∂C∂tm + (im + (1 − f )nKe Cim n −1 ) ∂C∂tim = m Dh

where is the ﬁrst-order biodegradation rate constant [T1]. An analytical solu-