Preface

Volume 80 contains five comprehensive reviews dealing with timely topics in the plant
and soil sciences. Chapter 1 is an important review on soil as a source of dust and
implications for human health. Dispersion mechanisms, characteristics of soil dust,
modes of exposure and health effects are discussed. Chapter 2 is a timely review on
ecological approaches for enhancing weed management in the semiarid Great Plains.
Discussions on herbicide resistance, changes in cropping practices, ecologically based
weed management, reducing weed density, and enhancing crop competitiveness are
included. Chapter 3 is a comprehensive overview of nutrient management practices for
improving nutrient use including liming, nitrogen, phosphorus, potassium and
micronutrients. Chapter 4 is a cutting-edge review on molecular breeding of
groundnut to enhance productivity and food security in the semi-arid tropics.
Discussions on genetic variability, conventional genetic improvement, biotechnology
applications, and opportunities for molecular breeding are included. Chapter 5 is a
contemporary overview of scale-dependent disperson in soils including stochastic and
numerical approaches.
I thank the authors for their fine contributions.
DONALD L. SPARKS

xi
 Contributors
Numbers in parentheses indicate the pages on which the authors’ contributions begin.

R. Anderson (33), USDA—Agricultural Research Service, 2923 Medary Ave.,

Brookings, SD 57006, USA
V. C. Baligar (63), USDA-ARS, Alternate Crops and Systems Research
Laboratory, Beltsville Agricultural Research Center, Room 342, Building 001,
10300 Baltimore Avenue, Beltsville, MD 20705-2350, USA
J. H. Crouch (153), International Crops Research Institute for Semi-Arid Tropics
(ICRISAT), Patancheru, PO 502324, A.P., India
S. L. Dwivedi (153), International Crops Research Institute for Semi-Arid Tropics
(ICRISAT), Patancheru, PO 502324, A.P., India
N. K. Fageria (63), National Rice and Bean Research Center of EMBRAPA, Caixa
Postal 179, Santo Antônio de Goiás, Goiás, CEP 75375-000, Brazil
M. E. Ferguson (153), International Crops Research Institute for Semi-Arid
Tropics (ICRISAT), Patancheru, PO 502324, A.P., India
K. Lee (1), Epidemiology and Preventive Medicine, University of California, One
Shields Avenue, Davis, CA 95616, USA
S. N. Nigam (153), International Crops Research Institute for Semi-Arid Tropics
(ICRISAT), Patancheru, PO 502324, A.P., India
A. H. Paterson (153), Plant Genome Mapping Laboratory, University of Georgia,
Room 162, Riverbend Research Center, 110 Riverbend Road, Athens, GA 30602,
USA
H. M. Selim (223), Louisiana State University, Department of Agronomy, 305 M
B Sturgis, Baton Rouge, LA 70810, USA
N. A. Slaton (63), University of Arkansas, 1366 W. Altheimer Dr., Fayetteville,
AR 72704, USA
J. Smith (1), Epidemiology and Preventive Medicine, University of California, One
Shields Avenue, Davis, CA 95616, USA
L. Zhou (223), 251 10th St NW #14-133, Atlanta, GA 30318-5615, USA

ix
 Contents
CONTRIBUTORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

SOIL AS A SOURCE OF DUST AND IMPLICATIONS

FOR HUMAN HEALTH
J. Smith and K. Lee
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
II. Dispersion Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
A. Wind Erosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
B. Agricultural Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
III. Characteristics of Soil Dust . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
A. Inorganic Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
B. Organic Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
C. Size Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
D. Contaminated Soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
IV. Exposure to Soil Dust . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
A. Occupational Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
B. Fugitive Dust Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
C. Direct Soil Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
V. Health Effects of Soil Dust . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
A. Health Effects of Agricultural Dust . . . . . . . . . . . . . . . . . . . . . . . . . . 21
B. Health Effects of Fugitive Dust . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
C. Protective Effects of Farm Environment . . . . . . . . . . . . . . . . . . . . . . 24
D. Reducing Exposure to Soil Dust . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

AN ECOLOGICAL APPROACH TO STRENGTHEN WEED

MANAGEMENT IN THE SEMIARID GREAT PLAINS
R. Anderson
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
A. Changes in Cropping Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
B. Concerns with Herbicide Resistance and Input Costs . . . . . . . . . . . . . 35
II. Ecologically Based Weed Management . . . . . . . . . . . . . . . . . . . . . . . . . 36
A. Reducing Weed Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
B. Enhancing Crop Competitiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
C. Opportunity to Integrate Ecological Processes
with Weed Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

v
vi CONTENTS

III. Reducing Weed Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

A. Rotations Comprised of Crops with Different Life Cycles . . . . . . . . 38
B. Tillage Minimizes Rotation Effect on Weed Density . . . . . . . . . . . . 43
IV. Enhancing Crop Competitiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
A. Winter Wheat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
B. Corn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
C. Sunflower . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
D. Proso Millet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
V. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
A. Rotations Based on a Cycle-of-Four Aid Weed Management
but also Accrue Other Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
B. Semiarid Paradox with Tillage and Herbicides . . . . . . . . . . . . . . . . . 57
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE

PRODUCTIVITY AND SUSTAINABILITY
N. K. Fageria, N. A. Slaton and V. C. Baligar
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
II. Soils Used for Rice Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
III. Management Practices to Improve Nutrient Use Efficiency . . . . . . . . . . 67
A. Liming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
B. Nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
C. Phosphorus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
D. Potassium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
E. Calcium, Magnesium, and Sulfur . . . . . . . . . . . . . . . . . . . . . . . . . . 107
F. Micronutrients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
IV. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136

MOLECULAR BREEDING OF GROUNDNUT FOR ENHANCED

PRODUCTIVITY AND FOOD SECURITY IN THE SEMI -ARID
TROPICS: OPPORTUNITIES AND CHALLENGES
S. L. Dwivedi, J. H. Crouch, S. N. Nigam, M. E. Ferguson
and A. H. Paterson
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
A. Crop Production and Uses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
B. Constraints to Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
 CONTENTS vii

II. Genetic Variability in Groundnut . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162

A. Wild Arachis Species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
B. Cultivated Groundnut . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
III. Highlights and Constraints of Conventional Genetic
Improvement in Groundnut . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
A. Breeding Methods and Cultivars . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
B. Successes and Limitations to Conventional Breeding . . . . . . . . . . . . 175
IV. Overview of Biotechnology Applications in Groundnut . . . . . . . . . . . . . 178
A. Molecular Genetic Diversity Analysis . . . . . . . . . . . . . . . . . . . . . . . 178
B. Molecular Genetic Linkage Mapping . . . . . . . . . . . . . . . . . . . . . . . 179
C. Progress in Model Systems and Comparative Mapping
Amongst Legumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
D. Wide Crosses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
E. Genetic Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
V. Opportunities for Molecular Breeding in Groundnut . . . . . . . . . . . . . . . 187
A. Current Understanding and Genetic Basis of
Economically Important Traits . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
B. Developing Appropriate PCR-based Markers . . . . . . . . . . . . . . . . . . 192
C. Mapping and Genetic Enhancement Strategies . . . . . . . . . . . . . . . . . 194
D. Marker-Assisted Gene Introgression from Wild
Arachis to Arachis hypogaea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
E. Marker-Assisted Backcross Breeding . . . . . . . . . . . . . . . . . . . . . . . 195
F. Prioritizing Traits for Marker-Assisted Selection . . . . . . . . . . . . . . . 197
VI. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201

SCALE -DEPENDENT DISPERSION IN SOILS: AN OVERVIEW

L. Zhou and H. M. Selim
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
A. Justification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
B. Laboratory Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
C. Field Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
D. Interpretations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
E. Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
II. Stochastic Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
A. Macrodispersivity Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
B. Fractal Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
C. Transfer Function Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
III. Analytical Solutions of CDE with Scale-Dependent
Dispersivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
A. Solutions of CDE with Time-Dependent Dispersivity . . . . . . . . . . . 239
B. Solutions of CDE with Distance-Dependent Dispersivity . . . . . . . . . 243
IV. Numerical Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
viii CONTENTS

V. Fractional Order Advection – Dispersion Equation (FADE) . . . . . . . . . . 255

A. Derivation of FADE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
B. Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
C. Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
VI. Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259

INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
 SOIL AS A SOURCE OF DUST AND
IMPLICATIONS FOR HUMAN HEALTH
Jodi L. Smith and Kiyoung Lee
Epidemiology and Preventive Medicine, University of California,
One Shields Avenue, Davis, California 95616, USA

I. Introduction
II. Dispersion Mechanisms
A. Wind Erosion
B. Agricultural Operations
III. Characteristics of Soil Dust
A. Inorganic Compounds
B. Organic Compounds
C. Size Distribution
D. Contaminated Soil
IV. Exposure to Soil Dust
A. Occupational Exposure
B. Fugitive Dust Exposure
C. Direct Soil Contact
V. Health Effects of Soil Dust
A. Health Effects of Agricultural Dust
B. Health Effects of Fugitive Dust
C. Protective Effects of Farm Environment
D. Reducing Exposure to Soil Dust
References

I. INTRODUCTION

Soil is a major part of the outer layer of the earth, which contains mineral and
organic materials. Our food is grown in the soil, and our shelters built on it. Soil is
the base upon which we build society (Singer and Warkentin, 1996). Good soil is
a critical resource for great civilizations, as it makes possible the shift from
nomadic societies to stable communities (Brady, 1974). Biological, chemical,
and physical agents can change soil conditions (Johnson, 1998). The contents and
1
Advances in Agronomy, Volume 80
Copyright q 2003 by Academic Press. All rights of reproduction in any form reserved
0065-2113/03$35.00
2 J.L. SMITH AND K. LEE

activities of soil are determined by a spectrum of interrelated physical and

chemical factors (Killham, 1994). Soil is the transition stage from rock back to
rock; it is in constant movement and under the influence of many forces (Lyon
et al., 1915).
In this constant movement, soil can become suspended in the atmosphere. Soil
dust is produced when energy is transferred to the particles that make up the soil
on the surface of the earth. Soil dust plays a large role in air quality, and dust from
soil is the second largest primary particle source after sea salt (Prospero et al.,
1983). Soil can be dispersed through the air by both natural and anthropogenic
mechanisms. Wind and agriculture are two of the primary forces that generate
soil dust, although construction, driving on unpaved roads, and other activities
may also have an effect.
When soil is suspended in the atmosphere as dust, it can be transported and
absorbed by the human body. This can result in health problems such as eye
irritation, respiratory disorders, pulmonary disease, and an increased risk of lung
and skin cancer (Rylander, 1986; Clausnitzer and Singer, 2000). People who
work in dusty environments, such as agriculture workers, construction workers,
and miners, are affected the most by soil dust exposure. However, as soil dust
increases the amount of particulate matter in the air, it can also influence the
health of the general population.
This paper reviews mechanisms of soil dust generation and exposure, the
characteristics of soil dust and some possible health effects of soil dust exposure.
Little research has focused on the health effects of general soil dust rather than
specific mineral or organic dusts. Few studies have looked at the health effects of
soil dust outside of the occupational realm.

II. DISPERSION MECHANISMS

Soil dust can be dispersed by either natural or anthropogenic mechanisms. The

chief natural mechanism of soil dust generation is wind erosion. There are three
processes by which wind erosion disperses soil dust: surface creep, saltation, and
suspension. The efficiency of each process is affected by the size of the soil
particles. The magnitude of wind erosion is influenced by a large number of
factors ranging from atmospheric conditions to soil composition. These factors
have either positive or negative impacts on the amount of dust released.
While anthropogenic dispersion mechanisms have less of an impact on large
open areas, they can have a significant impact on regional pollution and
occupational exposure. They are also easier to control than natural dispersion
mechanisms. Anthropogenic dispersion mechanisms include agricultural oper-
ations, construction, and driving on unpaved roads.
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 3

A. WIND EROSION

Wind is the main natural mechanism for soil dust generation (Gillette et al.,
1993). Wind erosion occurs when air currents move over the surface of the earth
and suspend particulates in the atmosphere. Depending on their size, these
particulates may fall to the ground shortly thereafter or they may be suspended
for weeks. Larger particles have greater settling velocities and therefore fall out
of the atmosphere faster than smaller particles. Particles that have aerodynamic
diameters of less than 20 mm can be suspended by rising air currents and
carried over long distances at much greater heights than coarser particles
(Gillette, 1977). These aerosols may remain in the atmosphere for weeks and, in
many cases, precipitation events are required to scrub these particles out of the
atmosphere.

1. Impacts of Wind Erosion

Soil dust is a major component of atmospheric particulates, though soil dust

concentrations vary widely over regions. At an estimated rate of generation of
182 £ 109 kg/year worldwide, soil dust represents 9% of the total global
particulate emissions from natural sources (Su, 1996). The rate at which soil dust
is generated can be the highest in arid regions, which cover more than a third of
the entire continental surface (Su, 1996). In the USA in 1989, wind erosion
generated an estimated 10.7 £ 109 kg/year of particulate matter with an
aerodynamic diameter of 10 mm or less (PM10) (US Environmental Protection
Agency (EPA), 1996).
But because wind can transport dust over great distances, wind-generated dust
is a problem that spans the globe. The smaller particles are suspended the longest
(Pye, 1987). For long-term suspension, particles must be smaller than 20 mm in
aerodynamic diameter (Pye, 1987). Soil dust generated in Africa can be
transported to the southeastern United States (Prospero, 1999). Soil dust from
China can reach the Hawaiian Islands and western North America (Shaw, 1980;
McKendry et al., 2001; Uno et al., 2001). Wind-generated soil dust can reduce
visibility, change the mineral content of soils, affect meteorological activity and
human health (Norboo et al., 1991). Of concern is the possibility that wind-borne
dust may contain toxic compounds in addition to natural soil dust. The suspension
time of particles depends on their size.
Dust storms are dramatic events in which a large quantity of soil dust is
injected into the atmosphere in a short period of time. Such storms are associated
with energetic meteorological events where wind speeds are high and there is a
fair amount of exposed soil. Dust concentrations of several hundred micrograms
per cubic meter are common in dust storms (Gillette et al., 1993), and the particle
4 J.L. SMITH AND K. LEE

size typically ranges from 1 to 20 mm in aerodynamic diameter (Gillette et al.,

1978).
Although most regions of the continental United States are susceptible to
airborne dust due to wind erosion, the southern Great Plains region has the
highest dust production (Orgill and Sehmel, 1976; Gillette and Hanson, 1989).
The Rocky Mountain region and the northern Great Plains region have the second
highest dust production levels while the northeastern region and Great Lakes
region have the lowest dust production levels. Relatively high dust production
levels can be observed in central and southern California, while the Pacific Coast
region of Washington, Oregon and northern California are among the lowest dust
producing regions in the nation.
When dust activity was studied in seven regions of the United States,
suspended dust occurred with the greatest frequency in the spring, except in the
Pacific Coast region (Orgill and Sehmel, 1976). This seasonal dust profile is due
to the strong winds associated with spring cyclonic and convective storm activity.
In the Pacific Coast region, the maximum dust frequency occurs during the fall
and early winter. All of the regions showed a definite decrease in dust during
summer. Visibility of less than 7 miles was used to define the occurrence of dusty
conditions. Orgill and Sehmel (1976) also found that the highest dust production
frequencies occur in the afternoon hours when the atmospheric boundary layer is
deep and mixing due to turbulence is the greatest.

2. Mechanisms

Wind erosion can generate soil dust by different processes: surface creep,
saltation, and suspension. Each of the three processes affects particles of different
sizes. Surface creep affects the largest of particles while suspension affects the
smallest.
Threshold friction velocity is an important concept in wind erosion. The
threshold friction velocity corresponds to the minimum wind stress needed to
overcome the forces holding soil particles in place (Gillette, 1988); it must be
exceeded to initiate wind erosion by surface creep or saltation. This is usually in
the range of 10 –50 m/s (Gillette et al., 1993).
As the wind blows, large particles of 500 –1000 mm in aerodynamic diameter
are dislodged from the surface and begin to roll and slide along the ground.
During this movement, the particles may strike other particles, where they impart
their momentum and cause the secondary particles to become airborne. This
process is surface creep. Surface creep is the primary wind erosion process
applicable to particles in the 500– 1000 mm range, because they are often too
heavy to become airborne (Cowherd, 1993). The dislodged secondary particles
have an upper size limit of about 70 mm in aerodynamic diameter (Pye, 1987).
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 5

In saltation, individual grains ranging from 75 to 500 mm in aerodynamic

diameter (Cowherd, 1993) are dislodged from the surface and pushed upwards.
The particles fall towards the earth under the effect of gravity and bounce
across the soil surface. When the bouncing grains strike the ground, they may
make contact with other particles protruding out of the surface. One or both
may then be launched upward and become airborne. Saltation is the dominant
process by which wind erosion produces aerosolized soil dust (Gillette and
Blifford, 1974).
The particles dislodged by the wind during saltation become airborne, but the
particles dislodged during surface creep do not. It is possible for the dislodged
particles to become airborne in saltation. Because they are smaller, they bounce
across the surface rather than roll or slide. During saltation, these bouncing
airborne particles do not rise much over a few feet (Cephil and Woodruff, 1963).
However, saltation is important in releasing particles with an aerodynamic
diameter less than or equal to 1 mm (PM1.0), which would normally need very
high wind velocities to become airborne by wind alone. Fine particles pose more
of a health risk than larger particles since they can reach the alveolar region of the
respiratory system, where deposition can cause respiratory damage.
In suspension, small particles follow air currents, rising up from the ground.
Once lifted off the ground, the particles are completely airborne, usually reaching
great heights (Cephil, 1945). This process affects particles that are less than
75 mm in aerodynamic diameter (Cowherd, 1993). In order for suspension to take
place without the presence of very high winds, saltation must first release fine soil
particles, which are then lifted by air currents. Suspension is the primary method
by which particles are transported long distances. In order to remove suspended
particles from the atmosphere, the wind must die down considerably or a
precipitation event must occur (Cephil, 1945).

3. Factors

Various factors can increase or decrease the amount of soil dust generated and
suspended in the air by wind erosion. The factors can be classified into three
groups: features of the surface material (composition, field length, vegetation,
and ground structure), energy expended onto the surface (wind speed, frequency,
and distribution), and atmospheric conditions (precipitation and soil moisture
content). Factors affecting wind erosion are interrelated, and the erodibility of the
soil can vary greatly depending on which factors are present.
Soil erodibility is associated with surface soil material and its texture. The
proportion of silt in a soil correlates inversely with erodibility, while the
proportion of sand correlates directly with erodibility (Cephil and Woodruff,
1963). When the texture of the surface material is fine, more fine particles can be
6 J.L. SMITH AND K. LEE

emitted. However, even if the surface texture is fine, fine particles will remain in
place when the surface structure is resistant to breakage (Gillette, 1977). Silts and
clay cohere and form clods after being wet. Sand grains are not as cohesive as
silts and clays and therefore do not cohere and form clods as well. Clod formation
reduces erosion because the compact mass is more resistant to wind than loose
soil. Clods also act to roughen the soil surface and decrease total erosion. The
most easily eroded soils consist of finely textured particles with aerodynamic
diameters of about 100 mm (Battan, 1966).
There is some evidence that the organic content of soils influences erodibility
(Cephil and Woodruff, 1963). The initial decomposition of organic material tends
to moisten the soil and therefore decreases erodibility. But after decomposition,
soils with high organic matter tend to be more susceptible to erosion.
In an field, the soil erosion rate increases with the distance the field extends in
the direction of the wind flow until reaching a maximum (Woodruff and
Siddoway, 1965). Therefore, the length of the exposed field in the direction of the
wind correlates with the amount of eroded soil dust. It may be possible to reduce
soil erosion by repositioning fields to reduce the length they extend in the average
direction of the wind.
Naturally occurring or planted vegetation offers protection from wind erosion,
as plants tend to anchor the soil (Chepil and Milne, 1941). When there is little
vegetation and the ground is dry, even moderate winds can cause particles to rise
to high altitudes (Battan, 1966). Crop type is a factor in erodibility as methods of
farming vary among crops (Green et al., 1990) and crops vary in the amount of
ground cover they provide. Crops that grow close to the ground offer more
erosion protection than those that grow taller. Crops that grow densely offer more
erosion protection than those that require space between plants.
Surface roughness affects soil erodibility. Wind velocity near the ground is
faster over a smooth surface than a rough surface. Therefore, a smooth surface is
more erodable than a rough surface, given that all other conditions are identical
(Cephil and Woodruff, 1963). This increased wind velocity causes an increased
chance of particle dislodgement. A smooth surface also results in less turbulence
at the surface, which would tend to result in a decrease in particle dislodgement.
However, this decrease in turbulence does not completely compensate for the
increase in erosion from the increased wind velocity over a smooth surface
(Chepil and Milne, 1941). Therefore, a rough surface is generally a better
protection against wind erosion.
Wind characteristics such as speed, frequency and distribution can affect soil
dust suspension, as can the minimum threshold friction velocity required to
initiate wind erosion. As different soil types are dominated by different particle
size distributions, the minimum threshold friction velocity varies with the type of
soil (Gillette et al., 1993). Extensive research has focused on determining the
threshold friction velocities of various types of soil (Cephil and Woodruff, 1963;
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 7

Gillette et al., 1980, 1982; Gillette, 1988). Sandy soils tend to have the lowest
threshold velocities, while loamy soils have the highest threshold velocities.
The more frequently the surface is disturbed by the wind, the more the
potential for erosion is increased, because the mitigating effects of crusts,
vegetation, and friable non-erodable elements are destroyed (Cowherd, 1993).
Frequent winds also expose fine soil particles that are easily eroded. Wind
frequency and distribution are affected by meteorological conditions and
topography. Major dust suspension in the United States can be caused by many
types of meteorological events, including convective weather systems, frontal
passages, cyclogenesis, low-level winds associated with upper-level jet streams,
diurnal winds, and mountain katabatic winds (Orgill and Sehmel, 1976).
Precipitation influences the moisture content of the soil, and its force on the
soil material releases erodable elements (Cephil and Woodruff, 1963). However,
the release of erodable materials by precipitation is probably not of great concern
since the increased moisture is a major deterrent to wind erosion. Dust emissions
from wind erosion are known to be strongly dependent on the moisture level of
the erodable material (Woodruff and Siddoway, 1965). Water forms cohesive
films between the individual grains of soil, and the fine particles tend to form a
crust that holds in moisture and resists erosion. The degree of protection afforded
by this crust depends on the thickness of the crust and its resistance to rupture
(Cowherd et al., 1985). In arid regions, extended periods of diminished rainfall
result in dry soils and a lack of vegetation that would have protected the surface
from erosion (Gillette and Hanson, 1989).
Temperature is another variable that affects the soil’s potential for erosion.
Increased air temperature reduces the soil’s moisture content. The relationship
between soil water content and air temperature was studied in California
(Clausnitzer and Singer, 2000). As the soil moisture content increased, the
respirable dust concentration decreased exponentially, and at midday tempera-
tures, the respirable dust concentration was 2 –3 times greater than it was during
cooler morning temperatures.

B. AGRICULTURAL OPERATIONS

Human activities, especially agriculture, are closely associated with soil dust
generation. In agricultural areas, seasonal variations in total suspended particles
reflect the cycles of agricultural activity (Green et al., 1990). Total suspended
particulates show bi-modal peaks that correspond with maximal farm activities in
the spring and early fall. In dust samples from Alberta, Canada, respirable dust
accounts for 50% or more of the total suspended particulates. Because fine dust
can travel short or long distances based on meteorological conditions,
agricultural activities can increase exposure to particulates for the general
population as well as agricultural workers.
8 J.L. SMITH AND K. LEE

The production of dust by agricultural operations is affected by the number of

such operations taking place in the dry season and by the temperature, the wind
speed, the tractor speed, the size of the equipment being used, and the degree of
contact the equipment has with the soil or crop (Clausnitzer and Singer, 1996).
Soil preparation and crop management activities are two types of agricultural
operations that can cause major dust emissions.

1. Soil Preparation

Activities aimed at preparing and maintaining soil are among the largest
anthropogenic producers of fugitive dust. Such activities include plowing,
grating, disking, mowing, and tilling. For many crops, land preparation
necessitates rigorous contact with the soil. Crops that are intensively cultivated
tend to result in higher dust concentrations (Clausnitzer and Singer, 1996).
Agricultural land is generally prepared for crops in the early spring and fall
months, which explains why there is a peak in dustiness for the country during the
spring months and a secondary peak during the fall (Orgill and Sehmel, 1976;
Green et al., 1990).
In a study of respirable dust production from agricultural operations,
respirable dust was collected at the farm implement for 29 farming operations
over a two-year period. Fields were arranged in seven crop systems, including
tomato, corn, wheat, a legume cover crop, and a fallow rotation. Land preparation
activities (including land planing, ripping, plowing, and finish disking) accounted
for 67% of farming operations and produced 82% of respirable dust (Clausnitzer
and Singer, 1997). In other words, land preparation accounted for a dispro-
portionate amount of the total respirable dust among every crop studied. In
contrast with cultivation activities, land preparation activities require more
contact between the implements and the soil, and they tend to occur when soil
moisture content is low.
During land preparation, soil is dropped to the surface and dust particles are
injected into the atmosphere by forced saltation. The total quantity of dust
released is proportional to the area tilled (Cowherd et al., 1985). Cowherd
reported that in depths of 0 – 4 in., dust emissions from tilling vary directly with
the silt content of the surface soil. The dust emission factor is proportional to the
percentage of silt in the surface soil raised to the 0.6 power. The silt content of
soil is determined by passing dry soil through a 200 mesh screen. The association
between soil silt content and dust emissions is also a function of the particle size
multiplier, a dimensionless fraction based on particle size. The particle size
multipliers are 1.0 for total particulate, 0.21 for particles less than 10 mm, and 0.1
for particles less than 2.5 mm. This indicates that the emission factor decreases as
particle size decreases.
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 9

Dust emissions also depend on how agricultural operations are performed.

Expending less energy on the soil surface generates a smaller amount of dust.
Different operations generated a wide range of respirable dust concentrations,
ranging between 0.33 mg/m3 from disking in corn stubble and 10.3 mg/m3 from
ripping soil (Clausnitzer and Singer, 1997). Although lower tractor speed reduced
personal exposure to respirable dust during land preparation (Nieuwenhuijsen
and Schenker, 1998; Nieuwenhuijsen et al., 1998), whether total emission
quantity is affected by lowering tractor speed is unclear. This study also
demonstrated that an enclosed tractor cab and an increase in relative
humidity reduced personal exposure to dust emissions during land preparation
activities.
As in wind erosion, agricultural dispersion of dust is affected by various
environmental conditions. Water suppresses dust by forming a cohesive film
between grains of surface materials. The soil moisture level depends on
precipitation and evaporation and is associated with wind speed, relative
humidity and temperature. Emission factors during land preparation activities
were generally lower when soil moisture was higher. They increased with
decreasing relative humidity, generally increased with increasing vertical
temperature differential, and were not correlated with wind speeds beyond the
range of 1– 5.5 m/s (Holmen et al., 2001). The effects of temperature and relative
humidity were also determined for agricultural operations other than land
preparation (Clausnitzer and Singer, 1996, 1997). Generally, soil moisture
content is lower during land preparation activities than during cultivation
activities (Clausnitzer and Singer, 1997), in part because land preparation usually
occurs during drier times of the year, before and after growing seasons.

2. Harvesting

Harvesting results in dust emissions but it is important to note that the type of
harvesting varies by crop. Mechanical harvesting may result in more dust
emissions than manual harvesting, which tends to disturb the soil less. Emissions
may be influenced by the condition of the crop, the harvesting mechanism used
and other environmental factors. When dust emissions were measured during the
harvesting of wheat, tomato, and corn crops, the corn harvest generated the
highest dust concentrations (Clausnitzer and Singer, 1996). For corn, tomato, and
wheat harvest, the concentrations of respirable dust were 5.6, 4.3, and 1.0 mg/m3,
respectively. The high dust concentrations during the corn harvest may be due to
the high plant biomass, the dryness of the plants at the time of harvest, and the
dust deposited on the leaves over the growing season. The concentrations during
the corn harvest were lower than 10.3 mg/m3 in land planning but higher than
1.0 mg/m3 in seeding operations.
10 J.L. SMITH AND K. LEE

A field study measured the ambient dust concentration from harvesting

operations (Clausnitzer and Singer, 1996). The ambient dust concentration
represents the total yield or emission strength of dust emissions. Although total
emission amounts may be less and plume size may be smaller from a manual
harvest than from a mechanical harvest, it should be noted that the distance
between the emission source and the worker is extremely small. Regardless of
how low the total emission strength is, agricultural workers can be exposed to a
significant amount of dust during a manual harvest. Geometric means of worker’s
respirable dust exposure during manual orange, grape, and peach harvesting were
1.14, 0.23, and 0.5 mg/m3, respectively (Popendorf et al., 1982; Lee et al., 2001).
These levels were similar to the ambient levels measured during a mechanical
wheat harvest.

3. Indirect Impacts of Agricultural Operations

Agricultural operations usually disturb the soil surface; each time that a
surface is disturbed, its potential to resist wind erosion is decreased. During
agricultural operations, the mitigating effects of crusts, vegetation, and friable
non-erodable elements are reduced or eliminated. Additionally, the fine soil
particles turned up during these operations are easily erodable (Cowherd, 1993).
Therefore, even agricultural activities that do not release a lot of dust can increase
the amount of dust released during wind erosion.

4. Other Human Activities

There are other human activities that cause soil dust dispersion, including
construction activities and driving on unpaved roads. Construction activities such
as dumping, dozing, grading, and scraping can increase the release of soil dust.
Construction also strips the land of vegetation that may decrease erosion.
Disturbing land causes an immediate acceleration in erosion (Wilshire, 1980),
which causes a short-term increase in dust generation that diminishes in time and
eventually subsides.
Driving on unpaved roads is a minor source of total atmospheric soil dust, but
it can be significant for local air pollution. On average, travel on an unpaved road
generates 1000 times more total dust than travel on a paved road over the same
distance (Evans and Cooper, 1980). Because of an uneven ratio of unpaved to
paved roads and different uses for the two types, the total amount of dust
emissions in the United States from unpaved roads was about 300 times that of
paved roads in 1976. Considering that more paved roads have been constructed in
the United States since then, the ratio may be lower now. However, it should be
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 11

noted that, in general, dust emissions from an unpaved road are likely higher than
emissions from a paved road. This can cause significant dust concentration levels
over a small area.
The amount of dust emitted by driving on unpaved roads was associated
with the silt content of the soil and with vehicle speed (Hall, 1981). The higher
the silt content, the more dust could be expected. The silt content of roads can
vary between 12 and 27%. Gravel roads, for example, contain about 12% silt, a
relatively low level. When vehicular speed increases on an unpaved road, more
dust can be generated. A typical four-wheel vehicle can generate 3.7 kg/km
driving at 60 km/h on a road with 12% silt content. Local precipitation was
not considered a significant factor in dust emissions from unpaved roads,
because compacted roads absorb less water in Arizona than an agricultural
field with a similar soil makeup. This factor may vary by region and with
road use.

III. CHARACTERISTICS OF SOIL DUST

Soil dust is composed of many agents of various particle sizes. It is a complex

mixture of organic and inorganic materials. Inorganic dusts, major components of
soil dust, are derived from a mineral source. Composition of organic compounds
can vary. The composition and properties of airborne dusts differ according to
farming patterns (Green et al., 1990), local geology and meteorological
conditions. Dust composition can be a marker as to its source (Clausnitzer and
Singer, 1999).

A. INORGANIC COMPOUNDS

As soil dust is generated from source soil, it can be assumed that the
mineralogy of soil dust reflects its source soil. Clausnitzer and Singer (1999)
tested this assumption by comparing the particle size and mineralogy of
respirable dust from agricultural activities to the particle size and mineralogy of
its source soil. Although the same major groups of minerals found in the dust
were found in the source soil, the composition of source soil and respirable dust
were proportionally different. Both the source soil and the dust contained quartz,
mica, plagioclase, and clay minerals. Quartz and mica were present in similar
amounts in both the respirable dust and the source soil. However, both the
plagioclase and clay minerals in the respirable dust were present in smaller
amounts than would be expected given the source soil. The reduced appearance
of clay in the respirable dust may have been due to the large particle size of clay
minerals.
12 J.L. SMITH AND K. LEE

The mineral content of soils varies with the geology of the land and any
additives introduced for agricultural or other purposes (Green et al., 1990). The
major inorganic components of soil dust are clay minerals and quartz (Guthrie,
1993). Clay particles in agricultural soils have a large surface area and can carry
organic materials as well as pesticides. Little is known about the frequency and
nature of the exposure to pesticides and organic materials adsorbed on the
particles (Giese and van Oss, 1993). In arid locations, calcium carbonate and
soluble salts may dominate the inorganic fraction of soils in the place of silicates.
In warm, humid climates, soils may have a greater proportion of oxides and
hydroxides of iron and aluminum than in drier climates (Schenker, 2000).
Silica refers to naturally occurring minerals composed principally of silicon
dioxide (SiO2). Quartz is one of the most common crystalline forms of silica and
can comprise as much as 90– 95% of the sand and silt fraction of soil (Klein,
1993). The highest health risks from soil dust are from silicates and quartz
(Schenker, 1998). Because quantities of silica can be high in soil dust, soil dust
should be seriously considered in silica risk evaluations (Gillette, 1997).
However, the quartz content of soils can be highly variable. In a study of mineral
dusts associated with farming, the quartz content of soil dust was found to vary
from 1 to 17% on a mass basis (Green et al., 1990).
Silicates are classified on the basis of how extensively the silica is
polymerized. The degree of polymerization will determine its resistance to
chemical weathering (Schenker, 2000). Quartz is abundant in soils because it is
resistant to weathering and therefore remains longer than minerals that weather
easily. Respirable quartz is commonly found in soil dust, although weathering
and chemical reactions may make it less fibrogenic than the freshly fractured
quartz found in occupational dust from quarrying and sandblasting (Castranova
et al., 1996).
Plants can take up silicic acid solved in water, and thus insoluble silica can be
precipitated in plant tissue. This is called biogenic silica. All soil-grown plants
contain silica, and their silica content can range from 0.1 to 10% of their dry
matter (Epstein, 1994). In rice straw, silica concentrations can be even higher—
up to 12% by weight. The silica is concentrated in the epidermal and vascular
tissue of rice. Such biogenic silica may occur in respirable dust particles with a
3:1 length to width aspect ratio. The aspect ratio is used to define a fiber in
occupational health. Biogenic silica fiber can be introduced to soils through the
burning or natural degradation of the plant biomass (Lawson et al., 1995).
Throughout all phases of rice farming, it was observed that agricultural workers
were exposed to airborne biogenic silica fiber. The highest exposure was
observed during field preparation, where the average concentration in the
respirable dust fraction was 6 fibers/cc for fibers greater than 5 mm in length.
The biogenic amorphous silica had a complex morphology. The health effects
of the biogenic silica fibers are unknown, but the high exposure levels of
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 13

agricultural workers and the fact that the silica is fibrous and has pointy ends may
justify further study.

B. ORGANIC COMPOUNDS

The organic components of soil can be extremely varied and are derived from
many sources, including plants, animals, and microorganisms. The organic
composition of soils depends on the type of vegetation and the microflora
growing in the soil (Dashek et al., 1986). Hay and grain are common sources of
organic compounds in dust. Dusts from grain include a mixture of organic and
inorganic particles, and the mixture varies by the type of grain, where and under
what conditions it is grown, and the method used to harvest it. Soil dust also
contains fungi and bacteria. Organic dust is often biologically active and can
cause respiratory disease.
The organic components of dust can depend on environmental factors like
acidity, aeration, water availability, and temperature (Lacey, 1994). Bacteria
require a slightly acidic pH that ranges from 5.5 to 8.5, while fungi can survive at
pHs between 2 and 8. Water availability is probably the most important factor in
determining the microflora of soil. As water decreases, fewer organisms can
grow. The temperature range of the soil helps to determine which species of
microorganisms will experience optimal growth. Soil temperature may be
affected both by the ambient temperature and by the heat from metabolic activity.
Endotoxin, an intrinsic part of the outer membrane of Gram negative bacteria,
is another common organic component of soil dust. Endotoxin consists of a
family of lipopolysaccharides. Endotoxin levels in soil can vary, and they are
higher in stables where animals are kept (Dutkiewicz, 1978). Several studies
suggest that endotoxin affects the level of serum IgE and interferes with the
development of atopy (Martinez, 1999).

C. SIZE DISTRIBUTION

The physical size, shape, and density of particles determine how they will be
transported in the human respiratory tract. Deposition sites within the
respiratory tract vary according to a number of factors, including the
aerodynamic diameter of the particle. Figure 1 shows the deposition of
particles of different aerodynamic sizes in various parts of the respiratory tract.
Most of the particles greater than 10 mm in aerodynamic diameter and about
60– 80% of the particles of 5– 10 mm are deposited in the nasopharyngeal
region. Small particles are breathed deeply into the pulmonary region and
therefore can be expected to present a greater risk than a comparable mass
concentration of larger particles.
14 J.L. SMITH AND K. LEE

Figure 1 Regional deposition prediction by the International Commission on Radiological

Protection Task Group on lung dynamics.

The size distribution of soil dust particles strongly resembles the size
distribution of the particles of soil from which they derived (Gillette et al., 1978).
Wind erosion results mostly in the suspension of particles with an aerodynamic
diameter of 1 mm or larger, but the size distribution can range from about 0.6 to
3.0 mm in aerodynamic diameter (Silva et al., 2000). Green et al. (1990) reported
that in mineral dusts associated with farming practices, the majority of particles
on a mass and number basis were within the respirable range of 4 mm or less in
aerodynamic diameter.
The size distribution of soil dust varies significantly over time and space. The
size distribution of particles close to the ground is usually similar to the size
distribution of particles available in the soil. With increasing distance and height,
there is a diminishing proportion of coarse to fine particles. Fine particles can be
transported farther and at higher altitudes than coarse particles. Particles smaller
than 20 mm are dominant in dust carried over great distances (Gillette, 1977).
Only a small amount of dust created by wind erosion is carried to altitudes as
high as 2 – 9 km (Gillette et al., 1978). Over a four-year period, dust samples
taken during dust storms in the southwestern United States showed consistent
particle size distributions and consistent oxygen isotope ratios in quartz. Oxygen
isotope ratios were determined by the ratio of 18O to 16O in the quartz molecules.
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 15

Quartz is the marker mineral, and similar oxygen isotope ratios indicate similar
compositions. Dust samples collected 1 m above the soil have a bimodal particle
size distribution with one mode in the 1– 30 mm range and the other in the
30– 100 mm range. Most particles are smaller than 20 mm based on the graphs.
In dust samples taken at high altitudes, only a single mode was detected, and the
particle sizes ranged between 1 and 20 – 30 mm. Therefore only a small portion of
the dust produced by wind erosion—the finer soil material—is carried to high
altitudes and over great distances.

D. CONTAMINATED SOIL

Agriculture uses various chemicals and soil additives. Excessive use and
inappropriate storage of additives such as pesticides and fertilizers influence the
composition of soil dust. Soil can act as a reservoir and carrier for toxic
chemicals, which can increase occupational exposure to toxic chemicals. And
soil dust can carry toxic substances into the home, where vulnerable groups such
as children and the elderly may be exposed.
Humans can be exposed to toxic chemicals in soil by various pathways,
including consuming food that has been grown in the contaminated soil or water
that has had contact with it, breathing in contaminated soil as suspended dust and
having direct skin contact with the contaminated soil. Agricultural workers are
regularly exposed to soil contaminated with pesticide residues. The pesticide-
containing dust can be dislodged from plant foliage when workers reenter an area
after pesticide application. The workers may then be exposed to toxins by dermal
contact with the dust or through inhalation (Popendorf and Spear, 1974; Spencer
et al., 1975).
House dust is another vehicle for toxic chemical exposure. When soil material
containing toxic chemicals is transported inside or around the home, infants and
toddlers are more susceptible to exposure. The soil materials accumulate most
efficiently in carpets, where infants and toddlers spend significant portions of
their time and engage in frequent mouthing activities.
Particle size is an important factor for dermal contact or inhalation by
suspension. When residential house dust was collected, 13 pesticides and all 10 of
the target polycyclic aromatic hydrocarbons (PAH) were detected in one or more
of seven size-fractionated samples (Lewis et al., 1999). Pyrethroid insecticides,
cis-permethrin and trans-permethrin were the most abundant pesticides.
Chlorpyrifos and carbaryl were also present in relatively high concentrations.
Of the PAHs, chrysene and benzo[b]fluoranthene were present in the highest
concentrations. The concentrations of most pesticides and PAHs generally
increased with decreasing particle size. This suggests that the chemicals may be
primarily attached to the surfaces of the particles. For the same volume, many
small particles give more surface area than fewer large particles. If chemicals
16 J.L. SMITH AND K. LEE

attach to particle surfaces, the chemical concentration will be higher in smaller

particles.

IV. EXPOSURE TO SOIL DUST

While soil dust exposure occurs principally in occupational settings, fugitive

and house dust can increase dust exposure in the general population. Inhalation of
soil dust carries health risks. Exposure level and the duration and frequency of
one’s exposure are all important determinants of how high one’s risk is. Since soil
dust can carry toxic substances, the health risks of inhaling, ingesting, or touching
it can come from agents carried on the dust as well as from the dust itself.

A. OCCUPATIONAL EXPOSURE

Occupational exposure is the most common type of soil dust exposure.

Agricultural workers have the greatest exposure to soil dust but some construction
workers may experience significant exposure as well. Agricultural dust exposure
varies greatly over different regions and time periods and depends on crops, farm
operations, environmental factors, and dust control measures. Agricultural dust
exposure is generally highest during open fieldwork done in arid climates.
However, significant exposure may occur during indoor agricultural activities
such as grain handling and animal management.
There are a few recent studies of occupational exposure to soil dust in
agriculture (Nieuwenhuijsen and Schenker, 1998; Nieuwenhuijsen et al., 1996,
1998, 1999; Lee et al., 2001). The studies are mainly from California, where arid
conditions cause higher levels of soil dust exposure. The studies are based on non-
randomly selected populations, and thus a population-based picture of
occupational dust exposure is missing. The characterization of dusts has generally
been limited to the total and respirable fractions, and speciation has primarily been
of the total inorganic fraction and respirable quartz.
Generally, agricultural dust exposure is extremely high and often exceeds the
American Conference of Governmental Industrial Hygienists’ threshold limit
value (TLVw) of 10 mg/m3 for inhalable nuisance dust. It should be noted that the
limit is based on nuisance dust, which assumes that the dust does not contain any
toxic components. Figure 2 shows inhalable and respirable dust exposure levels
from a few commodities. This figure is constructed from field measurements
taken in California. For all operations, the average inhalable dust levels were
significantly higher than the TLVw levels. When California farmers operated an
open cab tractor for land preparation, the geometric mean inhalable dust exposure
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 17

Figure 2 Average dust exposure levels from five selected commodities during various farming
operations. TLVw: threshold limit values for inhalable dust and respirable dust are 10 and 3 mg/m3,
respectively.

was 57.3 mg/m3 for land planing and 98.6 mg/m3 for disking. Exposure levels of
100 mg/m3 and greater have been reported (Nieuwenhuijsen et al., 1998).
Despite the high levels of inhalable dust, respirable dust levels were often close
to or well below the TLVw levels for respirable nuisance dust. Respirable dust is of
concern because it can reach the alveolar region of the respiratory system. In the
operations shown in Fig. 2, respirable fractions of dusts were generally 5 – 40%,
and total respirable dust concentrations commonly fell between 0.4 and 3 mg/m3.
During land planing and disking, respirable dusts measured using a cyclone
had geometric means of 0.46 and 0.58 mg/m3 (Nieuwenhuijsen et al., 1996).
A study of rice farming in California found average respirable dust concentrations
of 1.18 mg/m3 during harvest and 3.17 mg/m3 during field preparation
(Lawson et al., 1995).
When workers perform manual harvesting, they can be exposed to high levels
of a complex mixture of inorganic and organic dust (Lee et al., 2001). The
geometric means for exposure levels during manual orange harvesting were
39.7 mg/m3 for inhalable dust and 1.14 mg/m3 for respirable dust. The geometric
means for exposure levels during table grape harvesting were 3.5 mg/m3 for
inhalable dust and 0.23 mg/m3 for respirable dust. These exposure levels were
similar to those measured in peach orchards. Exposure levels of inhalable and
respirable dusts during hand harvest were 13 and 0.5 mg/m3, respectively
(Popendorf et al., 1982). Exposure levels during orange harvesting were
significantly higher than those during grape harvesting, exceeding the TLVw for
inhalable dust and respirable quartz. Considered independently, exposure
18 J.L. SMITH AND K. LEE

levels of inhalable dust and respirable quartz in orange harvesting were high
enough to cause respiratory health effects. The degree of vigorous contact with
foliage appeared to have a significant impact on exposure levels in manual
harvesting.
Levels of exposure to respirable quartz in agriculture commonly exceed
industrial standards (Schenker, 2000). During ground preparation, average
exposure levels to respirable quartz were 2 mg/m3 in an open cab tractor. The
level of exposure was 0.05 mg/m3 or less in a closed cab tractor (Gustafson et al.,
1978). But the quartz content of soil dust can vary by region: In Alberta, Canada,
crystalline silica was 0.8– 17.5% of dust samples (Green et al., 1990). Farmers
are often exposed to high respirable quartz levels due to high concentrations
produced during many agricultural operations. During manual harvesting, quartz
averaged 7 –20% of the total respirable dust. During a grape harvest, over 50% of
the samples exceeded the TLVw for respirable quartz (Popendorf et al., 1982).
During manual orange harvest, over 90% of the measurements exceeded the
TLVw for respirable quartz (Lee et al., 2001). Only about 15% of the samples
exceeded the TLVw for respirable quartz in the table grape harvest.
The type of agricultural operation being performed is one of the major
determinants of levels of dust exposure. In a study of personal dust exposure in
California agriculture (Nieuwenhuijsen and Schenker, 1998), the type of
operation alone accounted for 28% of the total variation in inhalable dust
exposure and 19% of the variation in respirable dust exposure. Ground
preparation activities, including land planing, disking, and rolling beds, resulted
in the highest levels of dust exposure, followed by machine tomato harvesting,
the feeding of cows, planting, and the mechanical harvest of fruit and nuts.
Milking resulted in the lowest level of exposure, with a geometric mean of
0.7 mg/m3 of inhalable dust. The lowest levels in milking were less than 1%
of the highest geometric mean of 98.6 mg/m3, which was observed during
disking.
When a tractor is used, differences in equipment, tractor speed, and climate
may affect the dust exposure levels. The presence of an enclosed cabin on the
tractor, higher relative humidity, and lower tractor speed were associated with a
decrease in personal levels of dust exposure (Nieuwenhuijsen and Schenker,
1998; Nieuwenhuijsen et al., 1998). In particular, an enclosed tractor cab reduced
the exposure of the operator to larger dust particles (50% cutoff , 9.8 mm)
60-fold and to smaller dust particles (50% cutoff , 3.5 –4 mm) four-fold.
Relative humidity can reduce inhalable dust five-fold, which supports the theory
that it is best to work in the early morning, when temperatures are cooler and
humidity is higher, to reduce dust exposure. Although reducing tractor speed
could lead up to a 10-fold reduction in inhalable dust exposure, the variation in
tractor speed is too small to practically achieve this high of a reduction. In a study
of silica fiber exposure during rice farming (Lawson et al., 1995), levels of silica
exposure were lower for those using tractors with cabs, those working on foot and
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 19

those working without vehicle traffic in the field. Personal silica exposure inside
the enclosed cabin ranged from 2 to 19% of outside levels. The variation in the
indoor/outdoor ratio suggests a potential infiltration of dust as well as occasional
opening of the cabin window, thereby increasing indoor dust concentrations.
While an enclosed tractor cab is the single most effective method to reduce
inorganic dust exposure, there is little data on the prevalence and use of enclosed
cabs. One population-based study in California found that 60% of farmers
reported that their primary tractor did not have an enclosed cab (Nieuwenhuijsen
et al., 1996). Also important is the fact that agricultural farm workers (e.g.,
migrant and seasonal farm workers) do not drive a tractor. They do not have
the protection of enclosed work environments, but there is little data on the
cumulative dust exposure in this population.

B. FUGITIVE DUST EXPOSURE

Non-occupational soil dust exposure occurs mostly from contact with wind-
blown dust, called fugitive dust. Exposure levels to fugitive dust vary widely by
location and time. Dust storms increase the amount of ambient particulate matter
and cause the general population to be exposed to particulates. In a study of the
effect of dust storms (Hefflin et al., 1994), it was found that naturally occurring
PM10 has an effect on the respiratory health of the general population. Hefflin
compared emergency room visits for respiratory disorders to daily PM10 levels
for the Tri-Cities area of Washington State in 1991. They found that emergency
room visits for bronchitis and sinusitis were correlated with the occurrence of
dust storms. As the area is free of industrial sources, the PM10 levels were
attributable to natural sources.
A dust storm is an event in which a significant amount of soil dust is suspended
for a relatively short period of time. Dust concentrations during a storm can vary,
however, there are few measurements of dust concentrations in actual storms.
One event in the Great Plains region of the United States had a 24-h average total
dust concentration of 843 mg/m3 (Hagen and Woodruff, 1973). Dust concen-
trations were measured during a dust storm in a semi-arid area of Washington
State (Claiborn et al., 2000). During the dust storm, both PM10 and PM2.5
concentrations were significantly increased. During a dust event, maximum 1-h
concentrations of PM10 and PM2.5 were 1879 and 127 mg/m3, respectively.
However, ratios of PM2.5 to PM10 were 0.30 during dust event days and 0.48 on
the non-dust event days. This may suggest an abundance of large particles during
a dust storm. For PM2.5, a significant portion of particulates ranged from 1.0 to
2.5 mm during dust storms.
Dust from paved roads is a complex mixture of soil dust, vehicle exhaust, tire
dust, brake dust, biological material, and other materials. Exposure occurs when
moving vehicles suspend the road dust in the atmosphere. Paved road dust was
20 J.L. SMITH AND K. LEE

found to be a source of allergen exposure for the general population (Miguel et al.,
1999). Estimates of total particulate matter emissions from paved roads are
7.0 £ 109 kg/year nationally and 0.602 £ 109 kg/year within California (US
Environmental Protection Agency (EPA), 1996). As a road ages, the surface
conditions deteriorate, and, as might be anticipated, the particulate emissions
increase. Freeways, highways, collector streets, and local streets differ in
emissions rates.

C. DIRECT SOIL CONTACT

Soil dust may come into direct or indirect contact with hands and objects, and
it may be ingested when someone puts their hand to their mouth. The body can
then absorb chemicals contained in the soil dust. Children are the principle at-risk
population for this sort of exposure as they are frequently involved in hand-to-
mouth activities. Soil ingestion was investigated using aluminum, silicon and
titanium as soil tracer elements. In the United States, children between the ages of
2 and 7 years ingest from 39 to 246 mg of soil per day, on the average, depending
on the tracer element analyzed (Davis et al., 1990). Children with pica behavior
may ingest much more. Spending more time outdoors, displaying mouthing
behavior and washing the face before eating are associated with high soil
ingestion levels. Washing the face before eating is thought to be associated with
high soil ingestion because children who have more contact with soil may have
an increased need to wash.
House dust is considered a significant source of exposure to pesticide residues
and other toxicants for small children and toddlers (Lewis et al., 1994). Dust can
drift into houses from yard soil or be transferred inside on clothes and shoes or by
pets, which means that pesticides and other toxicants in the soil can be
transferred into the house. The contaminants in house dust are not subject to
degradative environmental processes and may persist longer in the house than in
outdoor soil.
Pesticides can be among the contaminants in house dust and are of particular
concern in households where there are children. House dust pesticide residues
were measured in the homes of agricultural families (Simcox et al., 1995).
Significantly higher levels of organophosphorous pesticides were measured in
dust from farmer and farm worker homes than in reference homes. Organopho-
sphorous levels in household dust samples taken from farmer and farm worker
homes ranged from non-detectable to 17 ppm, while reference home levels
ranged from non-detectable to 0.82 ppm. Pesticide concentrations were
significantly higher in household dusts than in soils for all groups studied;
median household dust levels were 17 – 100 times higher than soil levels.
Proximity to spray areas seemed to be the predominant factor for elevated
pesticide concentration in the household dust. So families living near agricultural
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 21

areas may be at a greater risk of exposure to toxins from household dust, even
when they are not involved in agricultural processes.
Young children can easily come into direct contact with yard soil or house dust
as they spend a large portion of their time on the floor or ground. Children also
spend more time outdoors than adults. Children who frequently put their hands
and other things in their mouths can have greater levels of exposure because they
are ingesting soil or dust. And children may be more susceptible to environmental
contaminants because they absorb, excrete and metabolize them differently than
adults.
Direct contact with soil is not limited to children. Information on the
activities of 1200 Americans aged 18 or older was collected using a computer
assisted telephone interview. In response to a question about their previous day’s
activities, 20% of the American adult public reported having direct contact with
soil on a typical day (Robinson and Silvers, 2000). When they came in contact
with soil, the duration of their exposure was about 1.7 h per day. Men reported
more soil contact than women. However, no difference in soil contact was found
for age, employment status, race, education, or socioeconomic status. Higher
exposure was reported in the spring months, on weekends and in rural areas. For
adults, soil contact was associated with time spent outdoors.

V. HEALTH EFFECTS OF SOIL DUST

Since inhalation is a major route of exposure to soil dust, respiratory health

effects from soil dust are of great concern. Because soil dust is made up of a
complex mixture of inorganic and organic components, the health effects cannot
be attributed solely to either component. Organic dust is generally more
biologically active. Inorganic dust tends not to result in allergic reactions in the
lungs but rather acts as a mild, non-specific irritant, placing an added load on the
lungs’ clearance mechanisms (Donham, 1986). However, inorganic dust may
aggravate health effects from organic dusts (Schenker, 2000).

A. HEALTH EFFECTS OF AGRICULTURAL DUST

A large number of adverse respiratory health outcomes have been attributed to

agricultural dust exposure (Schenker, 1998). Dust exposure may affect the
airways and, depending on the antigenicity of the material and host susceptibility,
may result in asthma, an asthma-like syndrome, or chronic obstructive pulmonary
disease. Parenchymal disorders such as hypersensitivity pneumonitis may result
from exposure to fungal spores (Marx et al., 1990; do Pico, 1992). Exposure to
agricultural dusts composed primarily of inorganic components results in chronic
22 J.L. SMITH AND K. LEE

respiratory symptoms (e.g., chronic cough) and has been associated with an
impairment in pulmonary function consistent with restrictive lung function
(Gamsky et al., 1992a,b). It is important to note that dust in the agricultural
workplace is commonly a mixture of organic and inorganic dusts, and it is not
always possible (or valid) to attribute observed health effects to one component or
the other, particularly based on epidemiologic studies.
Agricultural dusts are the cause of the pulmonary pathology entity known as
“mixed-dust pneumoconiosis” (Churg and Green, 1988; Schenker, 1998).
Several case reports and case studies have observed this pathologic outcome in
farmers and farm workers (Sherwin et al., 1979; Fennerty et al., 1983; Glyseth
et al., 1984; Schenker, 1998). The pathologic lesion from mixed-dust
pneumoconiosis is distinct from silicosis, and the dusts contributing to its
occurrence are not well characterized. Chronic inhalation of common soil
minerals such as mica, kaolinite, and silica can cause pulmonary fibrosis. Silica
minerals, especially fibrous silica and crystalline silica and quartz, increase the
fibrogenicity of interstitial cells and can cause silicosis, silicotuberculosis, and
cancer. However, studies of the pathology of mixed-dust pneumoconiosis cases
in farmers have shown dust composition in the lung tissue that is similar to the
composition of agricultural soils with which the farmer worked (Sherwin et al.,
1979; Glyseth et al., 1984).
The organic components found in agricultural dusts are derived from organic
sources and are carbon based. The organic dusts often contain biologically active
proteins that may be allergenic and proinflammatory (Kirkhorn and Garry, 2000).
Diseases caused by organic dusts include acute inflammation of the bronchial
epithelium and chronic bronchitis, hypersensitive pneumonitis, asthma, and toxin
fever. These diseases do not include infections caused by microorganisms
multiplying in the lung.
It is difficult to identify specific individual agents from agricultural dust as the
cause of health effects. Such identification may be possible with toxicological
study, however, limited toxicological studies have been conducted on agricul-
tural dusts. One study measured the fibrogenic potential of dust by
intratracheally instilling field dust into the lungs of rats (Rajini et al., 1995).
The dusts were collected from vineyards and citrus orchards during harvest
operations. After the rats were exposed to the mixed dust, their lung responses
were evaluated by analyzing lung lavage fluid, lung collagen content,
histopathology, and cell kinetic studies. Significant cumulative labeling indices
were found 1 week after a single instillation of vineyard dust. The cell labeling
indices from citrus dust were not significant. The data suggested that dust
collected in vineyards has fibrogenic potential, while dusts from citrus orchards
are less biologically active. The study did not identify critical components of the
response. Since the study applied identical amounts of quartz in both dust
samples, the difference in fibrogenic potential may not be directly attributable to
the quartz content. The specific fraction of quartz or some other constituents of
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 23

the dust collected in vineyards may cause the fibrogenic response. This finding is
interesting because an epidemiological study demonstrated a greater restrictive
pulmonary response among grape workers than among citrus workers (Gamsky
et al., 1992a).
Several epidemiologic studies provide clear evidence of the increased
incidence and prevalence of respiratory symptoms and disease in agricultural
populations (Schenker, 1998). The increase in respiratory disease morbidity and
mortality in agricultural workers is of particular note because cigarette smoking
prevalence is lower in farming populations than in non-farming populations
(Sterling and Weinkam, 1978; Stellman et al., 1988; Gillum, 1994). This
divergence in respiratory disease mortality and cigarette smoking is well
demonstrated in the California Occupational Disease Mortality (COMS) study
(Singleton and Beaumont, 1989). In that statewide occupational mortality study,
deaths from chronic respiratory disease were significantly increased among
agricultural workers, but lung cancer deaths were significantly below the
statewide average.
An epidemiological cohort study has been conducted among California
farmers since 1993 (Schenker et al., 1994). The initial interview in 1993 asked
1947 California farmers about their individual occupational exposures and health
outcomes. The study determined the relationship between various farm-related
exposures and persistent wheeze, chronic bronchitis (phlegm for $ 3 months),
physician-diagnosed asthma and chronic cough. Persistent wheeze was reported
in 8.5% of the study population, chronic bronchitis in 3.7%, asthma in 7.8%, and
chronic cough in 4.1%. Among cigarette smokers, prevalence rates were
significantly higher for all respiratory symptoms. However, after adjusting for
smoking and other risk factors, it was evident that there was an independent,
dose-response relationship between self-reported dust exposure and persistent
wheeze and chronic cough (Schenker et al., 1994). The prevalence of persistent
wheeze increased from 4.8% (for those reporting 0% time in dusty job) to 9.1%
($ 50% time in dusty job) among non-smokers and from 17% (0% time in a dusty
job) to 27.6% ($ 50% time in a dusty job) among smokers. Similar increases
were observed for chronic bronchitis in both smokers (2.7 vs. 4.0%) and non-
smokers (9.4 vs. 15.5%).
In 1998, a follow-up study was conducted of the original cohort. Prevalence of
both chronic cough and chronic bronchitis increased over the five-year period
(4.2 – 8.5% for cough and 4.5 –9.4% for bronchitis). There were 80 incident cases
of chronic cough in this five-year period (5.9% of the cohort). Variables were
correlated directly with incidence of cough included being young, being female,
spending time working at a dusty job, smoking, and growing grapes. Chronic
bronchitis developed in 7.0% of the cohort ðn ¼ 94Þ: The incidence of bronchitis
was associated with hours per week spent farming, smoking, time spent at a dusty
job, and being male.
24 J.L. SMITH AND K. LEE

B. HEALTH EFFECTS OF FUGITIVE DUST

Fugitive dust may be similar to soil dust. The association between soil dust
and respiratory diseases was determined during a seasonal dust storm in
Washington State (Hefflin et al., 1994). Dust storms consist predominantly of soil
dust. During the 2-day dust storm, ambient PM10 concentrations exceeded
1000 mg/m3. This level is more than six times greater than the Environmental
Protection Agency’s 24-h PM10 National Ambient Air Quality Standard. The
number of daily emergency room visits for bronchitis was increased by 3.5% for
each 100 mg/m3 increase in PM10. The effects were observed only for emergency
room visits of people with pre-existing respiratory disease. This study revealed a
relatively small effect on the respiratory health of the population in general.
The effects of dust storms on mortality were also investigated (Schwartz et al.,
1999). The 24-h mean PM10 concentration during 17 dust storms between 1989
and 1995 was 263 mg/m3. The control dates selected were the same day of the
year in years without a dust storm. The 24-h mean PM10 level was 42 mg/m3 for
control days. The risk of mortality on dust storm days and one day after a dust
storm was not increased. The finding is contradictory to other studies of ambient
air pollution in which Schwartz estimates that a 20% increase in daily mortality
would be observed with an incremental PM10 exposure of 221 mg/m3 (Schwartz
et al., 1999). This study concluded that coarse particles (particles between 2.5 and
10 mm in aerodynamic diameter) from windblown soil dust are not associated
with mortality risk.
Although these two studies found little to indicate that exposure to windblown
dust has an impact on respiratory health, they are not clear evidence that soil dust
has little impact on respiratory health. It is possible that there is a delayed rise in
mortality and morbidity (Mage, 2000), but the studies restricted the analysis to
the day of the dust storm or the following day. In addition, people are likely stay
indoors during dust storms, and residential ventilation is minimal. Therefore,
personal exposure to dust is expected to be significantly lower than the level
measured outdoors.

C. PROTECTIVE EFFECTS OF FARM ENVIRONMENT

A recent series of epidemiological studies suggest that growing up on a farm

may protect a person against allergic sensitization and the development of
childhood allergic diseases. Early population studies showed lower prevalence
rates of allergic disease in rural populations vs. those in urban populations. The
difference was not explained by chronic air pollution in the urban areas. When a
rural community was studied, farmers’ children had significantly lower rates of
sneezing attacks and atopic sensitization during the pollen season than their peers
in the same community (Braun-Fahrlander et al., 1999). Several other studies
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 25

found that farming had a protective effect against the development of asthma, hay
fever, and atopic sensitization in children (Ernst and Cormier, 2000; Kilpelainen
et al., 2000; Reidler et al., 2000; Von Ehrenstein et al., 2000; Riedler and
Braun-Fahrlander, 2001). Adolescents and adults also experienced the protective
effect of growing up in a farm environment (Leynaert et al., 2001; Portengen
et al., 2002).
These findings are remarkable since farmers and their family members are
likely exposed to high levels of allergens. It is well known that some farm
environments provide high levels of exposure to airborne organic dusts, which
often contain a wide variety of antigenic materials and potential allergens.
Common allergens, such as allergens from pets and house dust mites, are usually
abundant in the farm home environment (Iversen et al., 1990; Parvaneh et al.,
1999). Outdoor allergens, such as pollens and mold spores, are also more
common in farming communities.
Several studies of potentially protective effects of farm environments
emphasize the importance of animal contact (Reidler et al., 2000; Von
Ehrenstein et al., 2000). Contact with livestock and poultry was found to explain
much of the association between farming and decreased atopy (Reidler et al.,
2000). Exposure to livestock has an inverse dose-dependent relationship with the
prevalence of atopic diseases (Von Ehrenstein et al., 2000). The protective effect
of contact with domestic pets was also determined (Ownby et al., 2002).
Exposure to two or more dogs or cats in the first year of life may reduce the risk of
sensitization to allergens during childhood. This points out the possible role of
endotoxin in developing allergies. Endotoxin is an intrinsic part of the outer
membrane of Gram negative bacteria, and it is abundant in stables and in farmers’
homes (Von Mutius et al., 2000).
Just as they are more likely to be exposed to biological allergens, children on a
farm are more likely to be exposed to soil dust than are children in urban areas.
One study found that exposure to soil in early life was not associated with lower
rates of asthma, eczema, and hay fever in young adults (Montgomery et al.,
1999). In the study, the possession of a garden was used to indicate soil exposure.
No positive relationship was observed between gardening and the development
of asthma, hay fever, and atopic sensitization in adults. However, the study may
have mistakenly equated having a garden with having high soil exposure. Further
studies are needed to determine the health effect of inorganic dust exposure in
early life.

D. REDUCING EXPOSURE TO SOIL DUST

There are various methods of controlling environmental and occupational

hazards. Source control is the most effective and often the most efficient. In
theory, source control means eliminating or finding a substitute for the pollution
26 J.L. SMITH AND K. LEE

source. The use of unleaded gasoline is a good example of source control. Lead in
the environment has decreased significantly since leaded gasoline was banned in
the United States. However, it is practically impossible to decrease soil dust by
eliminating or finding a substitute for soil.
Engineering control is the next best control measure. The preferred
engineering method is to reduce emissions of soil dust. Methods to reduce dust
emission include wetting the soil, planting ground cover, tilling regularly, or
using windbreaks or barriers. These methods will develop a surface barrier,
increase clodiness and roughen up the soil surface. Many methods have been
designed to reduce soil erosion and conserve soil moisture in agriculture: field
fallowing leaves vegetative residue to protect the soil surface, shelterbelts
provide a wind break, cover crops hold onto soil, deep plowing brings moist,
clay-rich soil up to the surface, irrigation increases soil moisture, and tilling
practices have been developed that retain vegetative residue from the previous
crop (Gillette, 1988). Wetting the soil may increase potential exposure to
biological agents if organic material is present.
In animal houses, engineering control measures include misting, adding
vegetable oil to feed, and sprinkling oil on the animals. Fats added to feeds in the
form of canola oil, rapeseed, and mineral oil decreased dust levels by 35– 60% in
a swine building (Takai et al., 1996). The use of canola oil sprinkled in a swine
room decreased dust, endotoxin levels, H2S, and ammonia (Zhang, 1997).
Although spraying and feeding oil are known to reduce organic dusts and gases,
they are not applicable for reducing soil dust.
Once soil dust is generated, separating humans from the dust can reduce their
exposure. The use of enclosed cabins on tractors is an example of separation that
can reduce the dust exposure of the tractor operator (Nieuwenhuijsen et al.,
1998). Appropriate cabin maintenance is important for such protection. While an
enclosed tractor cab is the single most effective intervention to reduce
occupational exposure to inorganic dust during soil preparation, a majority of
California farmers reported that their primary tractor did not have an enclosed
cab (Nieuwenhuijsen et al., 1996). Additionally, an enclosed cab will not reduce
the emission of soil dust and the eventual ambient concentration.
Ventilation with a proper mixing of fresh outdoor air can be a useful control
measure for indoor environments. Ventilation can improve air quality by either
mechanical or natural methods, and it can control indoor environments such as
animal houses. However, ventilation is not feasible for outdoor conditions, where
the majority of soil dust exposure occurs. Administrative controls such as
education and reduced exposure time can reduce exposure, although they cannot
reduce dust concentration. Education can also change certain behaviors, thereby
minimizing dust generation and dust exposure.
Personal protective equipment is the final measure used to control exposure,
but it should not be used as a long-term solution. It should be noted that use of a
respirator should be accompanied by proper training and an annual respirator fit
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 27

Figure 3 Use of personal protective equipment in California farmers.

test. Because environmental conditions during agricultural operations are usually

hot, wearing a respirator can be uncomfortable. The prevalence of respirator use
is often very low. In the California farmer health study, 1947 farmers were asked
questions about their use of protective equipment (Schenker et al., 1994). It was
found that subjects worked at a self-reported dusty job 10% of their farm time
(median). Less than 30% reported using dust filters more than half the time. A
dust filter was defined as a scarf over the nose and mouth or a disposable dust
mask or cartridge respirator. Most subjects reported using protection “rarely or
never” (Fig. 3). This was a study of farmers only. It is likely that the frequency of
personal protective equipment use is even lower among agricultural workers.

REFERENCES

Battan, L. J. (1966). In “The Unclean Sky: A Meteorologist Looks at Air Pollution.” Anchor Books,
Garden City, NY.
Brady, N. C. (1974). In “The Nature and Properites of Soils” (9th ed.). Macmillan, New York.
Braun-Fahrlander, C., Gassner, M., Grize, L., Neu, U., Sennhauser, F. H., Varonier, H. S., Vuille, J. C.,
Wuthrich, B., and SCARPOL team (1999). Prevalence of hay fever and allergic sensitizaton in
farmer’s children and their peers living in the same rural community. Clinical and Experimental
Allergy 29, 28–34.
Castranova, V., Dalal, N. S., and Vallyathan, V. (1996). Role of surface free radicals in the
pathogenicity of silica. In “Silica and Silica-induced Lung Disease.” (W. E. Wallace, Ed.),
pp. 91–105. CRC Press, Boca Raton.
Cephil, W. S. (1945). Dynamics of wind erosion: I. Nature of movement of soil by wind. Soil Science
60, 305–360.
Cephil, W. S., and Woodruff, N. P. (1963). The physics of wind erosion and its control. Advances in
Agronomy 15, 211– 302.
Chepil, W. S., and Milne, R. A. (1941). Erosion of soil in relation to roughness of surface. Soil Science
52, 417–433.
28 J.L. SMITH AND K. LEE

Churg, A., and Green, F. H. Y. (1988). In “Pathology of Occupational Lung Disease”. Igaku-Shoin,
New York.
Claiborn, C. C., Finn, D., Larson, T. V., and Koenig, J. Q. (2000). Windblown dust contributes to high
PM2.5 concentrations. Journal of the Air and Waste Management Association 50, 1440–1445.
Clausnitzer, H., and Singer, M. J. (1996). Respirable dust production from agricultural operations in
the Sacramento Valley, California. Journal of Environmental Quality V25, 877–884.
Clausnitzer, H., and Singer, M. J. (1997). Intensive land preparation emits respirable dust. California
Agriculture 51, 27–30.
Clausnitzer, H., and Singer, M. J. (1999). Mineralogy of agricultural source soil and respirable dust in
California. Journal of Environmental Quality V28, 1619–1629.
Clausnitzer, H., and Singer, M. J. (2000). Environmental influences on respirable dust production from
agricultural operations in California. Atmospheric Environment V34, 1739–1745.
Cowherd, C. (1993). Fugitive dust emissions. In “Aerosol Measurement: Principles, Techniques, and
Applications.” (P. A. B. Klaus Willeke, Ed.), pp. 640–658. Van Nostrand Reinhold, New York.
Cowherd, C., Jr., Muleski, G. E., Englehart, P. J., and Gillette, D. A. (1985). Rep. No. EPA/600/8-85/
002. In “Rapid Assessment of Exposure to Particulate Emissions from Surface Contamination
Sites.” US Environmental Protection Agency, Washington, DC.
Dashek, W. V., Olenchock, S. A., Mayfield, J. E., Wirtz, G. H., Wolz, D. E., and Young, C. A. (1986).
Carbohydrate and protein contents of grain dusts in relation of dust morphology. Environmental
Health Perspectives 66, 135 –143.
Davis, S., Waller, P., Buschbom, R., Ballou, J., and White, P. (1990). Quantitative estimates of soil
ingestion in normal children between the ages of 2 and 7 years: population-based estimates using
aluminum, silicon, and titanium as soil tracer elements. Archives of Environmental Health 45,
112–122.
do Pico, G. A. (1992). Hazardous exposure and lung disease among farm workers. Clinics in Chest
Medicine 13, 311–328.
Donham, K. J. (1986). Hazardous agents in agricultural dust and methods of evaluation. American
Journal of Industrial Medicine 10, 205–220.
Dutkiewicz, J. (1978). Exposure to dust-borne bacteria in agriculture. I. Environmental studies.
Archives of Environmental Health 33, 250 –259.
Epstein, E. (1994). The anomaly of silicon in plant biology. Proceedings of the National Academy of
Sciences of the United States of America 91, 11–17.
Ernst, P., and Cormier, Y. (2000). Relative scarcity of asthma and atopy among adolescence raised on
a farm. American Journal of Respiratory and Critical Care Medicine 161, 1563–1566.
Evans, J. S., and Cooper, C. W. (1980). An inventory of particulate emissions from open sources.
Journal of the Air Pollution Control Association 30, 1298– 1303.
Fennerty, A., Hunter, A. M., Smith, A. P., and Pooley, F. D. (1983). Silicosis in a Pakistani farmer.
British Medical Journal 287, 698 –699.
Gamsky, T. E., McCurdy, S. A., Samuels, S. J., and Schenker, M. B. (1992a). Reduced FVC among
California grape workers. American Review of Respiratory Disease 145, 257 –262.
Gamsky, T. E., Schenker, M. B., McCurdy, S. A., and Samuels, S. J. (1992b). Smoking, respiratory
symptoms, and pulmonary function among a population of hispanic farmworkers. Chest 101,
1361–1368.
Giese, R. F., Jr., and van Oss, C. J. (1993). The Surface Thermodynamic Properties of Silicates
and their Interactions with Biological Materials. In “Health Effects of Mineral Dusts.” (B. T.
Mossman, Ed.), Vol. 28, pp. 327 –346. Mineralogical Society of America, Washington, DC.
Gillette, D. (1977). Fine particulate emissions due to wind erosion. Transactions ASAE 20, 890– 897.
Gillette, D. A. (1988). Threshold friction velocities for dust production for agricultural soils. Journal
of Geophysical Research 93, 12645–12662.
Gillette, D. A. (1997). Soil derived dust as a source of silica. Journal of Exposure Analysis and
Environmental Epidemiology 7, 303–311.
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 29

Gillette, D. A., and Blifford, I. H., Jr. (1974). The influence of wind velocity on size distributions of
aerosols generated by wind erosion of soils. Journal of Geophysical Research 79, 4068–4075.
Gillette, D. A., and Hanson, K. J. (1989). Spatial and temporal variability of dust production caused by
wind erosion in the United States. Journal of Geophysical Research 94, 2197–2206.
Gillette, D. A., Clayton, R. N., Mayeda, T. K., Jackson, M. L., and Sridhar, K. (1978). Tropospheric
aerosols from some major dust storms of the southwestern United States. Journal of Applied
Meteorology 17, 832 –845.
Gillette, D. A., Adams, J., Endo, A., Smith, D., and Kihl, R. (1980). Threshold velocities for input of
soil particles into the air by desert soils. Journal of Geophysical Research 85, 5621–5630.
Gillette, D. A., Adams, J., Muhs, D., and Kihl, R. (1982). Threshold friction velocities and rupture
moduli for crusted desert soil for the input of soil particles into the air. Journal of Geophysical
Research 87, 9003–9015.
Gillette, D. A., Patterson, E. M., Prospero, J. M., and Jackson, M. L. (1993). Soil aerosols. In “Aerosol
Effects on Climate.” (S. G. Jennings, Ed.), pp. 73–109. University of Arizona Press, Tucson.
Gillum, R. F. (1994). Prevalence of cardiovascular and pulmonary diseases and risk factors by
region and urbanization in the United States. Journal of the National Medical Association 86,
105 –112.
Glyseth, B., Stettler, L., Mowè, G., Skaug, V., and Lexow, P. (1984). A striking deposition of mineral
particles in the lungs of a farmer: a case report. American Journal of Industrial Medicine 6,
231 –240.
Green, F. H. Y., Yoshida, K., Fick, G., Paul, J., Hugh, A., and Green, W. F. (1990). Characterization of
airborne mineral dusts associated with farming activities in rural Alberta, Canada. International
Archives of Occupational and Environmental Health 62, 423–430.
Gustafson, A., Eriksson, H. A., and Noren, O. (1978). Dust Concentrations During Operations with
Farm, Forest and Entrepreneur Machines (in Swedish). Special Report S 26. Swedish Institute of
Agricultural Engineering.
Guthrie, G. D., Jr. (1993). Mineral Characterization in Biological Studies. In “Health Effects of
Mineral Dusts.” (B. T. Mossman, Ed.), Vol. 28, pp. 251 –273. Mineralogical Society of America,
Washington, DC.
Hagen, L., and Woodruff, N. P. (1973). In “Particulate Loads Caused by Wind Erosion in the Great
Plains.” Air Pollution Control Association.
Hall, F. F. (1981). Visibility reductions from soil dust in the western United States. Atmospheric
Environment 15, 1929–1933.
Hefflin, B. J., Jalaludin, B., McClure, E., Cobb, N., Johnson, C. A., Jecha, L., and Etzel, R. A. (1994).
Surveillance for dust storms and respiratory diseases in Washington State, 1991. Archives of
Environmental Health 49, 170–174.
Holmen, B. A., James, T. A., Ashbaugh, L. L., and Flocchini, R. G. (2001). Lidar-assisted
measurement of PM10 emissions from agricultural tilling in California’s San Joaquin Valley—
part II: emission factors. Atmospheric Environment V35, 3265–3277.
Iversen, M., Korsgaard, J., Hallas, et al. (1990). Mite allergy and exposure to storage mites and house
dust mites in farmers. Clinical and Experimental Allergy 20, 211–219.
Johnson, D. L. (1998). A universal definition of soil. Quaternary International 51, 6– 7.
Killham, K. (1994). In “Soil Ecology.” Cambridge University Press, New York.
Kilpelainen, M., Terho, E. O., and Helenius, H. (2000). Farm environment in childhood prevents the
development of allergies. Clinical and Experimental Allergy 30, 201 –208.
Kirkhorn, S. R., and Garry, V. F. (2000). Agricultural lung diseases. Environmental Health
Perspectives 108, 705 –712.
Klein, C. (1993). Rocks, Minerals, and a Dusty World. In “Health Effects of Mineral Dusts.” (B. T.
Mossman, Ed.), Vol. 28, pp. 7–59. Mineralogical Society of America, Washington, DC.
Lacey, J. (1994). Microorganisms in organic dusts. In “Organic Dusts: Exposure, Effects, and
Prevention.” (R. J. Jacobs, Ed.), pp. 17–41. Lewis Publishers, Boca Raton.
30 J.L. SMITH AND K. LEE

Lawson, R. J., Schenker, M. B., McCurdy, S. A., Jenkins, B., Lischak, L. A., John, W., and Scales, D.
(1995). Exposure to amorphous silica fibers and other particulate matter during rice farming
operations. Applied Occupational and Environmental Hygiene 10, 677 –684.
Lee, K., Schenker, M. B., Lawson, R., Olenchock, S., Vallyathan, V., Southard, R., Thorne, P. S., and
Saiki, C. (2001). Dust exposures in manual harvest of California fruits. In “American Industrial
Hygiene Conference”, New Orleans, Louisiana, USA.
Lewis, R. G., Fortmann, R. C., and Camann, D. E. (1994). Evaluation of methods for monitoring the
potential exposure of small children to pesticides in residential environment. Archives of
Environmental Contamination and Toxicology 26, 37–46.
Lewis, R. G., Fotune, C. R., Willis, R. D., Camann, D. E., and Antley, J. T. (1999). Distribution of
pesticides and polycyclic aromatic hydrocarbons in house dust as a function of particle size.
Environmental Health Perspectives 107, 721–726.
Leynaert, B., Neukirch, C., Jarvis, D., Chinn, S., and Burney, P. (2001). Does living on a farm during
childhood protect against asthma, allergic rhinitis, and atopy in adulthood? American Journal of
Respiratory and Critical Care Medicine 164, 1829– 1834.
Lyon, T. L., Flippin, E. O., and Buckman, H. O. (1915). In “Soils, their Properties and Management.”
Macmillan, New York.
Mage, D. T. (2000). Comment on coarse particles and dust storm mortality. Environmental Health
Perspectives 108, A12.
Martinez, F. D. (1999). Maturation of immune responses at the beginning of asthma. Journal of
Allergy and Clinical Immunology 103, 355– 361.
Marx, J. J., Guernsey, J., Emanuel, D. A., Merchant, J. A., Morgan, D. P., and Kryda, M. (1990).
Cohort studies of immunologic lung disease among Wisconsin dairy farmers. American Journal
of Industrial Medicine 18, 363–368.
McKendry, I. G., Hacker, J. P., Stull, R., Sakiyama, S., Mignacca, D., and Reid, K. (2001). Long-range
transport of Asian dust to the Lower Fraser Valley, British Columbia, Canada. Journal of
Geophysical Research—Atmospheres 106, 18361–18370.
Miguel, A. G., Cass, G. R., Glovsky, M. M., and Weiss, J. (1999). Allergens in paved road dust and
airborne particles. Environmental Science and Technology 33, 4159–4168.
Montgomery, S. M., Twamley, S. L., Murch, S. H., Pounder, R. E., and Wakefield, A. J. (1999). Soil
exposure no protection against atopy. Immunology Today 20, 289–290.
Nieuwenhuijsen, M. J., and Schenker, M. B. (1998). Determinants of personal dust exposure during
field crop operations in California agriculture. American Industrial Hygiene Association Journal
59, 9– 13.
Nieuwenhuijsen, M. J., Schenker, M. B., Samuels, S. J., Farrar, J. A., and Green, S. S. (1996). Exposure
to dust, noise, and pesticides, their determinants, and the use of protective equipment among
California farm operators. Applied Occupational and Environmental Hygiene 11, 1217–1225.
Nieuwenhuijsen, M. J., Kruize, H., and Schenker, M. B. (1998). Exposure to dust and its particle size
distribution in California agriculture. American Industrial Hygiene Association Journal 59,
34–38.
Nieuwenhuijsen, M. J., Noderer, K. S., Schenker, M. B., Vallyathan, V., and Olenchock, S. (1999).
Personal exposure to dust, endotoxin and crystalline silica in California agriculture. The Annals of
Occupational Hygiene 43, 35 –42.
Norboo, T., Angchuk, P. T., Yahya, M., Kamat, S. R., Pooley, F. D., Corrin, B., Kerr, I. H., Bruce, N.,
and Ball, K. P. (1991). Silicosis in a Himalayan village population: role of environmental dust.
Thorax 46, 341–343.
Orgill, M. M., and Sehmel, G. A. (1976). Frequency and diurnal variation of dust storms in the
contiguous USA. Atmospheric Environment 10, 813 –825.
Ownby, D. R., Johnson, C. C., and Peterson, E. L. (2002). Exposure to dogs and cats in the first year of
life and risk of allergic sensitization at 6 to 7 years of age. Journal of the American Medical
Association 288, 963– 972.
 SOIL DUST AND IMPLICATIONS FOR HUMAN HEALTH 31

Parvaneh, S., Kronqvist, M., Johansson, E., and van HageHamsten, M. (1999). Exposure to an abundance
of cat (Fel d 1) and dog (Can f 1) allergens in Swedish farming households. Allergy 54, 229 –234.
Popendorf, W. J., and Spear, R. C. (1974). Preliminary survey of factors affecting the exposure of
harvesters to pesticide residues. American Industrial Hygiene Association Journal 35,
374 –380.
Popendorf, W. J., Pryor, A., and Wenk, H. R. (1982). Mineral dust in manual harvest operations.
Annals of the American Conference of Governmental Industrial Hygienists 2, 101 –115.
Portengen, L., Sigsgaard, T., Omland, O., Hjort, D., Heederik, D., and Doekes, G. (2002). Low
prevalence of atopy in young Danish farmers and farming students born and raised on a farm.
Clinical and Experimental Allergy 32, 247 –253.
Prospero, J. M. (1999). Long-range transport of mineral dust in the global atmosphere: impact of
African dust on the environment of the Southeastern United States. Proceedings of the National
Academy of Sciences of the United States of America 96, 3396–3403.
Prospero, J. M., Charlson, R. J., Mohnen, V., Jaenicke, R., Delany, A. C., Moyers, J., Zoller, W., and
Rahn, K. (1983). The atmospheric aerosol system: an overview. Reviews of Geophysics and
Space Physics 21, 1607–1629.
Pye, K. (1987). In “Aeolian Dust and Dust Deposits.” Academic Press, London.
Rajini, P., Last, J. A., McCurdy, S. A., Lawson, R., Southard, R. J., Allamneni, K. P., and Witschi, H.
(1995). Lung injury and fibrogenic response to dusts from citrus and grape harvests. Inhalation
Toxicology 7, 363–376.
Reidler, J., Eder, W., Oberfeld, G., and Schreuer, M. (2000). Austrian children living on a farm
have less hay fever, asthma and allergic sensitization. Clinical and Experimental Allergy 30,
194 –200.
Riedler, J., and Braun-Fahrlander, C. (2001). Exposure to farming in early life and development of
asthma and allergy: a cross-sectional survey. The Lancet 358, 1129–1133.
Robinson, J. P., and Silvers, A. (2000). Measuring potential exposure to environmental pollutants:
time spent with soil and time spent indoors. Journal of Exposure Analysis and Environmental
Epidemiology 10, 341– 354.
Rylander, R. (1986). Lung diseases caused by organic dusts in the farm environment. American
Journal of Industrial Medicine 10, 221 –227.
Schenker, M. B. (1998). Respiratory health hazards in agriculture. American Journal of Respiratory
and Critical Care Medicine 158, S1 –S76.
Schenker, M. B. (2000). Exposures and health effects from inorganic agricultural dusts.
Environmental Health Perspectives 108, 661 –664.
Schenker, M. B., Farrar, J. A., Green, R. S., McCurdy, S. A., and Lawson, B. (1994). Persistent
wheeze and dust exposure among California farm operators. American Journal of Respiratory
and Critical Care Medicine 149, A400.
Schwartz, J., Norris, G., Larson, T., Sheppard, L., Claiborne, C., and Koenig, J. (1999). Episodes of
high coarse particle concentrations are not associated with increased mortality. Environmental
Health Perspectives 107, 339–342.
Shaw, G. E. (1980). Transport of Asian desert aerosol to the Hawaiian Islands. Journal of Applied
Meteorology 19, 1254–1259.
Sherwin, R. P., Barman, M. L., and Abraham, J. L. (1979). Silicate pneumoconiosis of farm workers.
Laboratory Investigation 40, 576–582.
Silva, P. J., Carlin, R. A., and Prather, K. A. (2000). Single particle analysis of suspended soil dust
from southern California. Atmospheric Environment 34, 1811–1820.
Simcox, N. J., Fenske, R. A., Wolz, S. A., Lee, I.-C., and Kalman, D. A. (1995). Pesticides in
household dust and soil: exposure pathways for children of agricultural families. Environmental
Health Perspectives 103, 1126–1134.
Singer, M. J., and Warkentin, B. P. (1996). Soils in an environmental context: an American
perspective. Catena 27, 179– 189.
32 J.L. SMITH AND K. LEE

Singleton, J., and Beaumont, J. (1989). In “COMSII—California Occupational Mortality, 1979– 1981,
Adjusted for Smoking, Alcohol, and Socioeconomic Status.” California Department of Health
Services, Sacramento, CA.
Spencer, W. F., Cliath, M. M., Davis, K. R., Spear, R. C., and Popendorf, W. J. (1975). Persistence of
parathion and its oxidation to paraoxon on the soil surface as related to worker reentry into treated
crops. Bulletin of Environmental Contamination and Toxicology 14, 265–272.
Stellman, S. D., Boffetta, P., and Garfinkel, L. (1988). Smoking habits of 800,000 American men and
women in relation to their occupations. American Journal of Industrial Medicine 13, 43 –58.
Sterling, T. D., and Weinkam, J. J. (1978). Smoking patterns by occupation, industry, sex, and race.
Archives of Environmental Health 33, 313 –317.
Su, W. H. (1996). Dust and atmospheric aerosol. Resources, Conservation and Recycling 16, 1–14.
Takai, H., Jacobson, L., and Pederson, S. (1996). Reduction of dust concentrations and exposure in pig
buildings by adding animal fat in feed. Journal of Agricultural Engineering Research 63,
113–120.
Uno, I., Amano, H., Emori, S., Kinoshita, K., Matsui, I., and Sugimoto, N. (2001). Trans-Pacific
yellow sand transport observed in April 1998: a numerical simulation. Journal of Geophysical
Research—Atmospheres 106, 18331–18334.
US Environmental Protection Agency (EPA), (1996). Ambient Levels and Non-cancer Health Effects
of Inhaled Crystalline and Amorphous Silica. In “Rep. No. EPA/600/R-95/115.” Office of
Research and Development, National Center for Environmental Assessment, Research Triangle
Park, NC.
Von Ehrenstein, O. S., von Mutius, E., Illi, S., Hachmeister, A., and Von Kries, R. (2000). Reduced
risk of hay fever and asthma among children of farmers. Clinical and Experimental Allergy 30,
187–193.
Von Mutius, E., Braun-Fahrlander, C., Schierl, R., Riedler, J., Ehlermann, S., Maisch, S., Waser, M.,
and Nowak, D. (2000). Exposure to endotoxin or other bacterial components might protect
against the development of atopy. Clinical and Experimental Allergy 30, 1230–1234.
Wilshire, H. G. (1980). Human causes of accelerated wind erosion in California’s Deserts. In
“Thresholds in Geomorphology.” (J. D. Vitek, Ed.), pp. 415–434, produced by Dowden &
Culver, London, Boston and Stroudsburg, PA. Allen & Unwin, London.
Woodruff, N. P., and Siddoway, F. H. (1965). A wind erosion equation. Soil Science Society of
America Proceedings 29, 602–608.
Zhang, Y. (1997). Sprinkling oil to reduce dust, gases, and odor in swine buildings. Agricultural
Engineering Digest 42L, 1–8.
 NUTRIENT MANAGEMENT FOR
IMPROVING LOWLAND RICE
PRODUCTIVITY AND SUSTAINABILITY
N.K. Fageria,a N.A. Slatonb and V.C. Baligarc
a
National Rice and Bean Research Center of EMBRAPA, Caixa Postal 179, Santo
Antônio de Goiás, Goiás, CEP 75375-000, Brazil
b
University of Arkansas, 1366 W. Altheimer Dr. Fayetteville, Arkansas 72704, USA
c
USDA-ARS-Alternate Crops and Systems Research Laboratory, Beltsville
Agricultural Research Center, Beltsville, Maryland 20705-2350, USA

I. Introduction
II. Soils Used for Rice Production
III. Management Practices to Improve Nutrient Use Efficiency
A. Liming
B. Nitrogen
C. Phosphorus
D. Potassium
E. Calcium, Magnesium, and Sulfur
F. Micronutrients
IV. Conclusions
Acknowledgments
References

Rice (Oryza sativa L.) is an important food crop for a large proportion
of the world’s population. Total rice production will need to increase to
feed an increasing world population. Rice is produced under both upland
and lowland ecosystems with about 76% of the global rice produced from
irrigated-lowland rice systems. The anaerobic soil environment created
by flood-irrigation of lowland rice creates a unique and challenging
environment for the efficient management of soil and fertilizer nutrients.
Supplying essential nutrients in adequate rates, sources, application
methods, and application times are important factors that influence the
productivity and sustainability of rice. This review emphasizes our
current, research-based knowledge of N, P, K, Ca, Mg, S, B, Fe, Mn, and
Zn management in regards to the efficiency and sustainability of lowland
rice production and identifies where additional research is needed to
bridge information gaps. Our goal is to provide a comprehensive review
describing the nutritional problems, nutrient use efficiencies, and the
production strategies used for efficient nutrient use and production of
lowland rice. While the soils, climatic environments, cultivars, and
degree of mechanization may vary considerably among the rice

63
Advances in Agronomy, Volume 80
Copyright q 2003 by Academic Press. All rights of reproduction in any form reserved
0065-2113/03$35.00
64 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

producing regions of the world, the basic principles governing efficient

nutrient use by flood-irrigated rice are relatively constant. A summation
of best management practices should help scientists develop practical,
integrated recommendations that improve nutrient use efficiency in
lowland rice production systems. q 2003 Academic Press.

I. INTRODUCTION

Rice (Oryza sativa L.) is the staple food in the diet of about one-half of the
world’s population, who live mostly in developing countries, and is arguably
the most important crop worldwide. In 2001, rice was grown on 151.5 million
hectares, which represents approximately 23% of the world’s total area seeded to
cereals (FAO, 2001). Rice provides 35 – 60% of the dietary calories consumed by
nearly 3 billion people (Guerra et al., 1998). By the year 2025, it is estimated that
it will be necessary to produce about 60% more rice than what is currently pro-
duced to meet the food needs of a growing world population (Balasubramanian
et al., 1998). Enhancement of rice production and sustainability are important
features of grain production to benefit the world’s 3 billion people who depend on
rice for their livelihood and as their basic food. Adequate amounts of essential
nutrients are needed by modern rice cultivars with improved cultural practices to
obtain higher yields. In this context, efficient use of inputs is vital to safely
produce the additional food from limited resources with minimal impact on the
environment.
Rice, like barley (Hordeum vulgare L.), oats (Avena sativa L.), rye (Secale
cereale L.), and wheat (Triticum aestivum L.), belongs to the Gramineae family.
Rice was first domesticated about 10,000 years ago; however, its exact origin of
domestication is not known. The domestication of rice could have occurred
independently at several places in a broad belt from the foothills of the Himalayas
to Vietnam and southern China (Chang, 1975). The geographical dispersal and
selection pressures of farming led to a large number of varieties of Oryza sativa,
the Asian species. Another species, Oryza glaberrima, was later domesticated in
western Africa (Hargrove, 1988).
Before discussing nutrient management practices, the different types of
cultural systems used to produce rice should be defined since the majority of the
discussion here will focus on a specific production system. There is no true
consensus on the terminology used to describe the different rice growing systems
and environments. Upland, lowland, dryland, and wetland are often used to
describe the different production systems, but these general terms frequently
have regional connotations and may share some common characteristics.
The International Rice Research Institute (IRRI, 1984) classified rice into five
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 65

groups based on the seasonal water regime (deficit, excess, or optimal), drainage
(poor or good), air temperature (optimum or low), soils (normal or problem), and
topography (flat or undulating). These groups are: (i) irrigated lowland, (ii)
rainfed lowland, (iii) deep water, (iv) upland, and (v) tidal wetlands. In this
discussion, we primarily consider irrigated-lowland rice, which is cultivated on
relatively flat lands with water control so that rice is flooded for all or part of the
growing season.
The presence of floodwater for part or all of the growing season requires that
the rice root system is adapted to largely anaerobic soil conditions. The rice plant
has adapted to this environment by transporting oxygen from the aerial portions of
the plant to the root system via aerenchyma tissues (Yoshida, 1981). A secondary
adaptive mechanism is the development of an extensive lateral, fibrous root
system located in the surface 1 –2 mm of oxidized soil at the soil – water interface.
Oxygen diffusing through the water layer allows this zone of soil to remain
oxidized. For these reasons, flooded rice normally has a shallow, fibrous root
system (Wells et al., 1993). The aquatic environment not only influences the
development of the root system, but also alters the availability of several essential
nutrients, affects nutrient uptake and use efficiency, fertilization practices, and
makes rice especially unique among crop production systems. The nutrient
management practices discussed here will address the essential nutrients in terms
of their specific functions pertaining to rice growth in an aquatic environment,
nutrient deficiency, nutrient toxicity, and potential environmental implications.
Finally, we will review the fertilization practices (i.e., fertilizer sources, rates,
timing, and methods of application) that influence nutrient use efficiency and
provide insight on how these practices may be improved upon in the future.

II. SOILS USED FOR RICE PRODUCTION

Rice may be the world’s most diverse crop. Rice is grown in both dry and wet
conditions, over a wide range of latitudes, and across a wide range of soil,
climatic, and hydrological conditions. For example, rice tolerates water as deep
as 5 m in the flood-prone areas of south and southeast Asia and grows well in the
drought-prone upland areas of Asia, Latin America, and Africa. A major portion
of the rice produced in Brazil is grown under upland, or dryland, conditions much
like wheat or corn (Zea mays L.). Rice grows as far North as Czechoslovakia and
Manchuria, and as far South as Uruguay and New South Wales, Australia
(Hargrove, 1988). The geographic range of worldwide rice production is from the
equator to the temperate areas of northern Japan and southern Australia, and from
sea level to altitudes of more than 2500 m.
The pedogenetic and morphological characteristics of soils used to grow
rice also vary considerably. The soils used for rice production worldwide are
66 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

distributed over the 10 soil orders (Moormann, 1978; Hudnall, 1991). Moormann
(1978) summarized that, worldwide, rice is grown on all soil orders identified in
the soil classification system (USDA, 1975). Worldwide, the wide array of soils
used to produce rice results in an equally diverse assortment of management
practices implemented for successful rice production on these soils. Murthy
(1978) reported that the soils on which rice grows in India are so extraordinarily
varied that there is hardly a type of soil, including salt-affected soils, on which it
cannot be grown with some degree of success. In Brazil, flooded rice is mainly
grown on Alfisols, Vertisols, Inceptisols, Histosols, and Entisols (Moraes, 1999).
In Sri Lanka, rice is grown on Alfisols, Ultisols, Entisols, Inceptisols, and
Histosols (Panabokke, 1978). In Indonesia, the main rice soils are Entisols,
Inceptisols, Vertisols, Ultisols, and Alfisols (Soepraptohardjo and Suhardjo,
1978). Raymundo (1978) reported that in the Philippines the soils used for
wetland rice production are mainly Entisols, Inceptisols, Alfisols, and Vertisols.
In Europe, rice is planted on limited areas in Albania, Bulgaria, France, Greece,
Hungary, Italy, Portugal, Romania, Spain, and Yugoslavia where the predomi-
nate soil orders are Inceptisols, Entisols, and Vertisols (Matsuo et al., 1978). In
the United States, rice is grown primarily on Alfisols, Inceptisols, Mollisols, and
Vertisols (Flach and Slusher, 1978). However, in Florida, a small hectarage of
rice is produced on Histosols. Most of the soils used for rice production in the
United States and some other geographic areas have properties that make them
ideally suited for flood-irrigated rice. The soils are relatively young, contain
significant amounts of weatherable minerals, and have relatively high base
saturations despite the fact that some are in areas of high precipitation (Flach and
Slusher, 1978).
Soil parameters for optimum rice yields are optimum soil depth, compact
subsoil horizon, good soil moisture retention, good internal drainage, good
fertility, and a favorable soil structure. Clayey to loamy clay texture soils are
appropriate for lowland rice production. Permeable, coarse-textured soils are less
suitable for flood-irrigated rice production because they have low water or
nutrient holding capacities. In Brazil, there are about 35 million hectares of
poorly drained soils, known locally as “Varzea”, distributed throughout the
country. At present about 1.8 million hectares of these soils are cultivated,
primarily to lowland rice, during the rainy season (IICA, 2000). Generally,
Varzea soils have good initial soil fertility, but after 2– 3 years of cultivation, the
fertility level is known to decline (Fageria and Baligar, 1996). Farming systems
need to be developed with improved soil management technology to bring these
areas under successful crop production. A sufficient supply of nutrients is one of
the key factors required to improve crop yields and maintain sustainable
agricultural production on these soils. Flood-irrigated rice is an important crop
that needs to be included in the cropping system of these poorly drained areas
during the rainy seasons. During dry periods, other crops can be planted in
rotation, provided there is proper drainage. These soils generally have an
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 67

adequate natural water supply throughout the year, but are acidic and require
routine applications of lime if legumes are grown in rotation with rice.

III. MANAGEMENT PRACTICES TO IMPROVE

NUTRIENT USE EFFICIENCY

Modern production agriculture requires the implementation of efficient,

sustainable, and environmentally sound management practices. In this context,
fertilization is an important factor required to achieve optimum yields of all
crops, including rice (Fageria et al., 1997a). Fertilizer is, however, one of the
most expensive crop production inputs and, if used improperly, can pollute the
environment or decrease production efficiency. Increasing the rice yield per unit
area by the use of appropriate nutrient management practices has become an
essential component of modern rice production technology. Management
practices to improve the nutrient use efficiency of crops include management
of soil pH, use of the proper inorganic and organic nutrient sources, correct
method and rate of fertilizer application, appropriate application timing of
fertilizers, water management, pest management, and the use of high-yielding
cultivars adapted to a given environment.

A. LIMING

Soil acidity is a major factor limiting crop yield in vast areas of the world
(Shainberg et al., 1989; Farina et al., 2000). Liming is an effective and widely
used means of manipulating soil pH. Several acidity indices are used to diagnose
and correct the problem. These include pH, cation or base saturation, and
aluminum saturation. Crop response to soil pH is most commonly used to
determine the quantity of lime required to reduce the adverse effects of soil
acidity on productivity. The quantity of lime required depends on crop species,
and sometimes, the cultivar within species, soil clay content, soil organic matter
content, soil pH, concentration of soil Ca2þ, Mg2þ and Al3þ, and the quality of
liming material (Adams, 1984; Eckert, 1987).
The potential advantages or disadvantages of liming soils used for lowland
rice production are a controversial issue. The reduction processes that occur in
flooded soils cause the soil pH to increase in acid soils and decrease in alkaline
soils to near neutrality (Ponnamperuma, 1972; Reynolds et al., 1999). During the
first 2 –3 days after flooding an acid soil, the pH decreases due to biotic
transformation of easily degradable organic compounds to organic acids and CO2
(Motomura, 1962). After an extended period of anaerobic soil conditions, the pH
increase is related to the reduction of Fe3þ and Mn4þ oxides (Ponnamperuma,
1972). However, factors such as the initial soil pH, organic matter content, soil
68 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

texture, and soil iron content govern whether the soil pH increases or decreases
after flooding.
Patella (1976) reported a positive effect of liming on lowland rice yield in the
State of Rio Grande do Sul of Brazil when the initial soil Ca þ Mg content was
2.5 cmolc kg21. He also reported that liming improved N uptake by rice. Fageria
(2002) conducted a greenhouse experiment to evaluate crop tolerances to soil
acidity of an Inceptisol and found a significant ðP , 0:01Þ quadratic relationship
between the shoot weight of lowland rice and a range of soil basic cation
saturations (56 – 98%). Maximum shoot weights were produced at the lowest
basic cation saturation. Crop tolerance to soil acidity of the five crops used in this
study, in order from least sensitive to most sensitive, was rice . corn . soybean
[Glycine max (L.) Merr.] . common bean (Phaseolus vulgaris L.) . wheat. The
tolerance of rice to soil acidity, and the intolerance of soybean, common bean,
and wheat to soil acidity have been previously reported (Fageria et al., 1997a;
Fageria and Gheyi, 1999). In India, rice seldom shows significant, positive
responses from lime applications to acid soils (Goswami and Banerjee, 1978).
Liming is generally for the benefit of other crops grown in rotation with rice
because rice grows well in acid soils. If liming is necessary to optimize the
growth and yield of other crops in the rotation, it should be applied immediately
after the rice crop and prior to the production of other crops in the rotation, not
immediately ahead of the rice crop (Wilson et al., 2001). Research in Arkansas
has shown that rice yields may benefit from low to moderate rates of lime
application to soils with pH near or below 5.0 (Ntamatungiro et al., 1999). Wang
(1971) reported that 5 t lime ha21 applied to a submerged latosolic soil (pH 4.8)
low in base saturation increased rice grain yield by 23% and straw yield by 27%.
Ponnamperuma (1960) observed an increase in rice yields from lime applications
to the very acidic lateritic soils of Sri Lanka. Likewise, Abichandani and Patnaik
(1961) and Mandal et al. (1966) also increased rice yields by liming. The decision
to lime acid soils used for rice production should consider the crops grown in
rotation with rice. Management of soil acidity is often essential to improve or
maintain the yields of important legume crops such as soybean and common bean
(Fageria and Gheyi, 1999) and grain crops like wheat and corn (Z. mays L.)
produced on acidic lowland soils (Fageria and Baligar, 1999b).

B. NITROGEN

Nitrogen is usually the most yield-limiting nutrient in lowland rice production.

The exception is for rice grown on organic soils that have ample N release to
supply seasonal crop N requirements. Research has shown that flood-irrigated
rice can be one of the most efficient or inefficient crops in regards to N fertilizer
use efficiency. The relative efficiency of rice utilization of N fertilizer is directly
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 69

related to water management, rice growth stage at N application, N source, and

the chemical transformations that occur to N after it is applied to the soil.
Intensive agricultural production systems have increased the use of N fertilizer
in effort to produce and sustain high crop yields. Consequently, N losses into the
environment have also increased (Schmied et al., 2000). Environmental N losses
can occur from NO3 leaching, NH3 volatilization, surface runoff, nitrification,
and denitrification. Release of N2O from soils occurs during biological and
chemical denitrification (chemodenitrification) and contributes to global warm-
ing from the emission of greenhouse gasses. (Nelson, 1982; Hutchinson and
Davidson, 1993; Thorn and Mikita, 2000). Over the last 20 years the
concentration of N2O in the atmosphere has increased by , 0.25% yr21 (Elkins
and Rossen, 1989). Anthropogenic inputs of all forms of N into the global N cycle
are an estimated 145 Tg yr21 (Galloway et al., 1995). Of this total, N losses from
N fertilizers are estimated at , 80 Tg yr21. Thus, fertilizer usage is considered as
a major contributing factor of this increase, compared with other sources of N2O
such as biomass burning, fossil fuel combustion, deforestation, increasing
amounts of wet and dry NOx deposition, and NH4 in rainwater.
Raun and Johnson (1999) estimated the worldwide fertilizer N recovery
efficiency by cereal grains (rice, wheat, sorghum (Sorghum bicolor L.), millet
(Pennisetum glaucum (L.) R. Br.), barley, corn, oat, and rye) is only , 33%.
The 67% of unaccounted for N represented a US$15.9 billion annual loss of
N fertilizer. Thus, when N fertilizers applied to all crops, including rice, are
not used efficiently, crop production costs increase as more N fertilizer must
be used to compensate for losses; however, more importantly this lost N
contributes to air and water pollution. The development, demonstration, and
eventual grower acceptance of efficient N fertilization practices can reduce N
use, maintain or increase current rice yields, and minimize the potential for N
losses into the environment (Kundu et al., 1996).
Improving the N use efficiency of rice is an important goal in sustainable
production systems. Various management practices can be employed to balance
the supply of N required for optimum crop production while minimizing potential
losses into the environment. The successful adoption of efficient N management
practices for flood-irrigated rice requires an understanding of the relationships
among rice growth and development, the biological and chemical transform-
ations of fertilizer and native soil N, soil chemical and physical properties, and
management of irrigation water.

1. Chemistry in the Soil

Lowland flood-irrigated rice grows in a complex, dynamic soil environment

during the course of a complete growing season. Understanding nutrient use
efficiency requires that the seasonal soil environment be described. Most of
70 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

the soils used for flood-irrigated rice production are subjected to alternate
aerobic –anaerobic cycles because rice is drained for harvest and may be grown
in rotation with upland crops. The soil is generally flooded immediately before
rice is seeded (i.e., direct waterseeded culture); transplanted; or when seedlings,
established by direct, dry-seeded methods (i.e., broadcast or drill-seeded delayed
flood culture) reach the 5-leaf stage and begin to tiller. Once established, the flood
is commonly maintained until physiological maturity when it is drained for
harvest. In some instances, the flood is intentionally drained to allow the soil to
completely dry for specific management reasons (i.e., straighthead control and
pest control) or is lost due to lack of short-term water availability. Flooding an
oxidized soil affects the rice utilization efficiency of both native soil and fertilizer
N. Under aerobic conditions, if environmental conditions are not limiting, NH4 is
oxidized to NO3 almost as rapidly as it is formed (Schmidt, 1982). At the time of
flooding, nearly all the NO2 2
2 and NO3 present in the soil is lost to denitrification,
leaching, or both, depending on the soil properties (Patrick and Wyatt, 1964).
Thus, the mineralized organic N that undergoes nitrification under aerobic soil
conditions, during the off-season, or while upland rotation crops are grown is
ultimately lost to denitrification before significant uptake by flood-irrigated rice
occurs. One benefit of flooding is that organic N continues to be mineralized
under anaerobic conditions and remains as NH4, which is available for plant use
rather than being nitrified to NO3 and lost as N2. The organic N mineralized after
flooding plays an important role in determining the rate and times that fertilizer N
is required to produce optimum rice yields. Rice production systems that lack
controlled water management and undergo several alternating aerobic and
anaerobic cycles during a single season are highly inefficient in regards to native
soil N use (Patrick and Wyatt, 1964). The majority of fertilizer-N losses from all
flood-irrigated rice production systems are usually attributed to NH3 volatiliza-
tion and denitrification (Freney et al., 1990).
Once a soil is flooded, several distinct zones of varying oxygen status exist and
play important roles in determining soil and fertilizer N use efficiencies (Fig. 1).
The flood slows oxygen diffusion to the soil surface and prevents aerobic N
transformations from occurring in the majority of the soil beneath the oxygenated
soil surface – floodwater interface. The flood depth is generally 5– 20 cm. The
oxygenated floodwater – soil surface layer is often depicted as being about 1 cm
thick, but its depth varies with soil physical properties (Mikkelsen, 1987). Below

Figure 1 An illustration of the various N chemical forms, transformations and behavior in the
flooded soil environment in which rice is grown. Nitrogen sources are in blocks, N chemical forms
are in circles and the mechanisms responsible for the various N transformations or behavior are
located on the arrowed lines. (Reprinted with permission of R. J. Norman, C. E. Wilson, Jr. and
N. A. Slaton).
NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 71
72 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

the oxygenated soil surface layer, the soil is highly anaerobic except for the
oxidized root rhizosphere. The oxygenated soil surface is a dynamic layer where
N transformations occur that influence N uptake and the numerous N loss
pathways. The interaction of the anaerobic soil environment with N fertilizer
sources, application times, and placement will be further scrutinized when
fertilization practices are later reviewed, but will be mentioned briefly in
describing the various N loss mechanisms in the following sections.
Understanding the relationships among various N loss mechanisms, crop
nutrient uptake, and crop management practices is vitally important for adopting
management practices that improve N use efficiency. Ammonia-based fertilizers,
such as urea, are the most abundant forms of nitrogenous fertilizer applied to
agricultural soils (Eichner, 1990). The most fundamental N recommendation for
flood-irrigated rice is to use an ammonical form of N fertilizer. Ammonia N may
be lost through volatilization, fixed by clay minerals, or immobilized by soil
organic matter. Use of NO2 3 fertilizer sources for preflood N application is not
recommended because NO3-N is subject to denitrification after flooding and
seedling rice is sensitive to NO3 salts (Baser and Gilmour, 1982). Denitrification
is the anaerobic, microbial process that reduces NO3 to N2 gas which
subsequently escapes into the atmosphere. Denitrification occurs very rapidly
after a soil is flooded. Nitrate present at the time of flood establishment is lost
within 8 –10 days after flooding (Patrick and Wyatt, 1964). Nitrification may
occur in the oxidized soil – water interface from the upward movement of
mineralized soil NH4 and NH4 fertilizer. For these reasons, NO3-N contributes
very little towards the total N requirement of flood-irrigated rice. The goal of
preflood or early N applications is to keep the soil and fertilizer N in the NH4
form to prevent denitrification after flooding. The NHþ 4 is a cation which attaches
to the soils cation exchange complex and is held by the soil very close to its point
of placement. In comparison to NHþ 2
4 , NO3 is an anion and is mobile in the soil.
Thus, NO3 is also susceptible to leaching (on permeable, sandy soils) or may
possibly be transported in the general direction of water movement across a field.
Nitrate that is not leached out of the soil profile or absorbed by rice is quickly
denitrified upon its movement into the anaerobic soil zone.
In soil, N availability for plant uptake and leaching processes is related to the
NH4 adsorption and desorption capacity of the soil. Wang and Alva (2000)
reported that the potential NH4 buffering capacity and labile NH4 for sandy soils
were much lower than those for clay and silt loam soils. This means that NH4
fixation capacity of a soil should be taken into account when strategies for N
fertilization are developed to maximize N use efficiency. Further, the adsorption
and desorption of NH4 are also related to soil organic carbon. Organic C is
positively correlated with the potential NH4 buffering capacity and labile NH4
content for soils (Wang and Alva, 2000). Figure 1 summarizes the physical,
chemical and biological transformation, plant uptake, and loss pathways of N in a
flooded rice ecosystem.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 73

2. Functions and Deficiency Symptoms

Nitrogen is one of the major elements required for plant growth. It is a

constituent of numerous important compounds found in living cells, including
amino acids, proteins (enzymes), nucleic acids, and chlorophyll (Traore and
Maranville, 1999). Nitrogen fertilizers are routinely applied to rice, but N
deficiency symptoms are frequently observed because of the many loss
mechanisms that influence fertilizer N use efficiency and the amount of N
required to produce high yields. Nitrogen is a mobile element inside the plant;
hence, deficiency symptoms appear on the older leaves first. The lower leaves of
N deficient rice turn yellow, but the whole plant may become chlorotic if the N
deficiency is not corrected. The chlorosis begins near the tip of the leaf blade and
advances towards the leaf base until the whole blade is yellow. The dominant
symptom of N deficiency in rice is the development of the pale yellow color on
the lower leaves. Nitrogen deficiency reduces plant height, tillering, leaf area
index, leaf area duration, and crop photosynthetic rate which leads to lower
radiation interception and lower radiation use efficiency. Color photographs of
the N deficiency symptoms of rice have been presented by Mueller (1974),
Cheaney and Jennings (1975), Ishizuka (1978), Yoshida (1981), Fageria (1984),
and Fageria and Barbosa Filho (1994).

3. Forms of N Uptake
þ
Plants absorb N from the soil solution as NO2 3 or NH4 . Most plants can
absorb both forms of N equally; however, the form of N uptake is mainly
determined by its abundance and accessibility. In well-drained soils, NO3
dominates, while under anaerobic conditions or in cold climates NH4 is the
dominant form. In general, under agricultural conditions, soil NO3 concen-
trations range between 0.5 and 10 mM (7 – 140 mg NO3 kg21), while NH4
concentrations are 10 – 1000 times lower and reach the millimolar range only in
exceptional cases, such as after fertilization (Wiren et al., 1997). Although both
forms of N can be absorbed by rice, Fried et al. (1965) showed that excised
roots from rice seedlings absorbed NH4 5– 20 times faster than NO3, with the
rate being somewhat dependent on the solution pH. The majority of the
published literature suggests that rice prefers NH4 compared to NO3 (Mengel
and Viro, 1978). When NO3 is absorbed, it must be reduced for assimilation and
transport within the plant, with the reduction process requiring energy.
Takenaga (1995) suggested that rice may prefer one N form over another
based on plant growth stage. During vegetative growth, NH4 was absorbed more
effectively while NO3 is preferentially absorbed during reproductive growth.
Moristsugu (1980) concluded that rice shoot and root dry weights were higher
when rice was grown in an NH4-N culture solution compared to a NO3-N
74 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

culture solution. Rice that is flooded continuously, from early vegetative growth
until physiological maturity, does not absorb large amounts of NO3 because the
NO3 is rapidly lost via denitrification after flooding. For this reason, soil or
fertilizer NO3 is of little or no benefit to lowland rice. In most cases, an NH4
forming N fertilizer must be used along with management practices that largely
prevent nitrification to optimize N uptake efficiency.
Ammonium sulfate and urea [CO(NH2)2] are generally considered equally
effective fertilizer sources for rice (Bufogle et al., 1998; Gaudin and Dupuy,
1999). The initial soil pH, due its influence on potential NH3 volatilization losses,
and the length of time required to flood the soil may influence the efficiency of
uptake of these two N sources (Wells and Turner, 1984; Clark et al., 2001).
Ammonium sulfate may be more efficient when applied on alkaline soils or when
the N is not incorporated immediately by flooding. Urea, following application to
the soil, undergoes hydrolysis to eventually form NH4. However, under certain
soil and environmental conditions, N in the form of NH3, a gas, may be lost by
volatilization. Ammonia volatilization may be a pathway of significant N loss on
alkaline soils.
Marschner (1995) reported that calcifuges, or plants adapted to acid soils, and
plants adapted to low soil redox potential (e.g., lowland rice), prefer NH4. In
contrast, calcicoles, or plants with a preference for calcareous soils, utilize NO3
preferentially. De Datta (1981) reported that the N applied at planting should be
in the NH4 form. The N used as topdressing is less critical as NH4 and NO3 forms
appear to be equally effective (Wilson et al., 1994). When the crop is fully
established (i.e., panicle initiation), the NO2
3 form of N is rapidly taken up before
it can be leached down to the reduced soil layer where it could be lost through
denitrification. This may account for the equal performance of NO3 containing
fertilizers such as NH4NO3 as compared to other NH4 containing or forming N
sources such as ammonium sulfate or urea (De Datta, 1981).

4. Seasonal N Uptake and Partitioning

In order to understand the nutritional requirements of rice a general

knowledge of rice growth and development and its influence on the demand for
nutrients is required. Dry matter and grain yield production are perhaps the best
growth parameters to show how plant growth relates to nutritional requirements.
Between emergence and beginning tillering, a gradual linear increase in plant
dry matter occurs (De Datta, 1981; Fageria et al., 1997a). Dry matter increases
very rapidly between the onset of tillering and beginning internode movement
due to tiller formation and an increase in leaf size. This rapid growth produces a
sigmoidal shaped growth curve. In long-season cultivars, a plateau or
“midseason lag phase” may appear at the end of vegetative growth just prior
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 75

to panicle initiation, which is the onset of reproductive growth. The lag phase is
followed by another rapid increase in dry matter accumulation as the panicle
develops, emerges, and the grain filling process is completed. Very short-season
cultivars may not have a distinct lag phase resulting in a continuous rapid
growth curve from beginning tillering until the grain filling process is
completed.
For optimum growth and yield, rice requires that adequate N be available at the
beginning of the rapid growth (tillering) period. The number of panicles per unit
area is determined by either stand density or tiller development during vegetative
growth and is the first yield component determined (Stansel, 1975). By beginning
internode elongation, or panicle initiation, the maximum tiller number has been
reached. The second yield component, the potential number of grains per panicle,
is determined at panicle initiation and is influenced by the plants’ nutritional
status during vegetative growth. Wells and Faw (1978) showed that under
optimum stand densities increasing the N rate did not significantly increase the
number of tillers per unit area, but when stand density was constant, the number
of florets (potential grains) per panicle significantly increased with increasing N
rate. The number of blanks per panicle also increased with increasing N rate, but
grain yield increased too. Counce and Wells (1990) found that at extremely low
stand densities (, 30 seedlings per square meter) increasing the preflood N rate
stimulated tillering, increased the number of panicles, decreased the number of
grains per panicle, and increased grain yield. At low stand densities, the most
yield-limiting factor is panicles per unit area. Thus, increasing the preflood N rate
on fields with below optimum stands increases tillering (potential panicles) which
has the greatest impact on the number of grains per unit area. Grain weight, the
third and final yield component, is primarily determined by genetics and is not
influenced by N rate (Counce and Wells, 1990).
Rice grain yield is a function of the panicles per unit area, number of spikelets
per panicle, 1000-grain weight, and spikelet sterility or filled spikelets (Fageria
et al., 1997b). Therefore, it is very important to understand the management
practices that influence yield components and consequently grain yield. The
3-year average of these yield components, as affected by N rate, on a Brazialian
Inceptisol are presented in Fig. 2 (Fageria and Baligar, 2001). Application of N
rates up to 210 kg N ha21 significantly increased panicle length ðP , 0:01Þ and
the relationship between N rate and panicle length was linear (Fig. 2). The panicle
number per square meter and 1000-grain weight also increased significantly and
quadratically as N fertilizer application rate increased. Spikelet sterility,
however, decreased significantly and linearly with increasing N rates. Nitrogen
rate accounted for about 96% of the variation in panicle length, 91% of the
variation in panicles per square meter, 75% of the variation in spikelet sterility,
and 73% of the variation in 1000-grain weight. Yoshida (1981) also reported that
panicles per unit area, filled spikelet percentage, and 1000-grain weight were
major contributors to increased grain yield in modern high-yielding rice cultivars.
76 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Figure 2 Influence of N fertilizer rate on panicle length, panicle number per square meter,
spikelet sterility, and 1000-grain weight of lowand rice. Values are the average of 3 years of field
experimentation. (Reproduced with permission from Fageria, N. K. and Baligar, V. C. 2001. Lowland
rice response to nitrogen fertilization. Commun. Soil Sci. Plant Anal. 32: 1405–1429. (Copyright
Marcel Dekker, New York).).

Compared to the untreated control, application of 210 kg N ha21 increased

panicle length by 9%, increased panicle number per square meter by 24%,
increased 1000-grain weight by 6%, and decreased spikelet sterility by 14%. Low
spikelet sterility at high N application rates is considered one of the important
selection criteria for N responsive rice cultivars (Yoshida, 1972).
During the course of the growing season, whole plant tissue N concentration
gradually declines as plant dry matter increases, regardless of N rate (Table I).
Although tissue N concentration decreases, total N uptake increases as N fertilizer
rate increases (Table II). Approximately one-half of the total dry matter of a
mature rice crop is produced by panicle differentiation or the early boot stage
(Sims and Place, 1968; Guindo et al., 1994b). Total N uptake tends to follow the
same general pattern as dry matter accumulation. Under optimum fertilization,
one-half of the total N is absorbed before one-half of the total dry matter is
produced. The remaining one-half to one-third of the N, not in the plant by panicle
differentiation, is obtained either from topdressed N fertilizer, soil N released via
N mineralization, or both. The total N content of rice straw and grain at maturity
usually contains 20 – 40% more N than that supplied by fertilizer and represents
plant uptake of native soil N (Moore et al., 1981; Norman et al., 1992b).
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 77

Table I
Whole Plant Concentration of N in the Shoots of Flooded-Irrigated Rice Cultivar ‘Metica 1’
During the Growing Season. Concentrations are the Average of Three Crop Years
(Fageria, unpublished data, 2001)

Days after seeding (growth stage) (g N kg21)

N Fertilizer rate
(kg N ha21) 22 (IT) 35 (AT) 71 (IP) 97 (B) 112 (F) 140 (PM)

0 40 28 12 9 8 5
30 42 28 11 9 7 6
60 43 30 12 10 9 6
90 44 31 12 11 8 6
120 44 34 13 10 9 6
150 45 32 13 12 9 6
180 46 34 13 13 10 7
210 45 33 15 13 10 7

R2 0.95** 0.85** 0.84** 0.89** 0.72* 0.75**

*,**Significant at the 5 and 1% probability levels, respectively. IT, initiation of tillering; AT, active
tillering; IP, initiation of panicle; B, booting; F, flowering, PM, physiological maturity.

Table II
Whole Plant Content (Accumulation) of N in the Shoots of Flooded-Irrigated Rice Cultivar
‘Metica 1’ During the Growing Season. Contents are the Average of Three Crop Years
(Reproduced with Permission from Fageria, N. K. and Baligar, V. C. 2001. Lowland Rice
Response to Nitrogen Fertilization. Commun. Soil Sci. Plant Anal. 32: 1405– 1429.
Copyright Marcel Dekker, New York)

Days after seeding (growth stage) (kg N ha21)

N Fertilizer rate
(kg N ha21) 22 (IT) 35 (AT) 71 (IP) 97 (B) 112 (F) 140 (PM) Grain

0 13 24 35 50 57 28 36
30 13 26 41 62 66 37 41
60 15 37 45 82 95 40 55
90 17 33 52 88 89 41 55
120 17 42 57 91 122 48 66
150 19 40 63 113 113 52 67
180 16 44 67 137 135 59 71
210 16 39 87 130 130 66 74

R2 0.78** 0.85** 0.96** 0.95** 0.92** 0.98** 0.97**

**Significant at the 1% probability level. IT, initiation of tillering; AT, active tillering; IP, initiation of
panicle; B, booting; F, flowering, PM, physiological maturity.
78 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Nitrogen is a mobile element within the plant. Nitrogen stored in the stems and
bottom, older leaves can be translocated up to younger leaves when N is deficient
or to the developing panicle at heading. When individual leaves are analyzed for
N, a concentration gradient can be observed among the leaf positions from top to
bottom. The youngest (top) leaves have a higher N concentration than older
leaves (Westfall et al., 1973). At panicle differentiation, the top two leaf blades,
top two leaf sheaths, remaining bottom leaf blades, and remaining leaf sheaths
and stems contain near equal amounts of N (Norman et al., 1992b). The top two
leaf blades contain the largest amount of N. As rice development progresses
towards heading, N is first translocated from the bottom leaf sheaths and stems to
the developing panicle which may cause some senescence of the older leaves.
Once the panicle emerges, the total N content of all plant parts tends to gradually
decline at the same rate until physiological maturity. Based on total N content,
about 60– 70% of the aboveground N is located in the rice panicles at maturity
(Table II; Moore et al., 1981; Norman et al., 1992b). Norman et al. (1992b) found
that 86% of the total N in the panicles was translocated from rice leaf blades
(58%) and sheaths and stems (28%) to the panicles. The remaining 14% of the
panicle N was obtained from N uptake by the rice root system, presumably from
N released by mineralization from the flooded soil. Upon recognizing that N was
lost from the rice leaves during grain filling it was estimated that about 30% of the
panicle N was actually obtained from root uptake after heading. Regardless, the
majority of N residing in the rice grain at harvest is initially absorbed between
rice emergence and heading, stored in the vegetative plant tissues, and eventually
translocated to the panicle. For this reason, proper preflood N management and
efficient plant use of early N fertilizer is very important to high yield production.
The total N content of the straw and grain at maturity is equaled or surpassed
only by K and Si. At physiological maturity, the aboveground biomass, including
both straw and grain, of a healthy high-yielding rice crop may weigh between
16,000 and 28,000 kg ha21 and contain approximately 160 –280 kg N ha21
(Moore et al., 1981; Guindo et al., 1994a; Wilson et al., 1994).
The proportion of total plant N partitioned to the grain is called the N harvest
index (NHI). Nitrogen in the roots has little influence on the efficiency of N
partitioning (McNeal et al., 1966; Fawcett, 1980), the NHI ratio refers only to N
in the aboveground parts of the plant. A high NHI is associated with efficient
utilization of N (Fawcett and Frey, 1983; Rattunde and Frey, 1986) and grain
protein yields (Cox and Frey, 1978; Welch and Yong, 1980), but only rarely with
grain protein percentage (Peterson et al., 1975) and is inversely related to mean
straw yield (Rattunde and Frey, 1986). Selecting for high NHI may give
simultaneous improvement for grain yields and grain protein percentage (Fawcett
and Frey, 1983), or an increase in grain yield with constant grain protein
percentage (Loffler and Bush, 1982). The magnitude of NHI is influenced by
environmental factors. The efficiency of N translocation can be depressed by high
application rates of N (Fawcett and Frey, 1983). Dingkuhn et al. (1991) reported
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 79

NHI values ranging from 0.60 to 0.72 for three semidwarf rice cultivars differing
in growth duration. Guindo et al. (1994a,b) also reported NHI values of 0.58 and
0.62 in two lowland rice cultivars.
Mature rice straw has a N concentration of 5.0– 8.0 g N kg21 and the reported
C:N ratio ranges from 50:1 (Norman et al., 1990) to 71:1 (Eagle et al., 2001). The
C:N ratio of rice roots is slightly higher than straw at 98:1 (Norman et al., 1990).
A rice straw yield of 11,200– 16,800 kg ha21 contains 60 – 84 kg N ha21. In
comparison, soybean residue (stubble) remaining after harvest may weigh
5600– 6720 kg ha21 and contain 112 –224 kg N ha21 (Hanaway and Johonson,
1985; Norman et al., 1990). The N concentration of rough rice grain averages
about 10.0 g N kg21 (Moore et al., 1981).
Harvest index, the weight of harvested grain divided by the total aboveground
biomass (straw and grain), is an indication of how efficient a crop is at producing
grain. A crop harvest index could range from 0 to 1.0 depending on the crop and
the plant part that is harvested. A high harvest index is desired for rice. Published
harvest index values for rice usually range from 0.35 to 0.55 (Counce et al., 1992;
Norman et al., 1992a; Roberts et al., 1993; Wilson et al., 1994). The maximum
harvest indexes reported for rice are 0.60– 0.65 (Roberts et al., 1993). Advances
in pest management and breeding have increased the harvest index for rice.
Roberts et al. (1993) showed that semidwarf cultivars had a higher harvest index
than the taller cultivars that were grown before the development of semidwarf
cultivars. Although the biological yield between semidwarf and tall cultivars did
not vary, the harvest index decreased more rapidly for tall cultivars as N rate
increased. Norman et al. (1992a) found that harvest index increased when
preflood N application was delayed for more than 7 days past the normal
recommended time (5-leaf stage). Although, delayed fertilization and flooding
reduced straw yield it did not decrease grain yields.

5. Critical Level in Plant

The rate of nutrient uptake by rice changes during growth and development.
The rate of nutrient uptake is influenced by the nutrient; fertilizer application
source, time, and rate; nutrient availability in the soil; and plant growth
characteristics. The absorption of most nutrients is generally vigorous during
vegetative growth, but is limited by the root system. Thus, in general, nutrient
uptake rate reaches a peak before heading (Takenaga, 1995). Nitrogen, P, and K
are the first to reach the highest uptake rate in rice, while Mn and Si are the last in
this respect. By heading, rice has absorbed about 60– 80% of the total amount of
each nutrient to be absorbed (Ishizuka and Tanaka, 1952a).
The use of tissue N concentration of rice plants as a means of evaluating the
N nutritional status of the rice crop must be carefully interpreted. The N
concentration of rice tissues does not always increase as the rate of N fertilizer
80 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

increases. Nitrogen stimulates vegetative growth which dilutes N in the plant

tissue rather than increasing tissue N concentration. Plant N content is often a
more meaningful measurement than tissue N concentration. The N concentration
of rice straw decreases throughout the growing season as plant biomass increases
and the developing panicle acts as a strong sink for N and other nutrients. Fageria
and Baligar (2001) reported that the critical N concentration in the rice plant
increased quadratically with increasing N application rate and decreased with an
increase in plant age due to the dilution effect (Table I). Yoshida (1981) reported
that the tillering rate in rice increased linearly as leaf blade N concentration
increased up to 50 g N kg21. Tiller development stopped when leaf blade
N concentration was , 20 g N kg21. De Datta (1981) reviewed the critical N
concentrations proposed by various sources and suggested 25 g N kg21 in the leaf
blade of rice plants at the active tillering stage was the critical value. Wells et al.
(1993) suggested a critical range of 25– 32 g N kg21 in the Y-leaf at panicle
initiation. Fageria et al. (1997a) reported adequate N concentrations ranged from
26 to 42 g N kg21 in the uppermost mature leaves at heading.

6. Nitrogen Use Efficiency

Nutrient use efficiency by crops can be expressed and interpreted in a number

of ways. The majority of research on N fertilization of rice in the USA and many
other rice growing regions has focused on calibration of N fertilizer rates, plant
uptake and utilization, and the magnitude of N loss pathways among various N
fertilizer sources, application times, and application methods. Information
generated from these types of studies help to develop best management practices
for the production of high yields while minimizing N losses and costs associated
with N fertilization. Nutrient use efficiency can also be expressed in terms of
biomass or grain production efficiency per unit of nutrient uptake or application.
Regardless of how nutrient use efficiency is expressed, the ultimate goal of this
information is to enhance our knowledge of crop growth and production
efficiency.
Numerous research studies have been conducted to measure the fate (i.e., plant
uptake and N loss pathways) of N fertilizer under various cultural management
systems. These studies have focused on accounting for N fertilizer use efficiency
or uptake by rice; N lost via leaching, NH3 volatilization, and denitrification; and
microbial biomass use of N fertilizer in various cultural management systems,
straw management, environments, and crop rotations. Rice may be one of the
most efficient or inefficient crops in regards to N fertilizer use efficiency. The N
fertilizer source, the N application time, or both usually dictate at which end of the
efficiency spectrum N fertilizer use by rice resides. The recovery efficiency of N
fertilizer by rice generally ranges from 20 to 80% (Table III) with an average of
about 30 – 40% (Cassman et al., 1993). Recovery efficiency is generally defined
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 81
Table III
Nitrogen Recovery Efficiencies of Basal, Topdressed, or Basal and Topdressed N Applications from Various Geographic Regions

Recovery efficiency % of
applied 15N fertilizer Unaccounted for
N % of applied
15 15
Reference Cultural system N Application time Plant Root Soil N fertilizer

Bollich et al. (1994) Dry Basal þ Topdress 56 – 33 11

De Datta et al. (1988) Water Basal þ Topdress 41 35a 25
De Datta et al. (1988) Transplant Basal þ Topdress 33 35a 33
Norman et al. (1992b) Dry Basal 54 – 20 26
Norman et al. (1989) Dry Basal þ Topdress 57 – 14 29
Norman et al. (1989) Dry Basal 46 – 19 35
Norman et al. (1989) Dry Topdress 58 – 9 33
Patrick and Reddy (1978) Dry Basal 28 26a 46
Patrick and Reddy (1978) Dry Topdress 47 32a 21
Westcott et al. (1986) Water Topdress 48 – 8 44
Westcott et al. (1986) Dry Topdress 44 – 8 48
Westcott et al. (1986) Water Basal 42 – 12 46
Westcott et al. (1986) Dry Basal 30 – 10 60
Wilson et al. (1989) Dry Basal 42 11 22 25
Wilson et al. (1989) Dry Topdress 61 6 17 16
a
A single value denotes that soil and root N were grouped together.
82 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

as the total N accumulation in the aboveground biological yield (grain þ straw)

per unit of applied N fertilizer. Recovery efficiencies have been determined by
use of 15N labeled fertilizers and by the difference method. Fertilizer N recovery
measurements are affected by the method of determination, scope of sampling
(i.e., plant and soil), and the time of sampling. Fertilizer N recovery by difference
usually yields higher uptake efficiencies than studies using 15N labeled fertilizers
(Reddy and Patrick, 1976; Eagle et al., 2001). The differences between these two
methods may either be minimal (Reddy and Patrick, 1976) or significant (Reddy
and Patrick, 1976; Eagle et al., 2001). The difference method usually yields
higher recovery efficiencies because of the priming effect of N fertilization,
stimulation of N mineralization in the presence of fertilizer N, and root
exploration is greater in fertilized plots (Westerman and Kurtz, 1974; Olson,
1980; Olson and Swallow, 1984; Rao et al., 1991; Yamaguchi, 1991; Cassman
et al., 1993; Schnier, 1994). Regardless of the method of determination, the
balance of N fertilizer not accounted for by N recovery efficiency should not be
interpreted as N that is initially lost without utilization from the soil – plant
system. A portion of the unaccounted for N may certainly be lost to
denitrification, leaching, and ammonia volatilization, but a significant portion
of this N may be incorporated into the microbial biomass, reside in the rice root
system, or be lost from senescing rice leaves and should not be ignored when
performing N use efficiency studies.
When measured at plant maturity, N fertilizer uptake and recovery in the
aboveground biomass alone may overestimate fertilizer N losses and under-
estimate fertilizer recovery by as much as 60% when soil, root, and gaseous N
losses from senescing rice leaves are not accounted for. Wilson et al. (1989)
consistently recovered 11% of the total applied N fertilizer in the rice root system
at maturity. Furthermore, between 16 and 25% of the total applied N fertilizer has
been recovered in the soil organic fraction between panicle differentiation and
maturity (Wilson et al., 1989; Norman et al., 1992b; Bollich et al., 1994).
Measurement of N fertilizer recovery in the aboveground biomass at maturity
underestimates fertilizer N recovery by 14– 25%. Norman et al. (1992a) initially
observed that fertilizer N recovery was highest when measured at flowering and
declined during the grain fill process, presumably from gaseous N losses from the
senescing leaves. The decline in the total N content of aboveground biomass
during the ripening phase was later confirmed by Guindo et al. (1994a,b). The
magnitude of N loss from rice leaves is influenced by the cultivar and
environment (Stutte and da Silva, 1981). Even when all possible sources of plant
and soil N are considered, native soil and fertilizer N loss occurs in all flood-
irrigated rice cultural systems. Use of appropriate crop and fertilizer management
strategies is important to maximize N recovery efficiency by rice.
The fertilizer N recovery efficiency in the aboveground biomass of flood-
irrigated rice depends on the source, time of application, and method or
placement of N fertilizer. The time that N fertilizer is applied to rice varies
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 83

slightly among the various cultural management systems (e.g., transplanted,

direct-dry seed, or direct-water seed) used to produce flood-irrigated rice.
However, many of the principles that govern N use efficiency are the same.
Nitrogen fertilizer is typically applied in at least two or more split applications in
most production systems. The first application is often termed basal or preflood.
The basal application generally represents the largest proportion (50 – 75% of the
total N requirement) of N fertilizer applied in any single N fertilizer application
during the growing season because of its influence on grain yield potential. In
waterseeded and transplanted cultural systems, an NH4 forming N fertilizer
source is generally applied to the soil surface and mechanically incorporated
before the field is flooded, applied to the soil surface and incorporated by the
flood, or injected into the soil. In the dry-seeded, delayed flood system, the basal
N application is made during early vegetative growth (i.e., 5-leaf stage) and
followed by establishing the flood. Regardless of the cultural system, the most
fundamental components of basal N management are to use an ammonium
forming N fertilizer source, apply the N fertilizer on a dry soil surface, and
immediately flood the soil.
Proper management of the floodwater is critical for efficient uptake of basal
fertilizer N. Draining the flood before rice uptake of N may result in nitrification
resulting in leaching of NO3 and denitrification if the soil is reflooded before N
uptake is complete. In the dry-seeded, delayed flood system where basal N is
applied at the 5-leaf stage, maximum uptake is achieved in about 3 weeks (Wilson
et al., 1989). The time requirement for maximum uptake of basal N is likely to
increase as basal N rate increases, but has not been fully evaluated. The recovery
efficiency of basal N (urea) in aboveground rice biomass in the dry-seeded,
delayed flood system may approach 70% under optimum conditions (Norman
et al., 1992a; Wilson et al., 1994). Research consistently shows that recovery
efficiency of basal N in this system is 40 – 50% (Reddy and Patrick, 1976; Norman
et al., 1989, 1992b; Bollich et al., 1994; Guindo et al., 1994a,b). However, when
basal N is applied to a wet soil or the flood is mismanaged the recovery efficiency
declines. In the delayed flood system, the recovery efficiency of preplant
incorporated urea is only about 25 – 30%, which is considerably less efficient than
urea application to the soil surface immediately before flooding (Norman et al.,
1989). Fageria and Baligar (2001) found N recovery efficiencies ranging from 30
to 50% when basal N was applied before seeding and flooded 30 days after
seeding. The recovery efficiency decreased as the basal N rate increased
suggesting the N loss exceeded the rate of plant uptake after flooding. When basal
N is applied far in advance of flooding the concentration of soil NO3 increases
until flooding and is followed by rapid denitrification (Norman et al., 1988).
Considerably less research has been published on the recovery efficiency of
preplant incorporated or injected basal N applications in the waterseeded cultural
system. Most of the published research suggests that uptake of preplant, basal N,
applied as urea, by waterseeded rice is about 35– 40% (Eagle et al., 2001).
84 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Westcott et al. (1986) showed that the recovery efficiency of basal applied N
fertilizer of waterseeded rice (42%) rice was greater than that of dry-seeded rice
(30%). However, the majority of published research suggests that N recovery
efficiency between these two seeding methods is equal or greater for dry-seeded
rice. In general, recovery efficiency of basal N applied to 5-leaf rice in the
delayed flood system is more efficient than preplant incorporated basal N in
waterseeded systems because of the shorter interval between N application and
eventual plant uptake. De Datta et al. (1988) showed that basal N applications
made to transplanted rice have similar recovery efficiencies as those cited for
waterseeded rice. Regardless of the cultural system, broadcast basal N applica-
tions are relatively efficient as long as the N is incorporated immediately after
application with tillage or irrigation water and maintained in the NH4 form. The
amount of N lost to ammonia volatilization or that is nitrified and eventually lost
to denitrification increases as the time between application and flooding increases.
Numerous researchers have found that split applications of N fertilizer applied
into the flood are utilized more efficiently than basal applications (Reddy and
Patrick, 1976, 1978; Westcott et al., 1986; Wilson et al., 1998). However, the
recovery efficiency of N topdressed into the flood is dependent on the rice growth
stage at the time of application. Nitrogen fertilizer topdressed into the flood from
panicle initiation to the early boot stage is absorbed rapidly and efficiently by
rice, regardless of the cultural seeding method. The recovery efficiency of N
topdressed at panicle initiation ranges from 45 (Westcott et al., 1986) to 82%
(Wilson et al., 1994). The high recovery efficiency of topdressed N is due to the
application of relatively low rates of N fertilizer (, 1/3 of total seasonal N
fertilizer applied), a highly developed rice root system, and the high plant N
requirement at this stage of plant development. Wilson et al. (1989) showed that
maximum absorption of N fertilizer topdressed at panicle differentiation occurred
during the first 3 days after application. Both NH4-N and NO3-N are absorbed
efficiently when topdressed at the appropriate time, but NH4-N is still more
efficient (Wilson et al., 1994). Nitrogen fertilizer topdressed into the flood before
panicle initiation is less efficient because the plant root system is not fully
developed (Norman et al., 1993, 1994). The rate of N loss, via N loss pathways,
exceeds the rate of N uptake when N is topdressed into the flood when rice is at
the seedling or tillering stages.
Recovery efficiency primarily concerns the efficient use of fertilizer N as a
means of determining best management practices to produce high yields with a
minimum of fertilizer inputs and N losses into the environment. Various other
means are employed to examine the internal nutrient use efficiency by a crop and
include agronomic efficiency (AE, D kg grain kg21 fertilizer-nutrient added),
physiological efficiency (PE, D kg grain D kg21 nutrient uptake), agrophysiolo-
gical efficiency (APE, D kg grain D kg21 total nutrient uptake) and apparent
recovery efficiency (ARE, %) (Fageria, 1992). All four N use efficiencies shown
in Table IV were calculated using the formulas from Fageria et al. (1997a). The
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 85
Table IV
Nitrogen Use Efficiencies as Affected by N Fertilizer Rate (Reproduced with Permission from Fageria, N. K. and Baligar, V. C. 2001. Lowland rice
response to nitrogen fertilization. Commun. Soil Sci. Plant Anal. 32: 1405–1429. Copyright Marcel Dekker, New York)

Agronomic efficiency
N Fertilizer rate (Dkg grain kg21 N Physiological efficiency Agrophysiological efficiency Apparent recovery efficiency
(kg N ha21) fertilizer added) (Dkg grain Dkg21 N uptake) (Dkg grain Dkg21 N uptake) (%)

30 35 156 72 49
60 32 166 73 50
90 22 182 75 37
120 22 132 66 38
150 18 146 57 34
180 16 126 51 33
210 13 113 46 32
Average 23 146 63 39

R2 0.93** 0.62* 0.87** 0.82**

*,**Significant at the 5 and 1% probability levels, respectively.

86 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

use of appropriate management practices to attain high recovery efficiency of

applied nutrient sources is desirable when examining the internal nutrient use
efficiency by a crop.
In any experiment with a nutritional variable, plants grown at the lowest
nutrient concentrations will inevitably have the highest utilization quotient
because of a dilution effect (Jarrell and Beverly, 1981). All N use efficiencies tend
to decrease as N fertilizer rate increases (Fageria et al., 1997a; Carreres et al.,
2000), especially when the fertilizer rate required to produce maximum yields is
exceeded (Table IV). Fageria et al. (1997a) found that all N use efficiencies,
except PE, significantly decreased as N rate increased. In most experiments with
a nutritional variable, plants grown at the lowest nutrient concentrations will
inevitably have the highest utilization quotient because of this dilution effect
(Jarrell and Beverly, 1981).
The AE for N of lowland rice in the tropics ranges from 15 to 25 kg grain kg21
N fertilizer (Yoshida, 1981; Peng et al., 1996). The PE (113 – 182 kg dry
matter kg21 N) is higher than the APE (46 – 75 kg grain kg21 N uptake) because
total dry matter, rather than grain, is used in the calculation. Singh et al. (1998)
reported an APE of about 64 kg grain kg21 N uptake and an AE of 37 kg
grain kg21 N fertilizer for 20 lowland rice genotypes which are comparable to
values listed in Table IV.
The ARE of N by lowland rice is related to N losses from soil via
nitrification – denitrification, NH3 volatilization, leaching, or a combination of
these loss mechanisms (Craswell and Vlek, 1979). The efficiency of utilization
for grain production in the tropics is about 50 kg grain kg21 total N uptake and is
relatively constant regardless of rice yield (Yoshida, 1981). Wada et al. (1986)
suggested 15 – 17 kg N t21 grain was required to produce an average yield of
5 –6 t ha21 and 19 kg N t21 was needed for higher yields. However, the absolute
amount of N absorbed is not the only factor to be considered in high-yielding rice.
Rather, the N supply pattern and uptake process of N throughout the whole life
cycle of the plant are equally important. This is because the formation (number,
size, or extent) of each rice yield component is dependant on the amount of N
supply at each crucial stage for the respective yield components (Mae, 1997).

7. Nitrogen Fertilization Practices

Adequate, but not excessive, rates of N fertilizer must be applied to produce

high yields, minimize production costs, and reduce the potential for polluting the
environment. The required N rate is governed by yield level, soil properties,
organic matter content, cropping system, disease pressure, water management,
weed control, socioeconomic condition of farmers, and the price of rice. Figure 3
shows the response of lowland rice to N fertilization on a Brazilian Inceptisol
(Fageria and Baligar, 2001). Rice grain yield increased as N fertilizer rate
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 87

Figure 3 Response of flooded lowland rice to N fertilizer rate on a Brazilian Inceptisol.

(Reproduced with permission from Fageria, N. K. and Baligar, V. C. 2001. Lowland rice response
to nitrogen fertilization. Commun. Soil Sci. Plant Anal. 32: 1405– 1429. (Copyright Marcel Dekker,
New York).).

increased to a point showing a significant ðP , 0:01Þ quadratic response in each

of the 3 years. The 3-year average showed that the agronomic maximum grain
yield of 6465 kg ha21 was produced by application of 171 kg N ha21. Singh et al.
(1998) also reported that maximum average grain yield of 7700 kg ha21 of 20
lowland rice genotypes was obtained at 150 –200 kg N ha21 at the International
Rice Research Institute in the Philippines. However, in fertilization experiments
90% of the maximum yield is often considered as an economical rate (Fageria
et al., 1997a). Fageria and Baligar (2001) found that application of 84 kg N ha21
produced 90% of the maximum agronomic yield, but stated that most Brazilian
farmers apply only about 60 kg N ha21 due to economic reasons.
Maskina et al. (1988) and Aulakh et al. (2000) reported that flooded rice
responded to N rates up to 120 kg ha21 on sandy loam soils in India. In the
Philippines, Dobermann et al. (2000) reported 80– 100 kg N ha21 was used for
88 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

maximal yields in field experiments during the wet-season (rainy period), and
120 – 150 kg N ha21 was used during the dry-season. Two-thirds of the N was
incorporated into the soil at transplanting and one-third was topdressed before
panicle initiation. These authors also reported that N fertilization rates, from 1992
onward, were increased from 108 to 120 kg N ha21 during the wet-season and
from 190 to 216 kg N ha21 during the dry-season. The number of split applica-
tions was also changed from two to three or four to improve N use efficiency.
During the dry-season, an average 60 kg N ha21 was applied as basal, 60 kg
N ha21 at midtillering, 60 kg N ha21 at panicle initiation and 30– 35 kg N ha21
at flowering. In the wet-season, average N application rates were 40 kg N ha21
as basal, 40 kg N ha21 at midtillering, and 30 kg N ha21 at panicle initiation.
Berge and Riethoven (1997) reported, based on predictions from a N fertilization
model, that the optimum economic N rate for medium-duration cultivars was
100 – 150 kg N ha21 and 150 –200 kg N ha21 for short-duration cultivars. They
also reported that estimated yield levels were higher in medium-duration
cultivars, but the yield difference between the two types of cultivars decreased
with increasing N inputs. The simulations indicated that longer duration
genotypes absorbed more soil-derived N and were less efficient with fertilizer
N. In most Asian countries N fertilizer recommendations are provided for specific
soils, growing seasons, and geographic regions and include information on
the rates and times to apply N fertilizer (Cassman et al., 1998). In the USA,
N application rates range from about 120 to 180 kg N ha21 (Hill et al., 1992;
Norman et al., 1997; Wilson et al., 2001). The recommended N rate is usually
specific for cultivars, soil properties, previous crop, combinations of these
factors, and are determined by each rice-producing state.
Cassman et al. (1998) indicated that Philippine farmers did not adhere to some
N fertilizer use recommendations meant to increase N use efficiency. Farmers in
all rice producing areas occasionally use N fertilization methods that are known
to result in poor N use efficiency. This suggests that the extension, demonstration,
and successful adoption, or implementation, by growers of efficient, research-
based recommendations may be more limiting than our current knowledge of N
use management. In some highly mechanized areas, growers may be required to
manage nutrients, like N, efficiently to comply with nutrient management
regulations.
Nitrogen fertilizers are mostly applied in bands or broadcast. Urea and
ammonium sulfate are the two main sources of inorganic N fertilizer for lowland
rice. There has been an increasing interest in the use of livestock manures and
green manures in crop production, because soil organic matter is one of the most
important soil components for productive and sustainable agriculture as well as a
main nutrient source of C, N, P, and S (Allison, 1973). Use of organic manures
not only contributes to the nutrient pool, but also indirectly influences soil
chemical and physical properties. Increases in soil organic matter content occur
slowly, generally taking several years to detect (Wander et al., 1994), but can
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 89

have a dramatic effect on long-term productivity (Tiessen et al., 1994). The use
of livestock manure can increase rice yields and enhance the mineralization of
native N (Eneji et al., 2001). Fageria and Baligar (1996) showed that use of
pigeon pea (Cajanus cajan L. Millspaugh) as green manure, in combination with
inorganic fertilizers, increased lowland rice yields by 15% as compared to
inorganic fertilizers alone.

8. Use of N Efficient Genotypes

Nitrogen use efficiency has been defined in various ways and the definitions
generally account for the quantity of N accumulated in the plant, known as uptake
efficiency, and the quantity of N utilized for grain production, known as
utilization efficiency. The N utilization efficiency is more important than N
uptake efficiency when evaluating the genetic potential among cultivars for
efficient grain production, especially on soils that require high rates of N to
maximize yield (Moll et al., 1982). Utilization efficiency coupled with economic
yield is a desired characteristic in crop plants if minimum depletion of soil N is a
goal. Moll et al. (1982) recommended the development of genotypes with both
high uptake and utilization efficiencies.
Use of N efficient genotypes is an important N management strategy for flood-
irrigated rice. An ideal genotype could be the one that absorbs relatively large
amounts of N from soil and fertilizer, produces a high grain yield per unit of
absorbed N, and stores relatively little N in the straw (Isfan, 1993). Rice cultivars
can differ in root growth and morphology (Slaton et al., 1990) and nutrient uptake
rates (Teo et al., 1995). Fageria and Barbosa Filho (2001) evaluated N use
efficiency of eight lowland rice genotypes grown on an Inceptisol of Brazil.
Nitrogen uptake in grain, NHI, and N use efficiency were significantly different
among genotypes. Nitrogen harvest index is a measure of N partitioning in rice,
which provides an indication of how efficiently the plant utilized the acquired N
for grain production. Genotypes that produced the lowest grain yield had the
lowest NHI, whereas genotypes that produced the highest grain yields also had
the highest NHI. Genetic variability for NHI exists within the small grain
genotypes and high NHI is associated with efficient utilization of N (Rattunde and
Frey, 1986). Thus, the variation in NHI is a characteristic of genotype and this
trait may be useful in selecting rice genotypes for higher grain yield.
Nitrogen use efficiency varies significantly among lowland rice genotypes (De
Datta and Buresh, 1989; Fageria and Barbosa Filho, 2001). Many researchers
have found significant variations of N use efficiency among lowland rice
genotypes (Broadbent et al., 1987; Buresh et al., 1988; Fageria et al., 1997a).
Such differences may be related to many physiological processes such as
absorption, NO3 reduction efficiency, N remobilization, translocation, assimila-
tion, and storage (Isfan, 1993).
90 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Nutrient uptake efficiency and nutrient utilization efficiency are governed by

different physiological mechanisms (Baligar and Fageria, 1997, 1999; Baligar
et al., 2001). Plant physiological factors associated with N use efficiency include
(i) root proliferation; (ii) NO3 uptake efficiency; (iii) N translocation efficiency;
and (iv) nitrate reductase enzyme efficiency (Duncan and Carrow, 1999). Each
stage of N assimilation is under genetic control. Nitrate is reduced to NH3 by
nitrate (NR) and nitrite reductase (NiR). These metalloenzymes require
cofactors: molybdenum-pterin for NR and Fe-containing hydrochlorine (Sir-
oheme) for NiR. The NR enzyme is localized in the cytosol, whereas NiR is
found in leaf and root tissue plastids (Duncan and Carrow, 1999).
Rice hybrids have a 10– 15% yield advantage over conventional rice cultivars
(Yang and Sun, 1988). The yield advantage of hybrid rice is presumably related
to a greater total N uptake and internal use efficiency of N. The total N uptake by
hybrid rice is greater than that of conventional cultivars, especially from
transplanting to tillering and from panicle emergence to grain filling stages
(Yang, 1987). Hybrid rice takes up about 15– 20% of its total N after heading and
consistently responds to N fertilizer applications made at flowering compared to
only 6– 7% for conventional cultivars (Yang, 1987). Hybrid rice had a greater AE
than conventional rice (Lin and Yuan, 1980; Yang, 1987). This increased N use
efficiency is not due to greater internal N uptake in dry matter production (defined
as dry matter produced per unit N accumulated in the plant) (Yang et al., 1999).
The primary factors contributing to the higher N use efficiency of hybrid rice are
higher N recovery efficiency, greater root N absorption potential, greater shoot-N
use capacity, and greater N remobilization efficiency.

C. PHOSPHORUS

Phosphorus deficiency is the second most important nutritional disorder of

lowland rice, especially in the highly weathered acidic soils of the tropics that
contain large quantities of Al and Fe oxides (Wells et al., 1993; Fageria and
Baligar, 1996; Baligar and Fageria, 1997; Fageria et al., 1997a,b; Seneweera
and Conroy, 1997). The highly weathered tropical soils are primarily Oxisols
and Ultisols that have low total and available P content, and also have a high P
fixation capacity (Fageria et al., 1991). Plants rarely absorb more than 20% of the
total fertilizer P applied due to fixation (Friesen et al., 1997). The uptake of P by
plants is governed by the ability of a soil to supply P to plant roots and by the
desorption characteristics of the soil (Roy and De Datta, 1985). The supply of P to
plant roots depends on the concentration of inorganic P in the soil solution and on
the capacity of the soil to maintain this concentration.
Application of P is often essential for profitable agricultural production.
However, accumulation of soil P in excess of crop needs has the potential to
enrich surface runoff with P that can cause eutrophication. Eutrophication has
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 91

been identified as the main cause of impaired surface water quality (US
Environmental Protection Agency, 1996). Eutrophication restricts water use
for fisheries, recreation, industry, and drinking due to the increased growth of
undesirable algae and aquatic weeds and also due to oxygen shortages caused by
their death and decomposition (Abrams and Jarrell, 1995; Sharpley et al., 1999).
Hence, appropriate management of P is an important aspect not only for higher
rice yields, but also for environmental protection.

1. Chemistry in the Soil

The transformations and chemistry of P in flooded soils have been thoroughly

reviewed by Ponnamperuma (1972) and Sanyal and De Datta (1991); hence, only
the primary factors affecting P availability will be discussed briefly here. Soil
P can be divided into the organic and inorganic fractions with the inorganic P
fraction being recognized as the pool that controls P availability to plants. The
inorganic P fraction is made up of at least five basic categories including
aluminum phosphates (Al-P), iron phosphates (Fe-P), calcium phosphates (Ca-P),
reductant soluble or the occluded phosphates (RS-P) and the readily available
2
orthophosphate forms (PO32 22
4 , HPO4 , or H2PO4 ) of P (Chang and Jackson,
1957). The predominate form of orthophosphate present in the soil solution
changes with soil pH.
The reduction process that occurs in the soil following flooding normally
increases the P availability to rice. Therefore, on many soils P availability is not a
yield-limiting factor for rice. Rice yields may not respond to P fertilization, but
upland crops, like corn and soybean, growing on the same soil might show
dramatic responses to P fertilization. Phosphorus deficiency of upland crops
following rice in the rotation has been reported for several crops (Willett, 1982;
Wells et al., 1995). The alternate anaerobic– aerobic soil conditions reduce the
availability of P to upland crops following rice in the rotation. The soil’s P
sorption capacity and the bonding energy of P increases under alternate
anaerobic– aerobic conditions (Sanyal and De Datta, 1991). Flooding decreases
the crystallinity of ferrous hydroxides, which increases their sorption capacity,
increases the insoluble Fe-P fraction, and reduces P desorption.
Iron phosphates are the primary source of P to lowland rice because their
availability is quickly affected by the anaerobic conditions created by flooding
(Patrick and Mahapatra, 1968; Ponnamperuma, 1972; Goswami and Banerjee,
1978). Willett (1986, 1989) reported that P availability increases after flooding
from the (1) reductive dissolution of ferric oxides; (2) the liberation of sorbed-
and RS-P; (3) changes in soil pH that increase the solubility of Fe-, Al-, and Ca-P;
and (4) the desorption of surface P. The relatively insoluble ferric phosphates
are reduced to the more soluble ferrous phosphates resulting in hydrolysis
of P compounds. On acid soils, soil pH generally increases following soil
92 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

submergence and increases the solution concentrations of Fe and Al phosphates.

Flooding the soil usually reduces hydrated ferric oxides to ferrous hydroxides,
which releases part of the RS-P (Patrick and Mikkelsen, 1971).
Turner and Gilliam (1976) illustrated that flooding the soil significantly
influenced P availability on calcareous soils where Ca-P predominates over Fe-P.
They found that increased P availability in flooded or saturated soils occurred, at
least partially, from the decreased tortuosity which increased P diffusion rate by
10- to 100-fold. However, Sah and Mikkelsen (1986a) observed the Ca-P fraction
of some, but not all, clay soils in California increased for several weeks after
flooding from the formation and precipitation of insoluble Ca-P, which would
decrease the availability of P after flooding.
An interesting aspect of the P chemistry of flooded soils is that more P is
released from the soil into the solution under reduced conditions than under
oxidized conditions if the soil solution is initially low in phosphorus, but re-
duced soils also have a greater sorption capacity for P (Patrick and Reddy, 1978).
Khalid et al. (1979) and Roy and De Datta (1985) suggested that rice required
0.12 – 0.20 mg P L21 in the soil solution for optimum growth. Hossner et al.
(1973) concluded that the minimum soil solution P concentration required to
produce 90% rice yield was . 0.10 mg P L21.
Several forms of soil Fe are highly correlated with P sorption (Willett and
Higgins, 1978; Khalid et al., 1979). Soil P released under reduced conditions has
been related to oxalate-extractable Fe (Khalid et al., 1979; Shahandeh et al.,
1994). Fox and Kamprath (1970) and Evans and Smillie (1976) reported that soil
clay and available Fe content strongly influenced P sorption under aerobic soil
conditions. Ammonium oxalate extractable Fe was highly correlated with P
sorption under anaerobic conditions (Willett and Higgins 1978; Khalid et al.,
1979; Shahandeh et al., 1994). Shahandeh et al. (1994) reported that 84% of the
added P was sorbed under anaerobic conditions when ammonium oxalate extract-
able Fe was , 3000 mg Fe kg21. Oxalate extracts poorly crystalline Fe oxides
(Campbell and Schwertmann, 1984), which are the most reactive Fe oxides in the
soil because of their small size and high surface area (Shahandeh et al., 1994).
Although our understanding of P availability to rice grown under flooded soil
conditions has increased during the past 40 years, this remains an area that
requires additional research. The general relationships of P availability are well
characterized, but need to be better defined in regards to specific soil physical and
chemical properties and correlated to plant uptake of P.

2. Functions and Deficiency Symptoms

Phosphorus is one of the major essential nutrients needed for the growth and
reproduction of higher plants. Phosphorus is required for the synthesis of
phospholipids, nucleotides, adenosine triphosphate (ATP), glycophosphates,
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 93

and other phosphate esters. Plant growth and yield are dramatically reduced by P
deficiency because P is a component of high-energy compounds like ATP and is
an essential component of the genetic material required for seed production.
Phosphorus deficiency symptoms occur primarily on seedling rice at the onset of
tillering when rice begins to rapidly accumulate dry matter. Symptoms include
severe stunting with plants having erect and dark green leaves. Phosphorus
deficiency reduces seedling height, tiller number, stem diameter, leaf size, and
leaf duration. When P is deficient, cell and leaf expansions are retarded more than
chlorophyll formation. Thus, the chlorophyll content per unit leaf area increases,
but the photosynthetic efficiency per unit of chlorophyll decreases (Marschner,
1995). Phosphorus is not a constituent of chlorophyll; hence, the concentration of
chlorophyll of P deficient rice becomes comparatively high and the leaf color
changes from green to dark green. If a P deficiency persists, the older leaves may
turn an orangish color and desiccate from the leaf tip back towards the base. Rice
maturity can be delayed by as much as 10 – 12 days by P deficiency (Fageria,
1980). Phosphorus deficiency symptoms of many crops include a reddish or
purple tint on leaves due to the accumulation of anthocyanins (Hewitt, 1963).
However, the leaf purpling symptom has not been observed in P deficient rice
(Fageria and Barbosa Filho, 1994; Fageria and Gheyi, 1999). Phosphorus
deficient plants are more susceptible to some rice diseases. In Arkansas,
excessive brown spot (Bipolaris oryzae) is commonly observed on nutritionally
stressed rice. Phosphorus fertilization has also significantly increased rice root
growth, the number of panicles, and grain weight of rice grown on P deficient
soils (Fageria and Gheyi, 1999). When P is deficient, rice does not respond to the
application of N, K, or other fertilizers. Color photographs depicting typical P
deficiency symptoms of rice are available in various publications (Wallace, 1961;
Mueller, 1974; Cheaney and Jennings, 1975; Ishizuka, 1978; Yoshida, 1981;
Fageria, 1984; Bennett, 1993; Fageria and Barbosa Filho, 1994).

3. Critical Level in Plant and Uptake

Phosphorus is a mobile element inside the plant; hence, the P concentration of

individual leaves generally declines as leaf age increases. The top leaves have the
highest P concentration and the bottom leaves have the lowest P concentration,
especially when plant available P is limited (Westfall et al., 1973). Sims and
Place (1968) reported that tissue P concentration varied less across plant
development stages than N concentrations, which is generally true when P is not a
growth-limiting factor. During early vegetative growth, the P concentration in
rice tissue increases as P fertilizer application rates increase (Fig. 4). However,
the difference in tissue P concentrations between P fertilizer rates diminishes as
plant development progresses from the vegetative to reproductive growth stages.
94 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Figure 4 Phosphorus concentration in rice plant shoots at different growth stages in Brazil and
Arkansas, USA.

Whole plant P concentration has been observed to either increase or decrease

between the active tillering stage and panicle initiation depending on the soil pH,
P fertilization, or both. On acid soils (pH 5.0– 6.5) such as those represented
by the data from Brazil (Fageria, unpublished data, 1997) in Fig. 4, tissue P
concentrations tend to be high (2.5 – 4.5 g P kg21) during active tillering
and decrease or remain constant as rice progresses into reproductive growth. On
alkaline soils (pH . 7.0) as represented by data from Arkansas (Slaton,
unpublished data, 2001) in Fig. 4, whole plant P concentrations are low
(1.0 –2.0 g P kg21) at the onset of tillering, increase until panicle initiation, and
then are stable until flowering. These described trends are presumably related to
the availability of P after flooding as affected by P fertilization rate and the
predominate forms of soil P. Nevertheless, tissue P concentrations remain nearly
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 95

constant from panicle initiation until flowering on most soils. In studies

examining a range of P fertilizer rates, differences in tissue P concentrations are
seldom observed during this stage of reproductive growth, although plant P
content may vary if different P fertilizer rates were applied. Because differences
in tissue P concentration are not always evident, a critical P content at specific
rice growth stages, rather than concentration, should perhaps be used to evaluate
the P nutritional status of rice. After flowering the filling rice grain becomes a
strong sink for P and the straw P concentration declines.
A wide variation exists among the critical tissue P concentrations for rice
reported in the literature. Some of the variation can be attributed to differences
among the production practices, cultivars, soils, and environments common to
the various rice producing regions around the world. Other, more obvious,
differences are due to the growth stage, plant part sampled, or the growth
parameter used to establish the critical P concentration. In general, the critical
tissue P concentrations for rice during vegetative growth range from 1.0 to 2.0 g
P kg21. Yoshida (1981) reported that 2.0 g P kg21 in the Y-leaf (most recently
matured leaf) was needed to realize the maximum tillering rate. When leaf blade
P concentration was , 0.3 g P kg21 tillering did not occur. De Datta (1981)
suggested that 1.0 g P kg21 in the rice leaf blades at active tillering was the
critical concentration. Fageria et al. (1997a) reported that adequate P
concentrations in rice shoots ranged from 2.5 to 4.5 g kg21 at 75 days of age
(about panicle initiation). The decrease in P concentration during early plant
development is probably related to the rapid increase in dry matter accumulation,
which dilutes the P in the tissue until the P uptake rate can match dry matter
accumulation.
In general, it appears that Y-leaf or whole plant P concentrations during
vegetative growth . 2.0 g P kg21 are sufficient for optimum rice growth and
yield production, but even tissue P concentrations as low as 1.0 g P kg21 may be
adequate to produce maximum yields provided that P availability and tissue P
concentrations increase over time.
The concentration of P in rice straw during reproductive growth or maturity is
not commonly used to diagnose P deficiencies since P typically limits early
season vegetative growth. Slaton et al. (2001a) showed that flag-leaf P
concentrations at the late-boot stage are generally above 2.0 g P kg21 for high-
yielding rice in Arkansas. Similarly, Dobermann et al. (1998) suggested that the
rice flag-leaf P concentration should be . 1.7 g P kg21 for the production of
7 t ha21. In Asia, straw P concentrations , 0.6 g P kg21 at maturity indicate that
P was deficient (Dobermann et al., 1998). This critical concentration may be
appropriate for rice grown in Asia, but appears too low for rice grown in the USA.
Rough rice (grain þ hulls) P concentration averages about 3.0 g P kg21
(Nelson, 1980), but Dobermann et al. (1998) found that P concentration of rice
grain produced in Asia ranged from 1.5 to 2.5 g P kg21. The P concentration of
rice grain is relatively stable and is likely under genetic control, but varies
96 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

slightly among cultivars, which may partially explain the differences between
grain P concentrations cited by Nelson (1980) and Dobermann et al. (1998).
Batten et al. (2000) found that the P concentrations of brown rice of cultivars
from Australia and other countries averaged 3.3 and 3.5 g P kg21, respectively.
Moreover, they determined that brown rice of the various rice genotypes
analyzed differed in their mineral composition when grown under the same
environment and could often be grouped based on growth characteristics or their
country of origin. Brown rice P concentration was negatively correlated with
grain yield and harvest index, but positively correlated with the concentrations of
some other elements in brown rice. These findings are significant because it
shows the nutrient value of brown rice can be improved simply through breeding.
Total P uptake by lowland rice during the growing season increases with plant
development. The total, aboveground P uptake by a high-yielding rice crop may
approach 60 kg P ha21, but more commonly ranges from 25 to 50 kg P ha21 with
60 – 75% of the plant total P contained in the panicles at maturity. Seasonal P
uptake and dry matter accumulation tend to follow similar patterns. The
accumulation of P is closely related to plant age. Fageria et al. (1997b) showed
that plant age accounted for 70% of the variability in seasonal P accumulation.
On P deficient soils, total P uptake also increases as P fertilizer rate increases
(Table V). In the straw, P accumulates with increasing plant age until flowering.
After flowering, during the ripening period, straw P content decreases as P
is translocated from the straw to the developing grain. A significant portion of P
is either translocated from the root system or absorbed between flowering and

Table V
Phosphorus Accumulation in the Shoot and Grain of the Lowland Rice Cultivar ‘Javaé’ as
Influenced by Plant Age and P Application Rates on a Brazialian Inceptisol
(Fageria unpublished data, 1997)

Days after sowing (growth stage) (kg P ha21)

P Fertilizer rate 18 35 63 81 91 119 119

(kg P ha21) (IT) (AT) (IP) (B) (F) (PM) (grain) Total

0 0.20 0.6 4.6 9.9 12.8 4.5 16.4 20.1

87 0.33 1.1 6.4 13.8 16.5 4.6 19.0 23.6
175 0.43 1.5 7.4 13.5 18.1 6.5 20.1 26.6
262 0.53 2.1 7.8 15.3 22.4 8.7 25.1 33.7
350 0.60 2.3 8.0 14.4 18.4 6.4 25.0 31.4
437 0.61 1.7 7.0 13.0 19.0 7.5 23.1 30.6
Average 0.45 1.5 6.9 13.3 17.9 6.3 21.4 27.8

IT, initiation of tillering; AT, active tillering; IP, initiation of panicle; B, booting; F, flowering, PM,
physiological maturity. Phosphorus rates were broadcast.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 97

maturity to satisfy grain P requirements. About two-thirds of the total plant P at

maturity can be accounted for in the aboveground biomass at flowering (Table
V). The straw P content at maturity is only about one-third of the total P content at
flowering and the grain content at maturity. The average P harvest index [Grain P/
(Straw þ Grain P)] generally ranges from 0.60 to 0.75. The harvest of rice grain
removes a significant portion of the total P taken up during the growing season.
Thus, a major portion of P accumulated by the rice crop cannot be recycled for
use by crops following rice in the rotation even though the rice straw is
incorporated back into the soil.

4. Phosphorus Use Efficiency

Phosphorus use efficiencies of rice are generally higher than use efficiencies
for K and can be higher than those of N, depending on the soil. Phosphorus use
efficiencies calculated based on results obtained in field experiments conducted at
the National Rice and Bean Research Center of EMBRAPA are presented in
Table VI. All of the P use efficiencies for this soil were greater compared to the
same efficiencies for N and K (Baligar and Fageria, 1997). Dobermann et al.
(1998) reported the AE for P ranged from 0 to 114 kg grain kg21 P fertilizer
applied. Singh et al. (2000) reported that APE in lowland rice varied from 235 to
316 kg grain kg21 P. Similarly, Witt et al. (1999) reported an APE value of
385 kg grain kg21 P when all production factors were at normal levels. The
internal use efficiency of P as reported by Dobermann et al. (1998) ranged from
282 to 724 kg grain kg21 total P uptake. Approximately 4 –5 kg P is required to
produce 1000 kg of rough rice grain (Fageria et al., 1997b). The efficiency of P

Table VI
Phosphorus Use Efficiency by Lowland Rice Genotypes

Agronomic efficiency Physiological efficiency Agro-physiological efficiency

(Dkg grain kg21 (Dkg grain D kg21 (D kg grain D kg21 total
Genotype P fertilizer added) P uptake) P uptake)

Aliança 93 360 158

CAN 5751 79 533 226
CAN 6804 67 363 234
CAN 7238 98 407 187
CAN 7268 54 484 127
Metica1 95 491 252
Average 81 440 197

From Baligar and Fageria, 1997. Formulas used to calculate the different P use efficiencies are given in
section III.B.6
98 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

use varies among rice genotypes. Genetic variability among rice genotypes has
been reported by various authors (Ponnamperuma 1976a; Fageria, 1992; Fageria
and Baligar, 1993; Baligar and Fageria, 1997, 1999; Fageria et al., 1997a; Fageria
and Gheyi, 1999).

5. Phosphorus Fertilization Practices

Indiscriminate use of P fertilizer increases the cost of production, may reduce

rice yields on some soils, and can contribute to water quality problems. The
principal phosphate fertilizers commonly used to fertilize rice are the highly
water-soluble fertilizers like single and triple-super phosphates, diammonium
phosphate, and sometimes monoammonium phosphate.
Soils used for lowland rice production commonly have low soil test P values,
but may or may not respond to P fertilization (Shahandeh et al., 1994; Wilson
et al., 1999). Routine soil test methods that extract plant-available nutrients
may not provide a reliable estimate of the P available to lowland rice, but the
use of calibrated soil test data is still the best criteria for making P fertilizer
recommendations for most crops including rice. The anaerobic soil environment
of flood-irrigated rice significantly alters the availability of P and routine soil
test methods may not adequately represent nutrient availability to rice. One of
the most important factors affecting P fertilizer recommendations from
calibration studies is the method used to extract soil P. A number of different
extractants including Bray-1, Bray-2, Mehlich 1, Mehlich 3, and NaHCO3
(Olsens) have been used in different parts of the world to assess the status of
plant-available soil P (Kamprath and Watson, 1980; Sharply et al., 1994). Many
of these extractants tend to under- or over-estimate P availability to upland
plants (Kamprath and Watson, 1980) and their ability to accurately predict the P
fertilizer requirement of flood-irrigated rice are further compromised by the
anaerobic soil conditions used for its production. In the United States, each of
the primary rice producing states use a different extractant to estimate P
fertilizer requirements of rice (Norman et al., 2003). The various extractants are
used because no single extractant has shown a significant advantage for making
P recommendations on flood-irrigated rice, but the extractants have been
calibrated for the soils and upland crops grown in each of the states. Sanyal and
De Datta (1991) suggested that NaHCO3 (Olsen P) is perhaps the best routine
method for predicting rice response to P as it is better correlated with the
extraction of Fe-P.
Fageria et al. (1997b) evaluated the Mehlich 1 soil P test to predict lowland
rice response to P fertilization on a Brazilian Inceptisol. Based on relative grain
yield, ranges of Mehlich 1 soil test P were categorized as very low (0 –2.6 mg
P kg21), low (2.6 – 8.8 mg P kg21), medium (8.8 – 13.0 mg P kg21), or high
(. 13.0 mg P kg21). The amounts of broadcast P needed to increase soil P
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE… 99

concentration at very low, low, medium, and high soil classification categories
were 444, 292, 220, and 80 P ha21, respectively. A medium soil test P was
required for the production of . 95% relative yield. The rate of banded fertilizer
P required to produce maximum yield was 66 kg P ha21 for the very low and low
soil categories, 44 kg P ha21 for the medium category, and 22 kg P ha21 for the
high category.
Fageria et al. (1991) characterized the chemical properties of lowland rice soils
of Brazil and reported a 65-fold range in Mehlich 1 extractable P concentrations in
the 0- to 20-cm soil depth among 23 different municipalities. Similarly, Cochrane
et al. (1984) reported great variation in Bray 2 extractable P of lowland soils of
tropical America. Annual crop yield responses to P fertilization were expected on
40% of the soil samples that contained low (, 3 mg P kg21) to medium (3 – 7 mg
P kg21) concentrations of extractable P. Fageria (1980) also reported P
fertilization significantly increased the yield of lowland rice grown in the Goias
State of Brazil. Maximum grain yields were obtained when P was broadcast
applied at 175 kg P ha21 as triple superphosphate.
Dobermann et al. (2000) reported that at IRRI, 26 kg P ha21 is normally
applied to obtain maximum yield of flooded rice. When P fertilizer is
recommended in the USA, rates of 10– 40 kg P ha21 are normally sufficient to
produce maximum rice yields (Norman et al., 2003). Soils capable of fixing large
quantities of P required 97 (Fageria, 1980) to 175 kg P ha21 (Fageria et al.,
1997b) to produce optimum yields when P fertilizer was broadcast applied,
but only 22 – 44 kg P ha21 when P fertilizer was banded. Use of the most efficient
P application method to reduce the optimum P rate is critical to offset the cost
of P fertilizer and increase P fertilizer use efficiency. The P fertilizer rates
cited by Fageria et al. (1997b) required to produce maximum yields in Brazil
are relatively high and suggest that low soil P content and high P fixation
capacity may be associated with the high P requirement in Brazilian acidic
lowland soils.
The time and placement of fertilizer P is critical for optimum uptake,
especially on P deficient soils. Patrick et al. (1974) showed that P placed with the
rice seed during drilling was superior to broadcast application 2 weeks after
seeding. Rice yield declined as the time of P application was delayed. They also
generalized that broadcast preplant-incorporated P application would be equally
effective as P drilled with the seed. However, broadcast application of P at the
5-leaf stage increased tissue P concentration, P uptake, and grain yield more than
P broadcast applied to the soil surface at seeding on an alkaline soil in Arkansas
(McGee et al., 2002) suggesting that fixation potentially reduces P availability on
some soils in a very short time. The prevention of early season P deficiency is
critical for the production of high yields. When P is deficient, rice yield response
to P fertilizer declines as the time of P fertilization is delayed (Patrick et al., 1974;
Slaton et al., 1998).
100 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

D. POTASSIUM

Rice does not generally respond to K fertilization to the degree noted for either
N or P. Many soils used for the production of continuous rice or rice – wheat
rotations can be cropped for extended periods without needing supplemental K to
maintain crop production (Dobermann et al., 1996b). Rice grown in rotation with
legumes, like soybean, may require annual K inputs due to the greater K
requirement of the legume crop. Although annual K fertilization may not be
required, the aboveground K content of rice is equal to or greater than the plant N
content and greater than all other essential nutrients. Direct K fertilization of rice
has produced grain yield increases ranging from 0 to 47% with yield responses of
0 –10% being the norm (Dobermann et al., 1996b). Rice is highly efficient in
scavenging plant available soil K because of its fibrous root system and the
increase in K availability after flooding. However, on some soils, K deficiency of
rice may occur if rice and rotation crops are grown without regular applications of
K fertilizer to replace the K removed by the harvested crops. Prior to the early
1990s, K deficiencies of rice were rare in the rice producing areas of the USA.
However, K deficiency is now recognized as an annual problem on many
soils as rice and rotation crop yields have increased, soils have been mined of
K, and production practices have changed (Slaton et al., 1995; Williams and
Smith, 2001).
Many of the soils used for rice production throughout the world have low
cation exchange capacities or are highly weathered which makes them poor
reservoirs of plant available K. In highly weathered soils, the total soil K content
may be quite low because of K deficient parent materials and the climate. The
high rainfall and warm temperatures common to the tropical rice growing areas
have hastened the release and leaching of soil K over time (Tisdale et al., 1985).
In Arkansas, approximately 20% of the soil samples analyzed for rice or
irrigated-soybean, the primary crop rotated with rice, have soil K concentrations
, 200 kg Mehlich 3 K ha21 (DeLong et al., 2001). The chemistry of soil and
fertilizer reactions involving K is less complicated than the chemical and
biological reactions that occur when N and P fertilizers are added to the soil.
Potassium is not complexed into soil organic matter, susceptible to gaseous
losses, or subject to precipitate into forms that are not plant available. Essentially
all the soil K is associated with the soil mineral fraction (Foth and Ellis, 1988).
The availability of soil K is influenced by soil K concentration, soil texture, soil
pH, cation exchange capacity, temperature, soil moisture, soil aeration, yield
level, and root growth patterns. Unlike, N, P, and Zn, relatively few studies have
been published regarding efficient K fertilization practices, diagnosis of K
deficiency, and correction of K deficiency. Since De Datta and Mikkelsen (1985)
reviewed K nutrition of rice, research has begun to increase our knowledge of the
K nutritional requirements of rice.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…101

1. Chemistry in the Soil

The soil minerals that are considered the most important K sources in the soil
are the primary aluminosilicates that include K feldspars and micas and
secondary aluminosilicates like illite. Potassium is released by the weathering of
these minerals. The rate of weathering or K release depends on the mineral
properties and predominate environmental conditions (Tisdale et al., 1985). Soil
K can be divided into four fundamental categories including: (1) 0.1 –0.2% as
soil solution K, (2) 1– 2% as exchangeable K, (3) 1 –10% as nonexchangeable K,
and (4) 90 –98% as mineral K (Barber, 1995; Brady and Weil, 1996). Potassium
ions move from one category to another whenever the removal or addition of K
disturbs the equilibrium between these soil K pools. Equilibration between the
soil solution and exchangeable K pools is rapid and usually complete within
hours. However, the equilibration time between the nonexchangeable and
exchangeable K is much slower, requiring days or even months. The conversion
of K from the mineral form via weathering is extremely slow and varies among
the soil K minerals. The weathering process is extremely slow and has little
significance in supplying plants with K during a single season (Barber, 1995).
Most of the K used by plants during a single season comes from the soil solution
and exchangeable K pools and is supplemented by K released from the
nonexchangeable pool. The quantity of nonexchangeable Kþ in soils depends on
clay content and type of clay minerals (Sparks and Huang, 1985).
Potassium is considered relatively immobile in the soil and moves primarily
by diffusion in the soil – plant system, especially for upland crops. Teo et al.
(1995) found that diffusion, mass flow, and contact exchange accounted for
, 57.8, 42, and , 0.3%, respectively, of K uptake by flooded rice. Although
considered immobile, a significant amount of K can be lost via leaching on some
soils following displacement from the exchange complex after flooding.
Leaching is a significant problem in the humid tropical regions having acid
soils with low cation exchange capacities. Liming an acid soil to raise its pH can
reduce leaching losses of K because of the complementary ion effect and
increasing the soil CEC (Brady and Weil, 1996). Significant vertical K
movement, via leaching, from the 0– 20 cm soil depth to lower depths (20 –40,
40– 60 and 60– 80 cm) has been reported after harvesting only four lowland rice
crops on a Brazilian Inceptisol (Fageria et al., 1990c).
The soil solution concentration of K increases after flooding. Soil reduction
increases the soil solution concentrations of Fe, Mn and other soil cations, which
displace K from the cation exchange sites into the soil solution. The result is an
increase of K in the soil solution, where K can either be absorbed by rice plants
or, on permeable soils, leached to depths below the rice root system (Patrick et al.,
1986; Wells et al., 1993). The duration of increased availability of K has not been
adequately characterized across an array of soils, but is believed to be relatively
short lived. Preliminary data from Arkansas suggests that soil solution K
102 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

concentrations of a DeWitt silt loam (fine, smectitic, thermic, Typic Albaqualfs)

cropped to rice are highest within days after flooding and rapidly decline for 4– 5
weeks, due presumably to plant uptake, until an equilibrium is reached and
maintained until the flood is drained for harvest (Slaton, unpublished data, 2001).
The initial increase in soil solution is considered an advantage for plant uptake,
but also a disadvantage if the K is leached. Sandy soils high in organic matter and
reducible Fe and Mn generally have higher solution K concentrations than soils
with higher clay contents (Yoshida, 1981). The leaching loss of significant
amounts of soil K could result in K deficiency later in the growing season if
adequate K is not absorbed during vegetative growth or K is not supplemented
shortly before the onset of reproductive growth.

2. Functions and Deficiency Symptoms

Potassium deficiency symptoms initially appear on the lower, oldest rice

leaves because K is highly mobile in the plant. The onset of K deficiency of rice is
difficult to diagnose because the only initial difference between K sufficient and
deficient rice is the color of the lower leaves. This deficiency symptom can easily
be confused for N deficiency. Potassium deficiency symptoms include stunted
plants with little or no reduction in tillering, droopy and dark green upper leaves,
and chlorosis of the interveinal areas and margins of the lower leaves starting at
the leaf tip. Leaf tips will eventually die and turn brown with the progression of
severe K deficiency. The droopy leaves associated with K deficiency are not
always noted on K deficient rice because Na may substitute for K in osmotic
regulation functions. Potassium deficiency reduces grain size and weight
resulting in a direct yield loss. In the USA, K deficiency symptoms are seldom
observed during vegetative growth. More commonly, K deficiency symptoms are
observed during early reproductive growth beginning at panicle initiation.
Potassium deficient rice commonly has high levels of several diseases that
infest the leaves, stems, and panicles (Slaton et al., 1995). As K deficiency
progresses, rice usually develops severe disease infestation due to the plants’
reduced ability to resist infestation. Diseases that are normally insignificant
problems, such as brown leaf spot and stem rot (Sclerotium oryzae), may become
severe, in addition to common diseases such as rice blast (Pyricularia grisea).
The lodging associated with K deficient rice may be related to the increased
incidence and severity of stem diseases. Potassium is known to play an important
role in the lignification of vascular bundles, a factor that contributes to the higher
susceptibility to lodging and disease of K deficient plants. The yield loss of
K deficient rice may actually be a combination of losses from insufficient K
nutrition and losses caused directly by diseases that infest K deficient rice.
Limited research suggests that yields of rice low or deficient in K do not respond
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…103

to K fertilizer applications made after the panicle differentiation growth stage

(Slaton et al., 2001b). Thus, when K fertilizer is required for maximum yield
production it must be applied during vegetative growth. Research has not been
conducted to determine if a portion of the yield loss attributed to K deficiency can
be prevented or recovered by controlling or suppressing the diseases with timely
fungicide applications. Additional research efforts are needed to verify whether
application of K fertilizer to K deficient rice during the reproductive growth stage
is beneficial. Color photographs of K deficiency symptoms of rice are available in
various publications (Wallace, 1961; Mueller, 1974; Cheaney and Jennings,
1975; Ishizuka, 1978; Yoshida, 1981; Fageria, 1984; Bennett, 1993; Fageria and
Barbosa Filho, 1994; Wilson et al., 2001).

3. Critical Level in Plant and Uptake

The concentration and content of nutrients in plants are important parameters

used to evaluate the nutritional status of a crop. Knowledge of tissue nutrient
concentrations at specific crop growth stages allow for the diagnosis and
correction of nutritional disorders. The K nutritional status of plants is easily
determined by routine analysis of plant tissues. However, the critical K
concentrations of rice are not particularly well defined. Potassium has not
historically been a major yield-limiting nutrient in many rice-producing areas and
research efforts have been focused on other nutrients like N, P, and Zn. Thus, a
somewhat limited database concerning K nutrition of rice is available in the
literature.
As grain yield increases the demand for plant nutrients, particularly K, also
increases (Fageria et al., 1997a). The K concentration of whole rice plants
decreases with plant age (Sims and Place, 1968). The critical nutrient
concentration range required for optimum yields also decreases with plant age.
Tanaka et al. (1977) compared the physiological response of 17 seedling crop
species, including rice, to K nutrition. Rice was characterized as having a high
capacity to absorb and deplete K in the growth medium. Fageria et al. (1997a)
reported that tissue concentrations ranging from 15 to 40 g K kg21 in whole,
aboveground plant samples at 75 days of age (about panicle initiation) were
adequate for rice. During the vegetative growth phase, tillering stops when the K
concentration in the leaf blade is , 5.0 g K kg21 (Yoshida, 1981). Kiuchi and
Ishizaka (1961) reported that mature leaves should contain 20 g K kg21 at the
booting stage to maximize the number of grains per panicle and decrease spikelet
sterility. Fageria (1986) found K concentrations of 17 g K kg21 in the rice straw
and 2.6 g K kg21 in the rough rice grain were sufficient at maturity. Straw K
concentrations , 10 g K kg21 at maturity certainly indicate a K deficiency (De
Datta, 1981). Increasing the K concentration of rice stems and leaves is easily
104 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

done through the application of increasing rates of K fertilizer. Although luxury

uptake of K may occur in rice vegetation, the K concentration of rice seed
remains relatively constant regardless of K fertilization. The K concentration of
rice seed ranges from 2.5 to 3.0 g K kg21, which is roughly equal to the grain P
concentration (Fageria, 1986; Dobermann et al., 1998; Batten et al., 2000).
Dobermann et al. (1996b) showed that K concentration of rice grain was stable
and varied by less than 0.3 g K kg21 among K fertilizer treatments. Differences in
K nutrition or availability are exhibited by a wide range of K concentrations in
the rice straw, by reduced grain yields (grain K content), or both.
Rice absorbs the majority of its K during the vegetative and early reproductive
growth stages. Hirata (1995) reported that 75% of the total K uptake at maturity is
absorbed prior to the booting stage and almost no absorption occurs between
flowering and maturity. While it is generally agreed that the majority of K uptake
occurs before flowering, some K is certainly absorbed during the ripening period.
A major portion of the K absorbed before anthesis remains in the stems and
leaves (Hirata, 1995). Approximately 80 –90% of the aboveground K content of
rice is found in the leaves and stems at maturity (Ishizuka and Tanaka, 1952a;
Dobermann et al., 1996a; Wilson et al., 2001). The percentage of total K removed
in harvested grain declines when luxury consumption occurs (Fageria, 1986). The
grain concentrations of P and K are nearly equal in rice seeds, but, in contrast to
P, a comparatively small amount of the total plant K content is translocated and
stored in the grain. A single lowland rice crop may contain between 200 and
300 kg K ha21 at maturity (De Datta and Mikkelsen, 1985). If the rice straw is not
physically removed from the field, the majority of K is recycled back into the soil.
If the rice straw is removed K fertilization practices must be altered to prevent
from depleting the soil K. Rice hulls, which are separated from the grain during
the milling process, comprise about 20% (by weight) of the harvested rough
rice and have near equal K concentrations (, 2.8 g K kg21) as the brown
rice. Most research suggests that approximately 40 kg K (straw plus grain
content) is required to produce 1 t of rice grain (Fageria and Baligar, 1996;
Fageria et al., 1997b).

4. Potassium Use Efficiency

Nutrient absorption and utilization capacity of a crop variety are important

aspects for the improvement of crop yield and the reduction of production
costs. Dobermann and Fairhurst (2000) suggested that the recovery efficiency
of applied K fertilizer is about 50%, but the uptake efficiency is usually lower
when all K is basal applied or higher when K fertilizer is topdressed in two or
more split applications. Baligar and Fageria (1997) found that the recovery
efficiency of K fertilizer ranged from 37 to 58% and varied among lowland
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…105

rice genotypes (Table VII). The average fertilizer recovery for K is similar to
N (39%, Table III) and typically higher than for P (, 10%, De Datta, 1981)
although higher P fertilizer recovery efficiencies have been reported
(Baligar et al., 2001).
Rice yield responses to K fertilization are generally lower than observed for N
and P, especially on responsive soils. As such, the AE is much lower for K
compared with N and P. Dobermann et al. (1998) found the AE for K in five
Asian countries ranged from 0 to 26 kg grain kg21 K fertilizer applied; however,
AE values as high as 64 kg grain kg21 K fertilizer applied were reported by
Baligar et al. (2001). In comparison, the AE for P ranged from 0 to 114 kg
grain kg21 P fertilizer applied (Dobermann et al., 1998). The AE is dependent on
the K fertility status of the soil and encompasses a wide range of values across the
rice producing regions of the world. The internal use efficiency of K as reported
by Dobermann et al. (1998) ranged from 41 to 89 kg grain kg21 total K uptake
and is less variable among rice producing regions because it reflects the relative
grain production per unit of total K uptake rather than per unit of fertilizer K
absorbed (Dobermann et al., 1998). The K uptake requirements for rice yielding
between 4 and 8 t ha21 ranged from 17 to 30 kg K t21 grain produced
(Dobermann et al., 1996b). Rice genotypes also differ in their K use efficiencies
(Baligar et al., 2001). Fageria (unpublished data, 2001) found the APE among 14
rice genotypes ranged from 3.1 to 37.7 mg grain mg21 fertilizer K uptake. This
indicates that some rice genotypes have higher K requirements than others and
the possibility of developing rice cultivars that are adapted to soils with low K
availability exists.

Table VII
Potassium Use Efficiency Among Lowland Rice Genotypes

Agronomic Physiological Agro-physiological Recovery

efficiency efficiency efficiency efficiency
(Dkg grain kg21 (Dkg grain Dkg21 (Dkg grain Dkg21 (% fertilizer
Genotype K fertilizer added) K uptake) total K uptake) uptake)

Aliança 76 89 39 81
CAN 5751 64 119 51 58
CAN 6804 54 89 58 51
CAN 7238 80 100 48 75
CAN 7268 44 101 26 73
Metica 1 78 89 37 61
Average 66 98 43 67

From Baligar and Fageria, 1997. Formulas used to calculate the listed efficiencies are given in the
section III.B.6.
106 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

5. Potassium Fertilization Practices

The importance of K fertilization is increasing for lowland rice due to the

higher nutrient requirements of high-yielding modern cultivars, increased crop
removal of K by intensive agricultural production, and the role of K in the control
of diseases. Published data concerning rice response to K fertilization for lowland
rice is scarce compared to that of other macronutrients. Dobermann et al. (1996b)
reasoned that infrequent, significant yield rice yield responses to K fertilization
were due to high seasonal K inputs (7 –60 kg K ha21 yr21) in irrigation waters,
the release of nonexchangeable K, other nutritional limitations, or combinations
of these factors.
Fertilizer recommendations for immobile nutrients, like K, are made based on
soil test calibration studies, which need to be conducted on a routine basis.
Dobermann et al. (1996b) summarized critical soil concentrations of 1N
NH4OAc extractable K from around the world and found the range varied from
0.08 to 0.41 cmol K kg21 soil. In the USA, K fertilizer is usually recommended
for rice when exchangeable soil K is , 60 – 100 mg K kg21, regardless of the soil
texture or extractant. For example, in California, K fertilizer is recommended
when NH4OAc extractable K is , 60 mg K kg21 (Williams and Smith, 2001).
Fageria et al. (1990a) found that annual K fertilizer applications significantly
increased rice yields during the first and subsequent rice crops seeded on the same
plots on an Inceptisol in central Brazil. Rice did not respond to K fertilization
when soil test concentrations were . 50 mg K dm23 (Mehlich 1 extractable K;
Fageria, 1999b). Dobermann et al. (1996b) found that mixed-bed exchange resins
incubated for 2 weeks under flooded soil conditions were superior to K extracted
by 1N NH4OAc for prediction of K uptake by rice. They reasoned that resin
adsorbed K accounted for K diffusion rates, concentration of other cations that
influence K uptake by rice, and the K fertilization history of soils that resulted in
accumulation or depletion of rapid and slow released pools of soil K after
flooding. Dobermann et al. (2000) reported that at IRRI 50 kg K ha21 is normally
applied in field experiments to obtain maximum yields of flooded rice and is
representative of K fertilizer rates used to fertilize rice in other parts of the world.
The timing of K fertilizer application should consider two criteria: (1) the cost
of application and (2) maximizing fertilizer use efficiency by the crop.
Sometimes, K is applied in a band at the time of sowing to increase its
availability to seedling rice. In the USA and many other rice producing regions of
the world, K fertilizer is broadcast applied immediately before seeding, after
seeding, or split into multiple applications. In the humid tropical soils with low
cation exchange capacity and clay content the possibility of K loss via leaching
exists and K fertilizer is commonly broadcast applied as a topdressing, along with
N. Fageria (1991) reported that lowland rice yields were higher when the total K
fertilizer requirement was applied in split topdressed applications. Noguchi and
Sugawara (1976), Su (1976), and Santos et al. (1999), also reported the benefits
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…107

of split applications of K. Potassium application rates applied as topdressing

depend on the initial soil K concentration with higher K rates required as the
initial soil K concentration decreases. Preliminary data reported by Slaton et al.
(2001b) suggests that a single application of K fertilizer applied preplant,
preflood (5-leaf stage), or at the panicle initiation stage were sufficient to
maximize grain yields on a Calloway silt loam (Fine-silty, mixed, active, thermic
Glossaquic Fragiudalfs) in Arkansas. However, K fertilizer applied during the
boot stage did not increase yields above the untreated control. In general, a major
portion, and sometimes all, of the K fertilizer should be applied at or near the time
of seeding. A smaller portion of the total K fertilizer requirement should be
topdressed, along with N, to reduce the cost of application, on soils where
leaching losses of K are of concern.
Potassium chloride is the most common source of K for most annual crops,
including rice. Potassium chloride is highly effective and has the lowest cost per
unit of K. Potassium sulfate is equally effective as a K source and furnishes S, but
is also more expensive than KCl. Field experiments have shown little or no
difference in rice response to KCl, K2SO4, and KNO3 fertilizers in Arkansas
(Wilson et al., 1996).

6. Use of Efficient Genotypes

The identification and selection of nutrient efficient cultivars is considered one

of the most cost effective approaches for improving crop production in resource-
poor environments. Very little research has been conducted on finding genotypes
that exhibit high nutrient use efficiencies for K or other nutrients in the highly
mechanized agricultural producing areas. Increased fertilizer prices, depletion of
fertilizer resources, and the cultivation of marginal lands are excellent reasons for
finding crop genotypes with superior nutrient utilization efficiencies. Genetic
differences in K uptake and utilization efficiency among cultivars within crop
species, including rice, have been reported in several studies (Glass et al., 1981;
Siddiqi and Glass, 1981; Clark, 1990; Clark and Duncan, 1991; Fageria et al.,
1997a). Plants have different uptake potentials for nonexchangeable Kþ due to
different root morphologies (Mengel, 1985).

E. CALCIUM, MAGNESIUM, AND SULFUR

Calcium, Mg, and S are often referred to as the secondary elements. These
macronutrients are required in relatively large amounts for normal crop growth.
Rice tissue macronutrient concentrations generally follow the order, from high to
low, of N ¼ K . Ca . P . Mg . S. Calcium is usually the predominant soil
cation and is present in relatively large amounts especially on alkaline soils.
108 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Further, Ca is a nontoxic mineral nutrient, even in high concentrations, and is

very effective in detoxifying high concentrations of other mineral elements in
plants (Marschner, 1995). Calcium and Mg are much more stable in the soil than
S. Highly weathered Oxisols and Ultisols may be low in Ca, Mg, and plant
available S due to excessive leaching. Sandy soils may also contain low levels of
Ca, Mg, and S increasing the likelihood of crop deficiencies of these nutrients.
However, rice seldom suffers from Ca and Mg deficiencies. In comparison, S
deficiency has been reported from nearly all rice producing regions of the world
including Indonesia, Brazil, India, Bangladesh, Thailand, and the USA (De Datta,
1981; Wells et al., 1993). Yamaguchi (1997) reported that symptoms associated
with S deficiency of rice often occur in irrigated Vertisols of the lower Volta in
Ghana, Africa. Sulfur deficiency was a possible cause of these symptoms because
farmers applied urea and high-analysis NPK fertilizers that contained little S.
Blair et al. (1978) suggested that the low S content of most tropical soils was the
primary cause of S deficiency. Sulfur deficiency of rice has increased for
numerous reasons including: (1) increased crop removal of S via increased crop
yields; (2) use of fertilizers lacking S; (3) the reduced industrial emissions of S
lowering the input of atmospheric S; (4) reduction in soil organic matter; (5)
leaching and weathering processes; (6) erosion; and (7) crop management
practices (De Datta, 1981).

1. Chemistry in the Soil

Calcium and Mg are alkaline earth cations. Calcium is the dominant

exchangeable cation in most soils. Calcium and Mg concentrations in the soil
solution are in equilibrium with the exchangeable forms of these cations. The
degree of Ca and Mg saturation of the cation exchange sites, complimentary
cations, the nature of the bonding with exchange sites, and the concentration of
anions in the solution all interact to determine the Ca and Mg soil solution
concentrations (Barber, 1995). Exchangeable Ca and Mg usually account for
more than 60% of the exchangeable cations on soils at pH 5.5 or higher.
Exchangeable Al3þ, Hþ, Kþ, and Naþ occupy the majority of the remaining
exchange sites (Barber, 1995). Calcium (Ca2þ), Mg2þ, and SO22 4 ions are the
main forms absorbed by plants and mass flow is the principal mechanism by
which these nutrients move to plant roots. The availability and absorption of
Ca2þ, Mg2þ, and SO22 4 are governed by the soil pH, ion concentrations in the soil
solution, cation exchange capacity, organic matter content, type of soil colloid,
the ratio of these ions to other cations or anions, and many plant factors.
More than 90% of the total S in the A horizon of soils exists in the organic
form. The N:S ratio in surface soils is relatively constant and averages about
10:1.3 (Foth and Ellis, 1988). Plant available S in the rice root zone is quickly
depleted in many soils by plant uptake, SO4 leaching, and the reduction of SO4 to
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…109

sulfide (S22). Consequently, the soil SO4 concentration declines after flooding,
but may be accompanied by the accumulation of S22, which can be toxic to plants
and may also be lost from the soil as H2S gas. Thus, the availability of soil S
decreases as soil reduction proceeds. The rate of SO4 reduction in submerged
soils depends on a number of soil properties. In neutral and alkaline soils, SO4
21
concentrations as high as 15,000 mg SO22 4 kg may be reduced to zero within 6
weeks of submergence (Ponnamperuma, 1972). The reduction of SO4 begins at
an Eh of 2 0.15 to 2 0.2 V at pH 6.5– 7.0 (Takai, 1978).
Calcium and Mg concentrations in the soil solution are altered by the chemical
changes associated with flooding, but their ionic forms are not affected. In
contrast, the ionic forms of S undergo marked changes following flooding
(Patrick and Mikkelsen, 1971). Flooded soils frequently become sufficiently
reduced from a restricted oxygen supply and microbial activity to reduce SO22 4 to
S22. Since Fe3þ reduction to Fe2þ precedes SO22 4 reduction, Fe
2þ
is present in
the soil solution by the time S22 is produced. The formation of insoluble FeS may
prevent the formation of H2S and protect microorganisms and aquatic plants from
the toxic effects of H2S gas (Patrick and Reddy, 1978).

2. Functions and Deficiency Symptoms

Calcium accelerates the translocation of photosynthetic products in rice plants

(Kawasaki, 1995). Kawasaki (1995) reported that Ca stimulated the absorption of
P and K under certain ion concentration ranges in nutrient solutions. Among the
various physiological functions performed by Ca in higher plants, the most
important role is believed to be the maintenance of structure and the function of
biomembranes. Injury due to heavy metals is reduced by Ca because it lowers
heavy metal concentrations in rice plants (Kawasaki, 1995). Calcium is also a
promoter of normal root growth and development. In contrast to Mg2þ, which is a
strong activator of enzymes, Ca2þ increases the activity of only a few enzymes.
Hanson (1984) and Kirkby and Pilbeam (1984) presented comprehensive reviews
on the functions of Ca as a plant nutrient.
It is generally recognized that the most important physiological roles of Mg is
as a constituent of chlorophyll and an activator of numerous enzymes. A balance
between Ca and Mg concentrations within the plant is necessary for the
maintenance of normal metabolic processes. A high proportion of the total plant
Mg is involved in the regulation of cellular pH and the cation – anion balance.
Magnesium also functions as a bridging element for the aggregation of ribosome
subunits, which is a process necessary for protein synthesis.
Sulfur is an essential nutrient for all plants and animals because it is a
constituent of essential amino acids (e.g., cysteine and methionine), several
coenzymes (e.g., biotin, coenzyme A, thiamin pyrophosphate and lipoic acid),
thioredoxins, and sulpholipids (Zhao et al., 1997). There are many other
110 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

S-containing compounds in plants that are not essential, but may be involved in
defense mechanisms against pests and pathogens or contribute to the special taste
and odor of some plants (Bennett and Wallsgrove, 1994). Apart from the effects
on yield, the S nutrition of a crop often has a strong influence on food quality
because of its essential role in the synthesis of amino acids, proteins, and some
secondary metabolites (Zhao et al., 1997). A sufficient S supply to crops is also
important for the nutritional quality of legumes and the processing quality of
cereal grains (Randall and Wrigley, 1986).
Calcium deficiency in plants is typically expressed as cell death due to the
breakdown in critical functions such as selective membrane permeability and
dysfunction of cellular signaling mechanisms (O’Brien and Ferguson, 1997). In
Ca deficient rice the growing point becomes white and upper leaves roll and curl.
Magnesium deficiency of rice is characterized as pale green plants with
droopy and wavy leaves. Magnesium is relatively mobile in the plant resulting in
an interveinal chlorosis of the older leaves at the onset of deficiency. The
interveinal chlorosis gives the lower leaves an orangish-yellow color. Tillering
and plant height of Mg deficient rice are almost normal. The yellowing of the
outer leaf edge of Mg deficient plants is a characteristic difference between K and
Mg deficiency symptoms (Fageria and Barbosa Filho, 1994).
Sulfur deficiency symptoms are very similar to those described for
N. However, S has limited mobility in the plant and produces a relatively
uniform chlorosis of the plant. The general location of the chlorosis can be used
to distinguish between N and S deficiencies, especially during the early
development of symptoms. Sulfur deficiency is initially expressed as a chlorosis
of the younger leaves while N deficiency results in chlorotic older leaves.
Additionally, prolonged N deficiency results in a premature necrosis of the older
leaves, which is not characteristic of S (Wells et al., 1993). Sulfur deficient
seedlings are yellow to pale green. Sulfur deficiency largely affects the growth of
leaf blades as the reduction in the dry weight of leaf blades is larger than in stems
and roots. With S deficiency, the chlorophyll content of leaves decreases and
lowers the photosynthetic rate (Suzuki, 1995). In rice, severe S deficiency may
reduce the number of panicles, panicle length, and the number of spikelets per
panicle. In Arkansas, late-season S deficiency has recently been observed on a
number of fields with sandy loam, sandy clay, and clay soil textures (Slaton et al.,
2001a). These late-season symptoms generally appear on the top two or three
leaves as an interveinal chlorosis that begins near the leaf tips and proceeds
towards the leaf base shortly before the panicles exert from the boot.

3. Critical Level in Plant and Uptake

Rice plants generally have lower concentrations of Ca and Mg than other

crops, especially dicots (Kawasaki, 1995). Calcium and Mg concentrations in rice
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…111

tend to decrease during vegetative growth and then stabilize, near the time of
panicle initiation, for the remainder of the season (Table VIII). Calcium
concentration in the shoot is typically higher than that of Mg. However, Mg
concentration is generally higher in the rice grain than Ca presumably because it
is more mobile. Fageria et al. (1997a) reported that adequate concentration
ranges of Ca and Mg were 2.5 – 4.0 g Ca kg21 and 1.7 –3.0 g Mg kg21,
respectively, in the shoots of rice at 100 days of growth (about booting). De
Datta (1981) reported the critical concentrations of 1.5 g Ca kg21 and 1.0 g
Mg kg21 in rice straw at maturity.
Knowledge of nutrient accumulation by a crop during its growth cycle is
important to understand the nutrient requirements of the crop and know how
much of a nutrient is taken up and removed from the soil. Such information is
useful in maintaining the soil fertility at an adequate level for essential elements.
Data listed in Table IX shows the seasonal accumulation of Ca and Mg by
lowland rice at several different growth stages. Calcium accumulation was higher
than that of Mg in the straw, but the opposite was true for the grain. In this study,
lowland rice (straw þ grain) accumulated 6 kg Ca and 4 kg Mg t21 of grain
produced, but harvested grain removed nearly double the amount Mg (Table IX).
The critical S concentration in rice tissue, like that of N, varies with the stage
of plant development and part of the plant that is sampled. Wells et al. (1993)
reported that the critical concentration of S varies from approximately 2.5 g
S kg21 at tillering to 1.0 g S kg21 at heading. Yoshida (1981) reported that the
critical S concentrations in straw needed for maximum dry weight production

Table VIII
Whole Plant Concentrations of Ca and Mg in the Shoots and Grain of the Flood-Irrigated
Rice from Brazil (Fageria, unpublished 2001) and Arkansas (Slaton, unpublished data, 2001) at
Different Growth Stages

Brazila Arkansasb

Calcium Magnesium Calcium Magnesium

Growth Stage DAEc (g Ca kg21) (g Mg kg21) DAEc (g Ca kg21) (g Mg kg21)

Beginning tillering 22 4.0 2.7 – – –

Active tillering 35 4.0 2.3 55 8.3 3.0
Panicle initiation 71 3.2 1.8 81 5.8 3.0
Boot stage 97 3.0 1.9 96 5.5 3.0
Flowering 112 2.9 1.8 111 5.1 2.8
Mature-straw 140 3.3 1.8 145 5.3 2.6
Grain 140 0.8 1.2 146 0.24d 1.1d
a
The rice cultivar was ‘Metica 1’ and was grown on an Inceptisol.
b
In Arkansas, the rice cultivar was ‘Wells’ and was grown on a slighty calcareous Calloway silt loam.
c
DAE, days after emergence.
d
Harvested grain. Whole panicle concentration at maturity was 0.9 g Ca kg21 and 1.7 g g kg21.
 112
N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR
Table IX
Whole Plant Contents of Ca and Mg in the Shoots and Grain of the Flood-Irrigated Rice from Brazil (Fageria, unpublished data, 2001) and Arkansas
(Slaton, unpublished data, 2001) at Different Growth Stages

Brazila Arkansasb

Growth stage Calcium Magnesium Calcium Magnesium

DAEc (kg Ca ha21) (kg Mg ha21) DAEc (kg Ca ha21) (kg Mg ha21)

Beginning tillering 22 1.4 0.8 – – –

Active tillering 35 4.6 2.6 55 4.8 1.8
Panicle initiation 71 18.4 10.1 81 24.6 12.7
Boot stage 97 30.6 19.5 96 48.0 26.4
Flowering 112 37.3 24.1 111 60.3 33.1
Mature-straw 140 31.3 16.3 145 52.4 25.7
Grain 140 4.8 7.7 146 1.9d 8.6d

Total uptake/t grain – 5.7 3.8 – 6.7 4.2

Grain content, % of total – 13.3 32.1 – 3.5 25.1
a 21
The rice cultivar was ‘Metica 1’ and was grown on an Inceptisol. The total, aboveground dry matter produced was 15,647 kg ha with a grain yield of
6323 kg ha21.
b
In Arkansas, the rice cultivar was ‘Wells’ and was grown on a slightly calcareous Calloway silt loam. The total, aboveground dry matter produced was
17,640 kg ha21 with a grain yield of 8085 kg ha21.
c
DAE, days after emergence.
d
Ca and Mg concentrations in harvested rough rice grain.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…113

varied from 1.6 g S kg21 at tillering to 0.7 g S kg21 at maturity. The critical N:S
ratio in straw for maximum biomass production varies from 23 at active tillering
to 13 at maturity. Fageria et al. (1997a) reported that adequate concentrations of S
in the uppermost mature leaves at tillering were 2.0– 6.0 g S kg21. Suzuki (1995)
reported 1.0 g S kg21 as a critical level in the rice shoot at tillering and 0.55 g
S kg21 in rough rice grains. Wang (1976) concluded that the critical
concentration of S in rice straw should be 0.5 g S kg21 for optimum grain
yield. Slaton et al. (2001a) observed late-season S deficiency symptoms when
rice flag leaves, immediately before panicle emergence from the boot, contained
, 1.5 g S kg21. Rice grain S concentrations vary between 0.34 g S kg21 for S
deficient plants to 1.6 g S kg21 from plants that had no response to S application
(De Datta, 1981).
Wang et al. (1976) determined S uptake in the straw and grain of lowland rice
grown in the Amazon Basin in the State of Para, Brazil (Table X). Grain and
straw S contents increased as S and N fertilizer rates increased. Additionally, at
low rates of S fertilization, grain S content was greater than straw S content, but
straw and grain S contents were nearly equal at high rates of S fertilization. Wang
(1976) reported that lowland rice grain yields of 5 –7 t ha21 removed between 5
and 9 kg S ha21. The rate of S removal by lowland rice was affected by the
cultivar, S application rate, and N fertilization. In Arkansas, rice total S uptake at
maturity generally averages about 25 kg S ha21 with crop removal by harvested
grain representing about 30% of total plant uptake (Wilson et al., 2001).

4. Rate and Source of Application

Calcium and Mg deficiency can be corrected with the application of dolomitic

lime. The appropriate application rate should be based on lime recommendations,
which are discussed in the liming section. If only the soil Ca level is low, gypsum
(CaSO4) or CaCO3 can be applied to correct Ca deficiency. Zia et al. (1997)
reported that irrigated rice yield and N use efficiency were improved when urea N
was applied in combination with gypsum.
Wang (1976) reported that at least 10 kg S ha21 is required from fertilizer
for rice production on Brazilian lowland rice soils. These soils can tolerate very
high levels of S (as much as 1000 kg S ha21 in field and 2000 kg S ha21 in
pot experiments) without reducing grain yields. Either ammonium sulfate or
single superphosphate were good sources of S. Under pot culture, 10, 20, and
40 mg S kg21 applied once to the soil supported rice production for 1, 2, and
3 consecutive rice crops, respectively. Under field conditions, 27 kg S ha21
applied once supported two crops (Wang, 1976). Immobilization was considered
the major factor responsible for the reduced availability of residual S applied to
the previous crops. In the USA, a portion of the total N requirement is sometimes
 114
N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR
Table X
Sulfur Uptake (kg ha21) in Shoot and Grain of Lowland Rice (Average of Two Cultivars) Under Five Different S Rates and Two N rates
(After Wang et al., 1976)

Shoot Grain Total

S Rate
kg S ha21 60 kg N ha21 120 kg N ha21 60 kg N ha21 120 kg N ha21 60 kg N ha21 120 kg N ha21

Aboveground plant content (kg S ha21)

0 0.89 0.99 1.57 1.92 2.46 2.91
25 2.93 3.82 3.30 4.37 6.23 8.19
50 3.41 4.45 3.37 3.81 6.78 8.26
100 4.12 4.72 3.50 3.80 7.62 8.52
Mean 2.84 3.50 2.94 3.48 5.78 6.98
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…115

supplied by application of ammonium sulfate, which also supplies adequate S

when needed. Sulfur is primarily needed on permeable sandy soils or on highly
reduced clay soils that are continuously cropped to rice. Irrigation water
frequently contains adequate amounts of SO4 to supply seasonal crop
requirements.
Yamaguchi (1997) reported that S deficiency in lowland rice could be
corrected by the application of ammonium sulfate. Yamaguchi (1997) also
reported that a mixture of ammonium sulfate (24% S) and urea increased rice dry
matter production when the proportion of ammonium sulfate represented more
than 25% of the total N application. According to De Datta (1981), S is generally
supplied to rice as a component of other fertilizers such as ammonium sulfate
(24% S), single superphosphate (14% S), potassium sulfate (16 –18% S), or
agricultural gypsum (15 –18% S). Elemental S may also be used as a source of S
provided an adequate time interval is allowed for the oxidation of S into a plant
available form following its application.
Sulfur, like N, is subject to many chemical and biological reactions that
influence its availability when added to the soil. Sulfur fertilizer applications
should be timed based on the initial soil S status and the soil properties that
influence S availability during the course of the growing season. Routine soil
testing for S is seldom used as a guideline. Soil properties (i.e., drainage, texture,
and oxidation status) and field histories are used by growers as a means of
estimating the need for S fertilization. When the availability of S is initially low,
SO4 containing fertilizers should be applied at seeding or by the 5-leaf stage when
rapid plant growth and tillering begin. The application of SO4 containing
fertilizers may also be necessary during the reproductive growth phase (i.e.,
panicle initiation or early boot stage) to prevent late-season S deficiency on
highly permeable or reduced soils.

F. MICRONUTRIENTS

Micronutrients are also called minor or trace elements. Their concentrations in

plant tissues are present in small amounts relative to that of macronutrients. The
essential micronutrients are Zn, Cu, B, Fe, Mn, Mo, and Cl. Accumulation of
these micronutrients by plants generally follows the order of Cl . Mn . Fe .
Zn . B . Cu . Mo. This order may change among plant species and growth
conditions, but is generally correct for lowland rice.
Micronutrient deficiencies in crop plants are increasing because of (i)
increased micronutrient demands from intensive cropping practices and adoption
of high yielding cultivars which may have higher micronutrient demand, (ii)
enhanced production of crops on marginal soils that contain low levels of
available nutrients, (iii) increased use of high analysis fertilizers with low
amounts of micronutrient contamination, (iv) decreased use of animal manures,
116 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

composts, and crop residues, (v) cropping soils that are naturally low in
micronutrient reserves, and (vi) involvement of natural and anthropogenic factors
that limit adequate plant availability and create element imbalances (Fageria
et al., 2002). Deficiencies of B, Cu, Fe, Mn, and Zn have been reported in flooded
rice. Zinc is, by far, the most common micronutrient deficiency encountered
throughout the world’s rice producing regions. In the USA, and possibly other
rice growing regions of the world, Zn deficiency is far more common than all the
other essential elements except N. In addition to Fe deficiency, Fe toxicity is a
significant problem in some parts of the world. Deficiencies of B, Cu, and Mn are
not common, but have been observed in production fields or greenhouse studies.
Molybdenum deficiencies of rice have not been reported and although Cl is
essential for higher plants, its deficiency has not been reported in rice (Obata,
1995). Chloride toxicity in the form of salt injury is a more common problem to
some rice producing areas; however, Cl nutrition of rice is not discussed in this
review. This review will discuss the management practices and nutrient
requirements of four essential micronutrients in regards to lowland rice
production. A more detailed discussion of micronutrient chemistry in the soil
and functions in plant nutrition is available in reviews by Hodgson (1963),
Hodgson et al. (1966), Romheld and Marschner (1991), Mortvedt (1994), Fageria
(1999a), and Fageria et al. (2002).

1. Zinc

Zinc deficiency has been reported in various parts of the world for a large
number of annual crops including rice (Cakmak et al., 1998; Mandal et al., 2000;
Fageria, 2001). A global study by FAO, showed that about 30% of the cultivated
soils of the world are Zn deficient (Sillanpaa, 1982). Additionally, about 50% of
the soils used worldwide for cereal production contain low levels of plant-
available Zn (Graham et al., 1992; Welch, 1993). De Datta (1981) reported that
Zn deficiency is the second most serious nutritional disorder limiting the yield of
lowland rice in the Philippines. Zinc deficiency in crop plants reduces not only
grain yield, but also the nutritional quality of the grain. Consumption of large
quantities of cereal-based foods with low Zn concentrations, poor bioavailability
of Zn, or both is believed to be a major factor in the widespread occurrence of Zn
deficiency in humans (Welch, 1993). In Brazil, Zn deficiency has been reported
in upland as well as lowland rice (Fageria and Barbosa Filho, 1994; Fageria,
2001) and is related to low concentrations of Zn in the highly weathered soils
used for rice production and aggravated by high soil pH due to excessive lime
application (Fageria and Baligar, 1993; Fageria and Gheyi, 1999).

Chemistry in the soil. The availability of Zn to plants or its concentration in the

soil solution is regulated by sorption –desorption reactions at the surface of soil
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…117

colloids (Swift and McLaren, 1991). Desorption controls the amount and release
rate of Zn into the soil solution for plant uptake. Desorption of Zn into the soil
solution is controlled by the strength that Zn is adsorbed onto the surface of soil
colloids. Other forms of Zn are also associated with organic matter, carbonates
and oxide minerals, and Zn in primary and secondary minerals. Thus, the
availability of Zn is influenced by a number of soil characteristics including soil
pH; organic matter content; CaCO3 content; cation-exchange capacity; clay
content and mineralogy; and the quantity and types of Fe, Al, and Mn oxides
(Harter, 1991; Hazra and Mandal, 1996; Singh et al., 1997). After flooded soils
are drained, they contain relatively large amounts of amorphous Fe and Mn
oxides that have large surface areas and a greater adsorption capacity as
compared with their crystalline forms (Sah and Mikkelsen, 1986b; Quang and
Dufey, 1995). Zinc uptake by rice depends not only on the concentration of Zn in
the soil solution, but also on other factors, particularly the concentrations of Fe2þ
and Mn2þ present in the soil solution. High concentrations of Fe2þ and Mn2þ in
the soil solution antagonize Zn absorption (Sajwan and Lindsay, 1986; Mandal
et al., 2000). High soil pH and the presence of free CaCO3 decrease the
availability of Zn in soils. The solubility of soil Zn is highly pH dependent and
decreases 100-fold for each unit increase in pH (Tisdale et al., 1985). The uptake,
translocation, metabolism, and plant use of Zn is inhibited by high P availability
or high rates of P fertilizer applications (Lindsay, 1979).
Unlike the redox elements, Fe and Mn, the concentration of Zn in the soil
solution generally decreases with time after flooding; however, Zn concentrations
may increase briefly immediately after flooding (Mikkelsen and Kuo, 1976;
Gilmour, 1977a). A decrease in soil solution Zn concentration may be due to
precipitation of ZnFe2O4 from the increased Fe solubility after flooding (Sajwan
and Lindsay, 1986) or precipitation of ZnS under highly reduced soil conditions
(Kittrick, 1976).
Plant uptake of Zn depends not only upon the plant species, cultivar within
species, and plant age, but also upon the predominate forms of Zn in the soil (i.e.,
amount of Zn associated with water soluble and exchangeable Zn fractions).
Major factors affecting the availability of soil Zn include the soil pH, total soil
Zn, Zn fertilizer source, soil organic matter content, and soil texture (Chlopecka
and Adriano, 1996). Of these, soil pH extends the greatest influence over Zn
availability in most soils. Zinc deficiency is most likely to occur on coarse-
textured soils with high pH and low soil Zn, soils disturbed by land leveling, and
highly eroded soils (Westfall et al., 1971).

Functions and deficiency symptoms. Zinc is a cofactor for several enzymes that
are involved with N metabolism (e.g., glutamate dehydrogenase) and anaerobic
metabolism (e.g., alcohol dehydrogenase). The reduction of acetaldehyde to
ethanol in anaerobic metabolism requires alcohol dehydrogenase. The alcohol
dehydrogenase activity of seedling rice roots increases dramatically after
118 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

flooding and remains high for several weeks compared to rice seedlings that are
not flooded (Pedrazzini and McKee, 1984). When Zn is deficient, the activity of
alcohol dehydrogenase is depressed, anaerobic root metabolism decreases, and
the ability of the seedling rice to withstand anaerobic soil conditions is reduced
(Moore and Patrick, 1988). This is one reason why Zn deficiency symptoms are
more dramatic after flooding rather than before flooding. When Zn deficiency is
diagnosed, draining the flood is commonly recommended to aid in plant recovery
(Wilson et al., 2001). Removal of the flood allows seedlings to resume aerobic
respiration as oxygen is reintroduced into the soil. The activity of glutamate
dehydrogenase was not affected by Zn fertilization in studies conducted by
Moore and Patrick (1988).
Rice is considered susceptible to Zn deficiency. The symptoms associated
with Zn deficiency of rice are well documented. Zinc, like the other
micronutrients, is not very mobile within the plant; thus deficiency symptoms
are first observed in the youngest leaves. Zinc deficiency most commonly affects
seedling rice plants, but if the deficiency is mild and not corrected symptoms can
also affect plants in the reproductive growth phase. In the early stages of Zn
deficiency, the youngest leaves usually become chlorotic, especially at the leaf
base. As Zn deficiency progresses, the midribs and base of older leaves may also
turn yellow or pale green with brown blotches and streaks appearing on the lower
leaves (Yoshida, 1981). Brown spots usually develop near the tip of the leaf blade
as yellowing begins. Leaf collars may also be stacked as internode elongation is
inhibited (Wilson et al., 2001). Zinc deficiency tends to be more severe where
high rates of N and P are applied (Mueller, 1974). Zinc deficient rice plants do not
respond to N fertilization (Cheaney and Jennings, 1975). Adequate Zn levels in
the soil increase tillering and, consequently, the number of panicles per unit area
of lowland rice (Fageria, 2001). Application of high rates of P fertilizer is known
to aggravate Zn deficiency too. The major reasons for P induced Zn deficiency are
believed to be the formation of Zn phosphate in soil solutions and/or an inhibitory
effect of the excessive P on the metabolic functions of the Zn within the plant
(Shimada, 1995). Zinc deficiency symptoms of rice plants with color photographs
are shown by Mueller (1974), Cheaney and Jennings (1975), Yoshida (1981),
Fageria (1984), Wells et al. (1993), and Fageria and Barbosa Filho (1994).
In the direct seeded, delayed flood management system used in Arkansas, Zn
deficiency symptoms are visible within 2 or 3 days after the flood is established
on seedling rice and the severity of Zn deficiency increases with flood depth
(Wilson et al., 2001). When Zn deficiency is severe, the symptoms are also
visible before flooding. When Zn deficient rice is flooded and severe Zn
deficiency symptoms are expressed seedling death may occur if the flood is not
removed. For this reason, Zn deficiency can result in complete crop failure.
A mild Zn deficiency may not be expressed in highly visible symptoms, but rather
is characterized by slow growth. In such cases, plants may grow out of the
Zn deficiency or the more common symptoms described previously will be
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…119

expressed if the seedling Zn requirement is not satisfied. Mild Zn deficiencies

may be induced by cool temperatures and frequently disappear with warmer
temperatures. Rice yield losses due to Zn deficiency generally range from 10 to
60% in the untreated controls of research plots (Slaton et al., 2002). However,
very little yield loss may occur if Zn deficiency is recognized quickly and the
appropriate corrective actions are taken.

Critical level in plant and uptake. Zinc concentrations of plants typically range
from 30 to 100 mg Zn kg21, depending upon species (Shimada, 1995). Zinc
deficiency of rice occurs primarily on seedling and tillering rice, hence, most
studies initiated to evaluate critical tissue Zn concentrations have emphasized
this growth stage. Zinc deficiency of seedling rice is likely when leaf and/or
whole plant concentrations are , 15 mg Zn kg21 (Forno et al., 1975; Adriano,
1986). During the vegetative growth stages the plant part sampled is not critical
for rice. Although Zn is considered immobile in the plant, whole seedlings or
individual leaves have similar Zn concentrations (Gilmour, 1977b). Fageria et al.
(1997a) reported the Zn sufficiency range in rice shoots at tillering was 20 –
150 mg Zn kg21. Yoshida et al. (1973) developed plant tissue analysis criteria for
classifying the Zn nutritional status of rice. In their system whole seedling Zn
concentrations , 10, 10 – 15, 15 –20, and . 20 mg Zn kg21 are considered
deficient, probably deficient, low, and sufficient, respectively. Research from all
parts of the world agree that seedling Zn concentrations , 15 – 20 mg Zn kg21 are
low or deficient and require Zn fertilization for optimum rice growth.
The tissue concentration of other elements can also be useful in diagnosing Zn
deficiency. Zinc deficient rice tends to accumulate other divalent cations at the
expense of monovalent cations. Thus, the concentrations of Ca, Cu, Fe, Mg, and
Mn tend to be higher in Zn deficient rice, but tissue concentrations of K and N are
lower suggesting their uptake is inhibited in some way (Moore and Patrick,
1988).
The tissue concentration of Zn in the rice plant fluctuates during the growing
season. Whole plant tissue Zn concentration is generally highest after flooding,
decreases during tillering, and then, depending on the cultivar, may increase or
remain stable through ripening (Gilmour, 1977a; Wells, 1980). The decline in
tissue Zn concentration during tillering indicates that the rate of aboveground dry
matter accumulation exceeds that of Zn uptake by the developing root system.
The accumulation of Zn in the aboveground portion of flood-irrigated rice is
relatively slow during vegetative growth, reaches a maximum during late
vegetative to early reproductive growth, and then declines after anthesis
(Gilmour, 1977b). The rate of maximum Zn uptake and dry matter accumulation
occurs simultaneously. The rate of nutrient uptake is likely related to root growth
and development. Slaton et al. (1990) found that the maximum rice root length
was reached by the early boot stage, but the maximum root growth rate occurred
120 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

by panicle initiation, which corresponds to the same growth stage that Gilmour
(1977b) found maximum Zn uptake.
The Zn concentration of rice at which toxicity occurs is not well documented.
Marschner (1995) reported that the critical toxic concentration of Zn in leaves of
crop plants is . 400– 500 mg Zn kg21. Wells et al. (1993) indicated that rice is
very tolerant to Zn, with toxicity levels estimated at more than 1500 mg Zn kg21.
Fageria (2000a) reported a 10% reduction in rice shoot weight (critical toxic
level) when the Zn concentration in 42-day old plants was 673 mg Zn kg21.
Rattan and Shukla (1984) showed that application of excessive Zn fertilizer
decreased rice dry matter and estimated that tissue Zn concentrations at flowering
. 190 mg Zn kg21 were toxic. The availability of Zn can be reduced or stabilized
in the soil by the application of amendments such as lime, P, natural or synthetic
zeolities, apatities, glauconite, iron oxide containing materials, and alkaline
biosolids (Chlopecka and Adriano, 1996, 1997).

Zinc fertilization practices. Routine soil testing is a valuable tool that can be used
to assess the potential for Zn deficiency in crops. Sims and Johnson (1991)
reported that the critical soil Zn concentration range for most crops was between
0.5 to 2.0 mg Zn kg21 for DTPA and 0.5 –3.0 mg Zn kg21 for Mehlich 1. Most
research indicates that the critical soil test Zn concentrations for rice fall within
the ranges suggested by Sims and Johnson (1991). Fageria (1989) reported that
1.0 mg Zn kg21 of soil extracted by the Mehlich 1 method was the critical
concentration for lowland rice. Critical DTPA extractable soil Zn concentrations
of 0.8 mg Zn kg21 has been reported for Indian soils for lowland rice (Tiwari and
Dwivedi, 1994), whereas, 0.7 mg Zn kg21 (Sedberry et al., 1978) and 0.5 mg
Zn kg21 (Hill et al., 1992) have been suggested for rice in the USA. Sedbery et al.
(1980) and Wells (1980) both indicated that soil pH of silt loam soils was the best
predictor of rice response to Zn fertilization. However, their research was
conducted on soils that had not previously received applications of Zn fertilizer,
were uniformly low in Zn, and micronutrients were not commonly measured in
routine soil analysis. Thus, for a number of years Zn fertilizer recommendations
were based exclusively on soil texture and soil pH, which triggered the
recommendation to use Zn fertilizer on nearly every rice crop grown in the
rotation on alkaline soils. In Arkansas, Zn fertilizer recommendations for flooded
rice are now based on the soil pH, texture, and Mehlich 3 extractable Zn (Wilson
et al., 2001). Zinc fertilizer is recommended for rice grown on silt and sandy loam
soils having a pH . 6.0 and Mehlich 3 extractable Zn , 3.5 mg Zn kg21. Zinc
deficiencies are seldom observed on undisturbed clay soils in the USA. Precision
land leveling often exposes Zn deficient subsoils and Zn deficiency is
occasionally observed on leveled soils of all textures.
The most common commercially manufactured granular Zn fertilizers are Zn
sulfates, oxides, oxysulfates, lignosulfonates, and a number of organic chelated
materials like ZnEDTA and ZnHEDTA. Excellent reviews of the manufacturing
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…121

and properties of Zn fertilizers are given by Tisdale et al. (1985), Foth and Ellis
(1988), and Martens and Westermann (1991). Application of these Zn fertilizers
to rice is performed in a variety of methods depending on the production system.
Most commonly, relatively high rates of inorganic Zn fertilizers are applied to the
soil before seeding. Application of 5– 7 kg Zn ha21 as Zn sulfate was found to
correct Zn deficiency in lowland as well as upland rice grown on Inceptisols and
Oxisols in Brazil (Fageria and Barbosa Filho, 1994). Slightly higher rates of
11 kg Zn ha21 are typically recommended for soil application in the USA. In the
waterseeded system practiced in California, a surface, broadcast application of
either ZnSO4 or Zn lignosulfonate is recommended because the roots of rice
seedlings are positioned at or near the soil –water interface (Wells et al., 1993).
Highly water soluble ZnSO4 is generally regarded as the best Zn fertilizer used
to correct Zn deficiencies. Liscano et al. (2001) showed that the water solubility
of inorganic Zn fertilizers was highly correlated to Zn uptake by seedling rice in
greenhouse studies. They suggested a minimum of 40– 50% of a Zn fertilizer’s
total Zn content should be water-soluble to optimize Zn uptake. Amrani et al.
(1999) and Gangloff et al. (2002) reported similar results for corn. In general, the
water solubility of Zn sulfates and lignosulfonate sources is high and the water
solubility of Zn oxides and Zn oxysulfate sources is low to moderate. However, in
most cases, the Zn application rate is more critical than the water-soluble Zn
content of the fertilizer, but research data clearly shows that tissue Zn
concentration and total Zn uptake generally increase as water soluble Zn in a
fertilizer increases. The use of the water-soluble Zn criteria for selecting a Zn
fertilizer becomes more important as the severity of Zn deficiency for the
immediate crop increases. The recommended rates of soil applied Zn are about 20
times higher than the total crop uptake of Zn, but are required to obtain adequate
distribution of Zn fertilizer granules. The primary advantage of soil applied Zn
over other Zn fertilization methods that use much lower Zn application rates is
the residual benefit. A single Zn fertilizer application should provide adequate Zn
for several years before additional Zn fertilizer is needed to optimize grain yields.
The mobility of Zn in the soil following fertilization differs among Zn sources
and influences Zn uptake. When applied to the soil surface, water-soluble ZnSO4
showed much greater vertical, downward movement than the relatively water-
insoluble ZnO (Giordano and Mordvedt, 1972). The vertical mobility of Zn
supplied from ZnSO4 and Zn lignosulfonate fertilizer sources are similar with
both having greater mobility than ZnO and less mobility than ZnEDTA
(Mikkelsen and Brandon, 1975). The water solubility of the Zn fertilizer source,
the time allowed for vertical movement of the Zn, and the relative placement of
the Zn fertilizer are all important aspects to consider in Zn fertilization practices.
Granular Zn fertilizers do not need to be mechanically incorporated. Surface
applications of Zn before seedling emergence are equally effective as preplant
incorporated Zn (Giordano and Mordvedt, 1972; Slaton et al., 2001c). However,
granular Zn applications should not be made immediately before flooding
122 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

because the Zn is not positioned for immediate uptake. Generally, less than 5% of
the broadcast soil applied Zn fertilizer is taken up by flood-irrigated rice
(Giordano and Mordvedt, 1972).
The highly water-soluble Zn sulfates and chelated Zn sources may also be
sprayed in solutions to the soil or directly onto the seedling rice foliage before
flooding. Compared with granular Zn applications, the application of Zn
solutions has the advantage of uniform distribution. The chelated Zn fertilizer
sources are generally more expensive per unit of Zn and are applied at relatively
low rates (, 1 – 2 kg Zn ha21) so that little residual Zn is available for future
crops. Foliar Zn applications are usually employed only when the growing
crop shows deficiency symptoms (De Datta, 1981) or they allow for more flexi-
ble crop management in regards to production costs and application times
(Wilson et al., 2001).
In transplanted rice, dipping seedling roots in a 1% ZnO suspension has
prevented Zn deficiency (Yoshida, 1981). Abilay and De Datta (1978)
recommended coating pregerminated rice seeds with ZnO prior to seeding
followed by a foliar Zn application 5 –7 days before panicle initiation. Slaton
et al. (2001) showed that Zn application directly to rice seed at concentrations
between 2.2 and 5.7 g Zn kg21 seed was a low-cost alternative to the standard
recommendation of 11 kg Zn ha21 as a preplant incorporated inorganic Zn
fertilizer for dry-seeded rice. Earlier research also showed that the application of
low rates of Zn to rice seeds or dipping the roots of transplanted rice in a Zn
solution were effective alternatives to broadcast applications of Zn fertilizer
(Rush, 1972; Giordano and Mordvedt, 1973; Mengel et al., 1976; Haghighat and
Thompson, 1982). Zinc is also present in manures and other organic amendments
that, when readily available and soil applied at adequate rates, can supply crops
with sufficient Zn (Ye and Yang, 1997).
Lowering the pH of alkaline or calcareous soils, by application of acid-
forming fertilizers/amendments like elemental S, can improve Zn availability and
uptake by rice (Slaton, 1998). The effectiveness of acidifying the soil is often
limited by the economics and the practicality of soil pH reduction. Some soils
contain very high quantities of CaCO3 or are highly buffered and require very
high rates of acidic amendments to reduce pH. Additionally, the reduction of soil
pH may be only temporary as the soil pH may gradually increase over time and
return to near its initial value.
The literature shows that a number of Zn application methods, times, sources,
and rates are highly effective means of supplying Zn to rice in various production
systems. Routine soil testing used in conjunction with the application of suitable
Zn fertilizers at the proper rates and times is the best method to ensure that Zn
nutrition is not a yield-limiting factor for rice production. The most efficient
method of Zn fertilization is often dictated by the cultural production system, soil
conditions, economics, the availability of Zn fertilizers, or a combination of
several of these factors.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…123

Use of efficient genotypes. Crop species differ markedly in their ability to adapt to
or tolerate Zn deficient soils (Graham, 1983, 1984; Graham and Rengel, 1993;
Rengel, 1997). Among the cereal and grain crops, rice, sorghum, and corn are
classified as sensitive to Zn deficiency (Clark, 1990). The physiological
mechanisms responsible for crop tolerance to Zn deficiencies are not clearly
understood (Graham and Rengel, 1993; Rengel and Graham 1996; Cakmak et al.,
1998; Huang et al., 1996). Graham et al. (1997) reported significant differences in
the Zn concentration of rice grain among genotypes. These authors also reported
that several tests had been conducted to examine the effect of soil climatic factors
on the Zn content of grain. In rice, the filled grain trait was reasonably stable
across the range of environments in which rice was grown, as well as over
seasons, planting dates, and soil fertility. Hence, it is possible to select Zn
efficient genotypes for production in different agro-ecological regions. Numerous
other researchers have also established that differences exist among rice
genotypes to Zn deficiency (Ponnamperuma, 1976a; IRRI, 1977; Mahadevappa
et al., 1981; Bowen, 1986, 1987; Fageria, 2001). Rice genotypes showed
differences in Zn absorption that were not attributed to differences in root surface
area (Bowen, 1986, 1987). Rice genotypes showed marked differences in Vmax
(maximum ion uptake rate) and Km (Michaelis – Menten constant, equal to the
substrate ion concentration giving half the maximal rate of uptake) values.
Efficient genotypes increase Zn translocation from the roots to the shoot and
regulate Ca, Cu, Fe, Mg, and P transport in order to maintain balanced nutrient
ratios with respect to Zn (Cayton et al., 1985). Resistance to Zn deficiency
appeared to be controlled polygenically in rice and is thought to be a dominant
trait (Mahadevappa et al., 1981). Zinc deficiency has been a common mineral
nutrient problem in rice and genotypes have been screened extensively for this
disorder at the International Rice Research Institute (Clark, 1990).

2. Boron

Boron deficiency has been reported in at least 80 countries and 132 crop
species. It is estimated that about 15 million hectares are annually treated with B
fertilizers (Shorrocks, 1997). Plant species vary in B requirement with
dicotyledons generally requiring 3 –4 times more B than monocotyledons
(Bennett, 1993). A number of soil properties influence B availability to plants and
are reviewed by Fageria et al. (2002). Coarse-textured, low organic matter soils
located in humid regions are the most prone to B deficiency. The application of
lime to acid soils can also induce B deficiency because of increased B adsorption.
Boron deficiencies are not common to rice, but several environmental (i.e., high
rainfall), soil (i.e., low organic matter, texture, and pH), and rice production (i.e.,
flood-irrigation) factors common to many rice growing regions of the world hint
124 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

that B could very likely be a yield-limiting nutrient. Our lack of understanding of

the B nutritional requirements, seasonal patterns of B uptake, and B partitioning
within the rice plant demonstrates the need for increased research efforts for this
micronutrient.

Chemistry in the soil. The soil B pools can be categorized into the mineral,
adsorbed, and soil solution fractions. From the plant nutrition standpoint, the soil
solution fraction is of utmost importance. Boron is absorbed as H3BO3 (or
B(OH)3) and exists in the soil solution as an undissociated molecule. In
agricultural soils with a pH range of 5 – 9, undissociated H3BO3 is the
predominant species in the soil solution. The transport mechanisms of B from
the soil solution to plant roots are mass flow and diffusion with mass flow as the
dominant mechanism contributing to plant uptake of B.
Like most other micronutrients, the availability of B decreases as soil pH
increases. The reduction in B availability from increasing soil pH by liming is
caused by B adsorption by iron and aluminum hydroxides. Boron is adsorbed to
the surface of these precipitated Fe and Al hydroxides. Adsorption of B is very
pH dependent. Maximum adsorption by Al(OH)3 and Fe(OH)3 occurs in the soil
pH range of 7– 9 and corresponds to the soil pH range of lowest B availability
(Tisdale et al., 1985). Organic matter also adsorbs B and acts as a reservoir to
replenish soil solution B upon crop removal or loss via leaching. Replacement
by other anions or mineralization of the organic matter releases B (Foth and Ellis,
1988). Clay contents also influence B adsorption. Barber (1995) reported
that B adsorption by fine-textured soils is 2– 3 times greater than by coarse-
textured soils.
Despite this general knowledge, very little is known about the chemistry of B
in flooded soils. The concentration of B in the soil solution is believed to remain
more or less constant following soil submergence (Ponnamperuma, 1975). In
flooded soils, with pH buffered around neutrality, H3BO3 is the dominant species
in the soil solution. As with upland soils, the adsorption of B on Fe and Al oxides
(Sims and Bingham, 1968) seems to be an important mechanism in governing B
solubility in flooded soils (Patrick and Reddy, 1978).

Functions and deficiency symptoms. A thorough review of the literature suggests

that B deficiency symptoms of rice have not been documented in the field in any
rice-growing region of the world. However, deficiency symptoms have been
produced in nutrient solutions and induced in greenhouse studies. Boron is
relatively immobile in plant tissues and deficiency symptoms first appear in the
youngest growth. The tips of emerging leaves become white and rolled in B
deficient plants, which is similar to a symptom described for Ca deficiency. The
growing points may die in the case of severe B deficiency. Obata (1995) noted
that B deficiency also retarded root elongation of plants. The B requirement for
vegetative growth of plants, especially grasses, is very low, but the need for B
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…125

increases for seed production (Marschner, 1995). This is one possible reason why
B deficiency symptoms on rice leaves have not been documented in commercial
rice fields. Boron deficiency of rice may be expressed solely in the form of
reduced grain yields from floret sterility and may be mistakenly blamed on poor
environmental conditions during anthesis. This hypothesis is just speculation
since it has yet to be proven in replicated field trials. However, there is some
preliminary evidence that supports this theory. Okuda et al. (1961) observed that
the panicles of B deficient rice plants failed to exert from the boot. Likewise,
Dunn (1978) noted that B deficiency produced similar symptoms as those
associated with the physiological disorder straighthead that was induced by
arsenic toxicity in greenhouse studies. Very little research has been conducted to
verify or refute B deficiency as a possible cause of straighthead. Further evidence
supporting this theory is provided by the essential functions of B in plant growth.
Boron plays important roles in cell development and elongation, protein
synthesis, carbohydrate metabolism, pollen tube formation, and pollen viability
(Bennett, 1993; Marschner, 1995). In the United States, interest in the B nutrition
of rice has recently been stimulated by significant yield increases from direct B
fertilization of rice in Missouri (Dunn and Jones, 2002). Color photographs of B
deficiency symptoms of rice were published by Ishizuka (1978) and Fageria and
Barbosa Filho (1994).
Boron toxicity is also a major concern for rice since the soil and plant tissue B
sufficiency ranges, between deficient and toxic concentrations, are quite narrow.
The rice tissue concentration associated with B toxicity is 100 mg B kg21 in the
Y-leaf of tillering rice, 35 mg B kg21 in the shoots at panicle initiation, and
100 mg B kg21 in the straw at maturity (Dobermann and Fairhurst, 2000). The
plant part and the time of sampling are critical when using tissue analysis to
diagnose B toxicity because B tends to accumulate in the leaf tips and may be
leached from plant tissue by rain. Likewise, Dobermann and Fairhurst (2000)
cited hot-water soluble soil B concentrations . 5 mg B kg21 as potentially toxic.
Obata (1995) suggested a slightly higher hot-water soluble soil B concentration
of $ 10 mg B kg21 was associated with B toxicity of rice. Application of high
rates of B fertilizer may also induce B toxicity on soils that have low soil B
concentrations (Fageria, 2000b). Toxicity symptoms include scorching of the tips
and margins of older leaves and result in reduced dry matter and grain yield
production.

Critical level in plant and uptake. Boron contents vary widely among plant
species. Rice, together with wheat and barley, has a lower requirement for B than
do nongramineous crops (Obata, 1995). Boron requirement is higher during the
reproductive growth stages than during vegetative growth due its important
function in grain formation. Due to the lack of published research on B nutrition
of rice, very little data is available on the critical tissue B concentrations required
for the production of maximum rice yields. Dobermann and Fairhurst (2000)
126 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

suggested that the optimum B concentration range for the Y-leaf of tillering rice
was 6– 15 mg B kg21. Fageria et al. (1997a) suggested a critical concentration of
8 mg B kg21 in rice straw at maturity. Yu and Bell (1998) reported 18.5 mg
B kg21 in rice leaves and 8.9 mg B kg21 in rice stems were associated with
maximum rice yield production. They also reported that B deficiency in rice
occurred when B concentrations in the top mature leaves was , 7.3 mg B kg21.
In a review of the literature, Yu and Bell (1998) concluded that the sufficient B
concentration range of rice varied from 5 to 67 mg B kg21 depending on the plant
age and the part analyzed. Data in Table XI suggests that the B concentration of
whole rice plants is relatively constant during the growing season with total B
uptake increasing as plant age increased. Depending on the location, 22– 34% of
total aboveground B content was partitioned in the grain.

Boron fertilization practices. Soil analysis is sometimes used as a basis for

identifying potentially B deficient soils and has been used to make fertilizer
recommendations for some crops (Martens and Westermann, 1991; Sims and
Johnson, 1991). However, B deficiency of rice is yet to be recognized as a serious
yield-limiting factor and insight to critical soil test B concentrations for rice have
not been developed. Research conducted with other crops suggests that if positive
rice yield responses to B fertilization are found fertilizer recommendations can
potentially be based on routine soil analysis.

Table XI
Whole Plant B Concentrations and Contents of Shoots and Grain of Flood-Irrigated Rice from
Brazil (Fageria, unpublished data, 2001) and Arkansas (Slaton, unpublished data, 2001) at
Different Growth Stages

Brazila Arkansasb

Concentration Content Concentration Content

Growth stage DAEc (mg B kg21) (g B ha21) DAEc (mg B kg21) (g B ha21)

Beginning tillering 22 7.8 2.8

Active tillering 35 7.5 8.6 55 7.4 11
Panicle initiation 71 7.1 41.1 81 5.0 21
Boot stage 97 6.7 70.6 96 4.9 42
Flowering 112 6.9 92.5 111 5.3 63
Mature-straw 140 7.3 69.4 145 5.2 51
Grain 140 5.3 35.0 146 1.9d 15d
a
The rice cultivar was ‘Metica 1’ and was grown on an Inceptisol. Average yield (at maturity) was
9423 kg ha21 straw weight and 6389 kg ha21 grain yield.
b
In Arkansas, the rice cultivar was ‘Wells’ and was grown on a slightly calcareous Calloway silt loam.
Average yield (at maturity) was 9555 kg ha21 straw weight and 8085 kg ha21 grain yield.
c
DAE, Days after emergence.
d
Harvested grain. Whole panicle concentration at maturity was 3.8 g B kg21.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…127

Boron extracted with the hot-water procedure is generally considered the

standard extraction procedure to evaluate crop response to B. Sensitive crops are
likely to respond to B fertilization if the level of hot water-soluble B is less than
0.5 mg B kg21 soil (Sims and Johnson, 1991). A particular problem with B is the
narrow margin between deficiency and toxicity. Sensitive crops can be affected
by B toxicity at soil levels over 5.0 mg B kg21, so B fertilizer must be used with
caution. Sims and Johnson (1991) reported critical soil B levels of 0.1– 2.0 mg
B kg21, depending on yield level, soil type, pH, and organic matter content.

3. Iron

Rice is considered very susceptible to Fe deficiency, especially when grown

under upland conditions. Rice yield losses caused by Fe deficiency measured in
research plots range from 10 to 50% (Snyder and Jones, 1988, 1991). Ferrous Fe
(Fe2þ) is preferentially absorbed by plant root systems and is generally present in
high concentrations in reduced soils (Marschner, 1995). Thus, Fe deficiency does
not commonly occur in flooded rice due to the increase in Fe availability
associated with the anaerobic soil conditions used for its production. However, Fe
deficiency occurs primarily on seedling rice before flooding in some rice growing
regions of the world (Yoshida, 1981; Snyder and Jones, 1988). Mori et al. (1991)
suggested that seedling rice is highly susceptible to Fe deficiency because rice
roots produce comparatively low amounts of iron-chelating phytosiderophores
compared to other grass species. Soil conditions that limit Fe availability coupled
with plant limitations for obtaining Fe are the primary reasons why Fe deficiency
occurs.
In many rice-growing regions of the world, Fe toxicity rather than Fe
deficiency is the more common problem and is associated with inherent soil
properties (Yoshida, 1981; Fageria and Rabelo, 1987; Olaleye et al., 2001). Iron
toxicity has been reported as a significant problem in rice growing areas of
southeast Asia, Africa, and South Africa that occurs primarily on acid sulfate soils
(Ottow et al., 1983). Iron toxicity is believed to be caused by excessive Fe in the
soil solution or induced by deficiencies of other nutrients. The direct toxicity is
defined as excessive Fe absorption by the plant resulting from high soil solution
concentration of Fe (Howler, 1973). The indirect toxicity has been blamed on low
soil fertility. Specifically, low soil concentrations of Ca, Mg, K, and P have been
cited as the common factors among soils expressing Fe toxicity symptoms
(Benckiser et al., 1984). The name Fe toxicity is somewhat misleading in the sense
that, in many cases, the Fe toxicity is actually an effect rather than a cause. Howler
(1973) suggested that rice roots become coated with Fe oxide which reduces the
plants’ ability to absorb sufficient quantities of other plant nutrients that are
already present in low concentrations. The production of H2S, FeS, or both in
highly reduced flooded soils may contribute to Fe toxicity. Hydrogen sulfide and
128 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

FeS reduce the oxidizing capacity of the rice root system, thereby increasing the
susceptibility of the rice plant to Fe toxicity (Inada, 1966b; Tanaka et al., 1968).
Application of (NH4)2SO4 fertilizer has been noted to increase the incidence of Fe
toxicity (Inada, 1966a). Acidic soil conditions are likely to produce either type of
Fe toxicity. Severe Fe toxicity can cause significant rice yield reductions (Genon
et al., 1994). In general, soil solution Fe concentrations of flooded soils are very
high because of the anaerobic soil conditions, but rice is well adapted to this
flooded environment and normally is able to regulate Fe uptake.

Chemistry in the soil. Iron is a major constituent of most soils. Iron minerals
commonly found in soils include goethite (FeOOH), hematite (Fe2O3), pyrite
(FeS), siderite (FeCO3) and magnetite (Fe3O4). Therefore, Fe availability to
plants is affected by the soil properties like soil pH and redox that influence the
solubility of Fe containing minerals. The Fe concentration in the soil solution
decreases with an increase in soil pH. For each unit increase in soil pH there is a
1000-fold decrease in the solubility of Fe3þ and a 100-fold decrease in the
solubility of Fe2þ (Tisdale et al., 1985). The amount of Fe2þ increases rapidly at
redox potentials below 200 mV. Ponnamperuma (1976b) found that the soil Fe2þ
concentration increased to a peak ranging from 0.1 to 600 mg kg21 for several
soils shortly after submergence and then declined. Diffusion and mass flow are
believed to be the two mechanisms responsible for the movement of Fe from soil
to the root surface.
The chemistry of flooded soils is dominated more by Fe than by any other
redox element. The major reason for this dominance is the large amount of soil Fe
that can undergo reduction, which usually exceeds the total amount of other
redox elements by a factor of 10 or more (Patrick and Reddy, 1978). Under
submerged soil conditions, Fe3þ is reduced to Fe2þ by respiring microorganisms.
Although variable in composition, ferric oxyhydroxides in aerated soils can be
represented by the formula Fe(OH)3, which can undergo reduction (Patrick and
Reddy, 1978):

FeðOHÞ3 þ 3Hþ þ e2 $ Fe2þ þ 3H2 O

Although there is general agreement that the reduction of ferric compounds occur
as a result of the respiration of facultative anaerobic bacteria, it has not been
demonstrated conclusively that the reduction is brought about by enzymatic
transfer of electrons directly to Fe3þ, or that the reduction is an indirect chemical
reaction between bacterial metabolites and Fe3þ. In either case, it is likely that
complexing of Fe with organic chelates plays an important role in making the Fe
solution more reactive (Patrick and Reddy, 1978).

Functions and symptoms of Fe deficiency and toxicity. The concentration of Fe

in rice tissue is generally higher than all other micronutrients, except Mn, and
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…129

reflects the need for sufficient Fe nutrition for normal growth and seed
production. The functions of Fe in plant nutrition and growth processes are
thoroughly reviewed by Marschner (1995). Some of the essential functions of Fe
in plant growth are clearly exhibited by the deficiency symptoms expressed by
rice and other plants.
Iron deficiency is most common to upland rice production systems and seldom
occurs in lowland rice after flooding. However, many lowland rice production
systems are dry-seeded and later flooded at the onset of tillering; thus the period
between seeding and flooding is the time when Fe deficiency is most likely to
occur in lowland rice. Iron is not highly mobile within the plant and the youngest
leaves are the first to show Fe deficiency symptoms. At the onset of Fe deficiency,
symptoms begin as an interveinal chlorosis of the youngest leaves giving plants a
striped appearance. Further progression of the Fe deficiency gives seedlings a
uniform pale yellow to bleached white appearance (Snyder and Jones, 1988).
These symptoms have been noted to occur within 1 week after emergence on
Histosols used for rice production in the Florida Everglades. Iron deficiency
reduces seedling dry matter production, leaf chlorophyll content, panicle number
per unit area, and grain yield (Snyder and Jones, 1988). Plant tissue analysis used
to determine the total tissue Fe concentration has proven to be of little value for
diagnosing Fe deficiency of many plants, and rice is no exception (Mengel et al.,
1984). The total Fe concentration of rice seedlings showing Fe deficiency
symptoms is generally equal to that of seedlings without symptoms (Snyder and
Jones, 1988). Routine soil analysis is also of little value for predicting Fe
deficient soils, so diagnosis of Fe deficiency from visual symptoms or field
history is generally relied upon. Color photographs of rice plants with Fe defi-
ciency are available in publications by Ishizuka (1978), Yoshida (1981), Fageria
(1984), Fageria and Barbosa Filho (1994), and Dobermann and Fairhurst (2000).
Iron toxicity causes the older leaves to turn a yellow –orange color starting at
the leaf tip with symptoms proceeding towards the leaf base. The tips of the lower
leaves will eventually desiccate and give the plants a scorched appearance. Rice
roots commonly have a black coating of FeS, which can also be used to help
diagnose Fe toxicity. Toxic levels of Fe can induce deficiencies of K, P, Cu, and
Zn in rice (Fageria and Rabelo, 1987; Baruah and Nath, 1997). The degree of leaf
bronzing has been suggested to be a good measure of the severity of Fe toxicity in
flooded rice (IRRI, 1965).

Critical level in plant and uptake. Iron has a relatively wide sufficiency
concentration range in plant tissue between the proposed critical concentrations
for Fe deficiency and toxicity. At the tillering stage, the sufficiency range of Fe
concentrations in the leaf blades ranges from 70 to 300 mg Fe kg21 (Wells et al.,
1993). Iron deficiency or toxicity occurs at concentrations below or above this
sufficiency range. Fageria et al. (1997a) also reported a similar Fe sufficiency
concentration range, but noted the concentration varied depending on the plant
130 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

age and part analyzed. Dobermann and Fairhurst (2000) suggested the sufficient
Fe concentration range of the Y-leaf was 75 –150 mg Fe kg21 during vegetative
growth. The sufficient Fe concentration of the whole shoots was somewhat lower
at 60 –100 mg Fe kg21. Gilmour (1977a) found that whole plant and bottom leaf
Fe concentrations were significantly greater than that for the Y-leaf up to 69 days
after seeding. Similar to other nonmobile elements, the concentration of Fe
usually increases as leaf age increases. The use of tissue Fe concentrations is
useful only to the extent of establishing whether the total Fe concentration falls
within the suggested sufficiency range. The most promise for diagnosis of Fe
deficiency by tissue analysis is perhaps by analyzing for active Fe.
Iron toxicity is believed to occur when leaf blade total Fe concentrations
exceed 300 mg Fe kg21 (Tanaka et al., 1966). Dobermann and Fairhurst (2000)
indicated this critical concentration was specifically for the Y-leaf. The
importance of leaf age, the plant part sampled, and sample cleanliness cannot
be overemphasized for the diagnosis of Fe toxicity. The presence of toxic
concentrations of Fe may simply be the result of another nutrient deficiency. For
example, Zn deficient whole aboveground rice seedlings commonly contain Fe
concentrations . 300 mg Fe kg21. However, clean Y-leaf tissue is seldom above
this threshold. Whole plant samples of nutritionally healthy plant samples may
easily exceed 300 mg Fe kg21 from contamination of Fe precipitates on the rice
stems unless the tissues are thoroughly washed in a mild acid solution before
drying and analysis.
Representative Fe concentrations and total plant contents of lowland rice
grown on an Inceptisol of central Brazil and an Alfisol in Arkansas are listed in
Table XII. The concentration of Fe in the whole-aboveground tissue is rather high
during early tillering and then decreases during the growing season. Rice grain
generally contains about 30 mg Fe kg21 so only a fraction of the total plant Fe
content is actually removed in the harvested grain.

Iron fertilization practices. Soil analysis is not a highly effective means of

identifying Fe deficient soils. Sims and Johnson (1991) reported that for most
crops the critical soil Fe concentration range was 2.5 –5.0 mg kg21 of DTPA
extractable Fe, but is also influenced by soil pH. For example, Fe deficiency of
upland rice occurs on Brazilian Oxisols with Mehlich 1 extractable Fe of more
than 50 mg kg21 when the soil pH is raised above 6.0 by liming (Fageria, 2000c).
As a general rule, Fe deficiency is caused by soil chemical properties that affect
Fe availability and not by low Fe content. Organic soils may be the exception to
this rule. Snyder and Jones (1988) showed the soil ash content (i.e., weight)
following combustion of soil at 5508C was not a reliable indicator of seedling
chlorosis on Histosols in the Florida Everglades. However, the total iron content
of the soil ash was consistently lower from field areas that exhibited chlorosis.
Further, they observed the color of the ash ranged from white to yellow to a
yellow-red color. Soils with white-colored ash were associated with field areas
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…131

Table XII
Seasoncal Whole Plant Fe Concentration and Content in Rice Straw and Grain of Flood-
Irrigated Rice Grown in Arkansas (Slaton, unpublished data, 2001) and Brazil (Fageria,
unpublished data, 2001)

Brazila Arkansasb

Concentration Content Concentration Content

Growth stage DAEc (mg Fe kg21) (g Fe ha21) DAEc (mg Fe kg21) (g Fe ha21)

Beginning tillering 22 419 134 – – –

Active tillering 35 374 411 55 371 216
Panicle initiation 71 175 1045 81 152 638
Boot stage 97 155 1403 96 198 1535
Flowering 112 179 2202 111 171 1421
Mature-straw 140 448 3499 145 169 1653
Grain 140 81 265 146 27d 219d
a
The rice cultivar was ‘Metica 1’ and was grown on an Inceptisol.
b
The rice cultivar was ‘Wells’ and was grown on a slighty calcareous Calloway silt loam.
c
DAE, Days after emergence.
d
Harvested grain. Whole panicle Fe concentration at maturity was 99 mg Fe kg21.

that produced chlorotic rice and appeared to be a highly reliable means of

predicting Fe deficiency on these organic soils. Examination of the ash color
following soil combustion was highly correlated with rice response to Fe
fertilization on these organic soils.
The Fe deficiency problems common to the organic soils used to produce rice
in Florida initiated research to investigate the most efficient methods of Fe
fertilization. Water-soluble Fe fertilizers that are broadcast applied to the soil are
rapidly converted to insoluble forms that have limited effectiveness unless
applied at very high rates (Martens and Westermann, 1991). Snyder and Jones
(1988) found that application of 20 –30 kg Fe ha21, as water-soluble Fe (i.e.,
FeSO4·H2O and FeSO4·7H2O), with the seed was highly effective at preventing
seedling chlorosis and increasing rice yield. Foliar application of Fe solutions
varied in their effectiveness to prevent Fe chlorosis (Snyder and Jones, 1991).
Foliar applications of chelated Fe (1 kg Fe ha21) made 1 week after emergence
were equivalent to Fe drilled with the seed. Greenhouse studies also suggested
that water-soluble FeSO4 was equal to chelated Fe when applied in a timely
manner and with the proper water source. However, when foliar Fe was applied at
2 and 3 weeks after emergence, significant yield losses occurred. Thus, under
severe Fe deficiency, preventative measures, before or at the time of seeding, are
preferred over corrective measures that are initiated after deficiency symptoms
are expressed. Reducing soil pH to increase Fe availability has proven to be an
effective means of ameliorating Fe deficiencies of other crops (Martens and
132 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Westermann, 1991). However, soil acidification has not been thoroughly

evaluated on Fe deficient soils used for rice production to develop best
management practices. The majority of published research regarding Fe
fertilization of rice has been conducted on organic soils in Florida. We assume
the same fertilization rates and methods proven effective on organic soils would
be equally effective on Fe deficient mineral soils.
Fertilization practices used to ameliorate Fe toxicity include periodic surface
drainage to reduce soil solution Fe concentrations, liming to increase soil pH, and
the routine application of nutrients to maintain soil fertility. The Fe-excluding
ability of rice plants is lowered by deficiencies of P, K, Ca, and Mg (Obata, 1995).
In particular, K deficiency readily induces Fe toxicity. Use of tolerant rice
cultivars is another alternative of producing rice on Fe toxic soils (Fageria et al.,
1990b). Rice cultivars also show different susceptibilities to Fe toxicity
(Gunawarkena et al., 1982; Fageria and Rabelo, 1987; Wu et al., 1997).

4. Manganese

Manganese deficiencies and toxicities have been documented in some rice

growing regions, but are not common in most soils used for rice production.
Nutritional disorders related to Mn occur less frequently than those associated
with Fe. Rice grown on Oxisols and Histosols are most likely to experience
disorders related to Mn nutrition. Manganese deficiency may cause yield losses
that approach 100% from seedling mortality, but yield losses are usually less than
30% (Snyder et al., 1990).

Chemistry in the soil. Similar to Fe, Mn is widely distributed in soils, but largely
in the form of high-valency oxides, which are unavailable to plants. Manganese
oxides are the most common manganese minerals in soil and include pyrolusite
(MnO2), manganite (MnOOH), and hausmannite (Mn3O4). Soil manganese exists
in three oxidation states including Mn2þ, Mn3þ, and Mn4þ with Mn2þ being the
primary form absorbed by plants. The predominant Mn oxidation states in most
soils are Mn2þ and Mn4þ, with much more as Mn4þ than Mn2þ in aerated soils
(Barber, 1995). Small amounts of Mn2þ are present in soils as exchangeable ions
and as organic complexes and these constitute the source for plant uptake. It has
been reported that 80– 90% of the Mn in the soil solution is complexed with
organic matter (Foth and Ellis, 1988).
The concentration of Mn2þ in the soil is affected by soil pH and oxidation–
reduction reactions. Its solubility decreases 100-fold for each unit increase in pH
(Barber, 1995; Tisdale et al., 1985). In submerged soils Mn4þ is reduced to Mn2þ
due to oxygen depletion. The Mn2þ concentration in the soil solution increases
when redox potential (Eh ) values decrease. The reduction of Mn4þ occurs after
NO2 3 reduction and before Fe
3þ
reduction. Patrick and Reddy (1978) classified
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…133

soils based on redox potential as aerated or well-drained soils, þ 700 to

þ 500 mV; moderately reduced, þ 400 to þ 200 mV; reduced, þ 100 to
2 100 mV; and highly reduced, 2 100 to 2 300 mV. Under reduced soil con-
ditions, rice uptake of Mn increases. The reduction of Mn can be either chemical
or microbiological, although microbiological reduction is likely to predominate
in flooded rice soils that are at about pH 5.5– 6.0 (Patrick and Reddy, 1978). The
reduction of Mn can be explained from the reduction equation as:

MnO2 þ 4Hþ þ 2e2 $ Mn2þ þ 2H2 O

The adequate Mn concentration for rice growth in water culture experiments has
been reported as 0.1 –0.5 mg L21 (Shimada, 1995) and concentrations higher
than 10 mg L21 may be toxic (Yoshida, 1981).

Functions and symptoms of Mn deficiency and toxicity. Manganese is immobile

in the plant and Mn deficiency symptoms appear initially in the younger leaves.
Manganese deficient plants are chlorotic and develop an irregular yellow
mottling between the leaf veins. In the case of Mn toxicity, yellow spots generally
develop between leaf veins, extend to the interveinal areas, and eventually turn
brown as the toxicity develops (Shimada, 1995). In rice, Mn toxicity is rarely
observed because rice roots can effectively exclude Mn preventing its uptake
(Tanaka et al., 1975) and the rice plant is relatively tolerant to high Mn
concentrations. Excessive uptake of Mn by rice is generally suppressed
antagonistically by the coexistence of high concentrations of Fe2þ.
More than 60% of the Mn contained in higher plant leaves is found in the
chloroplast, and, along with Fe and Cu, performs vital roles in the electron
transport system (Obata, 1995). Manganese also functions as a cofactor to
activate enzymes such as dehydrogenases and hydrolyses in glycolysis system
and citric acid cycle and RNA polymerases in the chloroplasts. The protease
enzyme contained in rice seeds is activated by Mn (Horiguchi and Kitagishi,
1976). The most well-known and extensively studied function of Mn in green
plants is its involvement in photosynthetic O2 evolution (Marschner, 1995).

Critical level in plant and uptake. Manganese deficiency in rice occurs when the
Mn concentration in the plant tissue is less than 20 mg Mn kg21 (Wells et al.,
1993). The critical tissue concentration of Mn in most plants ranges from 10 to
20 mg Mn kg21 in mature leaves, and is surprisingly consistent regardless of
the plant species, cultivar, or the prevailing environmental conditions (Marschner,
1995). Fageria et al. (1997a) reported whole plant Mn concentrations of
30– 600 mg Mn kg21 at tillering were sufficient.
Rice can tolerate tissue levels of more than 2500 mg Mn kg21 without adverse
effects on either growth or grain yield (Wells et al., 1993). Cheng and Quellette
(1971) reported a critical, toxic tissue concentration for rice of 7000 mg
134 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Mn kg21. Yoshida (1981) reported that in many cases, a high Mn content in rice
tissues is frequently associated with high yields, possibly indicating that high Mn
content in the soil is associated with various favorable soil conditions.

Manganese fertilization practices. Soil analysis can be a useful guide for Mn

fertilization of annual crops. Sims and Johnson (1991) reported the critical level
of Mehlich 1 extractable Mn of 5 mg Mn kg21 at soil pH 6.0. These authors also
reported the critical soil Mn concentration range for most crops of 1 –5 mg
Mn kg21 by the DTPA extraction method. Both inorganic and organic Mn
fertilizers are used for correcting Mn deficiencies. The use of MnSO4 is so
common that it is usually included as the standard in research to determine the
efficacy of other Mn sources (Martens and Westermann, 1991). Manganese
deficiency is usually corrected by foliar application of Mn or by banding Mn with
an acid-forming fertilizer. If it is broadcast, a high rate is required. Martens and
Westermann (1991) reported broadcast rates varied from 6 to 45 kg Mn ha21 and
band application rates varied from 3 to 22 kg Mn ha21 for annual crops. If Mn
deficiency occurs, a foliar application of MnSO4 or MnCl2 can be effectively
adopted as control measures. The deficiency often occurs in neutral or alkaline
soils. Application of acidic fertilizers can decrease soil pH and improve Mn
uptake by plants as long as the acidic soil amendments are applied far enough in
advance to reduce the soil pH.
Manganese deficiencies of rice occur on high-pH organic soils used for rice
production in the Florida Everglades Agricultural Area (Snyder et al., 1990).
Manganese drilled with the rice seed at planting was the most effective method of
Mn fertilization on these organic soils. Soaking the rice seed in a Mn solution
before seeding and foliar sprays failed to significantly increase rice growth in
greenhouse studies. Flooding the soil in advance of seeding to allow for soil
reduction to increase the Mn concentration in the soil solution produced mixed
results in studies conducted by Snyder et al. (1990), but could possibly be used
as a cultural practice in waterseeded or transplanted systems to help alleviate
Mn deficiency if Mn fertilizer is not readily available. The Mn status of the main
crop also significantly affected the yield of the ratoon crop. Snyder et al.
(1990) recommended drilling 15 kg Mn ha21 as a water-soluble Mn source with
rice seed as the most economical and effective means of preventing Mn
deficiency of rice.

IV. CONCLUSIONS

Rice is the staple food of approximately one-half of the world’s population.

Total rice production will need to increase to help feed a large proportion of the
world’s increasing population. Increasing rice yields, especially in lowland
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…135

production systems, per unit of land area will be essential in order to meet this
challenge. Forty years ago, when the Green Revolution started, persuading rice
farmers to use modern varieties and accompanying fertilizer inputs was easy
because the results, in terms of yield increases, were often spectacular and
economically attractive. Further improvements in rice productivity, however, are
likely to be much more incremental and knowledge-based. Further yield
increases will mostly result from the positive interactions and integrated
management of various agronomic inputs including nutrients, pest control,
improved cultivars, and water management. Our knowledge of nutrient
management and behavior under the flooded soil conditions used to produce
lowland rice has progressed rapidly over the past several decades. Many nutrient
management issues still require significant improvement and will require a great
of deal of research. Our ability to predict the need for supplemental fertilization
of flood-irrigated rice, via routine plant and soil analysis, needs to be further
refined to optimize production and nutrient use efficiency.
Flooded rice soils are characterized by the absence of oxygen. Oxygen
diffusing into a flooded soil may be consumed by microbial respiration where it is
used as an electron acceptor in the chemical oxidation of reduced Fe2þ and Mn2þ,
the biological oxidation of NH4 and C, and the oxidation of sulfides. Upon
depletion of oxygen, oxidized compounds are reduced by anaerobic organisms in
the soil. Nitrate, NO2, the higher oxides of Mn, hydrated ferric oxide and SO4 will
be reduced if an energy source is available to the microorganisms. The principal
reduction reactions in flooded rice soils in sequence are O2/H2O, NO2 3 /N2,
Mn(IV, III)/Mn(II), Fe(III)/Fe(II), SO22 4 /H2S, and CO2/CH4.
Flooding the soil has a significant effect on the behavior of several essential
plant nutrients and on the growth and yield of rice. Some nutrients are increased
in availability to the crop, whereas others are subjected to greater fixation or loss
from the soil as a result of flooding. Changes in plant nutrient availability
resulting from flooding are due to biological oxidation– reduction processes
brought into play by the depletion of oxygen from the flooded soil. Flooding the
soil results in the potential loss of N through leaching and denitrification. When
managed properly a high level of N use efficiency can be obtained in flood-
irrigated rice systems. However, if N inputs are mismanaged N use efficiency can
be very low.
Saturation of the soil with water increases the availability of soil P to rice. This
has usually been attributed to the reduction of ferric phosphates to the more
soluble ferrous phosphates, and to the hydrolysis of P compounds. The increase in
soil pH of acidic soils as a result of submergence is also considered to aid in the
solution of Fe and Al phosphates. Potassium is less affected by flooding than are
N and P. Reduced soil conditions results in the displacement of exchangeable K
from the exchange complex into the soil solution. This process makes K more
available for uptake by rice, but may also enhance the potential for K leaching on
some soils. The availability of some nutrients, such as Ca and Mg, are changed
136 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

only to a limited extent by the biochemical changes associated with flooding.

Other nutrients, such as S, Fe, and Mn undergo marked changes following
waterlogging. The solubility, and presumably the availability, of Fe and Mn
should be increased as a result of the increased solubility of sparingly soluble
compounds, whereas B, Cu, and Zn availability may decrease due to adsorption
as soil pH increases in acid soils. Reduction of Fe3þ may increase the Fe2þ
concentration in the soil solution to the extent that toxicity can occur.
Modern production agriculture requires efficient, sustainable, and environ-
mentally sound management practices. Under these situations, increasing rice
yield per unit area through the use of appropriate nutrient management practices
has become an essential component of modern rice production technology.
Adoption of proper nutrient management strategies that include the use of
appropriate nutrient sources, rates, and application times used in conjunction with
high-yielding rice cultivars bred for high-nutrient efficiency may reduce
production costs and improve rice yields. Development of efficient nutrient
management recommendations for lowland rice production systems must
integrate our basic knowledge of soil physical properties, nutrient cycles,
biochemical transformation processes, and rice growth and nutrient uptake under
flooded soil conditions. It is clear that the irrigation water must be used to
manipulate these processes to produce nutrient management systems that are
agronomically and environmentally efficient. Many of these basic relationships
are well defined and already recommended in many rice-producing regions, but
require some refinement while others still require significant improvement. Our
challenge is to continue to incorporate new and emerging technologies into
practical management recommendations and effectively demonstrate their value
so that growers readily adopt them as routine.

ACKNOWLEDGMENTS

The authors are grateful to Dr C.D. Foy, USDA-ARS, Beltsville, for peer review
and giving useful comments on the manuscript.

REFERENCES

Abichandani, C. T., and Patnaik, S. (1961). Effect of lime application on availability of nitrogen and
rice yield on waterlogged soils. J. Indian Soc. Soil Sci. 9, 55 –61.
Abilay, W. P., Jr., and De Datta, S. K. (1978). Management practices for correcting zinc deficiency in
transplanted and direct seeded wetland rice. Philippp. J. Crop Sci. 3, 190 –194.
Abrams, M. M., and Jarrell, W. M. (1995). Soil phosphorus as a potential nonpoint source for elevated
stream phosphorus levels. J. Environ. Qual. 24, 132–138.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…137

Adams, F. (1984). Crop response to lime in the Southern United States. Agron. Monogr. No. 12. In “Soil
Acidity and Liming”. (F. Adams, Ed.), (2nd ed.), pp. 211–266. ASA, CSSA, SSSA, Madison, WI.
Adriano, D. C. (1986). In “Trace Elements in the Terrestrial Environment.” Springer, New York.
Allison, F. E. (1973). In “Soil Organic Matter and Its Role in Crop Production.” Elsevier Scientific
Publication Company, Amsterdam.
Amrani, M., Westfall, D. G., and Peterson, G. A. (1999). Influence of water solubility of granular zinc
fertilizers on plant uptake and growth. Plant Nutr. 22, 1815–1827.
Aulakh, M., Khera, T. S., Doran, J. W., Singh, K., and Singh, B. (2000). Yields and nitrogen dynamics
in a rice–wheat system using green manure and inorganic fertilizer. Soil Sci. Soc. Am. J. 64,
1867–1876.
Balasubramanian, V., Ladha, J. K., and Denning, G. L. (1998). In “Resource Management in Rice
Systems: Nutrients.” IRRI, Los Banos, Philippines.
Baligar, V. C., and Fageria, N. K. (1997). Nutrient use efficiency in acid soils: nutrient management
and plant use efficiency. In “Plant–Soil Interactions at Low pH: Sustainable Agriculture
and Forestry Production.” (A. C. Moniz, A. M. C. Furlani, N. K. Fageria, C. A. Rosolem and
H. Cantarells, Eds.), pp. 75– 95. Brazilian Soil Science Society, Campinas, Brazil.
Baligar, V. C., and Fageria, N. K. (1999). Plant nutrient efficiency: towards the second paradigma.
In “Soil Fertility, Biology, and Plant Nutrition Interrelationships.” (J. O. Siqueira, F. M. S.
Lopes, A. S. Lopes, L. R.G. Guilherme, V. Faquin, A. E. Furtini Neto and J. G. Carvalho, Eds.),
pp. 183 –204. Brazilian Soil Science Society, Lavras, Viçosa, Brazil.
Baligar, V. C., Fageria, N. K., and He, Z. L. (2001). Nutrient use efficiency in plants. Commun. Soil
Sci. Plant Anal. 32, 921–950.
Barber, S. A. (1995). In “Soil Nutrient Bioavailability: A Mechanistic Approach” (2nd ed.). Wiley,
New York.
Baruah, K. K., and Nath, B. C. (1997). Ion uptake, metabolism and yield of rice (Oryza sativa L.) at
excess iron in the growth medium. In “Plant Nutrition for Sustainable Food Production and
Environment.” (T. Ando, K. Fujita, T. Mae, H. Matsumoto, S. Mori and J. Sekiya, Eds.),
pp. 403–404. Kluwer Academic Publishers, Dordrecht, The Netherlands.
Baser, R. E., Gilmour, J. T., (1982). Tolerance of rice seedlings to potassium salts. Ark. Agric. Exp.
Stn. Bull. No. 860, Fayetteville.
Batten, G., Marr, K., Williams, R., and Farrell, T. (2000). Mineral concentrations in Australian and
overseas brown rice genotypes. Commun. Soil Sci. Plant Anal. 31, 2393–2400.
Bennett, W. F. (1993). Plant nutrient utilization and diagnostic plant symptoms. In “Nutrient
Deficiencies and Toxicities in Crop Plants.” (W. F. Bennett, Ed.), pp. 1–7. The American
Phytopathological Society, St Paul, MN.
Bennett, R. N., and Wallsgrove, R. M. (1994). Secondry metabolites in plant defense mechanisms.
New Phytol. 127, 617–633.
Benckiser, G., Ottow, J. C. G., Watanbe, I., and Santiago, S. (1984). The mechanism of excessive iron-
uptake (iron toxicity) of wetland rice. J. Plant Nutr. 7, 177– 185.
Berge, H. F. M., and Riethoven, J. J. M. (1997). Application of a simple rice nitrogen model. In
“Plant Nutrition for Sustainable Food Production and Environment.” (T. Ando, K. Fujita, T. Mae,
H. Matsumoto, S. Mori and J. Sekiya, Eds.), pp. 793 –798. Kluwer Academic Publishers,
Dordrecht.
Blair, G. J., Mamaril, C. P., and Momuat, E. (1978), IRRI Res. paper Ser. 21 In “Sulfur nutrition of
Wetland Rice”, 28p. IRRI, Los Banos, Philippines.
Bollich, P. K., Lindau, C. W., and Norman, R. J. (1994). Management of fertilizer nitrogen in dry-
seeded, delayed-flood rice. Aust. J. Exp. Agric. 34, 1007– 1012.
Bowen, J. E. (1986). Kinetics of zinc uptake by two rice cultivars. Plant Soil 94, 99– 107.
Bowen, J. E. (1987). Physiology of genotypic differences in Zn and Cu uptake in rice and tomato.
In “Genetic Aspects of Plant Mineral Nutrition.” (W. H. Gabelman and B. C. Loughman,
Eds.), pp. 413–423. Nijhoff, Dordrecht, The Netherlands.
138 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Brady, N. C., and Weil, R. R. (1996). In “The Nature and Properties of Soils” 11th ed., Prentice-Hall,
Upper Saddle River, New Jersey.
Broadbent, F. E., De Datta, S. K., and Laureles, E. V. (1987). Measurement of nitrogen utilization
efficiency in rice genotypes. Agron. J. 79, 786–791.
Bufogle, A., Bollich, P. K., Kovar, J. L., Lindau, C. W., and Macchiavellid, R. R. (1998). Comparison
of ammonium sulfate and urea as nitrogen sources in rice production. J. Plant Nutr. 21,
1601–1614.
Buresh, R. J., De Datta, S. K., Padilla, J. L., and Samson, M. I. (1988). Field evaluation of two urease
inhibitors with transplanted lowland rice. Agron. J. 80, 763–768.
Cakmak, I., Torun, B., Erenoglu, B., Ozturk, L., Marschner, H., Kalayci, M., Ekiz, H., and Yilmaz, A.
(1998). Morphological and physiological differences in the response of cereals to zinc deficiency.
Euphytica 100, 349 –357.
Campbell, J. M., and Schwertmann, U. (1984). Iron oxides mineralogy of placic horizons. J. Soil Sci.
35, 569 –582.
Carreres, R., Sendra, J., Ballesteros, R., and Garcia De De La Cuadra, J. (2000). Effects of preflood
nitrogen rate and midseason nitrogen timing on flooded rice. J. Agric. Sci. Cambridge 134,
379–390.
Cassman, K. G., Kropff, M. J., Gaunt, J., and Peng, S. (1993). Nitrogen use efficiency of rice
reconsidered: what are the key constraints? Plant Soil 155/156, 359–362.
Cassman, K. G., Peng, S., Olk, D. C., Ladha, J. K., Reichardt, W., Dobermann, A., and Singh, U.
(1998). Opportunities for increased nitrogen-use efficiency from improved resource management
in irrigated rice systems. Field Crops Res. 56, 7–39.
Cayton, M. T. C., Reyes, E. D., and Neue, H. U. (1985). Effect of zinc fertilization on the mineral
nutrition of rices differing in tolerance to zinc deficiency. Plant Soil 87, 319–327.
Chang, T. T. (1975). The origin, evolution, cultivation, dissemination, and diversification of Asian and
African rices. Euphytica 25, 425–441.
Chang, S. C., and Jackson, M. L. (1957). Fractionation of soil phosphorus. Soil Sci. 84, 133–144.
Cheaney, R. L., and Jennings, P. R. (1975). In “Field Problems of Rice in Latin America”, 90p. Centro
Internacional de Agricultura Tropical, Cali, Colombia.
Cheng, B. T., and Quellette, G. J. (1971). Manganese availability in soils. Soils Fert. 34, 589–595.
Chlopecka, A., and Adriano, D. C. (1996). Mimicked in situ stabilization of metals in a cropped soil:
bioavailability and chemical form of zinc. Environ. Sci. Technol. 30, 3294– 3303.
Chlopecka, A., and Adriano, D. C. (1997). Zinc uptake by plants on amended polluted soils. In “Plant
Nutrition for Sustainable Food Production and Environment.” (T. Ando, K. Fujita, T. Mae,
H. Matsumoto, S. Mori and J. Sekiya, Eds.), pp. 527–532. Kluwer Academic Publishers,
Dordrecht.
Clark, R. B. (1990). Physiology of cereals for mineral nutrient uptake, use, and efficiency. In “Crops as
Enhancers of Nutrient Use.” (V. C. Baligar and R. R. Duncan, Eds.), pp. 131– 209. Academic
Press, San Diego, CA.
Clark, R. B., and Duncan, R. R. (1991). Improvement of plant mineral nutrition through breeding.
Field Crops Res. 27, 219 –240.
Clark, S. D., Norman, R. J., Slaton, N. A., and Wilson, C. E., Jr. (2001). Influence of nitrogen fertilizer
source, application timing, and rate on grain yield of rice. Ark. Agric. Exp. Stn. Res. Ser. 485.
In “B.R. Wells Arkansas Rice Research Studies 2000.” (R. J. Norman and J. F. Meullenet, Eds.),
pp. 352 –357, Fayetteville, AR.
Cochrane, T. T., Sanchez, L. G., Azevedo, L. G. de, Porras, J. A. and Garver, C. L. (1984). Land in
tropical America. CIAT/EMBRAPA-CPAC, Cali, Colombia..
Counce, P. A., and Wells, B. R. (1990). Rice plant population density effect on early-season nitrogen
requirement. J. Prod. Agric. 3, 390– 393.
Counce, P. A., Wells, B. R., and Gravois, K. A. (1992). Yield and harvest-index responses to preflood
nitrogen fertilization at low rice plant populations. J. Prod. Agric. 5, 492–497.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…139

Cox, T. S., and Frey, K. J. (1978). Agronomy Abstract. In “Nitrogen harvest index in oats”, p. 50.
American Society of Agronomy, Madison, WI.
Craswell, E. T., and Vlek, P. L. G. (1979). Fate of fertilizer nitrogen applied to wetland rice. In
“Nitrogen and Rice.” ( IRRI, Ed.), pp. 175 –192. IRRI, Los Banos, Philippines.
De Datta, S. K. (1981). In “Principles and Practices of Rice Production.” Wiley, New York.
De Datta, S. K., and Buresh, R. J. (1989). Integrated nitrogen management in irrigated rice. Adv. Soil
Sci. 10, 143 –169.
De Datta, S. K., and Mikkelsen, D. S. (1985). Potassium nutrition of rice. In “Potassium in
Agriculture.” (R. D. Munson, Ed.), pp. 665–699. American Society of Agronomy, Madison, WI.
De Datta, S. K., Buresh, R. J., Samson, M. I., and Kai-Rong, W. (1988). Nitrogen use efficiency and
nitrogen-15 balances in broadcast-seeded flooded and transplanted rice. Soil Sci. Soc. Am. J. 52,
849 –855.
DeLong, R. E., Carroll, S. D., and Baker, W. H. (2001). Soil test and fertilizer sales data: summary for
the growing season 2000. Ark. Agric. Exp. Stn. Res. Ser. 463. In “Arkansas Soil Fertility Series
2000.” (R. J. Norman and S. L. Chapman, Eds.), pp. 1 –17, Fayetteville, AR.
Dingkuhn, M., Schnier, H. F., De Datta, S. K., Dorffling, K., and Javellana, C. (1991). Relationships
between ripening phase productivity and crop duration, canopy photosynthesis and senescence in
transplanted and direct-seeded lowland rice. Field Crops Res. 26, 327 –345.
Dobermann, A., and Fairhurst, T. (2000). In “Rice: Nutritional Disorders and Nutrient Management.”
Potash and Phosphate Institute of Canada/IRRI, Singapore/Los Banos.
Dobermann, A., Cassman, K. G., Sta. Cruz, P. C., Adviento, M. A., and Pampolino, M. F. (1996a).
Fertilizer inputs, nutrient balance, and soil nutrient-supplying power in intensive, irrigated rice
systems. I. Effective soil K-supplying capacity. Nutr. Cycling Agroecosyst. 46, 11– 21.
Dobermann, A., Sta. Cruz, P. C., and Cassman, K. G. (1996b). Fertilizer inputs, nutrient balance, and
soil nutrient-supplying power in intensive, irrigated rice systems. I. Potassium uptake and K
balance. Nutr. Cycling Agroecosyst. 46, 1–10.
Dobermann, A., Cassman, K. G., Mamaril, C. P., and Sheehy, J. E. (1998). Management of phosphorus,
potassium and sulfur in intensive, irrigated lowland rice. Field Crops Res. 56, 113– 138.
Dobermann, A., Dawe, D., Roetter, R. P., and Cassman, K. G. (2000). Reversal of rice yield decline in
a long-term continuous cropping experiment. Agron. J. 92, 633 –643.
Duncan, R. R., and Carrow, R. N. (1999). Turfgrass-molecular genetic improvement for abiotic/
edaphic stress environment. Adv. Agron. 67, 233 –306.
Dunn, R. J. (1978). A study of boron and arsenic as straighthead inducing agents of rice. M.S. Thesis.
University of Arkansas.
Dunn, D., Jones, S. (2001). Boron fertilization of rice (Available on-line at http://agebb.missouri.edu/
rice/research/00/pg10.htm (Verified 03 May 2002).
Eagle, A. J., Bird, J. E., Hill, J. A., Horwath, W. R., and van Kessel, C. (2001). Nitrogen dynamics and
fertilizer use efficiency in rice following straw incorporation and winter flooding. Agron. J. 93,
1346–1354.
Eckert, D. J. (1987). Soil test interpretations: basic cation saturation ratios and sufficiency levels. SSSA
Special Publication Number 21. In “Soil Testing: Sampling, Correlation, Calibration, and
Interpretation.” (J. R. Brown, Ed.), pp. 53–64, Madison, WI.
Eichner, M. J. (1990). Nitrous oxide emissions from fertilized soils: summary of available data.
J. Environ. Qual. 19, 272 –280.
Elkins, J. W., and Rossen, R. (1989). In “Summary Report 1988: Geophysical Monitoring for Climatic
Change.” National Oceanic Atmospheric Administration Environmental Research Laboratory,
Boulder, CO.
Eneji, A. E., Yamamoto, S., and Honna, T. (2001). Rice growth and nutrient uptake as affected by
livestock manure in four Japanese soils. J. Plant Nutr. 24, 333– 343.
Evans, L. K., and Smillie, G. W. (1976). Extractable iron and aluminum and their relationship to
phosphate retention in Irish soils. Ir. J. Agric. Res. 15, 65 –73.
140 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Fageria, N. K. (1980). Influence of phosphorus application on growth, yield and nutrient uptake by
irrigated rice. R. Bras. Ci. Solo 4, 26–31.
Fageria, N. K. (1984). In “Fertilization and Mineral Nutrition of Rice.” EMBRAPA-CNPAF/Editora
Campus, Rio de Janeiro.
Fageria, N. K. (1986). In “Final Report of the Project, Evaluation Potassium Need for Rice”, 53p.
National Rice and Bean Research Center of EMBRAPA, Goiania, Brazil.
Fageria, N. K. (1989). Tropical soils and physiological aspects of crop production. Brasilia:
EMBRAPA-DPU, Brasilia, EMBRAPA-CNPAF. Document, 18. 425p.
Fageria, N. K. (1991). Response of rice to fractional applied potassio in Brazil. Better Crops Int. 7, 19.
Fageria, N. K. (1992). In “Maximizing Crop Yields.” Marcel Dekker, New York.
Fageria, N. K. (1999a). Mineral nutrition. In “Rice Crop in Brazil.” (N. R. A. Vieira, A. B. Santos and
E. P. Santana, Eds.), pp. 172– 196. Embrapa Arroz e Feijão, Santo Antonio de Goias, Brazil.
Fageria, N. K. (1999b). Liming and fertilization. In “Rice Crop in Brazil.” (N. R. A. Vieira, A. B.
Santos and E. P. Santana, Eds.), pp. 329 –353. Embrapa Arroz e Feijão, Santo Antonio de Goias,
Brazil.
Fageria, N. K. (2000a). Adequate and toxic levels of zinc for rice, common bean, corn, soybean and
wheat production in cerrado soil. R. Bras. Eng. Agric. Ambiental 4, 390–395.
Fageria, N. K. (2000b). Adequate and toxic levels of boron for rice, common bean, corn, soybean and
wheat production in cerrado soil. R. Bras. Eng. Agric. Ambiental 4, 57–62.
Fageria, N. K. (2000c). Upland rice response to soil acidity in cerrdao soil. Pesq. Agropec. Bras. 35,
2303–2307.
Fageria, N. K. (2001). Screening method of lowland rice genotypes for zinc uptake efficiency. Scientia
Agricola 58, 623–626.
Fageria, N. K. (2002). Dry matter yield of common bean, lowland rice, corn, soybean, and wheat
at different basic cation saturation ratios in an acid soil. Commun. Soil Sci. Plant Anal. 33,
519–531.
Fageria, N. K., and Baligar, V. C. (1993). Proceedings of the Workshop on Adaptation of Plants to Soil
Stress. In “Screening crop genotypes for mineral stresses,” pp. 142 –159. University of Nebraska,
Lincoln, NE.
Fageria, N. K., and Baligar, V. C. (1996a). Response of lowland rice and common bean grown in
rotation to soil fertility levels on a varzea soil. Fert. Res. 45, 13– 20.
Fageria, N. K., and Baligar, V. C. (1999b). Growth and nutrient concentrations of common bean,
lowland rice, corn, soybean, and wheat at different soil pH on an Inceptisol. J. Plant Nutr. 22,
1495–1507.
Fageria, N. K., and Baligar, V. C. (2001). Lowland rice response to nitrogen fertilization. Commun.
Soil Sci. Plant Anal. 32, 1405–1429.
Fageria, N. K., and Barbosa Filho, M. P. (1994). In “Nutritional Deficiency in Rice: Identification and
Correction,” 36p. Embrapa Arroz e Feijão, Goiania, Brazil.
Fageria, N. K., and Barbosa Filho, M. P. (2001). Nitrogen use efficiency in lowland rice genotypes.
Commun. Soil Sci. Plant Anal. 32, 2079–2090.
Fageria, N. K., and Gheyi, H. R. (1999). In “Efficient Crop Production.” Federal University of Paraiba,
Campina Grande, Brazil.
Fageria, N. K., and Rabelo, N. A. (1987). Tolerance of rice cultivars to iron toxicity. J. Plant Nutr. 10,
653–661.
Fageria, N. K., Baligar, V. C., and Edwards, D. G. (1990a). Soil –plant nutrient relationships at low pH
stress. In “Crops as Enhancers of Nutrient Use.” (V. C. Baligar and R. R. Duncan, Eds.),
pp. 475 –507. Academic Press, San Diego, CA.
Fageria, N. K., Baligar, V. C., and Wright, R. J. (1990b). Iron nutrition of plants: an overview on the
chemistry and physiology of its deficiency and toxicity. Pesq. Agropec. Bras. 25, 553–570.
Fageria, N. K., Baligar, V. C., Wright, R. J., and Carvalho, J. R. P. (1990c). Lowland rice response to
potassium fertilization and its effect on N and P uptake. Fert. Res. 21, 157 –162.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…141

Fageria, N. K., Baligar, V. C., Wright, R. J., and Sousa, C. M. R. (1991). Characterization of physical
and chemical properties of varzea soils of Goias State of Brazil. Commun. Soil Sci. Plant Anal.
22, 1631–1646.
Fageria, N. K., Baligar, V. C., and Jones, C. A. (1997a). In “Growth and Mineral Nutrition of Field
Crops” 2nd ed., Marcel Dekker, New York, NY.
Fageria, N. K., Santos, A. B., and Baligar, V. C. (1997b). Phosphorus soil test calibration for lowland
rice on an Inceptisol. Agron. J. 89, 737–742.
Fageria, N. K., Baligar, V. C., and Clark, R. B. (2002). Micronutrients in crop production. Adv. Agron.
77, 185–268.
FAO (2001). FAOSTAT. Statistics database, (Online) (Subset Fertilizer within Agriculture database)
Available at http://apps.fao.org/ (verified 7 May 2002).
Farina, M. P., Channon, W. P., and Thibaud, G. R. (2000). A comparison of strategies for ameliorating
subsoil acidity: I. Long-term growth effects. Soil Sci. Soc. Am. J. 64, 646– 651.
Fawcett, J.A., (1980). Nitrogen harvest index variation within two Avena species. M.S. Thesis. Iowa
State University, Ames (OCLC 6827928).
Fawcett, J. A., and Frey, K. J. (1983). Associations among nitrogen harvest index and other traits
within two Avens species. Proc. Iowa Acad. Sci. 90, 150 –153.
Flach, K. W., and Slusher, D. F. (1978). Soils used for rice culture in the United States. In “Soils and
Rice.” ( IRRI, Ed.), pp. 199–215. IRRI, Los Banos, Philippines.
Forno, D. A., Yoshida, S., and Asher, C. J. (1975). Zinc deficiency in rice. I. Soil factors associated
with the deficiency. Plant Soil 42, 537– 550.
Foth, H. D., and Ellis, B. G. (1988). In “Soil Fertility.” Wiley, New York.
Fox, R. L., and Kamprath, E. J. (1970). Phosphate sorption isotherms for evaluating the phosphate
requirements of soils. Soil Sci. Soc. Am. Proc. 34, 902–907.
Freney, J. R., Trevitt, A. C. F., De Datta, S. K., Obcemea, W. N., and Real, J. G. (1990). The
interdependence of ammonia volatilization and denitrification as nitrogen loss processes in
flooded rice fields in the Philippines. Biol Fert. Soils 9, 31–36.
Fried, M., Zsoldos, F., Vose, D. B., and Shatokhin, I. L. (1965). Characterizing the NO3
and NH4 uptake process of rice plants by use of 15N labelled NH4NO3. Physiol. Plant 18,
313 –320.
Friesen, D. K., Rao, I. M., Thomas, R. J., Oberson, A., and Sanz, J. I. (1997). Phosphorus acquisition
and cycling in crop and pasture systems in low fertility tropical soils. In “Plant Nutrition for
Sustainable Food Production and Environment.” (T. Ando, K. Fujita, T. Mae, H. Matsumoto,
S. Mori and J. Sekiya, Eds.), pp. 493 –498. Kluwer Academic Publishers, Dordrecht.
Galloway, J. N., Schlesinger, W. H., Levy, H., Michaels, A., and Schnoor, J. L. (1995). Nitrogen
fixation: anthropogenic enhancement-environmental response. Global Biogeochem. Cycles 9,
235 –252.
Gangloff, W. J., Westfall, D. G., Peterson, G. A., and Mortvedt, J. J. (2002). Relative availability
coefficients of organic and inorganic Zn fertilizers. J. Plant Nutr. 25, 259 –274.
Gaudin, R., and Dupuy, J. (1999). Ammonical nutrition of transplanted rice fertilized with large urea
granules. Agron. J. 91, 33–36.
Genon, J. G., Hepcee, N., Duffy, J. E., Delvaux, B., and Hennebert, P. A. (1994). Iron and other
chemical soil constraints to rice in highlands swamps of Burundi. Plant Soil 166, 109–111.
Gilmour, J. T. (1977a). Micronutrient status of the rice plant. I. Plant and soil solution concentrations
as a function of time. Plant Soil 46, 549–557.
Gilmour, J. T. (1977b). Micronutrient status of the rice plant. II. Micronutrient uptake rate as a
function of time. Plant Soil 46, 558–564.
Giordano, P. M., and Mordvedt, J. J. (1972). Rice response to zinc in flooded and nonflooded soil.
Agron. J. 64, 521 –524.
Giordano, P. M., and Mordvedt, J. J. (1973). Zinc sources and methods of application for rice. Agron.
J. 65, 51–53.
142 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Glass, A. D. M., Siddiqi, M. Y., and Giles, K. I. (1981). Correlations between potassium uptake and
hydrogen efflux in barley varieties. Plant Physiol. 68, 457–459.
Goswami, N. N., and Banerjee, N. K. (1978). Soils and Rice. In “Phosphorus, potassium, and other
macroelements.” ( IRRI, Ed.), pp. 561–580. IRRI, Los Banos, Philippines.
Graham, R. D. (1983). Effects of nutrient stress on susceptibility of plants to disease with particular
reference to the trace elements. Adv. Bot. Res. 10, 221–276.
Graham, R. D. (1984). Breeding for nutritional characteristics in cereals. Adv. Plant Nutr. 1, 57– 102.
Graham, R. D., and Rengel, Z. (1993). Genotypic variation in zinc uptake and utilization by plants. In
“Zinc in Soils and Plants.” (A. D. Robson, Ed.), pp. 107 –118. Kluwer Academic Publishers,
Dordrecht, The Netherlands.
Graham, R. D., Ascher, J. S., and Hynes, S. C. (1992). Selecting zinc efficient cereal genotypes for
soils of low zinc status. Plant Soil 146, 241–250.
Graham, R. D., Senadhira, D., and Ortiz-Monasterio, I. (1997). A strategy for breeding staple-food
crops with high micronutrient density. In “Plant Nutrition for Sustainable Food Production and
Environment.” (T. Ando, K. Fujita, T. Mae, H. Matsumoto, S. Mori and J. Sekiya, Eds.),
pp. 933 –937. Kluwer Academic Publishers, Dordrecht.
Guerra, L. C., Bhuiyan, S. I., Tuong, T. P., and Baker, R. (1998). Discussion Paper Series No. 29. In
“Producing More Rice with Less Water from Irrigated Systems,” 19p. IRRI, Los Banos,
Philippines.
Guindo, D., Wells, B. R., and Norman, R. J. (1994a). Cultivar and nitrogen rate influence on nitrogen
uptake and partitioning in rice. Soil Sci. Soc. Am. J. 58, 840–845.
Guindo, D., Norman, R. J., and Wells, B. R. (1994b). Accumulation of fertilizer nitrogen-15 by rice at
different stages of development. Soil Sci. Soc. Am. J. 58, 410–415.
Gunawarkena, I., Virmani, S. S., and Sumo, F. J. (1982). Breeding rice for tolerance to iron toxicity.
Oryza 19, 5–12.
Haghighat, N. G., Thompson, L. F (1982). Zinc seed coating studies with rice. In the 19th
Proceedings of the Rice Technical Working Group, University of Arkansas, Hot Springs, AR,
23–25 February, 1982. Texas Agric. Exp. Stn. Texas A&M University, College Station, TX,
pp. 74 –75.
Hanaway, J. J., and Johonson, J. W. (1985). Potassium nutrition of soybean. In “Potassium in
Agriculture.” (R. D. Munson, Ed.), pp. 753–764. ASA, CSSA, and SSSAJ, Madison, WI.
Hanson, J. B. (1984). The function of calcium in plant nutrition. In “Advances in Plant Nutrition.”
(P. B. Tinker and A. Lauchli, Eds.), pp. 149– 208. Praeger, New York.
Hargrove, T. R. (1988). Rice production leaps forward. Span 30, 114–115.
Harter, R. D. (1991). Micronutrient adsorption–desorption reactions in soils. In “Micronutrients
in Agriculture.” 2nd ed., (J. J. Mortvedt, F. R. Fox, L. M. Shuman and R. M. Welch, Eds.),
pp. 59– 87. SSSA, Madison, WI.
Hazra, G. C., and Mandal, B. (1996). Desorption of adsorbed zinc in soils in relation to soil properties.
J. Indian Soc. Soil Sci. 44, 233–237.
Hewitt, E. J. (1963). The essential nutrients: requirements and interactions in plants. In “Plant
Physiology.” (F. C. Steward, Ed.), pp. 137 –360. Academic Press, New York.
Hill, J. E., Roberts, S. R., Brandon, D. M., Scardaci, S. C., Williams, J. F., Wick, C. M., Canevari, W.
M., and Weir, B. L. (1992). Publication 21498. In “Rice Production in California.” Division of
Agriculture and Natural Research, University of California, California.
Hirata, H. (1995). Science of Rice Plant: Physiology. In “Absorption and Metabolism of Potassium.”
(T. Matsuo, K. Kumazawa, R. Ishii, K. Ishihara and H. Hirata, Eds.), Vol. 2, pp. 383– 390. Food
and Agricultural Policy Research Center, Tokyo, Japan.
Hodgson, J. F. (1963). Chemistry of the micronutrient elements in soils. Advan. Agron. 15, 119– 159.
Hodgson, J. F., Lindsay, W. L., and Trierweiler, J. T. (1966). Micronutrient cation complexing in soil
solution. II. Complexing of zinc and copper in displaced solution from calcareous soils. Soil Sci.
Soc. Am. Proc. 30, 723–726.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…143

Horiguchi, T., and Kitagishi, K. (1976). Studies on rice seed protease: metal ion activation of rice seed
peptidase. J. Sci. Soil Manure, Japan 22, 73–80.
Hossner, L. R., Freeouf, J. A., and Folsom, B. L. (1973). Solution phosphorus concentration and
growth of rice (Oryza sativa L.) in flooded soils. Soil Sci. Soc. Am. Proc. 37, 405–408.
Howler, R. H. (1973). Iron-induced oranging disease of rice in relation to physico-chemical changes in
a flooded Oxisol. Soil Sci. Soc. Am. Proc. 37, 898–903.
Huang, C., Webb, M. J., and Graham, R. D. (1996). Pot size affects expression of Mn efficiency in
barley. Plant Soil 178, 205–208.
Hudnall, W. H. (1991). Taxonomy of acid rice growing soils of the tropics. In “Rice Production on
Acid Soils of the Tropics.” (P. Deturck and F. N. Ponnamperuma, Eds.), pp. 3–8. Institute of
Fundamental Studies, Kandy, Sri Lanka.
Hutchinson, G. L., and Davidson, E. A. (1993). Processes for production and consumption of gaseous
nitrogen oxides in soil. ASA Special Publication 55. In “Agriculture Ecosystem Effects on Trace
Gases and Global Climate Change.” (L. A. Harper, A. R. Mosier, J. M. Duxbury and D. E.
Rolston, Eds.), pp. 79–93. ASA, CSSA, and SSSA, Madison, WI.
IICA, (eds), (2000). In “Annual Reports for 1999,” 17p. IICA/PROCITROPICOS, Brasilia, Brazil.
Inada, K. (1966a). Studies on the bronzing disease of rice plant in Ceylon. I. Effect of field treatment
on bronzing occurrence and changes in leaf respiration induced by disease. Tropp. Agric (Ceylon)
122, 19–29.
Inada, K. (1966b). Studies on the bronzing disease of rice plant in Ceylon. II. Cause of the occurrence
of bronzing. Tropp. Agric (Ceylon) 125, 31 –46.
International Rice Research Institute, (eds), (1965). In “Annual Reports 1964,” 335p. International
Rice Research Institute, Los Banos, Philippines.
International Rice Research Institute, (eds), (1977). In “IRRI Annual Report for 1976.” International
Rice Research Institute, Los Banos, Philippines.
International Rice Research Institute, (eds), (1984). In “Terminology for Rice Growing Environ-
ments.” IRRI, Los Banos, Philippines.
Isfan, D. (1993). Genotypic variability for physiological efficiency index of nitrogen in oats. Plant Soil
154, 53–59.
Ishizuka, S. (1978). In “Nutrient Deficiencies of Crops,” 100p. Food and Fertilizer Technology
Center, Taipei, Taiwan.
Ishizuka, Y., and Tanaka, A. (1952a). Biochemical studies on the life history of rice plants.
I. Absorption and translocation of inorganic elements. J. Sci. Soil Manure Jpn 23, 23–28.
Jarrell, W. M., and Beverly, R. B. (1981). The dilution effect in plant nutrition studies. Adv. Agron. 34,
197 –224.
Kamprath, E. J., and Watson, M. E. (1980). Conventional soil and tissue tests for assessing the
phosphorus status of soils. In “The Role of Phosphorus in Agriculture.” (F. E. Khasawneh,
E. C. Sample and E. J. Kamprath, Eds.), pp. 433 –469. ASA, CSA, and SSSA, Madison, WI.
Kawasaki, T. (1995). Science of Rice Plant: Physiology. In “Metabolism and Physiology of Calcium
and Magnesium.” (T. Matsuo, K. Kumazawa, R. Ishii, K. Ishihara and H. Hirata, Eds.), Vol. 2,
pp. 391–395. Food and Agricultural Policy Research Center, Tokyo, Japan.
Khalid, R. A., Patrick, W. H., Jr., and Peterson, F. J. (1979). Relationship between rice yield and
soil phosphorus evaluated under aerobic and anaerobic conditions. Soil Sci. Plant Nutr. 25,
155 –164.
Kirkby, E. A., and Pilbeam, D. J. (1984). Calcium as a plant nutrient. Plant Cell Environ. 7, 397–405.
Kittrick, J. A. (1976). Control of Zn2þ in the soil solution by sphalerite. Soil Sci. Soc. Am. Proc. 40,
314 –317.
Kiuchi, T., and Ishizaka, H. (1961). Effects of nutrients on the yield constituting factors of rice. J. Soil
Sci. Manures Jpn 32, 198–202.
Kundu, D. K., Ladha, J. K., and Lapitan-de-Guzman, L. (1996). Tillalge depth influence on soil
nitrogen distribution and availability in a rice lowland. Soil Sci. Soc. Am. J. 60, 1153–1159.
144 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Lin, S. C., and Yuan, L. P. (1980). Hybrid rice breeding in China. In “Innovative Approaches to Rice
Breeding.” ( IRRI, Ed.), pp. 35 –37. IRRI, Manila, Philippines.
Lindsay, W. L. (1979). Zinc in soils and plant nutrition. Adv. Agron. 24, 147–181.
Liscano, J. F., Wilson, C. E., Jr., Norman, R. J., and Slaton, N. A. (2001). Ark. Agric. Exp. Stn Res.
Bull. No. 963. In “Zinc availability to rice from seven granular fertilizers”, Fayetteville, AR.
Loffler, C. M., and Bush, R. H. (1982). Selection for grain protein, grain yield, and nitrogen
partitioning efficiency in hard red spring wheat. Crop Sci. 22, 591–595.
Mae, T. (1997). Physiological nitrogen efficiency in rice: nitrogen utilization, photosynthesis, and
yield potential. In “Plant Nutrition for Sustainable Food Production and Environment.” (T. Ando,
K. Fujita, T. Mae, H. Matsumoto, S. Mori and J. Sekiya, Eds.), pp. 51–60. Kluwer Academic
Publishers, Dordrecht, The Netherlands.
Mahadevappa, M., Ikehashi, H., and Aurin, P. (1981). Screening rice genotypes for tolerance to
alkalinity and zinc deficiency. Euphytica 30, 253 –257.
Mandal, S. C., Sinha, H., and Prasad, C. R. (1966). Studies on liming of acidic red loam soils.
I. Response of crop to liming. J. Indian Soc. Soil Sci. 14, 127– 131.
Mandal, B., Hazra, G. C., and Mandal, L. N. (2000). Soil management influences on zinc desorption
for rice and maize nutrition. Soil Sci. Soc. Am. J. 64, 1699–1705.
Marschner, H. (1995). In “Mineral Nutrition of Higher Plants” 2nd ed., Academic Press, New York.
Martens, D. C., and Westermann, D. T. (1991). Fertilizer applications for correcting micro-
nutrient deficiencies. In “Micronutrient in Agriculture.” 2nd ed., (J. J. Mortvedt, F. R. Cox,
L. M. Shuman and R. M. Welch, Eds.), pp. 549–592. Soil Science Society of America,
Madison, WI.
Maskina, M. S., Singh, B., Singh, Y., and Meelu, O. P. (1988). Fertilizer requirement of rice–wheat
and maize– wheat rotations on coarse-textured soils amended with farmyard manure. Fert. Res.
17, 153 –164.
Matsuo, H., Pecrot, A. J., and Riquier, J. (1978). Rice soils of Europe. In “Soils and Rice.”
(International Rice Research Institute, Ed.), pp. 191 –198. IRRI, Los Banos, Philippines.
McGee, J. B., Slaton, N. A., DeLong, R. E., Wilson, C. E., Jr., Norman, R. J., (2002). Rate and time of
phosphorus fertilization on efficiency of phosphorus uptake by rice. In the 29th Proceedings of the
Rice Technical Working Group, University of Arkansas, Little Rock, AR, 24– 27 February, 2002.
Agricultural Center, Louisana State University, Crowley, LA.
McNeal, F. H., Berg, M. A., and Watson, C. A. (1966). Nitrogen and dry matter in five spring wheat
varieties at successive stages of development. Agron. J. 58, 605–608.
Mengel, K. (1985). Dynamics and availability of major nutrients in soils. Adv. Soil Sci. 2, 65– 131.
Mengel, K., and Viro, M. (1978). The significance of plant energy status for the uptake and
incorporation of NH4-N by young rice plants. Soil Sci. Plant Nutr. 24, 407 –416.
Mengel, D. B., Leonards, W. J., Sedberry, J. E., Jr (1976). Effect of zinc oxide seed treatment on
stand establishment, leaf zinc concentrations and yield of Saturn rice. In 68th Annu. Prog. Repp.
Rice Expp. Stn. Crowley, LA. Louisiana State Univ. Agric. Exp. Stn. Baton Rouge, LA,
pp. 62 –63.
Mengel, K., Bubl, W., and Scherer, H. W. (1984). Iron distribution in vine leaves with HCO2 3 induced
chlorosis. J. Plant Nutr. 7, 715–724.
Mikkelsen, D. S. (1987). Nitrogen budgets in flooded soils used for rice production. Plant Soil 100,
71–97.
Mikkelsen, D. S., and Brandon, D. M. (1975). Zinc deficiency in California rice. Calif. Agric. 29,
8–9.
Mikkelsen, D. S., and Kuo, S. (1976). Zinc fertilization and behavior in flooded soils. In “The Fertility
of Paddy Soils and Fertilizer Application for Rice.” ( Food and Fertilizer Technology Center,
Ed.), pp. 170 –196. Food and Fertilizer Technology Center, Taipei, Taiwan.
Moll, R. H., Kamprath, E. J., and Jackson, W. A. (1982). Analysis and interpretation of factors which
contribute to efficiency of nitrogen utilization. Agron. J. 74, 562–564.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…145

Moore, P. A., Jr., and Patrick, W. H., Jr. (1988). Effect of zinc deficiency on alcohol dehydrogenase
activity and nutrient uptake in rice. Agron. J. 80, 882–885.
Moore, P. A., Jr., Gilmour, J. T., and Wells, B. R. (1981). Seasonal patterns of growth and soil nitrogen
uptake by rice. Soil Sci. Soc. Am. J. 45, 875–879.
Moraes, J. F. V. (1999). Soils. In “Rice Culture in Brazil.” (N. R. A. Vieira, A. B. Santos and E. P.
Santana, Eds.), pp. 88–115. Embrapa Arroz e Feijão, Santo Antônio de Goiás, Brazil.
Mori, S., Nishazawa, N., Hayashi, H., Chino, M., Yoshimura, E., and Ishihara, J. (1991). Why are
youngrice plants highly susceptible to iron deficiency? Plant Soil 130, 143–156.
Moristsugu, M (1980). Comparative plant nutritional studies on the effect of ammonium-N and
nitrate-N on plant growth, with special emphasis on the nutrient uptake under stabilized pH of
culture solution. PhD Thesis. Kyoto University, pp. 52–54.
Mortvedt, J. J. (1994). Need for controlled availability of micronutrient fertilizers. Fert. Res. 38,
213 –221.
Motomura, S. (1962). Effect of organic matters on the formation of ferrous iron in soils. Soil Sci. Plant
Nutr. 8, 20–29.
Moormann, F. R. (1978). Morphology and classification of soils on which rice is grown. In “Soils and
Rice.” (International Rice Research Institute, Ed.), pp. 255 –272. IRRI, Los Banos, Philippines.
Mueller, K. E. (1974). In “Field Problems of Tropical Rice,” 95pp. International Rice Research
Institute, Los Banos, Philippines.
Murthy, R. S. (1978). Rice soils of India. In “Soils and Rice.” (International Rice Research Institute,
Ed.), pp. 3–17. IRRI, Los Banos, Philippines.
Nelson, L. E. (1980). Phosphorus nutrition of cotton, peanuts, rice, sugarcane, and tobacco. In “The
Role of Phosphorus in Agriculture.” (F. E. Khasawneh, E. C. Sample and E. J. Kamprath, Eds.),
pp. 693–736. ASA, CSSA, and SSSA, Madison, WI.
Nelson, D. W. (1982). Gaseous losses of nitrogen other than through denitrification. Agronomy
Monograph 22. In “Nitrogen in Agricultural Soils.” (F. J. Stevenson, Ed.), pp. 327 –363. ASA,
CSSA, and SSSA, Madison, WI.
Noguchi, Y., and Sugawara, T. (1976). In “Potassium and Japonica rice.” International Potash
Institute, Berne.
Norman, R. J., Wells, B. R., and Helms, R. S. (1988). Effect of nitrogen source, application time and
dicyandiamide on rice yields. J. Fert. Issues 5, 76–82.
Norman, R. J., Wells, B. R., and Moldenhauer, K. A. K. (1989). Effect of application method and
dicyandiamide on urea-nitrogen 15 recovery in rice. Soil Sci. Soc. Am. J. 53, 1269– 1274.
Norman, R. J., Gilmour, J. T., and Wells, B. R. (1990). Mineralization of nitrogen from nitrogen-15
labeled crop residues and utilization by rice. Soil Sci. Soc. Am. J. 54, 1351–1356.
Norman, R. J., Guindo, D., Wells, B. R., and Wilson, C. E., Jr. (1992a). Seasonal accumulation and
partitioning of nitrogen-15 in rice. Soil Sci. Soc. Am. J. 56, 1521–1527.
Norman, R. J., Helms, R. S., and Wells, B. R. (1992b). Influence of delaying flood and preflood
nitrogen application on dry-seeded rice. Fert. Res. 32, 55–59.
Norman, R. J., Wolf, D. C., Wells, B. R., Helms, R. S., and Slaton, N. A. (1993). Influence of
application time and soil moisture condition on yield and recovery of fertilizer 15N in dry-seeded
rice. Ark. Agri. Exp. Sta. Res. Ser. 425. In “Arkansas Soil Fertility Studies 1992.” (W. E. Sabbe,
Ed.), pp. 7–10, Fayetteville, AR.
Norman, R. J., Wells, B. R., Helms, R. S., Wilson, C. E., Jr., Slaton, N. A., and Beyrouty, C. A. (1994).
Influence of split applying the preflood nitrogen fertilizer on rice growth and accumulation and
partitioning of nitrogen by the rice plant. Ark. Agri. Exp. Stn. Res. Ser. 439. In “Arkansas Rice
Research Studies 1993.” (B. R. Wells, Ed.), pp. 138 –145, Fayetteville, AR.
Norman, R. J., Wilson, C. E., Jr., Slaton, N. A., Gravois, K. A., and Moldenhauer, K. A.K. (1997).
Grain yield response of new rice cultivar/varieties to nitrogen fertilization. Ark. Agric. Exp. Stn.
Res. Ser. 456. In “B.R. Wells Arkansas Rice Research Studies 1996.” (R. J. Norman and
T. H. Johnston, Eds.), pp. 125–129, Fayetteville, AR.
146 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Norman, R. J., Wilson, C. E., Jr., and Slaton, N. A. (2003). Soil fertilization and mineral nutrition in US
mechanized rice culture. In “Rice: Origin, History, Production, and Technology.” (C. W. Smith,
Ed.), pp. 331 –411. Wiley, New York.
Ntamatungiro, S., Slaton, N. A., Wilson, C. E., Jr., Daniels, M., Robinson, J. R., Ashlock, L., and
Windham, T. (1999). Effects of lime, phosphorus, and zinc application, on rice and soybean
production. Ark. Agric. Exp. Stn. Res. Ser. 468. In “B.R. Wells Rice Research Studies 1998.” (R.
J. Norman and T. H. Johnston, Eds.), pp. 268–276, Fayetteville, AR.
Obata, H. (1995). Physiological functions of micro essential elements. In “Science of Rice Plant:
Physiology.” (T. Matsuo, K. Kumazawa, R. Ishii, K. Ishihara and H. Hirata, Eds.), Vol. 2,
pp. 402 –419. Food and Agricultural Policy Research Center, Tokyo, Japan.
O’Brien, I. E.W., and Ferguson, I. B. (1997). Calcium signalling in programmed cell death in plants.
In “Plant Nutrition for Sustainable Food Production and Environment.” (T. Ando, K. Fujita,
T. Matsumoto, H. Matsumoto, S. Mori and J. Sekiya, Eds.), pp. 99 –103. Kluwer Academic
Publishers, Dordrecht.
Okuda, A., Hori, S., and Ida, S. (1961). Boron nutrition in higher plants. I. A method of growing boron
deficient plants. J. Sci. Soil Manure, Japan 32, 153 –157.
Olaleye, A. O., Tabi, F. O., Ogunkunle, A. O., Singh, B. N., and Sahrawat, K. L. (2001). Effect of toxic
iron concentrations on the growth of lowland rice. J. Plant Nutr. 24, 441– 457.
Olson, R. V. (1980). Fate of tagged nitrogen fertilizer applied to irrigated corn. Soil Sci. Soc. Am. J. 44,
514–517.
Olson, R. V., and Swallow, C. W. (1984). Fate of labeled nitrogen applied to winter wheat for five
years. Soil Sci. Soc. Am. J. 48, 583–586.
Ottow, J. C.G., Benckiser, G., Watanabe, I., and Santiago, S. (1983). Multiple nutritional soil stress as
the prerequisite for iron toxicity of wetland rice (Oryza sativa L.). Tropp. Agric (Trinidad) 60,
102–106.
Panabokke, C. R. (1978). Rice soils of Sri Lanka. In “Soils and Rice.” (International Rice Research
Institute, Ed.), pp. 19– 33. IRRI, Los Banos, Philippines.
Patella, J. F. (1976). In “Rice in Flooded Soils: Adequate Use of Fertilizers,” 76p. Nobel, São Paulo,
Brazil.
Patrick, W. H., Jr., and Mahapatra, I. C. (1968). Transformation and availability to rice of nitrogen and
phosphorus in waterlogged soils. Adv. Agron. 20, 323–359.
Patrick, W. H., Jr., and Mikkelsen, D. S. (1971). Plant nutrient behavior in flooded soil. In “Fertilizer
Technology and Use.” 2nd ed., (R. A. Olson, Ed.), pp. 187–215. Soil Science Society of
America, Madison, WI.
Patrick, W. H., Jr., and Reddy, C. N. (1978). Chemical changes in rice soils. In “Soils and Rice.”
(International Rice Research Institute, Ed.), pp. 361–379. International Rice Research Institute,
Los Banos, Philippines.
Patrick, W. H., Jr., and Wyatt, R. (1964). Soil nitrogen loss as a result of alternate submergence and
drying. Soil Sci. Soc. Am. J. 28, 647– 653.
Patrick, W. H., Jr., Peterson, F. J., and Wilson, F. E. (1974). Response of lowland rice to time and
method of application of phosphate. Agron J. 66, 459 –460.
Patrick, W. H., Jr., Mikkelsen, D. S., and Wells, B. R. (1986). Plant nutrient behavior in flooded soils.
In “Fertilizer Technology and Use.” (O. P. Englested, Ed.), (3rd ed.), pp. 197– 228. Soil Science
Society of America, Madison, WI.
Pedrazzini, F. R., and McKee, K. L. (1984). Effect of flooding on activities of soil dehydro-
genases and alcohol dehydrogenase in rice (Oryza sativa L.) roots. Soil Sci. Plant Nutr. 30,
359–366.
Peng, S., Garcia, F. V., Gines, H. C., Laza, R. C., Samson, M. I., Sanico, A. L., Visperas, R. M., and
Cassman, K. G. (1996). Nitrogen use efficiency of irrigated tropical rice established by broadcast
wet-seeding and transplanting. Fert. Res. 45, 123– 134.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…147

Peterson, D. M., Schrader, L. E., Cataldo, D. A., Young, V. L., and Smith, D. (1975). Assimilation and
remobilization of nitrogen and carbohydrate in oats, especially as related to groat protein
concentration. Can. J. Plant Sci. 55, 19 –28.
Ponnamperuma, F. N. (1960). Fertilizer experiments in farmers fields in Ceylon. Tropp. Agric. 116,
253 –267.
Ponnamperuma, F. N. (1972). The chemistry of submerged soils. Adv. Agron. 24, 29–96.
Ponnamperuma, F. N. (1975). Micronutrient limitations in acid tropical rice soils. In “Soil
Management in Tropical America.” (E. Bornemisza and A. Alvarado, Eds.), pp. 330–347. North
Carolina State University, Raleigh, NC.
Ponnamperuma, F. N. (1976). Screening rice for tolerance to mineral stresses. In “Plant Adaptation to
Mineral Stress in Problem Soils.” (M. J. Wright, Ed.), pp. 341–353. Cornell University Press,
Ithaca, NY.
Ponnamperuma, F. N. (1976). Specific soil chemical characteristics for rice production in Asia. IRRI
Res. Paper Ser. 2, 18.
Quang, V. D., and Dufey, J. E. (1995). Effect of temperature and flooding duration on phosphate
sorption in an acid sulphate soil from Vietnam. Eur. J. Soil Sci. 46, 641 –647.
Randall, P. J., and Wrigley, C. W. (1986). Effects of sulfur supply on the yield, composition, and
quality of grain from cereals, oilseeds, and legumes. Adv. Cereals Sci. Tech. 8, 171– 206.
Rao, A. C.S., Smith, J. L., Papendick, R. L., and Parr, J. F. (1991). Influence of added nitrogen
interactions in estimating recovery efficiency of labeled nitrogen. Soil Sci. Soc. Am. J. 55,
1616–1621.
Rattan, R. K., and Shukla, L. M. (1984). Critical limits of deficiency and toxicity of zinc in paddy in a
typic ustipsamment. Commun. Soil Sci. Plant Anal. 15, 1041–1050.
Rattunde, H. F., and Frey, K. J. (1986). Nitrogen harvest index in oats: its repeatability and association
with adaptation. Crop Sci. 26, 606 –610.
Raun, W., and Johnson, G. V. (1999). Improving nitrogen use efficiency for cereal production. Agron.
J. 91, 357–363.
Raymundo, M. E. (1978). Rice soils of the Philippines. In “Soils and Rice.” (International Rice
Research Institute, Ed.), pp. 115 –133. IRRI, Los Banos, Philippines.
Reddy, C. N., and Patrick, W. H., Jr. (1976). Yield and nitrogen utilization by rice as affected by
method and time of application of labeled nitrogen. Agron. J. 68, 965 –969.
Rengel, Z. (1997). Root exudation and microflora population in rhizosphere of crop genotypes
differing in tolerance to micronutrient deficiency. Plant Soil 196, 255– 260.
Rengel, R. D., and Graham, R. D. (1996). Uptake of zinc from chelate buffered nutrient solutions by
wheat genotypes differing in Zn efficiency. J. Exp. Bot. 47, 217 –226.
Reynolds, J. G., Naylor, D. V., and Fendorf, S. E. (1999). Arsenic sorption in phosphate-amended soils
during flooding and subsequent aeration. Soil Sci. Soc. Am. J. 63, 1149–1156.
Roberts, S. R., Hill, J. E., Brandon, D. M., Miller, B. C., Scardacai, S. C., Wick, C. M., and Williams,
J. F. (1993). Biological yield and harvest index in rice: nitrogen response of tall and semidwarf
cultivars. J. Prod. Agric. 6, 585–588.
Romheld, V., and Marschner, H. (1991). Function of macronutrients in plants. In “Micronutrients
in Agriculture.” (J. J. Mortvedt, F. R. Cox, L. M. Shuman and R. M. Welch, Eds.), (2nd ed.),
pp. 297–328. Soil Science Society of America, Madison, WI.
Roy, A. C., and De Datta, S. K. (1985). Phosphate sorption isotherms for evaluating phosphorus
requirement of wetland rice soils. Plant Soil 86, 185–196.
Rush, M. C. (1972). Effects of seed treatment with four zinc sources on stand and yield of Saturn rice.
In 64th Annual Prog. Rep. Rice Exp. Stn, Crowley, LA. Louisiana State Univ. Agric. Exp. Stn.
Baton Rouge, pp. 269–270.
Sah, R. N., and Mikkelsen, D. S. (1986a). Transformations of inorganic phosphorus during the
flooding and draining cycles of soil. Soil Sci. Soc. Am. J. 50, 62–67.
148 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Sah, R. N., and Mikkelsen, D. S. (1986b). Sorption and bioavailability of phosphorus during the
drainage period of flooded-drained soils. Plant Soil 92, 265–278.
Sajwan, K. S., and Lindsay, W. L. (1986). Effects of soil redox on zinc deficiency in paddy rice. Soil
Sci. Soc. Am. J. 50, 1264–1269.
Santos, A. B., Fageria, N. K., Stone, L. F., and Santos, C. (1999). Water and potassium management in
the irrigated rice. Pesq. Agropec. Brasileira 34, 565–573.
Sanyal, S. K., and De Datta, S. K. (1991). Chemistry of phosphorus transformations in soil. Adv. Soil
Sci. 16, 1–120.
Schmidt, E. L. (1982). Nitrification in soil. Agronomy Monograph 22. In “Nitrogen in Agriculture
Soils.” (F. J. Stevenson, Ed.), pp. 253 –288. ASA, CSSA, and SSSA, Madison, WI.
Schmied, B., Abbaspour, K., and Schulin, R. (2000). Inverse estimation of parameters in a nitrogen
model using field data. Soil Sci. Soc. Am. J. 64, 533 –542.
Schnier, H. F. (1994). Nitrogen-15 recovery fraction in flooded tropical rice as affected by added
nitrogen interaction. Eur. J. Agron. 3, 161–167.
Sedberry, J. E., Jr. Schilling, P. G., Wilson, F. E., Peterson, F. J (1978). Diagnosis and correction of
zinc problems in rice production. St. Univ. Agri. Exp. Sta. Bul. 708. Baton Rouge, LA.
Sedbery, J. E., Jr., Peterson, F. J., Wilson, F. E., Mengel, D. B., Schilling, P. E., and Brupbacher, R. H.
(1980). Influence of soil reaction and applications of zinc on yields and zinc contents of rice
plants. Commun. Soil Sci. Plant Anal. 11, 283–295.
Seneweera, S. P., and Conroy, J. P. (1997). Growth, grain yield and quality of rice (Oryza sativa L.) in
response to elevated CO2 and phosphorus nutrition. In “Plant Nutrition for Sustainable Food
Production and Environment.” (T. Ando, K. Fujita, T. Mae, H. Matsumoto, S. Mori and J. Sekiya,
Eds.), pp. 873–878. Kluwer Academic Publishers, Dordrecht, The Netherlands.
Shahandeh, H., Hossner, L. R., and Turner, F. T. (1994). Phosphorus relationships in flooded rice soils
with low extractable phosphorus. Soil Sci. Soc. Am. J. 58, 1184–1189.
Shainberg, I., Sumner, M. E., Miller, W. P., Farina, M. P.W., Pavan, M. A., and Fey, M. V. (1989). Use
of gypsum on soils: a review. Adv. Soil Sci. 9, 1–111.
Sharply, A. N., Sims, J. T., and Pierzynski, G. M. (1994). Innovative soil phosphorus availability
indices: assessing inorganic phosphorus. SSSA Special Publication 40. In “Soil Testing: Pros-
pects for Improving Nutrient Recommendations.” (L. P. Wilding, Ed.), pp. 115 –142. SSSA,
Madison, WI.
Sharpley, A. N., Daniel, T., Sims, T., Lemunyon, J., Stevens, R., and Parry, R. (1999). Agricultural
Research Service, ARS-149. In “Agricultural Phosphorus and Eutrophication,” 42pp. US
Department of Agriculture.
Shimada, N. (1995). Science of Rice Plant: Physiology. In “Deficiency and Excess of Micronutrient
Elements.” (T. Matsuo, K. Kumazawa, R. Ishii, K. Ishihara and H. Hirata, Eds.), Vol. 2,
pp. 412 –419. Food and Agricultural Policy Research Center, Tokyo.
Shorrocks, V. M. (1997). The occurrence and correction of boron deficiency. In “Boron in Soils and
Plants: Reviews.” (B. Dell, P. H. Brown and R. W. Bell, Eds.), pp. 121– 148. Kluwer Academic
Publishers, Dordrecht, The Netherlands.
Siddiqi, M. Y., and Glass, A. D. M. (1981). Utilization index: a modified approach to the estimation
and comparison of nutrient utilization efficiency in plants. J. Plant Nutr. 4, 289–302.
Sillanpaa, M. (1982). Soils Bulletin 48. In “Micronutrients and nutrient status soils: a global study.”
FAO, Rome.
Sims, J. R., and Bingham, F. T. (1968). Retention of boron by layer silicates, and soil materials. II.
Sesquioxides. Soil Sci. Soc. Am. Proc. 32, 364– 369.
Sims, J. T., and Johnson, G. V. (1991). Micronutrient soil tests. In “Micronutrients in Agriculture.”
(J. J. Mortvedt, F. R. Cox, L. M. Shuman and R. M. Welch, Eds.), (2nd ed.), pp. 427– 476. Soil
Science Society of America, Madison, WI.
Sims, J. L., and Place, G. A. (1968). Growth and nutrient uptake of rice at different growth stages and
nutrient levels. Agron. J. 60, 692–696.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…149

Singh, D., McLaren, R. G., and Cameron, K. C. (1997). Desorption of native and added
zinc from a range of New Zealand soils in relation to soil properties. Aust. J. Soil Res. 35,
1253–1266.
Singh, U., Ladha, J. K., Castillo, E. G., Punzalan, G., Tirol-Padre, A., and Duqueza, M. (1998).
Genotypic variation in nitrogen use efficiency in medium and long duration rice. Field Crops Res.
58, 35–53.
Singh, Y., Dobermann, A., Singh, B., Bronson, K. F., and Khind, C. S. (2000). Optimal phosphorus
management strategies for wheat–rice cropping on a loamy sand. Soil Sci. Soc. Am. J. 64,
1413–1422.
Slaton, N. A. (1998). The influence of elemental sulfur amendments on soil chemical properties
and rice growth. PhD Dissertation. University of Arkansas, Fayetteville, Dissertation Abstract
AAT-983-9310.
Slaton, N. A., Beyrouty, C. A., Wells, B. R., Norman, R. J., and Gbur, E. E. (1990). Root growth and
distribution of two short-season rice genotypes. Plant Soil 121, 269– 278.
Slaton, N. A., Cartwright, C. D., and Wilson, C. E., Jr. (1995). Potassium deficiency and plant diseases
observed in rice fields. Better Crops 79, 12–14.
Slaton, N. A., Wilson, C. E., Jr., Ntamatungiro, S., and Norman, R. J. (1998). Rice response to
phosphorus application timing. Better Crops 82, 10–12.
Slaton, N. A., Cartwright, R. D., Wilson, C. E., Jr., and Norman, R. J. (2001a). Symptoms and
diagnosis of late-season sulfur deficiency of rice in Arkansas. Ark. Agric. Exp. Stn. Res. Ser. 485.
In “B.R. Wells Rice Research Studies 2000.” (R. J. Norman and J. F. Meullenet, Eds.), pp.
388 –394, Fayetteville, AR.
Slaton, N. A., Norman, R. J., Boothe, D. L., Ntamatungiro, S., Clark, S. D., Wilson, C. E., Jr., and
DeLong, R. E. (2001b). Potassium nutrition of rice: summary of 2000 research studies. Ark.
Agric. Exp. Stn. Res. Ser. 485. In “B.R. Wells Rice Research Studies 2000.” (R. J. Norman and J.
F. Meullenet, Eds.), pp. 395–404, Fayetteville, AR.
Slaton, N. A., Wilson, C. E., Jr., Ntamatungiro, S., Norman, R. J., and Boothe, D. L. (2001c).
Evaluation of zinc seed treatment for rice. Agron. J. 93, 152–157.
Slaton, N. A., Wilson, C. E., Jr., Norman, R. J., and Gbur, E. E., Jr. (2002). Development of a critical
Mehlich 3 soil test zinc concentration for rice in Arkansas. Commun. Soil Sci. Plant Anal. 33,
2759–2770.
Snyder, G. H., and Jones, D. B. (1988). Prediction and prevention of iron-related rice seedling
chlorosis on everglades Histosols. Soil Sci. Soc. Am. J. 52, 1043–1046.
Snyder, G. H., and Jones, D. B. (1991). Post-emergence treatment of iron-related rice-seedling
chlorosis. Plant Soil 138, 313 –317.
Snyder, G. H., Jones, D. B., and Coale, F. J. (1990). Occurrence and correction of manganese
deficiency in Histosol-grown rice. Soil Sci. Soc. Am. J. 54, 1634–1638.
Soepraptohardjo, M., and Suhardjo, H. (1978). Rice soils of Indonesia. In “Soils and Rice.”
(International Rice Research Institute, Ed.), pp. 99–113. IRRI, Los Banos, Philippines.
Sparks, D. L., and Huang, P. M. (1985). Physical chemistry of soil potassium. In “Potassium in
Agriculture.” (R. D. Munson, Ed.), pp. 201–276. American Society of Agronomy, Madison, WI.
Stansel, J. W. (1975). In Six Decades of Rice Research in Texas, Agri. Exp. Sta. Res. Mono. 4. In “The
Rice Plant—Its Development and Yield,” pp. 9–21. Texas A&M University, Texas.
Stutte, C. A., and da Silva, P. R. F. (1981). Nitrogen volatilization from rice leaves. I. Effect of
genotype and air temperature. Crop Sci. 21, 596 –600.
Su, N. R. (1976). The Fertility of Paddy Soils and Fertilizer Application for Rice. In
“Potassium fertilization of rice,” pp. 117 –148. Food and Fertilizer Technology Center, Taipei,
Taiwan.
Suzuki, A. (1995). Science of Rice Plant: Physiology. In “Metabolism and physiology of sulfur.”
(T. Matsuo, K. Kumazawa, R. Ishii, K. Ishihara and H. Hirata, Eds.), Vol. 2, pp. 395–401. Food
and Agricultural Policy Research Center, Tokyo.
150 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

Swift, R. S., and Mclaren, R. G. (1991). Micronutrient sorption by soils and colloids. In “Interactions
at the Soil Colloid– Soil Solution Interface.” (G. H. Bolt et al., Ed.), pp. 257 –292. Kluwer
Academic Publishers, Dordrecht, The Netherlands.
Takai, Y. (1978). An oxidation – reduction process in the soil under submerged conditions.
In “Pedology on Paddy Soils.” (K. Kawaguchi, Ed.), pp. 23 –55. Kodansha, Tokyo.
Takenaga, H. (1995). Science of Rice Plant: Physiology. In “Internal Factors in Relation to Nutrient
Absorption.” (T. Matsuo, K. Kumazawa, R. Ishii, K. Ishihara and H. Hirata, Eds.), Vol. 2,
pp. 294 –309. Food and Agricultural Policy Research Center, Tokyo.
Tanaka, A., Loe, R., and Navasero, S. A. (1966). Some mechanism involved in the development of
iron toxicity symptoms in the rice plant. Soil Sci. Plant Nutr. 12, 158– 162.
Tanaka, A., Mulleriyawa, R. P., and Yasu, T. (1968). Possibility of hydrogen sulfide induced iron
toxicity of the rice plant. Soil Sci. Plant Nutr. 14, 106.
Tanaka, A., Tadano, T., and Fujita, H. (1975). Comparative studies on plant nutrition: adaptability to
heavy metals among crop species. I. Adaptability to manganese. J. Soil Sci. Manure 46, 425– 430.
Tanaka, A., Tadano, T., and Akiyama, Y. (1977). Comparative studies on plant nutrition: adaptability
to potassium. J. Soil Sci. Jpn 48, 175–180.
Teo, Y. H., Beyrouty, C. A., Norman, R. J., and Gbur, E. E. (1995). Nutrient uptake relationship to root
characteristics of rice. Plant Soil 171, 297– 302.
Thorn, K. A., and Mikita, M. A. (2000). Nitrite fixation by humic substances: nitrogen-15 nuclear
magnetic resonance evidence for potential intermediates in chemodenitrification. Soil Sci. Soc.
Am. J. 64, 568 –582.
Tiessen, H., Cuevas, E., and Chacon, P. (1994). The role of soil organic matter in sustaining soil
fertility. Nature 371, 783– 785.
Tisdale, S. L., Nelson, W. L., and Beaton, J. D. (1985). In “Soil Fertility and Fertilizers” (4th ed.).
Macmillan, New York.
Tiwari, K. N., and Dwivedi, B. S. (1994). Fertilizer Zn needs of rice (Oryza sativa L.) as influenced by
native soil Zn in Udic Ustochrepts of the Indo-Ganetic plains. Tropp. Agric (Trinidad) 71, 17 –21.
Traore, A., and Maranville, J. W. (1999). Nitrate reductase activity and diverse grain sorghum
genotypes and its relationship to nitrogen use efficiency. Agron. J. 91, 863–869.
Turner, F. T., and Gilliam, J. W. (1976). Increased P processes in flooded soils. Plant Soil 45,
365–377.
USDA (United States Department of Agriculture), (eds), (1975). USDA Agriculture Handbook 436. In
“Soil taxonomy: a basic system of soil classification for making and interpreting soil surveys,”
754p. US Government Printing Office, Washington, DC.
US Environmental Protection Agency (1996). Environmental indicators of water quality in the United
States, EPA 841-R-96-002.
Wada, G., Shoji, S., and Mae, T. (1986). Relation between nitrogen absorption and growth and yield of
rice plants. JARQ 20, 135 –145.
Wallace, T. (1961). In “The Diagnosis of Mineral Deficiencies in Plants by Visual Symptom” (2nd
ed.). Chemical Publishing, New York.
Wander, M. M., Traina, S. J., Stinner, B. R., and Peters, S. E. (1994). Organic and conventional
management effects on soil biologically active organic matter pools. Soil Sci. Soc. Am. J. 58,
1130–1139.
Wang, T. S. L. (1971). Effect of CaCO3, CaSiO3, and organic manure on the growth and yield of rice
on flooded acid latosolic soil. J. Taiwan Agric. Res. 20, 47–55.
Wang, C. H. (1976). Sulphur fertilization of rice. In “The Fertility of Paddy Soils and Fertilizer
Application for Rice.” (Food and Fertilizer Technology Center, Ed.). Food and Fertilizer
Technology Center, Taipei, Taiwan.
Wang, F. L., and Alva, A. K. (2000). Ammonium adsorption and desorption in sandy soils. Soil Sci.
Soc. Am. J. 64, 1669–1674.
 NUTRIENT MANAGEMENT FOR IMPROVING LOWLAND RICE…151

Wang, C. H., Liem, T. H., and Mikkelsen, D. S. (1976). In “Sulfur Deficiency—A Limiting Factor in
Rice Production in the Lower Amazon Basin. II. Sulfur Requirement for Rice Production.” IRI
Research Institute, New York.
Welch, R. M. (1993). Zinc concentrations and forms in plants for humans and animals. In “Zinc in Soil
and Plants.” (A. D. Robson, Ed.), pp. 183–195. Kluwer Academic Publishers, Dordrecht, The
Netherlands.
Welch, R. W., and Yong, Y. Y. (1980). The effects of variety and nitrogen fertilizer on protein
production in oats. J. Sci. Food Agric. 31, 541–548.
Wells, B. R. (1980). Ark. Agric. Exp. Stn. Res. Bull. No. 848. In “Zinc Nutrition of Rice Growing on
Arkansas Soils.” University of Arkansas, Fayettville, AR.
Wells, B. R., and Faw, W. F. (1978). Short-statured rice response to seeding and N rate. Agron. J. 70,
477 –480.
Wells, B. R., and Turner, F. T. (1984). Nitrogen use in flooded rice soils. In “Nitrogen Use in Crop
Production.” (R. D. Hauck, Ed.), pp. 349 –362. ASA, CSSA, and SSSA, Madison, WI.
Wells, B. R., Huey, B. A., Norman, R. J., and Helms, R. S. (1993). Rice. In “Nutrient Deficiencies and
Toxicities in Crop Plants.” (W. F. Bennett, Ed.), pp. 15 –19. The American Phytopathological
Society, St Paul, MN.
Wells, B. R., Bacon, R. K., Dilday, R., Kelly, J. T., and Dickson, P. A. (1995). Response of wheat
following rice to fall fertilization. Ark. Agric. Exp. Stn Res. Ser. 443. In “Arkansas Soil Fertility
Studies 1994.” (W. E. Sabbe, Ed.), pp. 16 –20, Fayetteville, AR.
Westcott, M. P., Brandon, D. M., Lindau, C. W., and Patrick, W. H., Jr. (1986). Effects of seeding
method and time of fertilization on urea-nitrogen-15 recovery in rice. Agron. J. 78, 474–478.
Westerman, R. L., and Kurtz, L. T. (1974). Isotopic and non-isotopic estimation of fertilizer and
nitrogen uptake by sudan grass in field experiments. Soil Sci. Soc. Am. Proc. 38, 107 –109.
Westfall, D. G., Anderson, W. B., and Hodges, R. J. (1971). Iron and zinc response of chlorotic rice
grown on calcareous soils. Agron. J. 63, 702 –705.
Westfall, D. G., Flinchum, W. T., and Stansel, J. W. (1973). Distribution of nutrients in the rice plant
and effect of two nitrogen levels. Agron. J. 65, 236–238.
Willett, I. R. (1982). Phosphorus availability in soils subjected to short periods of flooding and drying.
Aust. J. Soil Res. 20, 131– 138.
Willett, I. R. (1986). Phosphorus dynamics in relation to redox processes in flooded soils. Transactions
of 13th Congress International Soil Science Society, Hamburg, Vol. 6, pp. 748 –755.
Willett, I. R. (1989). Causes and predictions of changes in extractable phosphorus during flooding.
Aust. J. Soil Res. 27, 45–54.
Willett, I. R., and Higgins, M. L. (1978). Phosphorus sorption by reduced and reoxidized rice soils.
Aust. J. Soil Res. 16, 319– 326.
Williams, J., and Smith, S. G. (2001). Correcting potassium deficiency can reduce rice stem diseases.
Better Crops 85, 7 –9.
Wilson, C. E., Jr., Norman, R. J., and Wells, B. R. (1989). Seasonal uptake patterns of fertilizer
nitrogen applied in split application to rice. Soil Sci. Soc. Am. J. 53, 1884–1887.
Wilson, C. E., Jr., Wells, B. R., and Norman, R. J. (1994). Fertilizer nitrogen uptake by rice from
urea –ammonium nitrate solution vs. granular urea. Soil Sci. Soc. Am. J. 58, 1825–1828.
Wilson, C. E., Jr., Slaton, N. A., Dickson, P. A., and Norman, R. J. (1996). Phosphorus fertilizer
management for rice grown on alkaline soils. Ark. Agric. Exp. Stn. Res. Ser. 453. In “Arkansas
Rice Research Studies 1995.” (R. J. Norman and B. R. Wells, Eds.), pp. 196–200, Fayetteville,
AR.
Wilson, C. E., Jr., Bollich, P. K., and Norman, R. J. (1998). Nitrogen application timing effects on
nitrogen efficiency of dry-seeded rice. Soil Sci. Soc. Am. J. 62, 959–964.
Wilson, C. E., Jr., Slaton, N. A., Ntamatungiro, S., and Norman, R. J. (1999). Phosphorus
fertilizer management for rice produced on alkaline soils. Ark. Agric. Exp. Stn. Res. Ser. 468.
152 N. K. FAGERIA, N. A. SLATON AND V. C. BALIGAR

In “B.R. Wells Rice Research Series 1998.” (R. J. Norman and T. H. Johnston, Eds.),
pp. 310 –316, Fayetteville, AR.
Wilson, C. E., Jr., Slaton, N. A., Norman, R. J., and Miller, D. M. (2001). Efficient use of fertilizer. Ark
Coop. Ext. Serv. Misc. Publ. No. 192. In “Rice Production Handbook.” (N. A. Slaton, Ed.),
pp. 51– 74, Litttle Rock, AR.
Wiren, N. V., Gazzarrini, S., and Frommer, W. B. (1997). Regulation of mineral nitrogen uptake
in plants. In “Plant Nutrition for Sustainable Food Production and Environment.” (T. Ando,
K. Fujita, T. Mae, H. Matsumoto, S. Mori and J. Sekiya, Eds.), pp. 41–49. Kluwer Academic
Publishers, Dordrecht, The Netherlands.
Witt, C., Dobermann, A., Abdulrachman, S., Gines, H. C., Wang, G. H., Nagrajan, R.,
Satawathananont, S., Son, T. T., Tan, P. S., Tiem, L. V., Simbahan, G. C., and Olk, D. C.
(1999). Internal nutrient efficiencies of irrigated lowland rice in tropical and subtropical Asia.
Field Crops Res. 63, 113 –138.
Wu, P., Luo, A., Zhu, J., Yang, N. H., and Senadhira, D. (1997). Molecular markers linked to genes
underlying seedling tolerance for ferrous iron toxicity. In “Plant Nutrition for Sustainable Food
Production and Environment.” (T. Ando, K. Fujita, T. Mae, H. Matsumoto, S. Mori and J. Sekiya,
Eds.), pp. 789–792. Kluwer Academic Publishers, Dordrecht, The Netherlands.
Yamaguchi, J. (1991). Fertilizer-nitrogen absorption determined by the 15N isotopic and difference
methods. Jpn. Agric. Res. Q. 25, 93–100.
Yamaguchi, J. (1997). Sulfur status of rice and lowland soils in West Africa. In “Plant Nutrition
for Sustainable Food Production and Environment.” (T. Ando, K. Fujita, T. Mae, H. Matsumoto,
S. Mori and J. Sekiya, Eds.), pp. 813– 814. Kluwer Academic Publishers, Dordrecht, The
Netherlands.
Yang, X. (1987). Physiological mechanisms of nitrogen efficiency in hybrid rice. PhD Dissertation.
Zhejiang Agricultural University, Hangzhou, China.
Yang, X., and Sun, X. (1988). Physiological characteristics of F1 hybrid rice roots. In “Hybrid Rice.”
(IRRI, Ed.), pp. 159–164. IRRI, Los Banos, Philippines.
Yang, X., Zhang, J., and Ni, W. (1999). International Rice Research Notes. In “Characteristics of
Nitrogen Nutrition in Hybrid Rice,” pp. 245–248. IRRI, Los Banos, Philippines.
Ye, Z., and Yang, Y. (1997). Decomposition characteristics of organic manure in soil and influence of
manure application on extractable Fe and Zn. In “Plant Nutrition for Sustainable Food Production
and Environment.” (T. Ando, K. Fujita, T. Mae, H. Matsumoto, S. Mori and J. Sekiya, Eds.),
pp. 597 –598. Kluwer Academic Publishers, Dordrecht, The Netherlands.
Yoshida, S. (1972). Physiological aspects of grain yield. Annu. Rev. Plant Physiol. 23, 437–464.
Yoshida, S. (1981). In “Fundamentals of Rice Crop Science.” IRRI, Los Banos, Philippines.
Yoshida, S., Ahn, J. S., and Forno, D. A. (1973). Occurence, diagnosis and correction of zinc
deficiency of lowland rice. Soil Sci. Plant Nutr. 19, 83–93.
Yu, X., and Bell, P. F. (1998). Nutrient deficiency symptoms and boron uptake mechanisms of rice.
J. Plant Nutr. 21, 2077–2088.
Zhao, F. J., Withers, P. J. A., Evans, E. J., Monagham, J., Salmin, S. E., Shewry, P. R., and McGrath,
S. P. (1997). Sulphur nutrition: an important factor for the quality of wheat and rapeseed. In “Plant
Nutrition for Sustainable Food Production and Environment.” (T. Ando, K. Fujita, T. Mae,
H. Matsumoto, S. Mori and J. Sekiya, Eds.), pp. 917–922. Kluwer Academic Publishers,
Dordrecht, The Netherlands.
Zia, M. S., Aslam, M., Ali, A., and Saeed, Z. (1997). Fertilizer management and nitrogen use
efficiency for irrigated rice grown on calcareous-alkaline soils. In “Plant Nutrition for Sustainable
Food Production and Environment.” (T. Ando, K. Fujita, T. Mae, H. Matsumoto, S. Mori and
J. Sekiya, Eds.), pp. 815–816. Kluwer Academic Publishers, Dordrecht, The Netherlands.
 MOLECULAR BREEDING OF
GROUNDNUT FOR ENHANCED
PRODUCTIVITY AND FOOD
SECURITY IN THE SEMI- ARID TROPICS:
OPPORTUNITIES AND CHALLENGES
S.L. Dwivedi,1 J.H. Crouch,1 S.N. Nigam,1 M.E. Ferguson1 and
A.H. Paterson2
1
International Crops Research Institute for the Semi-Arid Tropics (ICRISAT),
Patancheru, PO 502324, A.P., India
2
Plant Genome Mapping Laboratory, University of Georgia, Rm. 162, Riverbend
Research Center, 110 Riverbend Road, Athens, GA 30602, USA

1. Introduction
A. Crop Production and Uses
B. Constraints to Production
II. Genetic Variability in Groundnut
A. Wild Arachis Species
B. Cultivated Groundnut
III. Highlights and Constraints of Conventional Genetic
Improvement in Groundnut
A. Breeding Methods and Cultivars
B. Successes and Limitations to Conventional Breeding
IV. Overview of Biotechnology Applications in Groundnut
A. Molecular Genetic Diversity Analysis
B. Molecular Genetic Linkage Mapping
C. Progress in Model Systems and Comparative Mapping
Amongst Legumes
D. Wide Crosses
E. Genetic Transformation
V. Opportunities for Molecular Breeding in Groundnut
A. Current Understanding and Genetic Basis of Economically
Important Traits
B. Developing Appropriate PCR-based Markers
C. Mapping and Genetic Enhancement Strategies
D. Marker-Assisted Gene Introgression from Wild Arachis to
Arachis hypogaea
E. Marker-Assisted Backcross Breeding
F. Prioritizing Traits for Marker-Assisted Selection

153
Advances in Agronomy, Volume 80
Copyright q 2003 by Academic Press. All rights of reproduction in any form reserved
0065-2113/03$35.00
154 S.L. DWIVEDI ET AL.

VI. Conclusion
Acknowledgments
References

About 94% of the world groundnut (Arachis hypogaea L.) production

comes from the rainfed crop grown largely by resource-poor farmers.
Several biotic and abiotic stresses limit groundnut productivity, together
causing annual yield losses of over US $ 3.2 billion. The Arachis species
harbor genes capable of improving both seed yield and quality in addition
to imparting high levels of resistance to diseases and insect pests. Many of
the wild Arachis species are not cross compatible with cultivated
groundnut. However, efforts to overcome incompatibility in wide crosses
have started to liberate resistance genes in interspecific progenies. But
these progenies carry a lot of linkage drag. Marker-assisted backcross
breeding should minimize the linkage drag as it greatly facilitates
monitoring of introgressed chromosome segments carrying beneficial
genes from wild Arachis to cultivated groundnut. Transgenic groundnuts
with resistance or tolerance to biotic and abiotic stresses have been
produced and are in various stages of characterization under containment
and/or controlled field conditions. Once favorable genes are introduced
into cultivated groundnut through wide crossing and/or genetic trans-
formation techniques, these genes will become ideal candidates for
marker-accelerated introgression.
DNA marker based genetic linkage map should enable breeders to
effectively pyramid genes for good seed quality and resistance to biotic and
abiotic stresses into agronomically enhanced breeding populations in a
much shorter time than would be possible by conventional techniques. To
date 110 SSR markers detected genetic variation in a diverse array of 24
groundnut landraces. However, substantial efforts are still required to
develop sufficient PCR-based markers, particularly SSRs and SNPs, for
the construction of high-density genetic linkage map and for the routine
application in the molecular breeding in groundnut. The use of automated
technologies will become increasingly important for large-scale germ-
plasm characterization and realistic scale marker-assisted selection in
groundnut. An international legume genomics initiative has been formed
between USA Universities and the International Agricultural Research
Centers of the Eco-Regional Alliance on legumes to translate the benefits
of the “consensus legume genome” for rapid impacts on the genetic
improvement of tropical legumes. q 2003 Academic Press.

1. INTRODUCTION

A. CROP PRODUCTION AND USES

Cultivated groundnut, also known as peanut (Arachis hypogaea L.), is grown

on nearly 24 million hectares between latitudes 408 N and 408 S with a total
 MOLECULAR BREEDING OF GROUNDNUT 155

global production of 34.5 million tones (FAO, 2000) Although originating in

South America, the vast majority of groundnut is produced in Asia and Africa:
Asia 68% (23 Mt), Africa 24% (8 Mt). The remaining 8% (3.5 Mt) comes from
North America, the Caribbean, Europe and Oceania. Approximately 94% of
groundnut is produced in the developing world, mostly under rainfed conditions.
Fig. 1 shows the proportion of groundnut area and production in each of the major
groundnut growing regions of the world. The major groundnut producing
countries are China, India, Indonesia, Myanmar, and Vietnam in Asia; Nigeria,
Sudan, Democratic Republic of Congo, Chad, Mozambique, Zimbabwe, Burkina
Faso, Uganda, and Mali in Africa; USA in North America; and Argentina, Brazil,
and Mexico in Latin America and the Caribbean (Table I).
The average (FAO, 2000) yield of groundnut in Africa is 0.9 t ha21 which is
markedly lower than groundnut yields in Asia (1.7 t ha21) and in Latin America
and the Caribbean (1.8 t ha21), while yields are by far the highest in North
America (2.9 t ha21) and China (3.1 t ha21). The largest groundnut acreage in
Asia occurs in India. However, India ranks below China in total production, as its
average yield is 1.0 t ha21. The key factors contributing to higher yields in China
are (1) introduction of improved varieties presently covering 90% of the total
groundnut area, (2) adoption of improved cultural practices including crop rota-
tion and polythene film mulching, (3) rewards to groundnut growers for producing
higher yields, and (4) national policies for price support systems and marketing
opportunities (Shuren et al., 1996). In contrast, groundnut yields in Africa are
very low with many countries reporting yields as low as 0.5 –0.8 t ha21. Although
the Latin American and the Caribbean regions contribute only 3.4% of the world
groundnut production, high yields of 2.2 t ha21 in Argentina and 1.8 t ha21 in
Brazil have been reported.
Groundnut (33 Mt) is one of the world’s most important oilseeds crop, along
side soybean (154 Mt), cottonseed (52 Mt), rapeseed (42 Mt), and sunflower
(29 Mt) (FAO, 1999). It is also a rich source of edible oil and vegetable protein

Figure 1 The average groundnut land area and yield production (expressed as %) of the major
groundnut growing regions of the world for the period from 1998 to 2000.
156 S.L. DWIVEDI ET AL.

Table I
The Major Groundnut Producing Countries in Asia, Africa, North America, and Latin America
with the Caribbean and Proportionate Contribution (%) to the Global Groundnut Area and
Production

Region Country Area in ’000 haa (%) Production in ’000 ta (%)

Asia India 7207 30.5 7017 20.9

China 4297 18.2 13243 39.5
Indonesia 650 2.8 973 2.9
Myanmar 489 2.1 580 1.7
Vietnam 255 1.1 341 1.0
Africa Nigeria 2643 11.2 2700 8.1
Sudan 1417 6.0 934 2.8
Congo 525 2.2 397 1.2
Chad 419 1.8 471 1.4
Zimbabwe 227 1.0 121 0.4
Burkina Faso 215 0.9 215 0.6
Uganda 198 0.8 139 0.4
Mali 160 0.7 150 0.4
North America USA 569 2.4 1675 5.0

Latin America and the Argentina 311 1.3 658 2.0

Caribbean
Brazil 101 0.4 185 0.6
Mexico 92 0.4 134 0.4
Developing countries 22919 97 31522 94
Developed countries 709 3 1941 6
World 23628 33493
a
Average of 1998– 2000 figures of FAO data (http://www.fao.org).

(Weiss, 1983). Approximately 53% of the total global production of groundnut is

crushed for high quality edible oil, 32% for confectionery consumption, and the
remaining 15% is used for feed and seed production (FAO, 2000). However,
there has been a gradual shift away from the use of groundnut as oil and meals to
confectionery consumption, particularly in Asia, Latin America, and the
Caribbean (Freeman et al., 1999). Figure 2 shows the proportion of in-shell
groundnut production used for oil extraction, confectionery, and feed/seed uses in
the major groundnut producing regions in the world.
The cake remaining after groundnut oil extraction can be used in human food
or incorporated into animal feeds (Savage and Keenan, 1994). Groundnut haulms
constitute approximately 45% of the total plant biomass, and provide excellent
forage for cattle. Haulms are rich in protein and more palatable than many other
fodders (Cook and Crosthwaite, 1994). Wild Arachis species (Arachis pintoi
and A. glabrata) have been used for pasture improvement in North America,
 MOLECULAR BREEDING OF GROUNDNUT 157

Figure 2 Average in-shell groundnut usage patterns in major groundnut producing regions in the
world during the period from 1997 to 1999.

Central America and Mexico, South America, and Australia (Kerridge and
Hardy, 1994). The greater adaptability of rhizoma perennial peanut (A. glabrata)
to the tropical environment, and its high yield when harvested for hay, give it the
potential of becoming one of the most important forages in the tropics (Ruiz et al.,
2000). The nutritive values of A. glabrata cultivar Florigraze and A. pintoi are
higher than that of most tropical forage legumes of commercial importance
(Kerridge and Hardy, 1994).
Figure 3 shows the distribution of market shares for shelled groundnut seeds
entering international trade from the groundnut producing regions of the world

Figure 3 The average distribution of international trade market share for shelled groundnut seeds
exported from the major groundnut producing regions of the world for the period from 1997 to 1999.
158 S.L. DWIVEDI ET AL.

(FAO, 2000). The major exporting countries are China, India, and Vietnam in
Asia; South Africa, Sudan, Zimbabwe, and Senegal in Africa; USA in North
America; and Argentina in Latin America and the Caribbean.

B. CONSTRAINTS TO PRODUCTION

Groundnut is extensively grown in the semi-arid tropics (SAT) by resource-

poor farmers where many abiotic and biotic factors limit its productivity and seed
quality. The major abiotic factors affecting groundnut production include
drought, low availability of phosphorus especially under acidic soil conditions,
and non-availability of iron in calcareous soils. The major biotic constraints to
groundnut production are

Diseases Fungi Rust (Puccinia arachidis Speg.), early leaf spot (ELS)
(Cercospora arachidicola Hori), and late leaf spot (LLS)
[Phaeoisariopsis personata (Berk. and Curtis) Deighton]
Virus Groundnut rosette disease (GRD), peanut clump virus
(PCV), peanut bud necrosis virus (PBNV), and tomato
spotted wilt virus (TSWV)
Bacterial Bacterial wilt [Burkholderia solanacearum (E.F. Smith)
Yabuuchi et al. ]
Nematodes Meloidogyne, Scutellonema, Pratylenchus, Helicotylenchus,
Aphelenchoides, Telotylenchus, and Paralongidorus
species
Insect pests Field pests Leaf miner [Aproaerema modicella (Deventer)], army worm
(Spodoptera litura Fab.), corn earworm (Helicoverpa
armigera Hubner), lesser corn stock borer (Elasmopalpus
lignosellus Zeller), southern corn rootworm (Diabrotica
undecimpuctata howardi Barber), thrips (Frankliniella
and Scirtothrips species), jassids (Empoasca kerri Pruthi),
aphids (Aphis craccivora Koch.), and termites
(Microtermesand Odontotermes species)
Storage pests Bruchid (Caryedon serratus Olivier), red flour beetle
(Tribolium castaneum Herbst), rice moth (Corcyra
caphalonica Stainton), and pod-sucking bug
(Elasmolomus (Aphanus) sordidus Fab.)

Rust, early leaf spot, and late leaf spot are widely distributed foliar diseases of
groundnut (Subrahmanyam et al., 1984; 1985c; Waliyar, 1991). Groundnut
rosette disease is the most destructive disease of groundnut in sub-Saharan
Africa. It is not present in Asia or in Latin America or the Caribbean. The two
main forms of the disease are chlorotic and green rosette (Hayes, 1932; Smart,
1961; Hull and Adams, 1968). Chlorotic rosette is the most common in southern,
 MOLECULAR BREEDING OF GROUNDNUT 159

eastern, and central Africa whereas green rosette is the most common in West
Africa (Subrahmanyam et al., 1977; 1991). There are three agents that interact to
produce rosette disease syndrome in groundnut: groundnut rosette virus (GRV),
groundnut rosette assister virus (GRAV), and satellite RNA (sat RNA) (Bock
et al., 1990). GRV is transmitted by aphids but only from the plants that also
contain GRAV. GRAV is not mechanically transmissible and causes no apparent
symptoms in groundnut. The sat RNA, which is dependent on GRV for
multiplication and on GRAV for aphid transmission, is largely responsible for
rosette symptoms (Murant et al., 1988). Variation in sat RNA has been correlated
with the different forms of rosette disease (Murant and Kumar, 1990). Peanut
clump virus is an economically important soil-borne virus disease of groundnut in
West Africa (Thouvenel et al., 1988). It has an extremely wide host range
including monocots (Reddy et al., 1988), and is transmitted by the fungus vector
Polymyxa graminis Lendingham (Ratna et al., 1991). The two isolates of peanut
clump virus, Indian peanut clump virus (IPCV) and West African peanut clump
virus (WAPCV), are not serologically related (Reddy et al., 1983). Peanut bud
necrosis virus is prevalent in south Asia (Reddy et al., 1995) and tomato spotted
wilt virus predominates in North America (Reddy et al., 1991). Root-knot
diseases caused by Meloidogyne species of nematode are widely distributed in
Asia, Australia, and North America. The widely distributed nematode species
causing substantial damage in groundnut in Africa are Scutellonema, Praty-
lenchus, Helicotylenchus, Aphelenchoides, Telotylenchus, and Paralongidorus
(Sharma et al., 1991; 1992). Bacterial wilt of groundnut is prevalent in South East
Asia, the Far East, and Uganda (Hayward, 1990). It also infects many other crop
plants including potato (Solanum tuberosum L.), tomato (Lycopersicon
esculentum Mill.), tobacco (Nicotiana sps.), pepper (Capsicum sps.), eggplant
(Solanum melongena L.), and ginger (Zingiber officinale Rosc.). The species is
highly heterogenous (Bradbury, 1986). Isolates are classified into five races based
on host range (Buddenhagen and Kelman, 1964; He et al., 1983), and into five
biovars based on biochemical characteristics (Hayward, 1964; He et al., 1983).
Race 1 isolates cause wilt in groundnut and in many other leguminous and
solanaceous plants. Biovar 1 isolates cause wilt of groundnut in the USA; biovar
3 and to a lesser extent biovar 4 isolates cause wilt of groundnut in Asia and
Africa (Hayward, 1991). Aflatoxins are a serious quality problem because they
are carcinogenic and immunosuppressive agents. Their presence, therefore,
influences marketing of groundnut kernels as well as cake. Aflatoxins are
produced by Aspergillus flavus Link ex Fries. The harmful effects of aflatoxin
contaminated confectionery and groundnut cake have been reported (Mehan
et al., 1991). Aflatoxin contamination in food and livestock feed is particularly
severe in the developing countries of Africa, and South and South East Asia.
Unlike the diseases listed above, insects are occasional pests of groundnut, and
their distribution is erratic and localized even within regions. The only groundnut
insect pests of economic significance are leaf miner in South Asia, armyworm in
160 S.L. DWIVEDI ET AL.

South East Asia, and termite in Africa (Wightman et al., 1990). The major pests
in North America are corn earworm, lesser corn stock borer, and southern corn
rootworm (Campbell and Wynne, 1980). The important vectors of groundnut
virus diseases are T. palmi for peanut bud necrosis (Wightman et al., 1995),
F. occidentalis and F. fusca for tomato spotted wilt virus (Culbreath et al., 1992),
and Aphis crassivora for groundnut rosette virus (GRV) (Hull and Adams, 1968).
However, thrips, jassids, and aphids are not themselves considered economically
important pests of groundnut.
Table II lists the important abiotic and biotic constraints to groundnut
production in major groundnut producing regions of the world. These abiotic and

Table II
Constraints to Groundnut Production in South Asia, South East Asia, Southern and Eastern
Africa, Western and Central Africa, North America, and Latin America and the Caribbean
Regions

South South East Southern and Western and North Latin America
Constraints Asia Asia Eastern Africa Central Africa America and Caribbean

Rust p p p p p p
Early leaf spot p p p p
Late leaf spot p p p p p p
Aflatoxin p p p p p
Drought p p p p p
Groundnut rosette p p
virus
Nematodes p p
Spodoptera p p
Termites p p
Sclerotium rolfsii p p
Sclerotinia minor p p
Pythium myriotylum p p
Acidic soils p p
Low temperature p p
Bacterial wilt p
Peanut bud necrosis p
virus
Tomato spotted p
wilt virus
Peanut clump virus p
Leaf minor p
Rhizoctonia solani p
Corn earworm p
Lesser corn stock p
borer
Southern corn p
rootworm
 MOLECULAR BREEDING OF GROUNDNUT 161

biotic stresses often occur in combinations and their severity and extent of
distribution vary with cropping systems, growing seasons, and regions. The
estimates of the global annual yield losses caused by these stresses and the
economic value that could be brought by genetic amelioration of these in
the groundnut crop are projected in Table III. High yielding, well-adapted
cultivars with multiple resistances to biotic stresses and tolerances to abiotic
stresses would provide enhanced and sustainable groundnut production to
subsistence farmers in the SAT regions. Enhanced pest and diseases resistance
would allow reduced agrochemical use while resistance to aflatoxin contami-
nation would facilitate the production of food and feed products with reduced
health risks that would be accepted for international trade.
As in other crops, weeds compete with groundnut for soil moisture, nutrients,
and light and may thereby dramatically reduce yields (Wilcut et al., 1995). They
also harbor pests and diseases, and serve as alternate hosts. Weed competition is
most severe during early crop growth stages because of the slow initial growth of
the groundnut. Weeds can be effectively controlled by the application of
herbicides coupled with one to two weedings at critical groundnut growth stages.
However, this approach is both labor intensive and expensive. Moreover, the
effectiveness of chemical weed control depends on environmental conditions, the
physiological stage of the crop and weeds, soil type, moisture, organic matter,
clay content and pH, and atmospheric temperature and humidity. Thus, the use of
herbicides has been very limited in rain-fed groundnut in the SAT. An alternative
strategy to minimize losses due to weed competition is to introduce genes for
herbicide tolerance and/or early vigour into groundnut.

Table III
Economic Values of Yield Losses Associated with Abiotic and Biotic Stresses and Potential Gains
that can be Realized by Genetic Enhancement in Groundnuta

Yield loss Potential yield gain by genetic

Trait (US$ m) enhancement (US$ m)

Rust 467.0 242.0

Early Leaf Spot 326.0 82.0
Late leaf spot 599.0 300.0
Aflatoxin 264.0 62.0
Groundnut rosette virus 156.0 121.0
Peanut clump virus 38.0 22.0
Peanut bud necrosis virus 89.0 45.0
Leaf miner 164.0 82.0
Spodoptera 97.0 –
Drought 520.0 208.0
Low yield potential 388.0 388.0
Lack of adaptability 44.0 44.0
a
ICRISAT Medium Term Plan, 1994–98. Volume 1, Main Report.
162 S.L. DWIVEDI ET AL.

II. GENETIC VARIABILITY IN GROUNDNUT

The origin of genus Arachis, its taxonomy, cytogenetics and genomes

relationships, the botanical classification, and reproductive development have
been extensively covered in a recent review article by Holbrook and Stalker
(2003). They also discussed in brief about the domestication of groundnut
across continents, the methods and sampling techniques used by plant explorers
to collect germplasm, the status of germplasm maintained at USDA Plant
Introduction Station at Griffin, Georgia, USA, the descriptors used for
characterizing accessions and methodologies to develop core collection, the
levels of variability discovered in core collection for seed quality and biotic
stresses, the preservation and regeneration techniques to maintain cultivated
and wild Arachis species, the difficulties associated with field collection of new
germplasm as well for the exchange of those germplasm that were collected
after adoption of Convention on Biological Diversity treaty, the memorandum
of understanding signed between ICRISAT and USDA to facilitate germplasm
exchange, and the needs to collect additional wild Arachis species in eastern
Bolivia and northwestern Paraguay and cultivated groundnuts from Columbia,
Venenzuela, Uruguay, and possibly from Bolivia, and the impact of use of
genetic resources in cultivar development that benefited to USA peanut farmers
more than US $ 200 million annually. In the following section we discuss the
status of wild Arachis species and cultivated groundnut accessions maintained
in ICRISAT gene bank, the core collection developed involving 15,000
accessions and its significance to the breeders, the variability reported for
major abiotic and biotic stresses and seed quality traits in cultivated and wild
Arachis germplasm, the successful crosses reported between wild Arachis
species, the gene introgression from wild Arachis in to cultivated groundnut,
and finally the impact of plant introductions in developing elite groundnut
germplasm/cultivars at ICRISAT that were either released for cultivation or
used as an improved source by breeders in national breeding programs around
the world.

A. WILD ARACHIS SPECIES

A database of wild Arachis species, compiled at ICRISAT, is now available

at http://www.icrisat.org/text/research/grep/homepage/groundnut/arachis/start.
htm. Gregory and Gregory (1979) reported 296 successful cross combinations
that resulted in 223 intrasectional hybrids in the then sections Arachis,
Erectoides, Rhizomatosae, Caulorrhizae, Extranervosae, Triseminalae, and
Ambinervosae and 73 inter-sectional hybrids involving Arachis with
Erectoides and Rhizomatosae; Erectoides with Rhizomatosae, Caulorrhizae,
 MOLECULAR BREEDING OF GROUNDNUT 163

and Ambinervosae; and Ambinervosae with Extranervosae. Krapovikas and

Gregory (1994) further report successful intra-sectional hybrids in 8 of the 9
sections in the genus Arachis. They could not produce hybrids involving species
within section Trierectoides. They also reported intersectional hybrids
involving Trierectoides with Erectoides and Procumbentes; Erectoides with
Heteranthae, Caulorrhizae, Procumbentes, Rhizomatosae, and Arachis; Extra-
nervosae with Heteranthae; Caulorhizae with Procumbentes; and Rhizomatosae
with Arachis and Procumbentes. Hybrids involving section Arachis with
Rhizomatosae, Heteranthae, Procumbentes, and Erectoides have also been
successful at ICRISAT (Mallikarjuna and Bramel, 2001; Mallikarjuna, 2002).
Both pre- and post-zygotic hybridization barriers have been shown to restrict
crossing between Arachis species. These barriers are most severe when
accessions from tertiary and quaternary gene pools are crossed with A.
hypogaea, but such barriers may also be expressed in crosses with certain
accessions of the secondary gene pool.
Wild Arachis species harbor a range of genes conferring resistance to pests
and diseases (Table IV), oil and protein contents, and oleic (O)/linoleic (L) fatty
acid ratios (Table V). Some genotypes show very high levels of resistance to rust,
ELS, LLS, nematodes, GRD, PBNV, thrips, jassids, leaf miner, Spodoptera, and
aphids. Accessions belonging to 13 species in section Arachis show wide
variation for most of the morphological traits reported (Chandran and Pandya,
2000).

B. CULTIVATED GROUNDNUT

Over 15,000 accessions of cultivated groundnut, including 6351 landraces,

from 92 countries are housed at ICRISAT (India). They differ for many
vegetative, reproductive, physiological, and biochemical traits including their
reactions to abiotic and biotic stresses (Singh and Nigam, 1997; Rajgopal
et al., 1997; Upadhyaya et al., 2001e; 2003). The Arachis genepool includes
sources of resistance to rust, ELS, LLS, GRD, PBNV, A. flavus, bacterial
wilt, leaf miner, Spodoptera, jassids, thrips, and iron chlorosis, and tolerance
to low and high temperature and drought as well as sources of photoperiod
insensitivity (Table IV), and variation in total sugars, oil and protein
contents, O/L ratio, and for flavor attributes (Table V). Genotypic differences
in root hair density and/or root growth in groundnut have been associated
with high phosphorus (P) acquisition in P deficient soils (Wissuwa and Ae,
2001). However, much of this variability remains poorly understood and
under-utilized in genetic enhancement efforts mainly because of the large
number of accessions in the ex situ collections, lack of data on the extent
of the diversity present in them for specific characteristics, and high geno-
type (G) £ environment (E) interactions for traits of economic importance
 164
Table IV
Sources of Resistance to Abiotic and Biotic Stresses Reported in A. hypogaea and Wild Arachis Species in Groundnut

Resistance in A. hypogaea germplasm Resistance in Wild Arachis species

Reaction to abiotic and Reaction to abiotic

Trait biotic stresses Reference and biotic stresses Reference

Rust 1.7–5.0a Singh et al., 1997 Ib-HRc Subrahmanyam et al., 1983b;

Stalker and Simpson, 1995;
Pande and Rao, 2001
Early leaf spot 3.6–6.3a Singh et al., 1997; Stalker et al., 2002a,b; I-HR Subrahmanyam et al., 1985b;

S.L. DWIVEDI ET AL.

Stalker and Lynch, 2002 Stalker and Simpson, 1995;
Upadhyaya et al., 2001f
Late leaf spot 2.9–5.0a Singh et al., 1997; Stalker et al., 2002b I-HR Subrahmanyam et al., 1985a,b;
Stalker and Simpson, 1995;
Pande and Rao, 2001
Peanut bud 13.6–23.7% Dwivedi et al., 1995b , 20% Reddy et al., 2000a
necrosis disease
Groundnut Rosette 1.0–12.0% Subrahmanyam et al., 1998; HR Stalker and Simpson, 1995;
disease Olorunju et al., 2001 Subrahmanyam et al., 2001
Nematode MRe Holbrook and Noe, 1992; Rd Baltensperger et al., 1986;
Holbrook et al., 1996; 1998; Holbrook and Noe, 1990;
Timper et al., 2000; Stalker et al., 2002a Nelson et al., 1989;
Sharma et al., 1999
A. flavus Seed , 15.0% Mehan, 1989 #5% Thakur et al., 2000
infection
Aflatoxin , 10 mg g21 seed Mehan et al., 1986 22–110 mg g21 seed Mehan et al., 1992;
contamination Thakur et al., 2000
Bacterial wilt 1–25% Singh et al., 1997; Pande et al., 1998 Resistance not reported
in wild Arachis species
Table IV (continued)

Resistance in A. hypogaea germplasm Resistance in Wild Arachis species

Reaction to abiotic and Reaction to abiotic

Trait biotic stresses Reference and biotic stresses Reference

MOLECULAR BREEDING OF GROUNDNUT

Thrips R–MR Lynch, 1990; Wightman et al., 1990; R Stalker and Campbell, 1983;
Campbell and Wynne, 1980; Dwivedi et al., Stalker and Simpson, 1995;
1996b; Rao and Wightman, 1999; Lynch and Mack, 1995;
Amin et al., 1985; Upadhyaya et al., 2001e Amin, 1985
Jassids R–MR Lynch, 1990; Wightman et al., 1990; R Stalker and Campbell, 1983;
Campbell and Wynne, 1980; Stalker and Simpson, 1995;
Campbell et al., 1971; Rao and Lynch and Mack, 1995;
Wightman 1999; Amin et al., 1985; Amin, 1985
Dwivedi et al., 1995a; Dwivedi et al.,
1993b; Stalker et al., 2002a;
Stalker and Lynch, 2002; Upadhyaya
et al., 2001e
Leaf miner MR Wightman and Rao 1994; Anderson R–HR Wightman and Rao, 1994;
et al., 1990b; Dwivedi et al., 1993b; Lynch and Mack, 1995
Rao and Wightman, 1999
Spodoptera MR Dwivedi et al., 1993b; Campbell and R–HR Wightman and Rao, 1994;
Wynne, 1980; Rao and Wightman, Lynch and Mack, 1995
1999; Stalker and Lynch, 2002
Aphids MR-HR Padagham et al., 1990; Wightman and R–HR Wightman and Rao, 1994;
Rao 1994; Minja et al., 1999; Stalker and Simpson, 1995;
Rao and Wightman, 1999 Lynch and Mack, 1995;
Amin, 1985
Southern corn R Stalker et al., 2002a; Stalker and
rootworm Lynch, 2002
(continued on next page)

165
 166
Table IV (continued)
Resistance in A. hypogaea germplasm Resistance in Wild Arachis species

Reaction to abiotic and Reaction to abiotic

Trait biotic stresses Reference and biotic stresses Reference

Corn earworm R Stalker et al., 2002a; Stalker and

Lynch 2002
Temperature Tolerance to high Nageshwar Rao et al., 1989; Wild Arachis species not evaluated for high temperature
temperature Greenberg et al., 1992; Chauhan and
Senboku, 1997; Wheeler et al., 1997;

S.L. DWIVEDI ET AL.

Talwar et al., 1999; Craufurd et al.,
1999, 2002; Prasad et al., 1999
Tolerance to low Upadhyaya et al., 2001d Wild Arachis species not evaluated for low temperature
temperature
Photoperiod Insensitive Nigam et al., 1997 Wild Arachis species not evaluated for photoperiod
sensitivity
a
Scored on 1–9 scale where 1, no disease; 2 ¼ ,5% disease severity; 3, 6–10% disease severity; 4, 11 –20% disease severity; 5, 21 –30% disease severity;
6, 31–40% disease severity; 7, 41–60% disease severity; 8, 61–80% disease severity; 9, 81–100% disease severity.
b
I, Immune.
c
HR, Highly resistant.
d
R, Resistant.
e
MR, Moderately resistant.
 MOLECULAR BREEDING OF GROUNDNUT
Table V
Variability for Oil and Protein Contents and O/L Ratio in Arachis hypogaea and Wild Arachis Groundnut Germplasm

A. hypogaea Wild Arachis

Trait Variability reported Reference Variability reported Reference

Oil (%) 45 –56 Ahmed and Young, 1982; 42–63 Cherry, 1977; Stalker et al., 1989;
Dwivedi et al., 1993a Jambunathan et al., 1993; Grosso et al., 2000
Protein (%) 23 –29 Young and Hamons, 1973; 17–31 Cherry, 1977; Jambumnathan et al., 1993;
Dwivedi et al., 1993a Grosso et al., 2000
O/L ratio 0.95–40.0 Norden et al., 1987; 0.35–0.23 Stalker et al., 1989; Grosso et al., 2000
Dwivedi et al., 1993a, 1998b;
Gorbet and Knauft, 1997;
Harch et al., 1995
Total sugars 23169 –44795 Pattee et al., 2000
Sweet flavor 2.33– 4.12 ifu Pattee et al., 1998
Bitter flavor 2.43– 4.46 ifu
Roasted flavor 3.75– 5.22 ifu

167
168 S.L. DWIVEDI ET AL.

(Tai and Hammons, 1978; Wynne and Isleib, 1978; Shorter and Hammons,
1985; Dwivedi et al., 1993a; Coffelt et al., 1993). Upadhyaya et al. (2001e)
developed a core collection of 1704 groundnut accessions consisting of 584
(34.3%) accessions from subsp. fastigiata var. vulgaris, 299 (17.5%) from
subsp. fastigiata var. fastigiata, 27 (1.6%) from subsp. fastigiata var.
peruviana, 6 (0.4%) from subsp. fastigiata var. aequitoriana, 784 (46.0%)
from subsp. hypogaea var. hypogaea, and 4 (0.2%) from subsp. hypogaea
var. hirsuta, and arrayed these accessions in 23 clusters. When this core
collection is evaluated for traits of economic importance including resistance
to abiotic and biotic stresses, it should provide breeders with opportunities to
further broaden the genetic base of the crop by integrating diverse germplasm
into their breeding programs. In some countries particularly in India and
Vietnam, tolerance to cold temperature is required as the low temperature
prevailing during the planting time results in delayed germination and a
reduced growth rate thus delaying the harvest. When Upadyaya et al. (2001d)
evaluated their core collection for tolerance to low temperature (128C), they
identified 343 accessions with 81 –100% germination compared to 43%
germination in control cultivar ICGS 44 (ICGV 87128). Botanically these
accessions represented subsp. aequatoriana, fastigiata, peruviana, vulgaris,
and hypogaea.
The oil content of dried groundnut seeds is reported to vary from 44 to 56%,
while protein content ranges from 22 to 30%. Groundnuts are also a rich source of
minerals (phosphorus, calcium, magnesium, and potassium) and vitamins (E, K,
and B group) (Savage and Keenan, 1994). Seed size, shape, color, oil and protein
contents, fatty acid and amino acid composition, taste, and flavor are important
quality traits in groundnut. Oleic (O), linoleic (L), and palmitic fatty acids,
together, account for over 80% of the total fat in groundnut seeds (Dwivedi et al.,
1993a). Considerable genetic variation has been reported for pod size and shape,
seed size and shape, seed color, oil content, and fatty acid composition (Norden
et al., 1987; Dwivedi et al., 1989; 1998b; Branch et al., 1990; Singh et al., 1998;
Upadhyaya et al., 2001f).
Plant breeders in the USA have registered 62 Arachis germplasm lines
possessing genes for resistance to biotic and abiotic stresses, and for seed quality
traits for use in breeding programs (Isleib and Wynne, 1992, Table VI). Of these,
27 were introduced germplasm. However, because of stringent industry and
market demands, US plant breeders use only those accessions that conform to
market and industry standards. This has resulted in a narrowing of the genetic
base of released cultivars there. ICRISAT breeders have used 78 plant
introductions to develop 73 elite germplasm lines. Of these, 41 have been
released for cultivation in 19 countries, and the remainder possesses genes for
early maturity, seed dormancy, seed quality, photoperiod insensitivity, and
resistance to rust, ELS, LLS, thrips, jassids, leaf miner, Spodoptera, PBNV, iron
chlorosis, aflatoxin, and tolerance to drought (Table VI), and these elite
 Table VI
List of Cultivars and Elite Germplasm with Specific Characteristics Developed by Conventional Breeding in Groundnut

Area of adaptation and other useful

Trait incorporated Identity traits identified Reference

MOLECULAR BREEDING OF GROUNDNUT

Resistance to rust ICGV 87160 Released in India and Myanmar; Reddy et al., 1992
and/or LLS tolerant to leaf miner,
peanut bud necrosis disease (PBND),
and drought
ICGV 86590 Released in India; tolerant to PBND, Reddy et al., 1993
stem and pod rots, and Spodoptera
ICGV-SM 85048 Released in Mauritius; resistant to web Niagm et al., 1998b
blotch (Phoma arachidicola)
ICGV-SM 86715 Released in Mauritius; resistant to pepper Moss et al., 1998
spot (Leptosphaerulina crassiasca)
ICGV 87165 Resistant to bacterial wilt, leaf miner, Moss et al., 1997
and Spodoptera
ICGV 86699 Resistant to stem and pod rots, tolerant Reddy et al., 1996
to PBND and Spodoptera
ICGV 87157 Tolerant to PBND, stem and pod rots, Niagm et al., 1992b
leaf minor, and mid-season drought
Southern runner Released in USA Gorbet et al., 1987
Yue You 223 Released in China Liang et al., 1999b
TxAG 6 and TxAG 7 Resistant to ELS and nematodes Simpson et al., 1993
GP-NCWS 1, GP-NCWS 2, Interspecific derivatives resistant to Stalker and Beute, 1993
GP-NCWS 3, and GP-NCWS 4 late leaf spot
Zhonghua 117 Released in China, moderate resistance Guiying et al., 1995
to rust and bacterial wilt and tolerant
to acid soils
(continued on next page)

169
 170
Table VI (continued)
Area of adaptation and other useful
Trait incorporated Identity traits identified Reference

VRI Gn 5 Released in India, resistant to rust and Vindhiyavarman and

late leaf spot Mohammed, 2001
Sylvia (ICGV 93207) Released in Mauritius Reddy et al., 2001a
Venus (ICGV 87853) Released in Mauritius Reddy et al., 2000
Bacterial wilt E Hua 5, Lu Hua 3, Yue You 5, Released in China; resistant to Liang, 1998; Liao et al., 1998;
Yue You 79, Yue You 92, rust and root rot in Liang et al., 1999a
Yue You 200, Yue You 202 –35, You 202– 35 and Yue You 79

S.L. DWIVEDI ET AL.

Yue You 256, Yue You 589, and to rust in Yue You 256
Gui You 28, Wu You 4,
Zhong Hua 2, and 93–81
Badak, Biawak, Jepara, Released in Indonesia Machmud and Rais, 1994
Komodo, Landak,
Mahesa, Simpai, Trenggiling,
and Zebra
Soil born diseases VG P9 Resistant to Sclerotinia minor, Coffelt et al., 1994a
Sclerotinia blight, and
Cylindrocladium crotalariae
Tamrun 98 Released in USA, resistant to Sclerotinia minor Simpson et al., 2000
Southwest Runner Released in USA, resistant to Sclerotinia minor Kirby et al., 1998
Resistant to thrips, jassids,ICGV 86031 Photoperiod insensitive and resistance to PBNV Dwivedi et al., 1993b
leaf miner, and Spodoptera and iron chlorosis
Resistant to jassids ICGV# 86252, 86393, 86455, High density of trichomes on leaves Dwivedi et al., 1995a
and 86462
Resistant to peanut bud ICGV 86388 Resistant to jassids and thrips Dwivedi et al., 1996b
necrosis virus
Georgia Green Released in USA, resistant to Tomato Branch, 1996
Spotted Wilt Virus
Table VI (continued)
Area of adaptation and other useful
Trait incorporated Identity traits identified Reference

MOLECULAR BREEDING OF GROUNDNUT

Resistant to rosette virus RG 1, KH 149A, KH 241D, 69– 101, 28–206 and 69–101 released in Senegal Bockelee-Morvan, 1983;
RMP 12, RMP 91, 28–206, and and KH 149A, KH 241D, RMP 12 and Mauboussin et al., 1970;
ICGV-SM 90704 RMP 91 in Burkina Faso van der Merwe et al., 2001
Resistant to thrips and jassids ALR 2 Released in India; resistant to stem rot and Varman et al., 1998
tolerant to rust and LLS
Resistant to seed infection ICGV# 88145 and 89104 Support low aflatoxin production Rao et al., 1995
by A. flavus
Streeton Released in Australia; support low aflatoxin Cruickshank et al., 2000
production
ICGV# 91278, 91283, and 91284 Resistant to natural seed infection and Upadhyaya et al., 2001c
in vitro seed colonization by A. flavus
Tolerant to drought ICGV 87354 Resistant to rust Reddy et al., 2001b
55–437, 73 –30, 47 –10, T3.3, 47–10, 55–437, 73 –30, and 73 –33 Bockelee-Morvan, 1983
TS 32–1, and 73 –33 released in Senegal and T3.3 and
TS 32–1 in Burkina Faso
High yield per se ICGS 1 Released in India; tolerant to mid-season Nigam et al., 1991b
drought and PBND
ICGV 87121 Released in India; tolerant to mid-season Nigam et al., 1992
drought
ICGV 87141 Released in India; tolerant to mid-season Nigam et al., 1991a
drought and PBND
ICGV 86325 Released in India; tolerant to PBND Dwivedi et al., 1996a
Sinkarzei Released in India; tolerant to mid-season Nigam et al., 1993
drought and PBND
(continued on next page)

171
 172
Table VI (continued)

Area of adaptation and other useful

Trait incorporated Identity traits identified Reference

CG 7 (ICGV-SM Released in Malawi, Uganda, and Zambia; Nigam et al., 1995a

83708) superior oil quality and large seed size
ICGV-SM 83005 Released in Zambia Nigam et al., 1998a
ICGS 11 Released in India and Srilanka; Nigam et al., 1990b
Tolerant to PBND and end-of-season drought,
insensitive to photoperiod

S.L. DWIVEDI ET AL.

ICGV 87128 Released in India; wide adaptability; tolerant Nigam et al., 1990a
to PBND, and mid- and end-of-seasons drought
ICGV 87187 Released in India; tolerant to end-of-season drought Nigam et al., 1992a
and PBND; insensitive to photoperiod
ICGS 35 Released in Korea Nigam et al., 1994a
Jeokwantangkong Released in Korea, resistant to pod rot Youn-Sup et al., 2000
Luhua 15 Released in China; resistant to web blotch Zhengchao et al., 1997

Early maturity ICGV 86143 Released in India and Vietnam Upadhyaya et al., 1997b
ICGV 86015 Released in Pakistan, Nepal, and Vietnam Nigam et al., 1995b
ICGV# 92196, 92206, 92234, Upadhyaya et al., 1998
and 92243
VA 93B Released in USA for early maturity Coffelt et al., 1994b
ICGV 94361 Early maturity and moderate resistance to rust Upadhyaya et al., 2001b
Andru 93 Released in USA, early maturing into Virginia Gorbet and Knauft, 1995
market type
Kadiri 4 Released in India Reddy et al., 1998
Nonghua 22 Released in China, resistant to leaf spots and tolerant Gao et al., 1996
to drought
Table VI (continued)

Area of adaptation and other useful

Trait incorporated Identity traits identified Reference

Fresh seed dormancy ICGV# 86155, 86156, 86158, Early maturity Upadhyaya et al., 1997a
87378, and 87921

MOLECULAR BREEDING OF GROUNDNUT

ICGV 93470 Early maturity with fresh seed dormancy Upadhyaya et al., 2001a
Large seed size ICGV 86564 Released in Srilanka; high oil content and O/L ratio Dwivedi et al., 1994
ICGV# 88438, 89214, and 91098 Released in Cyprus; high oil content, tolerant to Hadjichristodoulou et al., 1997
lime induced iron chlorosis
ICGV# 96230 and 96234 High O/L ratio Dwivedi et al., 1998b
Huayu 16 High oil and protein contents, and resistant Zhengchao and Qingshu, 2000
to root rot (Macrophomina phaseolina) and
tolerant to peanut stripe virus
Huayu 17 Low oil Shanlin et al., 2000
Georgia Browne Released in USA for confectionery or candy use Branch, 1994
VA-C 92R Released in USA, high seed calcium content Mozingo et al., 1994
VA 98R Released in USA, bright pod color and pink seed Mozingo et al., 2000
Gregory Released in USA Isleib et al., 1999
Georgia Bold Released in USA, resistant to TSWV Branch, 1998
Georgia Valencia Released in USA, 3–4 seeded large pods Branch, 2001
BRS 151 Amendoim L7 Released in northeast region in Brazil, dos Santos, 1998
early maturing type
High O/L ratio SunOleic 95 R Released in USA Gorbet and Knauft, 1997
VG P 10 Released in USA, early maturity with large Coffelt and Mozingo, 1998
seed size
Georgia Hi-O/L peanut Large seed size and resistant to TSWV Branch, 2000
SunOleic 97R Released in USA Gorbet and Knauft, 2000

173
174 S.L. DWIVEDI ET AL.

germplasm are widely used by NARS breeding programs to transfer these traits
into locally adapted cultivars.

III. HIGHLIGHTS AND CONSTRAINTS

OF CONVENTIONAL GENETIC IMPROVEMENT
IN GROUNDNUT

A. BREEDING METHODS AND CULTIVARS

The most commonly used breeding methods in groundnut are (1) pedigree
selection, (2) bulk-pedigree selection, and (3) single-seed descent. Backcross
breeding has not been used extensively as most of the economically important
traits in groundnut are quantitatively inherited (Wynne and Gregory, 1981;
Knauft and Wynne, 1995). Often, breeders make single crosses to generate
variability. However, with increased emphasis on multiple resistance breeding,
emphasis is now focused on complex crosses followed by intercrossing of
segregants to bring the desired improvement into breeding populations. While
selection for resistance to insect pests and diseases is practiced in early
generations, selection for yield and yield component traits is delayed to later
generations. Recurrent selection has also been used for continued genetic
enhancement in groundnut (Guok et al., 1986; Halward et al., 1991b).
Over 276 groundnut cultivars were released between 1920 and 2000 for
cultivation in various countries in Asia, Africa, and the Americas. Each has
specific adaptation to its respective region of production and cropping system
(Isleib et al., 1994; Godoy and Giandana, 1992; Table VI). Breeding for high seed
yield has caused changes in dry matter allocation. More recently developed
cultivars have reduced vegetative mass, shorter main stem length, and greater
reproductive allocation (partition more of their daily assimilate to fruit) than
those developed previously (as predicted by Duncan et al., 1978). Further studies
on reproductive efficiency (RE) revealed that high yield in more recently released
cultivars appears to be related more to total flower production than to RE, and
therefore, future increases in seed yield might be accomplished by developing
cultivars with a combination of high RE, harvest index, and total flower count
(Coffelt et al., 1989). A yearly genetic gain of nearly 15 kg per hectare has been
reported for large-seeded Virginia type cultivars released from the 1950s to the
1970s in the USA (Mozingo et al., 1987). The highest yielding cultivars
developed during the 1950s, 1960s, and 1970s had an average yield increase of
3.4%, 10.2%, and 18.5%, respectively, over the standard NC 4. However, since
the 1970s there has been increased emphasis on improving pest resistance and
quality traits so that the yield potential of cultivars released since that time has not
surpassed those of the highest yielding cultivars released during the 1970s.
 MOLECULAR BREEDING OF GROUNDNUT 175

B. SUCCESSES AND LIMITATIONS TO CONVENTIONAL BREEDING

1. Disease Resistance

Cultivars resistant to rust, bacterial wilt, and groundnut rosette disease (GRD)
have been developed (Table VI). Bacterial wilt resistant cultivars are grown in
South East Asia and the Far East and GRD resistant cultivars are grown in Africa
on large acreage. However, many cultivars continue to be susceptible to rust, early
leaf spot (ELS) and/or late leaf spot (LLS). A few cultivars with moderate
resistance to rust and LLS have been released in China, India, Mauritius, and the
USA. The adoption of rust and LLS resistant cultivars among SAT farmers has
been low mainly because of their relatively long duration and low shelling outturn
(proportion of seeds to pods; also referred to as shelling percentage or meat
content). Progress in ELS and LLS resistance breeding has been limited by the
absence of high levels of resistance in cultivated groundnut and the linkage of
resistance with long duration, lower partitioning and with undesirable pod (highly
reticulated, constricted, prominently ridged and conspicuously beaked pods with
thick shells) and seed (purple or blotched seed color) characteristics (Wynne et al.,
1991; Singh et al., 1997). In contrast, several wild Arachis species show a very high
level of resistance to ELS and LLS. They also possess very small and catenate
pods. The success in transferring ELS and LLS resistance from wild Arachis
species to cultivated groundnut has been limited mainly because of cross
compatibility barriers, the linkage of resistance with many undesirable pod
characteristics, and the long periods of time required for developing stable
tetraploid interspecific derivatives. Despite these obstacles, a few interspecifc
derivatives possessing high levels of resistance to ELS, LLS, and nematodes have
been developed (Table VI). A nematode resistant cultivar, Coan, derived from an
interspecific cross, has been released for cultivation in the USA. There has been
some progress toward developing elite groundnut germplasm resistant to seed
infection and/or aflatoxin production by A. flavus, and tolerance to peanut bud
necrosis virus (Table VI). “Streeton” has been released for commercial cultivation
in Australia because of its excellent yield, grade stability, and lower susceptible to
aflatoxin contamination under drought conditions. There are only a few examples
of multiple resistances incorporated into elite groundnut germplasm (Table VI).

2. Insect Pest Resistance

Resistance to thrips and jassids and tolerance to leaf miner and Spodoptera has
been successfully transferred into improved genetic backgrounds (Table VI). A
few interspecific derivatives possessing high levels of resistance to southern corn
rootworm, corn earworm, Spodoptera, and jassids have been developed from
interspecifc crosses in the USA (Table VI). However, reduced vulnerability to
one or more of these pests has not been the primary criterion for release of any
176 S.L. DWIVEDI ET AL.

improved groundnut cultivar in SAT regions to date. Lack of a high level of

resistance to leafminer and Spodoptera in cultivated groundnut, and difficulties in
conducting reliable resistance screens under field conditions are the main reasons
for the slow progress in developing improved germplasm with resistance to these
pests. Many of the wild Arachis species, in contrast, possess a high degree of
resistance to leafminer and Spodoptera (Table IV). However, these are not
readily cross compatible with cultivated groundnut. A transgenic approach might
be the best option to incorporate genes for resistance to leafminer and Spodoptera
into cultivated groundnut, provided that genes conferring reasonable levels of
resistance can be identified and isolated.

3. Abiotic Stress Tolerance

Success in breeding drought tolerant groundnut genotypes using conventional

plant breeding methodologies has been limited. However, recent studies revealed
that genotypic variation for physiological traits (specific leaf area, water use
efficiency, amount of water transpired, transpiration efficiency, and harvest
index) under drought stress conditions are associated with drought tolerance
(Nageswara Rao et al., 1993; Nageswara Rao and Wright, 1994; Wright et al.,
1994; Craufurd et al., 1999; Nageswara Rao and Nigam, 2001). These traits are,
now, being used to select for drought tolerance in groundnut. A few drought
tolerant cultivars have been released in West Africa (Table VI). Elite groundnut
germplasm with tolerance to mid-season and/or end-of-season drought stress has
been developed at ICRISAT (Table VI).

4. Quality Traits

Seed size, oil content, and oil quality as measured by variation in the O/L ratio
are important seed quality traits in groundnut. For confectionery and other means
of direct consumption, groundnuts with large seed size, low oil content, and a
high O/L ratio are preferred. Oils with high O/L ratios are less prone to oxidation
and the development of unfavorable flavors. Groundnut seeds with high O/L
ratios have long product stability and shelf-life (James and Young, 1983; Branch
et al., 1990). Oil content and O/L ratio are highly influenced by G £ E interaction
(Dwivedi et al., 1993a). Seed size is not an important trait for oil types but
varieties with high oil content and a high O/L ratio are preferred. Excellent
progress has been made in developing high-yielding breeding lines with large
100-seed mass (. 80 g) (Table VI). However, these are late-maturing types and
many have high oil contents-a trait not preferred for edible groundnut. Success in
combining large-seed size, high O/L ratio, and early maturity in genotypes
belonging to subsp. fastigiata has been very limited. Although in the late 1990s,
 MOLECULAR BREEDING OF GROUNDNUT 177

US-based researchers reported large genetic variability for O/L ratio (Table V),
success in improving the chemistry of oil quality has been limited. This is partly
due to the inaccessibility of germplasm with unique oil profiles (such as F435).
Recently, improved groundnut cultivars with high O/L ratios have been released
in the USA (Table VI).

5. Adaptation and Yield

Photoperiod, temperature, and photoperiod by temperature interactions

influence the partitioning (also known as harvest index) and therefore the
adaptation of groundnut genotypes to new environments (Nigam et al., 1994b;
1998c). The breeding environment under which selection is conducted among
segregating populations strongly influences the yield adaptability of the selected
groundnut genotypes (Branch and Hildebrand, 1989). Genotypes with large seeds
and/or resistance to pests and diseases are, in general, sensitive to photoperiod
whereas early maturing types are least affected by variation in photoperiod (Flohr
et al., 1990; Nigam et al., 1997). Groundnut is grown on a wide range of soils,
and strong soil type £ genotype interaction suggests specific varietal adaptation
for soil types (Nageswara Rao et al., 1992).
Crop duration also plays an important role in yield and adaptation of
genotypes. Early maturing cultivars are suitable for areas where the growing
season is short, end-of-season droughts or early frosts are common, low
temperature at sowing resulted delayed germination and slow growth, and the
crop is grown in after rice with residual moisture. Many breeding programs
including ICRISAT’s developed several cultivars with a potential yield of
3 t ha21 and a 90 day maturity (Table VI). However, most of the early maturing
cultivars have small seeds (30 –40 g 10021 seeds), possess no seed dormancy,
and are highly susceptible to pests and diseases. Some progress has been made in
efforts to combine early maturity with relatively large seed size (50 g 10021
seeds), 2– 3 weeks of fresh seed dormancy, tolerance to cold temperature, and
moderate resistance to rust and late leaf spot (Table VI). A short period of seed
dormancy is beneficial as it helps to reduce losses associated with low
germination if there is rain at harvest and proper care has not been taken to
fast dry the groundnut pods. Substantial progress has been made towards
developing medium- and late-maturing cultivars adapted to rainfed and/or post-
rainy irrigated high input situations. In Zimbabwe and China, some of these
varieties produced over 9.0 t ha21 pod yield (Smart, 1978; Yanhao and Caibin,
1990). However, there is a wide gap between realized yields at the farm level
(world average yield 1.0 t ha21) in SAT regions when compared to the average
yields (3.0 t ha21) in China, the USA and the highest yields reported from China
and Zimbabwe. There is therefore a need to incorporate multiple stress resistance
into improved genetic backgrounds, even if it requires some sacrifice in yield
178 S.L. DWIVEDI ET AL.

potential, in order to narrow the gap between realized yield and potential yield
(Nigam, 2000).

IV. OVERVIEW OF BIOTECHNOLOGY APPLICATIONS

IN GROUNDNUT

A. MOLECULAR GENETIC DIVERSITY ANALYSIS

Assessment of genetic diversity and development of genetic linkage maps are

important steps in the development of molecular breeding programs. Only very
low levels of molecular genetic polymorphism have been detected among
cultivated groundnut accessions using isozymes, restriction fragment length
polymorphism (RFLP), random amplified polymorphic DNA (RAPD), DNA
amplification fingerprinting (DAF), and amplified fragment length polymorphism
(AFLP) markers (Grieshammer and Wynne, 1990; Kochert et al., 1991; Bhagwat
et al., 1997; He and Prakash, 1997; Subramanian et al., 2000). Similarly, Hopkins
et al. (1999) have found only six simple sequence repeat (SSR) markers that
detected polymorphisms amongst cultivated groundnut. However, Dwivedi et al.
(2001) detected upto 41% genetic dissimilarity in RAPD profiles among 26
cultivated groundnut accessions. In contrast, abundant DNA marker polymorph-
isms have been detected between wild species in section Arachis (Halward et al.,
1991a; 1992; Paik-Ro et al., 1992; Lanham et al., 1992). This supports the
hypothesis that A. hypogaea may have originated from a single hybridization
event followed by chromosome doubling, with very little subsequent introgres-
sion from related diploid species (Young et al., 1996).
Assessment of molecular diversity should facilitate the identification of
agronomically valuable and diverse germplasm for use in linkage mapping and
genetic enhancement of specific traits in groundnut. Agronomically superior
germplasm lines with relatively high level of DNA marker polymorphism have
been identified at ICRISAT. This should facilitate the mapping of many
important agronomic traits including ICG 405, ICG 1705, ICG 6284, and TMV 2
for early leaf spot (ELS); ICGV 99001, ICGV 99004, and TMV 2 for late leaf
spot (LLS); ICGV 99003, ICGV 99005, and TMV 2 for rust; ICG 6323, ICG
6466, ICG 11044, and JL 24 for groundnut rosette disease (GRD); CSMG 84-1,
TAG 24, ICGV 86031, ICGV 87128, TMV 2 NLM, and Chico for drought; ICG
7893, ICG 15222, and Chico for bacterial wilt; and U 4-7-5, 55-437, and J 11 for
resistance to seed infection and/or aflatoxin production by A. flavus. ICG 405,
ICG 1705, ICG 6284, ICG 7893, ICG 11044, and 55-437 originated from South
America, ICG 6323, ICG 6466, and J 11 from Africa, ICG 15222 from China, and
U4-7-5 from North America. ICGV 99001, ICGV 99003, ICGV 99004, and ICGV
99005 are derivatives from interspecific hybridization made at ICRISAT. The
 MOLECULAR BREEDING OF GROUNDNUT 179

highly susceptible accessions included are TMV 2 for rust, ELS, and LLS; JL 24
for GRD; and Chico for bacterial wilt. The drought tolerant accessions show wide
variation in specific leaf area, partitioning, and water-use efficiency.

B. MOLECULAR GENETIC LINKAGE MAPPING

The groundnut genome is nearly 20 times larger than Arabidopsis thaliana,

and 2 –6 times larger than Oryza sativa, Medicago truncatula, M. sativa,
Phaseolus vulgaris, Sorghum bicolor, Lycopersicon esculentum, Solanum
tuberosum, Ipomoea batata, and Glycine max. However, the groundnut genome
is of a size similar to Gossypium hirsutum, Zea mays, and Helianthus annuus,
and smaller than Pisum sativum, Lensculinaris esculenta, Hordeum vulgare,
Avena sativa, and Triticum aestivum (Table VII). Variation in genome size
among accessions of A. hypogaea ð2n ¼ 4x ¼ 40Þ and A. duranensis ð2n ¼ 2x ¼
20Þ (Singh et al., 1996), and between A. hypogaea and A. monticola (Temsch and
Greilhuber, 2000) has also been reported. Genome size variation in groundnut has
not been related to ecological or evolutionary factors. Variation in genome size is
generally the result of differences in the amount of repetitive DNA and ploidy
level (Flavell et al., 1974).
The first RFLP-based genetic linkage map of groundnut, with a total map
distance of approximately 1063 cM, was constructed using an F2 population
derived from an interspecifc cross between two related diploid species
(A. stenosperma and A. cardenasii) in section Arachis (Halward et al., 1993).
Burow et al. (2001) subsequently reported the RFLP-based tetraploid genetic
linkage map of groundnut derived from a BC1 population of TxAG 6 with
Florunner. TxAG 6 was derived by crossing the A-genome diploid hybrid from
A. cardenasii (GKP-10017, PI 262141) £ A. chacoensis (GKP-10602, PI 276235)
as male parent on to the B-genome species A. batizocoi (K-9484, PI 298639) as
female parent. The resulting tri-species hybrid was chromosome doubled to
produce fertile amphiploids. Three hundered and seventy RFLP loci were ulti-
mately mapped to 23 linkage groups with a total map distance of approximately
2210 cM. This map is unique in that the donor parent is a synthetic polyploid
created by crossing three diploid species. These RFLP loci will detect alleles
in populations involving crosses between wild Arachis species or between
A. hypogaea £ wild Arachis species crosses. They are unlikely to detect alleles in
A. hypogaea £ A. hypogaea crosses. Holbrook and Stalker (2003) reviewed the
progress achieved in (1) identifying RAPD and RFLP markers linked with
root-knot nematode and southern corn rootworm damage and for components of
resistance to leaf spots in interspecific hybrid with A. cardenasii in the pedigree,
(2) markers associated with cylindrocladium black rot resistance and sporulation
to C. arachidicola in a hypogaea cross, and (3) utility of these markers to monitor
180 S.L. DWIVEDI ET AL.

Table VII
Chromosome Number (2n ¼ 2x) and Genome Size Variation Among Major Cereals, Legumes,
Oilseeds, and Tuber Crops (http://www.nalusda.gov/pgdic/tables/nucdna.html)

Crop
Chromosome Genome size
Ploidy Common name Scientific name number (2n ¼ 2x) (Mbp/1C)

Diploid Arabidopsis Arabidopsis thaliana 10 145

Medicago Medicago truncatula 16 454–526
Rice Oryza sativa sps. Indica 24 419–463
Rice Oryza sativa sps. 24 415–439
japonica
Black mustard Brassica nigra 16 468
Turnip rape Brassica compestris sps. 20 468–516
oleifera
Turnip Brassica compestris sps 20 511
rapifera
Pakchoi Brassica compestris sps. 20 507
chinensis
White mustard Brassica hirta 24 492
Urdbean Vigna mungo 22 574
Moongbean Vigna radiata 22 579
Cowpea Vigna unguiculata 22 613
Lima bean Phaseolus lunatus 22 622
French bean Phaseolus vulgaris 22 637
Runner bean Phaseolus coccineus 22 709
Chickpea Cicer aerietinum 16 738
Jowar Sorghum bicolor 20 748, 772
Brown mustard Brassica juncea 36 1105
Tepary bean Phaseolus acutifolius 22 647
Sunflower Helianthus annuus 34 2871–3189
Garedn pea Pisum sativum 14 3947, 4397
Lentil Lensculinaris esculenta 14 4063
Barley Hordeum vulgare 14 4873
Tomato Lycopersicon 24 907–1000
esculentum
Soybean Glycine max 40 1115
Rapseed Brassica napus 38 1129–1235
Tobacco Nicotiana tabaccum 48 4221–4646
Oat Avena sativa 42 11315
Tetrapolid Groundnut Arachis hypogaea 40 2813
Cotton Gossypium hirsutum 52 2118, 2374
Alfalfa Medicago sativa 32, 16 1510
Potato Solanum tuberosum 48, 24, 72 1597–1862
Maize Zea mays 20 2292–2716
Hexaploid Wheat Triticum aestivum 42 15966
Sweetpotato Ipomoea batatas 90 1597
 Table VIII
Total Map Distance (cM), Number of Markers Mapped, and Average Marker Distance of Groundnut DNA Marker vis a vis Those Reported in Cereals,
Legumes, and Oilseeds Crops

MOLECULAR BREEDING OF GROUNDNUT

Total map distance (cM) Average marker
Crop Mapping population and # markers mapped distance (cM) Reference

Oryza sativa An F2 population of the cross 1521.6 cM and 2275 markers 0.67 Harushima et al., 1998
between sub sps japonoca and indica
Phaseolus vulgaris RILs from a cross between 1226 cM and 563 markers 2.18 Freyre et al., 1998
BAT93 £ Jalo EEP 558 (BJ)
Sorghum bicolor RILs from a cross between 1406 cM and 470 markers 2.99 Bhattramakki et al., 2000b
BTx623 and IS620C
Helianthus annuus An F2 population of the cross 1326 cM and 400 markers 3.31 Gedil et al., 2001
between HA370 and HA372
Glycine max RILs from a cross between BSR-101 3441 cM and 840 markers 2.90 Cregan et al., 1999
and PI437.654
Vigna unguiculata RILs from a cross between 972 cM and 181 markers 5.37 Menendez et al., 1997
IT84S-2049 and 524B
Arachis hypogaea BC1 population of a cross between 2210 cM and 370 markers 5.97 Burow et al., 2001
synthetic amphidiploid (TxAG-6)
{[A. batizocoi £ (A. cardenasii £
A. digoi)]4x} and Florunner
Cicer arietinum RILs from an interspecific cross between 981.6 cM and 116 markers 8.46 Santra et al., 2000
Cicer arietinum (FLIP84-92C)
and Cicer reticulatum

181
182 S.L. DWIVEDI ET AL.

the introgression of nematode resistance in wild species chromosome segments

from A. cardenasii in A. hypogaea.
High-density genetic linkage maps are theoretically useful to detect markers
tightly linked to quantitative trait loci (QTL) with economically important traits
(Paterson et al., 1988; Lander and Botstein, 1989), to clone gene(s) by
chromosome walking (Wicking and Williamson, 1991), and to facilitate marker-
assisted selection of desirable genes in breeding programs (Burr et al., 1983;
Tanksley et al., 1989). The average marker distance in the current RFLP based
tetraploid genetic linkage map of cultivated groundnut is greater than those
reported for Oryza sativa, Phaseolus vulgaris, Sorghum bicolor, Helianthus
annuus, Glycine max, and Vigna unguiculata but smaller than Cicer arietinum
(Table VIII). There is a need to saturate the groundnut genetic linkage map with
PCR-based co-dominant markers, preferably SSRs, in order to provide sufficient
markers for routine marker-assisted breeding.

C. PROGRESS IN MODEL SYSTEMS AND COMPARATIVE MAPPING

AMONGST LEGUMES

1. Marker Systems and Linkage Mapping in Soybean

Among the legumes, soybean has been the most widely studied crop for
development of suitable marker assays for assessment of genetic diversity,
marker-trait relationships, identifying genes/quantitative trait loci (QTL)
associated with useful traits, and constructing genetic linkage maps for map-
based cloning of genes for the targeted genetic enhancement in soybean. Since
1990, a large number of reports have been published on the use of DNA markers
in assessing genetic diversity for identifying diverse germplasm in soybean yield
improvement (Narvel et al., 2000; Brown-Guedira et al., 2000; Concibido et al.,
2003). There are also a considerable number of publications using the full range
of available types of marker assay to map the genes underlying a wide range of
biotic constraints (Concibido et al., 2003; www.gsfgg.uiuc.edu/invited/2_1_01.
pdf) and agronomic traits (Lee et al., 1996a,b; Orf et al., 1999; Sebolt et al., 2000;
Concibido et al., 2003), and the use of such maps for marker-assisted selection to
map-based cloning of genes for the genetic enhancement in soybean (Polzin et al.,
1994; Keim et al., 1997; Cregan et al., 1999).
Soybean yields in the past 75 years in USA have more than tripled from
12 bushels per acre in 1924 to 40 bushels per acre in recent years, and at least
half of it is attributable solely to genetic improvement through breeding.
However, a greater emphasis have now been placed on use of genomics to bring
rapid genetic enhancement in soybean yields (http://129.186.26.94/genomics/
soybean_genomics.html) that should enable US growers to make the
soybean production globally competitive and meet the ever increasing energy
 MOLECULAR BREEDING OF GROUNDNUT 183

(good quality oil and protein) demands of world population. The focused research
priorities for soybean improvement in USA are use of DNA markers,
transformation, structural genomics, functional genomics, and bio-informatics
technologies for increasing the genetic potential of soybean crop. The US
researchers have developed a time bound (in most of the cases 3 –5 years)
targeted action plan to provide large number of DNA markers (2000 SSR and
10,000 SNP markers in public domain); characterize allelic variation in major
candidate genes; improve the efficiency of transformation by 5 – 10 folds and
generate technology to precisely deliver DNA; develop transgenic screens to
elucidate gene function; tag 80% of the genes; develop and integrate the genetic,
physical, and transcript maps; assign biological function to identified genes; use
comparative genomics to understand soybean interaction with pathogens and
symbionts; and identify bioinformatics needs of the soybean genomic program
(http://129.186.26.94/genomics/soybean_genomics.html). The successful appli-
cation of biotechnology-assisted breeding of soybean provides considerable
direct and indirect support for similar progress in other legume crops.

2. Model Systems and Comparative Mapping

The family Leguminoseae consists of three subfamilies: Caesalpinioideae,

Mimosoideae, and Papilionoideae (Raven and Polhill, 1981; Herendeen et al.,
1992). Within the Papilionoideae, three evolutionary lineages are represented by
the beans (common beans, cowpea, and soybean), the cool season legumes (lentil,
pea, chickpea, and alfalfa), and groundnut (Stylosanthes). The close phylogenetic
relationship between these species suggests that a comparative genomics
approach will be useful to define the common attributes of this legume
subfamily. Thus, knowledge of genome structure and gene function gained from
the intensive study of model legume species such as Glycine, Medicago and Lotus
should enable more effective research in other legumes. With this in mind, an
international legume genomics initiative has been formed between USA
Universities and the International Agricultural Research Centers of the Eco-
Regional Alliance on legumes to translate the benefits of the “consensus legume
genome” for rapid impacts on the genetic improvement of tropical legumes. For
example, researchers will be able to determine if genes for drought resistance in
two legume species share a common origin, or if they are derived from different
genetic determinants. Alternatively, having intensively characterized the nature
and location of genes for a given trait in a model species, it may then be easy to
identify similar genes in another lesser studied crop. Such information will allow
leap-frogging progress in the genetic improvement of lesser studied crops and
may lead to rapid and cost effective means for breeders to carry out trait-based
mining of large germplasm collections.
184 S.L. DWIVEDI ET AL.

Comparative mapping studies in cereals have demonstrated that gene content

and orders are highly conserved between different species. Integration of the
genetic maps of rice, foxtail millet, sugarcane, sorghum, maize, the triticeae
cereals, and oats into a single synthesis reveals that some chromosome
arrangements characterize taxonomic groups, while others have arisen during
or after speciation (Devos and Gale, 1997). The linear organization of genes
among nine species in the grass family, differing in basic chromosome numbers
(5 –12) and nuclear DNA amount (400 – 6000 Mb), can be described in 25 “rice
linkage blocks” (Gale and Davos, 1998). Elucidation of the organization of the
economically important grasses with large genomes such as maize will to a
greater or lesser extent be predicted from sequence analysis of smaller genomes
such as rice. Synteny studies will be greatly aided by knowledge of the entire
sequence of Arabidopsis and in due course Medicago and Lotus. Examples of
conserved collinearity between Arabidopsis and Brassica (Kowalski et al., 1994;
Cavell et al., 1998; Quiros et al., 2001; Ryder et al., 2001), between Arabidopsis
and tomato (Ku et al., 2000, 2001), between tomato and potato (Tanksley et al.,
1992), between Arabidopsis and soybean (Grant et al., 2000), and between the
dicot genome of Arabidopsis with monocots such as sorghum and rice has been
reported (Paterson et al., 1996; van Dodeweerd et al., 1999; Mayer et al., 2001).
Lee et al. (2001) have suggested the use of Arabidopsis as a “bridge species” to
resolve the genome evolution among dicots. They not only reported conservation
of large regions of the genomes in soybean, Phaseolus vulgaris, and Vigna
radiata but these conserved regions were also relatively conserved in
Arabidopsis. They also suggested that there is conservation of blocks of DNA
between species as distantly related as legumes and brassicas, representing 90
million years of divergence. Cross-species, cross-genera, and cross-kingdom
comparisons are, therefore, providing critical information for understanding how
genes are structured, how gene structure relates to gene function, and how
changes in DNA have given rise to the biological diversity on the planet
(McCouch, 1998).
Preliminary comparative mapping studies have been conducted between
soybean and cowpea (Maughan et al., 1996), pea and lentil (Weeden et al., 1992),
pea and chickpea (Simon and Muehlbauer, 1997), mungbean and cowpea
(Menancio-Hautea et al., 1993; Fatokun et al., 1993), mungbean, common bean,
and soybean (Boutin et al., 1995), azuki bean and rice bean (Kaga et al., 2000),
and mungbean and lablab (Humphry et al., 2002). Localized synteny between
Arabidopsis and distantly related dicot crops (Paterson et al., 1996; Ku et al.,
2000) suggests that it may be possible to utilize progress in Arabidopsis and
Medicago genomes to enhance molecular breeding efforts in groundnut. For
example, the reported synteny between a segment of tomato chromosome 2 and
Arabidopsis chromosome 4 has been used to identify several expressed sequence
tags (ESTs) including TX680 that cosegregate with ovate fruit shape in tomato
(Ku et al., 2001). Conservation of the genome microstructure between
 MOLECULAR BREEDING OF GROUNDNUT 185

Arabidopsis and rice (22 of the 56 genes identified in the rice genome segement
were also represented in the corresponding Arabidopsis genome segment, with at
least five genes present, in conserved order, in each segment) can be identified
even between monocot and dicot species (Mayers et al., 2001). These reports
clearly demonstrate that rich sources of new markers can be obtained at relatively
low cost by mining public sequence databases (Ku et al., 2001). Thus, using the
identified position or known sequence of important genes in model species it may
be possible to quickly locate genes of similar function in lesser- studied crops.
Such approaches promise to dramatically enhance progress in molecular
breeding of groundnut.
Researchers have recently adopted Medicago truncatula as a model legume
particularly for the study of plant – microbe interactions (e.g., symbiotic nitrogen
fixation, mycorrhizal and legume – pathogen interactions) that cannot be studied
in Arabidopsis. It is the native to Mediterranean Basin, exhibits tolerance to
drought and salinity, and can be grown in a wide range of soil and environmental
conditions (Barker et al., 1990; Cook et al., 1997). The key attributes of this
species include: a small diploid genome (haploid chromosome number 8 and
genome size of about 5 £ 108 bp/1C), self-fertility nature, prolific seed
production, rapid generation cycling, and ease of transformation using
Agrobacterium tumefaciens and regenerated to yield fertile transgenic plants
(Cook, 1999). It has numerous ecotypes that exhibited wide diversity for growth
habit, flowering time, and disease resistance. It has been recognized as a potential
model crop for comparative mapping and syntenic relationships with Arabidopsis
and other legume crops (Cook, 1999).

D. WIDE CROSSES

Wild Arachis species harbor genes for resistance to many abiotic and biotic
stresses (Table IV), and for seed quality traits (Table V). Many of the wild species
are not cross compatible with A. hypogaea, and the major barrier for gene
introgression to A. hypogaea is post-zygotic failure of embryo development.
However, diploid species of section Arachis and the cultivated tetraploid A.
hypogaea can be crossed at the same ploidy level, reducing sterility in hybrids.
Strategic approaches to introgress genes from wild diploid species to A. hypogaea
include (1) interploidy crosses [between A. hypogaea (AABB genomes) and wild
diploid species (AA or BB genomes)], (2) artificial polyploidization (crosses
between A. hypogaea and autotetraploid wild species with either AA or BB
genomes), and (3) resynthesis (crosses between A. hypogaea and amphidiploid
wild species containing both AA and BB genomes or only the AA genome),
followed by recurrent backcrossing to A. hypogaea genotypes (Stalker and Moss,
1987; Singh et al., 1991; Simpson, 2001). These crossing schemes can be
expected to facilitate interspecific chromosome pairing that can result in different
186 S.L. DWIVEDI ET AL.

frequencies of inter- and intra-genomic recombinations. Using these techniques,

several interspecific tetraploid derivatives have been developed with the aim of
introgressing genes for resistance to rust, ELS, LLS, nematodes, southern corn
rootworm, corn earworm, Spodoptera, and jassids (Gardner and Stalker, 1983;
Moss, 1985; Singh, 1986a,b; Stalker and Moss, 1987; Singh et al., 1991; Simpson
et al., 1993; Stalker and Lynch, 2002; Stalker et al., 2002a,b). Simpson and Starr
(2001) released the first root-knot nematode-resistant peanut cultivar (Coan) in
USA that contains a pest resistant gene from A. cardenasii. However, this has
allowed only slow progress in transferring resistance genes from wild Arachis to
A. hypogaea in improved genetic backgrounds because of problems associated
with linkage drag. Exploitation of alien germplasm in the genus Arachis has so
far only in the primary and secondary gene pools. Use of an aneuploid series in
cultivated groundnut improvement might enhance the utilization of diploid
species of section Arachis from the secondary gene pool. The possibilities of
alien gene transfer from the tertiary gene pool within the accessible limit of A.
hypogaea also exist by using bridge species, in vitro fertilization and hormone
treatment, protoplast fusion, and plant regeneration techniques (Singh et al.,
1991).

E. GENETIC TRANSFORMATION

Sharma et al. (2000) reviewed the prospects for transgenic resistance and
concluded that with the advent of genetic transformation techniques, it has
become possible to clone and insert genes (d-endotoxins from Bacillus
thuringiensis (Bt), protease inhibitors, and enzymes and plant lectins) into crop
plants to confer resistance to insect pests. Holbrook and Stalker (2003) reviewed
the progress achieved in development of an efficient tissue culture and
transformation systems to introduce foreign DNA into groundnut, and the
transgenic plants developed having genes for resistance to Tomato Spotted Wilt
Virus and lesser cornstalk borer in USA. Transgenic groundnuts with IPCVcp or
IPCVreplicase, GRAVcp, and rice chitenase genes have been produced at
ICRISAT, and these are in various stages of characterization under containment
glasshouse and/or controlled field conditions (ICRISAT, 2001). The first
products of transgenic plants with IPCVcp gene are being evaluated for
resistance to peanut clump virus (PCV) under field conditions during 2002 rainy
season at Patancheru, India. A new initiative with Japan International Research
Center for Agricultural Sciences has been taken up to use their constructs
(rd29A:DREB1A) carrying drought responsive elements (DRE) of Arabidopsis
into Arachis for inducing drought resistance in groundnut. The putative
transformants obtained in Agrobacterium-mediated transformation are being
characterized for presence and expression of the introduced genes, and the
confirmed transgenic groundnut plants will be later on evaluated for their
 MOLECULAR BREEDING OF GROUNDNUT 187

response to drought stress conditions (Sharma and Lavanya, 2002). It is hoped

that once favorable genes for resistance to leaf miner, Spodoptera, groundnut
rosette assistor virus (GRAV), drought, and aflatoxin have been introduced into
cultivated groundnut accessions through wide crosses and/or genetic transform-
ation, these genes will become ideal candidates for marker-accelerated
introgression in support of adaptive breeding across the world.

V. OPPORTUNITIES FOR MOLECULAR BREEDING IN

GROUNDNUT

A. CURRENT UNDERSTANDING AND GENETIC BASIS OF

ECONOMICALLY IMPORTANT TRAITS

1. Rust

The characterized sources of resistance to rust in A. hypogaea exhibit

component mechanisms that reduce the rate of disease development. Thus
resistant lines have increased pathogen incubation periods, decreased infection
frequencies, and reduced pustule sizes, spore production, and spore germinability
(Subrahmanyam et al., 1983a; Mehan et al., 1994). In contrast, the characterized
sources of resistance in wild Arachis species and their interspecific derivatives
have more dramatic effects on the pathogen. In particular, uredosori on these
accessions are observed to be small (containing very few uredospores), slightly
depressed, and do not rupture to release their uredospores (Subrahmanyam et al.,
1983b). Resistance to rust in A. hypogaea is conferred either by a few recessive
genes (Knauft, 1987; Kalekar et al., 1984; Paramasivam et al., 1990) or
predominantly controlled by additive, dominance, and additive £ additive and
additive £ dominance genetic effects (Reddy et al., 1987; Varman et al., 1991).
In addition, partial dominance is reported in some diploid species (Singh et al.,
1984).

2. Early Leaf Spot (ELS)

Incubation period, infection frequency, lesion diameter, and defoliation are

important components of resistance to ELS. The resistant germplasm accessions
have longer incubation periods, reduced sporulation rates, lesion diameters,
infection frequencies, and less defoliation (Nevill, 1981; Waliyar et al., 1993;
http://www.icrisat.org/text/research/grep/homepage/annualreport/annualreport.
htm). Resistance to ELS is quantitative and controlled by additive, dominance,
and additive £ additive genetic effects (Kornegay et al., 1980; Hamid et al., 1981;
188 S.L. DWIVEDI ET AL.

Anderson et al., 1986; Green and Wynne, 1987). Maternal effects and/or
cytoplasmic factors have also been reported (Coffelt and Porter, 1986; Kornegay
et al., 1980; Sharief et al., 1978).

3. Late Leaf Spot (LLS)

Resistance to LLS is partial and is similar to the “slow rusting” type of

resistance. Sporulation rate, lesion size, lesion number, and latent period are
important components that contribute to a desired field score for LLS (Chiteka
et al., 1988; Anderson et al., 1990a). Resistant genotypes have longer incubation
periods, fewer lesions, and lower sporulation rates than susceptible genotypes
(Nevill, 1981). Both two-gene (Tiwari et al., 1984) and five-locus recessive
genetic models (Nevill, 1982) have been reported for resistance to LLS.
Recessive genes for resistance have been reported in crosses involving cultivated
groundnut and wild Arachis species (Sharief et al., 1978). Other studies report
predominantly additive genetic variance for most of the components of resistance
to LLS (Kornegay et al., 1980; Hamid et al., 1981; Anderson et al., 1986; Jogloy
et al., 1987).

4. Groundnut Rosette Disease (GRD)

All three agents [groundnut rosette virus (GRV), groundnut rosette assister
virus (GRAV), and satellite RNA (sat RNA)] should be present in the plant on
which the vector (Aphis crassivora) feeds for effective transmission of disease by
the vector (Bock et al., 1990). An efficient reverse transcription-polymerase
chain reaction (RT-PCR) assay has been developed that allows the detection of
each of the three components of the rosette virus syndrome (Naidu et al., 1998).
GRV resistance is controlled by two independent recessive genes in crosses
between resistant (RG 1 and RMP 40) and susceptible (JL 24, ICGM 48, and
Mani Pintar) germplasm in groundnut (Nigam and Bock, 1990). However,
dominant monogenic resistance to rosette was also reported in a cross between
RMP 12 and M 1204.781 (Olorunju et al., 1992) while resistance to aphids in
ICG 12991 appears to be recessive in nature (van der Merwa, pers. comm.). GRV
resistance seems to be effective against both chlorotic and green rosette.

5. Bacterial Wilt

Resistance to bacterial wilt in groundnut is a function of the duration of

the latent period, degree of vascular browning, hypersensitive reaction showing
partial symptoms, and rate of wilting. The late-maturing Virginia runner
 MOLECULAR BREEDING OF GROUNDNUT 189

(subsp. hypogaea var. hypogaea) and Dragon (subsp. hypogaea var. hirsuta)
types have longer latent periods than the early maturing Spanish (subsp.
fastigiata var. vulgaris) and Valencia (subsp. fastigiata var. fastigiata) types. A
large number of land races of Dragon types from south China are reported to be
resistant to bacterial wilt (Duan et al., 1993). Resistance has also been
correlated with specific root characteristics in some genotypes. The susceptible
genotypes tend to have a long and strong main root while the resistant
genotypes tend to have long lateral roots (Liao et al., 1992). Resistance to
bacterial wilt has been reported to be recessive or partially dominant in crosses
between resistant (Xiekangqing, Taishan Sanlirou, You 112, and Taishan
Zhenzhu) and susceptible (Honghua No. 1, E Hua No. 3, Heyou No. 4, and
Youguo) genotypes (Wang et al., 1985; Liao et al., 1986). Three major genes
have been reported to confer resistance to bacterial wilt in groundnut
accessions Xiekangqing, Taishan Sanlirou, Taishan Zhengzhu, and Hong Hua
1 (Liao et al., 1990). However, both cytoplasmic and nuclear factors have been
reported to confer resistance to bacterial wilt in some Dragon types.

6. Aflatoxin Contamination

Groundnuts are prone to aflatoxin contamination by A. flavus. Drought during

pod formation substantially increases the level of aflatoxin contamination. It was
reported that pre-harvest infection by A. flavus requires a drought period of
30– 50 days and a mean soil temperature of 29– 318C in the podding zone
(Cole et al., 1989; 1995). The susceptibility of groundnut to aflatoxin
contamination is related to lower water activity (0.80 –0.95) in the kernel and
favorable temperature (25 – 328C) for growth of A. flavus (Schearer et al., 1999).
As the kernel moisture content decreases under end-of-season drought, protection
from natural defense mechanisms is lost and the kernel becomes vulnerable to
colonization by A. flavus and aflatoxin contamination. Because of the high
correlation between seed moisture and pre-harvest aflatoxin contamination
(Dorner et al., 1989), there is the possibility to select for reduced pre-harvest
aflatoxin contamination by identifying germplasm with the capacity to maintain
high kernel water activity during severe drought stress. The drought tolerant
lines, PI 145681 and Tifton 8, support less pre-harvest aflatoxin contamination
than drought-intolerant line, Florunner (Holbrook et al., 2000).
Resistance to A. flavus in groundnut is reported to operate independently in at
least three tissues: pod, seed coat, and cotyledons (Mixon, 1986). Resistance to
pod infection is confered by pod wall structure and the presence of a wax layer
while resistance to seed invasion and colonization is correlated with thickness
and density of palisade cell layers and absence of fissures and cavities. However,
seed coat resistance is effective only in intact seed testa. Phenolics have also been
implicated in imparting resistance to seed infection (Pettit et al., 1989).
190 S.L. DWIVEDI ET AL.

Resistance to seed colonization by A. flavus, aflatoxin production, and pre-harvest

infection in crosses AR 4 £ NC 7 and GFA 2 £ NC 7 were controlled by different
genes all with low heritabilities (Utomo et al., 1990). However, Mixon (1976)
reported a high broad-sense heritability for percentage seed colonization in cross
PI 337409 £ PI 331326.
Lipoxygenase (LOX) enzymes and their products could play a role in the
Aspergillus- seed interaction. The C6 –C12 products of the LOX pathway inhibit
Aspergillus spore germination (Doehlert et al., 1993; Zeringue et al., 1996) and
methyl jasmonate inhibits aflatoxin biosynthesis but not fungal growth (Good-
rich-Tanrikulu et al., 1995). The 9S- and 13S-hydroperoxides differentially affect
Aspergillus mycotoxin biosynthesis (Burow et al., 1997; Gardner et al., 1998)
and these same hydroperoxides act as Aspergillus sporulation factors (Calvo
et al., 1999), suggesting that LOX isozymes play a role in regulating Aspergillus
infection and aflatoxin contamination in oil seeds crops. Burow et al. (2000)
cloned and characterized a peanut seed lipoxygenase gene, PnLOX1. This gene
encodes a 98 kDa protein highly similar in sequence and biochemical properties
to soybean LOX2. The gene is highly induced by Aspergillus infection and the
active protein produces a mixture of 9S- and 13S-hydroperoxides. PnLOX1 is an
organ-specific gene expressed in immature cotyledons but is highly induced by
methyl jasmonate, wounding, and Aspergillus infection in mature cotyledons.
Some of the cloned genes of aflatoxin biosynthetic pathway can be effectively
utilized to induce resistance to aflatoxin production.

7. Drought

A number of physiological mechanisms have been correlated with genotypic

differences in yield under drought stress including variation in transpiration,
water-use efficiency (WUE), and partitioning under end-of-season drought
stress (Nageswara Rao et al., 1993). Variation in WUE arises mainly from
genotypic differences in water use. Carbon isotope discrimination (D) can be
used to select genotypes with improved WUE under field drought stress
conditions. However, analysis of D is not economic particularly when to analyse
a large number of plants in segregating generations. A strong relationship
between WUE and specific leaf area (SLA) and between D and SLA revealed
that genotypes with thicker leaves had greater WUE (Wright et al., 1994). SLA
could, therefore, be used as a rapid and inexpensive indirect selection criterion
for WUE to facilitate selection for end-of-season drought tolerant genotypes
(Nageswara Rao and Wright, 1994). However, there appears to be a negative
relationship between WUE and partitioning under end-of-season drought stress
conditions suggesting that selection for high WUE might enhance groundnut
dry matter production under stress but not necessarily improve pod yield
(Wright et al., 1994; Nageswara Rao and Wright, 1994). SLA is also highly
 MOLECULAR BREEDING OF GROUNDNUT 191

influenced by G £ E interaction. Additive genetic effect for D, and both additive

and additive £ additive epistasis effects for SLA (Jayalakshmi et al., 1999;
Nigam et al., 2001) and partitioning (Dwivedi et al., 1998a; Nigam et al., 2001)
are reported. Variation in root characteristics and the ability of roots to
extract water from deeper layers of the soil profile have also been reported
(Ketring, 1984).
Drought stress triggers a number of physiological and developmental changes
associated with selective increase or decrease in the biosynthesis of a number of
distinct proteins that alter enzyme activity. The changes in protein profile are due
to changes in transcription rate, RNA stability, post-transcriptional control, and
protein turnover (Smirhoff and Colombe, 1989). Several genes responding to
dehydration at the transcriptional level have been reported in plant species
(Skriver and Mundy, 1990; Shinozaki and Yamaguchi-Shinozaki, 1996; Bray,
1997; Oliver et al., 1998; Tabaeizadeh, 1998). Using RT-PCR, Jain et al. (2001)
reported 43 peanut transcripts (mRNA) responsive to drought (PTRD) and these
show quantitative variation in their levels and duration of expression in tolerant
(PI 145681) and susceptible (Florunner) groundnut genotypes. PTRD-1, -10, and
-16 are completely suppressed due to prolonged drought in the tolerant genotype
indicating these transcripts may be used as markers along with other
morphological characters such as large root system and visual stress ratings for
screening genotypes with drought tolerant characteristics in groundnut (Ketring,
1984; Rucker et al., 1995; Holbrook et al., 2000).

8. Seed Quality Traits

One hundred-seed mass, oil content, and oleic (O) and linoleic (L) fatty acid
ratio are important seed quality traits in groundnut. Oil content is quantitatively
inherited trait (Layrisse et al., 1980; Makne and Bhale, 1987). Several studies
involving high oleic acid groundnuts revealed that high oleic acid is controlled by
two duplicate recessive genes, and one of the recessive alleles occurs with high
frequency in US peanut breeding populations whereas the other allele is rare
(Holbrook and Stalker, 2003). Oleic acid content is also reported to be influenced
by additive and additive £ additive genetic effects (Layrisse et al., 1980; Moore
and Knauft, 1989; Mercer et al., 1990; Upadhyaya and Nigam, 1999b). Jung et al.
(2000a) reported that high oleate groundnut resulted from reduction in the
activity or transcript level of microsomal oleoyl-PC desaturase. They isolated
two non-allelic but homoeologus genes, ahFAD2A and ahFAD2B, from the
developing peanut seed with a normal oleate seeds. ahFAD2A is expressed in
both normal and high oleate seeds. Reduction in ahFAD2B transcript levels in the
developing seeds is correlated with high oleate trait. Further studies revealed that
a mutation in ahFAD2A and a significant reduction in levels of the ahFAD2B
transcript together cause the high oleate phenotype, and expression of one gene
192 S.L. DWIVEDI ET AL.

encoding a functional enzyme appears to be sufficient for the normal oleate

phenotype (Jung et al., 2000b). Hundred seed mass is a quantitatively inherited
trait controlled by additive, dominance, and epistatic effects (Wynne et al., 1970;
Garet, 1976; Sandhu and Khera, 1976; Layrisee et al., 1980; Arunachalam et al.,
1985; Upadhyaya and Nigam, 1998).

9. Yield, Maturity, and Adaptation

Many agronomically important traits in groundnut appear to be quantitatively

inherited (reviewed by Murthy and Reddy, 1993). Additive, non-additive, and
epistatic genetic effects are reported for early maturity, pod yield, pods and seeds
per plant, pod length and width, seed length and width, shelling outturn, and
sound mature seeds (Parker et al., 1970; Wynne et al., 1970; 1975; Garet, 1976;
Sandhu and Khera, 1976; Gibori et al., 1978; Isleib et al., 1978; Layrisse et al.,
1980; Sangha and Labana, 1982; Arunachalam et al., 1985; Swe and Branch,
1986; Dwivedi et al., 1989; Upadhyaya and Nigam, 1998). Response to
photoperiod is controlled by additive gene action in some crosses and partial
dominance to dominance in others (Nigam et al., 1997). However, some
agronomically important traits have been reported to have a simple genetic basis.
For example, days to first flower is controlled by a single gene with additive gene
action (Upadhyaya and Nigam, 1994). Although, three independent genes with
complete dominance at each locus appear to control the number of days to the
accumulation of 25 flowers. Similarly, fresh seed dormancy in a cross between
dormant (ICGV 86158 and ICGV 87378) and non-dormant (JL 24) genotypes is
conferred by the dominant allele of a single gene (Upadhyaya and Nigam,
1999a).
Heterosis is reported in crosses between the subspecific groups of groundnut
for biomass, pod and seed yield, pod and seed size, pod and seed number per
plant, shelling outturn, and 100-seed mass (Wynne et al., 1970; Garet, 1976;
Layrisee et al., 1980; Isleib and Wynne, 1983; Swe and Branch, 1986; Dwivedi
et al., 1989), and its magnitude is linearly related to genetic divergence of the
parents (Isleib and Wynne, 1983; Arunachalam et al., 1982; 1984). Pod yield in
groundnut is a function of crop growth rate, reproductive duration, and
partitioning. However, the low heritability of these traits suggests that
conventional selection for them during early segregating generations will not
be very effective (Ntare and Williams, 1998).

B. DEVELOPING APPROPRIATE PCR-BASED MARKERS

Recent advances in the development of PCR-based marker protocols have

revolutionized genetic analysis and opened new possibilities in the study of
 MOLECULAR BREEDING OF GROUNDNUT 193

complex traits in crop plants. The hybridization-based co-dominant markers

(RFLP) and PCR-based dominant-markers (RAPD and AFLP) in many crops
have been superseded by co-dominant PCR-based markers (SSR). However,
when screening cultivated groundnut accessions with SSR markers (Hopkins
et al., 1999) polymorphisms were rarely found. In contrast, RAPD and RFLP
markers associated with resistance to nematodes have been reported in
interspecific crosses in groundnut (Burow et al., 1996; Choi et al., 1999).
However, both RAPD and RFLP technologies have their own limitations for
applications in large-scale marker-assisted breeding programs. The AFLP assay
has been frequently used in diversity and mapping studies in many crop plants.
However, effort to convert AFLP marker into simple co-dominant PCR markers
has met with mixed success, is laborist, expensive and time consuming.
The low level of detectable molecular genetic variation among cultivated
groundnut greatly constrains progress in molecular breeding of this crop. The
RFLP-based tetraploid map developed by Burow et al. (2001), based on an
interspecific cross, is likely to be useful in terms of locating specific genes of
interest in this interspecific cross. However, the markers themselves may be of
mixed value in molecular breeding programs as their linkage to loci of interest
may be lost as a result of different recombination patterns in cultivated crosses.
Clearly, there is a need to use assays that are more likely to reveal
polymorphisms, such as microsatellitic markers (SSR) and single nucleotide
polymorphisms (SNP). A collaborative project between ICRISAT and University
of Georgia (USA) has recently generated 192 SSR primer pairs which produce
scorable amplification products in cultivated groundnut from genomic libraries of
the groundnut cultivar, Florunner. To date 110 SSR markers reveal genetic
variation in a diverse array of 24 groundnut landraces (ME Ferguson, ICRISAT,
pers. comm.). Recent developments in SNP technology indicate that in the near
future, additional options may be available for rapid identification of large
numbers of polymorphic markers (Kanazin et al., 2002). SNPs comprise the
largest set of sequence variants in most organisms (Kwok et al., 1996; Kruglyak,
1997). SNPs are biallelic markers but occur very frequently within the genome,
their mutation rate is low, capable of high throughput genotyping, and are often
linked to genes (Kwok and Gu, 1999). For example a map containing 1.42 million
SNPs distributed throughout the human genome have been constructed, with an
average density of one SNP every 1.9 kb (The International SNP Map Working
Group, 2001). SNPs have also been reported in crop plants such as Arabidopsis
(Cho et al., 1999; Drenkard et al., 2000), barley (Schmitz et al., 2000; Kota et al.,
2001a,b), common bean (Melotto and Kelly, 2001), groundnut (Lopez et al.,
2000), maize (Bhattramakki et al., 2000a; Tenaillon et al., 2001), rice (Ayres
et al., 1997; Larkin and Park, 1999), and soybean (Coryell et al., 1999; Meksem
et al., 2001). SNPs map with a resolution of 3.5 cM have also been reported in A.
thaliana that has been used to map Eds16 gene, located at 7 cM interval on the
bottom of chromosome 1 between markers SNP 177 and SNP 231, involved in
194 S.L. DWIVEDI ET AL.

the defence response to the fungal pathogen Erysiphe orontii (Cho et al., 1999).
The generation of denser biallelic maps should allow high-throughput
indentification of both monogenic and polygenic traits and thus effectively
removing the rate-limiting nature of high-resolution mapping from the study of
biological processes (Cho et al., 1999).

C. MAPPING AND GENETIC ENHANCEMENT STRATEGIES

Marker-assisted selection (MAS) offers great promise for improving the

efficiency of conventional plant breeding. Molecular markers are especially
advantageous for traits where conventional phenotypic selection is difficult,
expensive or lacks accuracy or precision (Crouch, 2001). This includes
resistance to certain pathogens and insect pests plus tolerance to abiotic
stresses, quality parameters, and complex agronomic traits with low heritab-
ilities. The essential requirements for developing marker-assisted selection
systems are (1) availability of germplasm with substantially contrasting
phenotypes for the traits of interest, (2) highly accurate and precise screening
techniques for phenotyping mapping populations for the trait of interest,
(3) identification of flanking marker(s) closely associated with the loci of
interest and the flanking regions on either side, and (4) simple robust PCR-
based marker technology to facilitate rapid and cost effective screening of large
breeding populations.
Molecular marker studies using near-isogenic lines (NILs) (Muehlbauer et al.,
1988), bulked segregant analysis (Michelmore et al., 1991), and recombinant
inbred lines (RILs) (Burr et al., 1988) have accelerated the mapping of many
genes in different plant species. Advanced backcross QTL analysis has been
proposed for the simultaneous discovery and transfer of valuable QTL from
unadapted and wild germplasm into elite breeding lines (Tanksley and Nelson,
1996). This approach is effective for QTL from the donor line which have
dominant, partially dominant or over-dominant gene action, and allows the
generation of elite NIL for specific QTL for rapid variety development and
reduced linkage drag around targeted QTL.
Trait heritability, the proportion of additive genetic variance explained by the
marker loci affecting the trait, the selection method used, and the effective
population size influence the selection efficiency of both conventional and
marker-assisted breeding programs. MAS is equally effective for characters with
low heritability when additive (Lande and Thompson, 1990) or non-additive
(Gimelfarb and Lande, 1994) genetic variance are associated with the marker
loci. The effectiveness of MAS decreases as the linkage distances between
marker and target QTL increases. Linkage disequilibria between the marker loci
and QTL, that condition trait expression, also influence the effectiveness of MAS
(Lande and Thompson, 1990).
 MOLECULAR BREEDING OF GROUNDNUT 195

D. MARKER- ASSISTED GENE INTROGRESSION FROM

WILD ARACHIS TO ARACHIS HYPOGAEA

Wild Arachis species and exotic germplasm are usually agronomically inferior
to modern cultivars. However, reports in rice and tomato suggest that wild
germplasm may contain alleles capable of improving both yield and seed quality
of elite cultivars (Xiao et al., 1996; Tanksley and McCouch, 1997). However, the
effects of these alleles is often masked by the presence of deleterious genes at
nearby loci. Advanced backcross populations and molecular genetic tools can be
used to exploit the genetic potential of wild species for enhancing yield, seed
quality, and resistance to diseases of elite genetic materials (Tanksley and
Nelson, 1996). Whereas resistance to rust, ELS, LLS, and nematodes has been
successfully transferred into A. hypogaea from wild Arachis species, only early
attempts have been made to tap favorable genes from wild species for enhancing
yield and seed quality in groundnut. There is a need to exploit these, along with
disease resistance genes, to develop interspecific derivatives for enhanced
yield, seed quality, and resistance to abiotic and biotic stresses in groundnut.
MAS and marker-accelerated backcross breeding promise to dramatically
improve the efficiency and success for rapid transfer of alien chromosome
segments containing genes for yield, seed quality, and resistance to pests and
diseases as it minimizes the deleterious linkage drag that often a problem while
transferring genes from wild species or exotic germplasm by conventional
breeding techniques.

E. MARKER- ASSISTED BACKCROSS BREEDING

Marker-assisted backcrosss breeding facilitates gene introgression from a

“donor” line into the genomic background of a “recipient” line. Molecular
markers can be used to assess the presence of the introgressed genes (“foreground
selection”) and to accelerate the return to the recipient parent genome
(“background selection”). Over the past decade a number of important simulation
studies have been conducted to ascertain conditions under which MAS could be
competitive with conventional phenotypic selection. Frisch et al. (1999a)
determined the optimal positioning of flanking markers and minimum number of
individuals required to obtain, with a specific probability of success, at least one
desired individual when backcrossing to transfer a target allele. Their study
revealed that the length of the carrier chromosome, the chromosomal position of
the target locus, its distance to the flanking marker loci, and the number of
individuals evaluated influenced the efficiency of marker-assisted backcrossing.
Frisch et al. (1999b) then compared various selection strategies with regard to the
proportion of the recurrent parent genome (RPG) recovered and the number of
marker data points (MDP) required in a backcross program designed for
196 S.L. DWIVEDI ET AL.

introgression of one target allele from a donor line into a recipient line. They
concluded that increasing population sizes from generation BC1 to BC3, in
comparison to a constant population sizes across all generations, reduce the
number of required MDP by as much as 50% without affecting the proportion of
RPG. A four-stage selection approach, emphasizing in the first generations,
selection for recombinants on the carrier chromosome of the target allele,
reduced the required number of MDP by as much as 75% in comparison to a
selection index taking into account all markers across the genome. Frisch and
Melchinger (2001a) reported marker-assisted backcross strategy for the
simultaneous introgression of two genes with respect to RPG recovered and
the number of MDP required. Their simulation study, using data from published
genetic linkage map consisting of 80 markers and assuming selection for
dominant target genes in maize, revealed reduction in the number of back cross
generations from six to three can be attained with 1000 –1500 MDP for unlinked
as well for linked target locus. Small population sizes in early generations and
large population sizes in advanced generations require less MDP than constant or
decreasing population sizes while attaining the same RPG content. Frisch and
Melchinger (2001b) further demonstrated the use of marker-assisted backcross
breeding for introgression of a recessive target gene from a donor into the genetic
background of a recipient line by foreground selection combined with
background selection for reducing the donor chromosome segment around the
target gene.
Hospital and Charcosset (1997) provided a general framework for the
optimization of the use of molecular markers in backcross breeding programs
aimed at introducing one to several superior QTL into a recipient line. Using at
least three markers per QTL allows a good control of the donor chromosome
segment over several generations. When several target alleles are monitored
simultaneously, background selection among the limited number of individuals
resulting from the foreground selection step accelerates the increase in genomic
similarity with the recurrent parent with only limited increase in the cost. These
flanking markers should cover , 10 –20 cM around the estimated position of the
gene to ensure that allele frequency does not decline in later generations
(Visscher et al., 1996). Hospital et al. (1997) found that the relative efficiency of
MAS over purely phenotypic selection in the first generation increases with (1)
larger population sizes, (2) lower trait heritabilities, and (3) higher type-I error
risk. However, at low heritability the response to MAS is more variable than
response to phenotypic selection. The MAS may become less efficient than
phenotypic selection in long term as the rate of fixation of QTL with large effects
in early generations is balanced by a higher rate of fixation of unfavourable alleles
at QTL with small effects in later generations. MAS efficiency therefore depends
on the genetic determinism of the trait. Alternating generations of MAS and
conventional phenotypic selection appeared to offer the best improvement in
genetic gain per unit time in applied breeding programs. Sen and Churchill
 MOLECULAR BREEDING OF GROUNDNUT 197

(2001) developed simple Monte Carlo algorithm to implement Bayesian QTL

analysis for the genetic analysis of QTLs in an inbred line cross. This algorithm
simulates multiple version of complete genotype information on a genomewide
grid of locations using information in the marker genotype data. Weights are
assigned to the simulated genotypes to capture information in the phenotype data,
and the weighted genotypes are used to approximate quantities needed for
statistical inference of QTL locations and effect sizes. In this approach only
weights are recomputed as the analyst considers different candidate models. This
approach allows the analyst to focus on modeling and model comparisons, and
can accommodate multiple interacting QTL, non-normal and multivariate
phenotypes, covariates, missing genotype data, and genotyping errors in any
type of inbred line cross.

F. PRIORITIZING TRAITS FOR MARKER- ASSISTED SELECTION

The major constraints to groundnut productivity have been discussed in

Section 1B. Table IX summarizes the traits of economic importance and the
suggested conventional and non-conventional techniques for genetic enhance-
ment in groundnut. For many traits, adequate and cost effective progress is being
made through traditional approaches. Traits for which MAS is not justified
include maturity, pod yield, pod size and shape (except in situations wherein
resistance to pests and diseases is linked with undesirable traits), seeds per pod,
seed color, shelling outturn, sound mature seeds, 100-seed weight, and seed
dormancy as well resistance to rust and bacterial wilt. There is a large pool of
genetic variation reported for these traits in cultivated groundnut gemplasm
(Singh and Nigam, 1997; Rajgopal et al., 1997; Upadhyaya et al., 2001e; Table
IV) that are easy to exploit through conventional breeding techniques. Several
cultivars with these characteristics have been developed and are commercially
grown in semi-arid tropics (Table VI).
In general, traits that justify the cost and time required to develop and apply
MAS system, will include those that are difficult or expensive to score, traits that
are associated with deleterious linkage drag, traits that are controlled by different
genetic mechanisms such as GRD (GRV, GRAV, and sat RNA) or traits where
the application of DNA markers will allow breeders to address new goals. For
traits such as ELS, LLS, nematodes, leafminer, and Spodoptera there are only
low to moderate levels of resistance (or tolerance) available in cultivated
groundnut. In contrast, many wild Arachis species show a very high degree of
resistance to these diseases and pests (Table IV). However, the resistant wild
species are often sexually incompatible with cultivated groundnut. Efforts to
overcome incompatibility in wide crosses for transferring resistance genes from
the tertiary gene pool of genus Arachis by using non-conventional techniques
have had limited success but are beginning to liberate useful interspecific
 198
Table IX
Traits and Breeding Strategy Suggested for Rapid and Cost Effective Genetic Enhancement in Groundnut

Wide Crosses þ
Conventional Marker-assisted Marker Assisted Genetic
Trait Breeding Selection Backcross transformation Genetic basis

Maturity † Polygenic, refer page 192

Pod yield † Polygenic, refer page 192
Pod size and shape † Polygenic, refer page 192
100-seed weight † Polygenic, refer page 192
Shelling outturn † Polygenic, refer page 192

S.L. DWIVEDI ET AL.

Sound mature seeds † Polygenic, refer page 192
Seed dormancy † Monogenic, refer page 192
O/L ratioa † Oligogenic, refer page 191–192
Aflatoxin † Polygenic, refer page 189–190
Drought † † Polygenic, refer page 190–191
Leaf miner † Not known
Spodoptera † Not known
Rust † Oligogenic, refer page 187
ELSb † Polygenic, refer page 187–188
LLSc † Both oligo- and poly-genic, refer page 188
Bacterial wilt † Oligogenic, refer page 188–189
GRDd † † Mono- and diagenic refer page 188
PBNDe † Not known
a
O/L, Oleic/Linoleic fatty acid ratio.
b
ELS, Early leaf spot.
c
LLS, Late leaf spot.
d
GRD, Groundnut rosette disease.
e
PBND, Peanut bud necrosis disease.
 MOLECULAR BREEDING OF GROUNDNUT 199

progenies (N Mallikarjuna, ICRISAT, pers. comm.). Marker-assisted back cross

breeding is therefore suggested for rapid transfer of resistant gene(s) from wild
Arachis to cultivated groundnut. Efforts to select for high levels of resistance to
PBNV or TSWV have received with limited success by conventional breeding
techniques. Similarly most of the GRD resistant accessions of cultivated
groundnut germplasm are resistant to GRV with few exceptions that are resistant
to aphids but susceptible to GRV, and none resistant to GRAV. Several wild
Arachis species, on contrary, are reported to possess high level of resistance to
GRAV, sat RNA, PBNV, and TSWV. An efficient tissue culture and
transformation system has been reported in groundnut. Wide crosses and/or
genetic transformation are therefore suggested to introduce genes for resistance
to these diseases into cultivated groundnut. Wild Arachis species are no better
than levels of resistance reported for aflatoxin in cultivated groundnut accessions.
Thus, a transgenic approach may be the most effective option to introduce genes
for resistance to aflatoxin in groundnut.
Traits associated with seed quality (as measured by O/L ratio: higher the ratio
better the shelf-life of the groundnut products) and drought tolerance (specific
leaf area, total transpiration, water use efficiency, and partitioning) are difficult
and uneconomic to measure in large segregating generations. They are also
substantially influenced by genotype-by-environment interaction. Thus, breeding
progress in these traits by conventional techniques has had limited success. MAS
may therefore be a highly justified option for indirect selection for high O/L ratio
and drought tolerance in groundnut.

VI. CONCLUSION

Groundnut is extensively grown in the semi-arid tropics (SAT) by resource-

poor farmers. Several abiotic and biotic stresses limit groundnut productivity and
affect its seed quality. Drought and temperature among abiotic stresses and rust,
early leaf spot (ELS), late leaf spot (LLS), and aflatoxin among biotic stresses are
the global constraints to groundnut production and adversely influence seed
quality. Regionally, groundnut rosette disease (GRD) in Africa; bacterial wilt,
leafminer, Spodoptera, and peanut bud necrosis disease (PBND) in South and/or
South East Asia; corn earworm, lesser corn stock borer, southern corn rootworm,
Sclerotium, nematodes, and tomato spotted wilt virus (TSWV) in North America;
and low calcium and phosphorus availability in acidic soils in Latin America and
Caribbean are important constraints to groundnut production. These stresses
together cause annual yield losses exceeding US $ 3.2 billion, and probably half of
this could be recovered through genetic enhancement in groundnut (ICRISAT,
1994).
200 S.L. DWIVEDI ET AL.

Researchers have made excellent progress toward developing elite groundnut

germplasm/cultivars with specific traits. However, there has been only limited
success in introgressing good seed quality and resistance to the above mentioned
constraints into an elite genetic backgrounds. The major constraints to rapid
genetic enhancement include (1) disease resistance genes are often closely linked
with loci conferring undesirable pod and seed characteristics, (2) disease resistant
germplasm are late maturing types, have lower partitioning, and are sensitive to
photoperiod than agronomically elite susceptible materials, (3) large genotype-
by-environment interactions for traits of economic importance, and (4) limited
gene introgression from wild Arachis species to cultivated groundnut. The
application of DNA markers will allow breeders to break non-pleiotropic
associations and pyramid genes for resistance to abiotic and biotic stresses with
improved productivity and seed quality of groundnut.
Assessment of genetic diversity and development of a saturated genetic
linkage map are important steps in the development of molecular marker-assisted
breeding programs. There are over 15,000 cultivated groundnut accessions
maintained in ICRISAT gene bank. ICRISAT scientists have developed a core
collection consisting of 1704 accessions, and this core when further evaluated
could provide new sources of variation for use in breeding programs. The Arachis
species harbor genes capable of improving both seed yield and quality in addition
to high levels of resistance to ELS, LLS, nematodes, leaf miner, and Spodoptera.
Many of the wild Arachis species are not cross compatible with cultivated
groundnut. However, efforts to overcome incompatibility in wide crosses, by
using non-conventional techniques, have started to liberate interspecific
progenies with high levels of resistance to leaf spots, nematodes, Spodoptera,
and leafminer. Marker-assisted backcross breeding should minimize the linkage
drag as it greatly facilitate monitoring of introgressed chromosome segments
carrying beneficial genes from wild Arachis to cultivated groundnut. An efficient
tissue culture and transformation system has been developed, and transgenic
groundnut plants with IPCVcp or replicase, GRAVcp, and rice chitenase genes
have been produced that are in various stages of characterization under
containment glasshouse and/or field conditions at ICRISAT. Transgenic
approach may be the best option to introduce genes for resistance to aflatoxin
as conventional breeding has failed to enhance the level of resistance beyond that
present in cultivated groundnut germplasm. For traits such as GRAV, PBNV, and
TSWV the use of wide hybridization and/or genetic transformation may be the
most efficient strategy to introduce resistance genes into cultivated groundnut.
Once favorable genes are introduced into cultivated groundnut through wide
crossing and/or genetic transformation techniques, these genes will become ideal
candidates for marker-accelerated introgression.
DNA marker based genetic linkage map should enable breeders to effectively
pyramid genes for good seed quality (high O/L ratio and resistance to aflatoxin)
and resistance to ELS, LLS, aflatoxin, nematodes, leafminer, Spodoptera and
 MOLECULAR BREEDING OF GROUNDNUT 201

tolerance to drought into agronomically enhanced breeding populations in a much

shorter time than would be possible by conventional techniques. Recombinant
inbred lines (RILs) mapping populations are being developed to map the genes
underlying most of these traits. Meanwhile, substantial efforts are still required to
develop sufficient PCR-based markers (particularly SSR and SNP markers) for the
construction of high-density genetic linkage map and for the routine application in
the molecular breeding of abiotic stress tolerance, biotic stress resistance, yield,
and seed quality in groundnut. The use of automated technologies will become
increasingly important for large-scale germplasm characterization and realistic
scale marker-assisted genetic enhancement in groundnut.

ACKNOWLEDGMENTS

The authors wish to thank C.T. Hash, D.V.R. Reddy, R.P. Thakur, K.K. Sharma,
N. Mallikarjuna, P.J.A. ven der Merwe, E. Mace (ICRISAT), R.L. Jarret
(USDA), and R. Ortiz (IITA) for their feedback and discussions during the
development of this manuscript. Assistance provided by T. Ravindrakumar,
D. Yadgiri, and K.D.V. Prasad for formatting the text and figures is gratefully
acknowledged.

REFERENCES

Ahmad, M. S., and Young, C. T. (1982). Composition, quality and flavor of peanut. In “Peanut Science
and Technology” (H. E. Pattee and C. T. Young, Eds.), pp. 653 –688, Yokohama, USA.
Amin, P. W. (1985). Resistance of wild species of groundnut to insect and mite pests. In “Proc. Int.
Workshop on Cytogenetics of Arachis.” (J. P. Moss and S. D. Feakin, Eds.), pp. 57–60.
ICRISAT, Patancheru, 502324, AP, India.
Amin, P. W., Singh, K. N., Dwivedi, S. L., and Rao, V. R. (1985). Sources of resistance to the jassids
(Empoasca kerri Pruthi), thrips [(Frankliniella schultzei (Trybom)] and termites (Odontotermis
sp.) in groundnut (Arachis hypogaea L.). Peanut Sci. 12, 58–60.
Anderson, W. F., Wynne, J. C., and Green, C. C. (1986). Potential for incorporation of early and late
leaf spot resistance in peanut. Plant Breed. 97, 163– 170.
Anderson, W. F., Beute, M. K., Wynne, J. C., and Wongkaew, S. E. (1990a). Statistical procedure for
assessment of resistance in a multiple foliar disease complex of peanut. Phytopathology 80,
1451–1459.
Anderson, W. F., Patanothai, A., Wynne, J. C., and Gibbons, R. W. (1990b). Assessment of diallel
cross for multiple foliar pest resistance in peanut. Oleagineux 45, 373 –378.
Arunachalam, V., Bandyopadhyay, A., Nigam, S. N., and Gibbons, R. W. (1982). Heterotic potential
of single crosses in groundnut (Arachis hypogaea L.). Oleagineux 37, 415–420.
Arunachalam, V., Bandyopadhyay, A., Nigam, S. N., and Gibbons, R. W. (1984). Heterosis in relation
to genetic divergence and specific combining ability in groundnut (A. hypogaea L.). Euphytica
33, 33–39.
202 S.L. DWIVEDI ET AL.

Arunachalam, V., Bandyopadhyay, A., and Rao, M. V. K. (1985). Performance of three-way crosses in
groundnut. Indian J. Agric. Sci. 55, 75– 81.
Ayres, N. M., McClung, A. M., Larkin, P. D., Bligh, H. F. J., Jones, C. A., and Park, W. D. (1997).
Microsatellite and single nucleotide polymorphism differentiate apparent amylose class in an
exotic pedigree of US rice germplasm. Theor. Appl. Genet. 94, 773– 781.
Baltensperger, D. D., Prine, G. M., and Dunn, R. A. (1986). Root-knot nematode resistance in Arachis
glabrata. Peanut Sci. 13, 78– 80.
Barker, D. G., Bianchi, S., London, F., Dattee, Y., Duc, G., Essad, S., Flament, P., Gallusci, P., Genier,
G., Guy, P., Muel, X., Tourneur, J., Denarie, J., and Huguet, T. (1990). Medicago truncatula, a
model plant for studying the molecular genetics of the Rhizobium-legume symbiosis. Plant Mol.
Biol. 8, 40 –49.
Bhagwat, A., Krishna, T. G., and Bhatia, C. R. (1997). RAPD analysis of induced mutants of
groundnut (A. hypogaea L.). J. Genet. 76, 201–208.
Bhattramakki, D., Ching, A., Dolan, M., Register, J., Tingey, S., and Rafalski, A. (2000a). Single
nucleotide polymorphisms (SNPs) in corn: early lessons. Maize-Genetics-Cooperation-Newsletter
74, pp. 54.
Bhattramakki, D., Dong, J., Chhabra, A. K., and Hart, G. E. (2000b). An integrated SSR and RFLP
linkage map of Sorghum bicolor (L.) Moench. Genome 43, 988–1002.
Bock, K. R., Murant, A. F., and Rajeshwari, R. (1990). The nature of resistance in groundnut to rosette
disease. Ann. Appl. Biol. 117, 379 –384.
Bockelee-Morvan, A. (1983). Les differentes varieties d’arachide: Repartition-geographique et
climatique, disponibilite. Oleaginuex 38, 73–76.
Boutin, S. R., Young, N. D., Olson, T. C., Yu, Z.-H., Shoemaker, R. C., and Vallejos, C. E. (1995).
Genome conservation among three legume genera detected with DNA markers. Genome 38,
928–937.
Bradbury, J. F. (1986). Guide to plant pathogenic bacteria. Kew, UK. pp 332.
Branch, W. D. (1994). Registration of Georgia Browne peanut. Crop Sci. 34, 1125–1126.
Branch, W. D. (1996). Registration of Georgia Green peanut. Crop Sci. 36, 806.
Branch, W. D. (1998). Registration of Georgia Bold peanut. Crop Sci. 38, 895–896.
Branch, W. D. (2000). Registration of Georgia Hi-O/L peanut. Crop Sci. 40, 1823– 1824.
Branch, W. D. (2001). Registration of Georgia Valencia peanut. Crop Sci. 41, 2002–2003.
Branch, W. D., and Hildebrand, G. L. (1989). Pod yield comparison of pure-line peanut selections
simultaneously developed from Georgia and Zimbabwe breeding programs. Plant Breed. 102,
260–263.
Branch, W. D., Takayama, T., and Chinan, M. S. (1990). Fatty acid variation among U.S. runner-type
peanut cultivars. J. Amer. Oil Chem. Soc. 67, 591–593.
Bray, E. A. (1997). Plant responses to water deficit. Trends Plant Sci. 2, 48 –54.
Brown-Guedira, G. L., Thompson, J. A., Nelson, R. L., and Warburton, M. L. (2000). Evaluation of
genetic diversity of soybean introductions and north American ancestors using RAPD and SSR
markers. Crop Sci. 40, 815–823.
Buddenhagen, I. W., and Kelman, A. (1964). Biological and physiological aspects of bacterial wilt
caused by Pseudomonas solanacearum. Ann. Rev. Phytopathology 2, 203–230.
Burow, M. D., Simpson, C. E., Paterson, A. H., and Starr, J. L. (1996). Identification of peanut
(Arachis hypogaea L.) RAPD markers diagnostic of root-knot nematode (Meloidogyne arenaria
(Neal) Chitwood) resistance. Mol. Breed. 2, 369–379.
Burow, G. B., Nesbitt, T. C., Dunlap, J., and Keller, N. P. (1997). Seed lipoxygenase products
modulate Aspergillus mycotoxin biosynthesis. Mol. Plant-Microbe Interact. 10, 380 –387.
Burow, G. B., Gardner, H. W., and Keller, N. P. (2000). A peanut seed lipoxygenase responsive to
Aspergillus colonization. Plant Mol. Biol. 42, 689–701.
Burow, M. D., Simpson, C. E., Starr, J. L., and Paterson, A. H. (2001). Transmission genetics of
chromatin from a synthetic amphiploid in cultivated peanut (A. hypogaea L.): Broadening the
gene pool of a monophyletic polyploid species. Genetics 159, 823–837.
 MOLECULAR BREEDING OF GROUNDNUT 203

Burr, B., Evola, S. V., Burr, F. A., and Beckmann, J. S. (1983). The application of restriction
fragment length polymorphism to plant breeding. In “Genetic Engineering” (J. K. Setlow and
A. Hollaender, Eds.), Vol. 5, pp. 45–49. Plenum Press, New York.
Burr, B., Burr, F. A., Thompson, K. H., Albertson, M. C., and Stubber, C. W. (1988). Gene mapping
with recombinant inbreds in maize. Genetics 118, 519–526.
Calvo, A., Hinze, L., Gardner, H. W., and Keller, N. P. (1999). Sporogenic effect of polyunsaturated
fatty acids on Aspergillus spp. development. Appl. Environ. Microbiol. 65, 3668–3673.
Campbell, W. V., and Wynne, J. C. (1980). Resistance of groundnuts to insects and mites. In “Proc.
Int. Workshop on Groundnuts” (R. W. Gibbons, Ed.), pp. 149 –157. ICRISAT, Patancheru,
India.
Campbell, W. V., Emery, D. A., and Gregory, W. C. (1971). Registration of GP-NC343 peanut
germplasm. Crop Sci. 11, 605.
Cavell, A. C., Lydiate, D. J., Parkin, I. A. P., Dean, C., and Trick, M. (1998). Collinearity between a 30
centimorgan segment of Arabidopsis thaliana chromosome 4 and duplicated regions within the
Brassica napus genome. Genome 41, 62– 69.
Chandran, K., and Pandya, S. M. (2000). Morphological characterization of Arachis species of section
Arachis. Plant Genet. Resour. Newsletter 121, 38 –41.
Chauhan, Y. S., and Senboku, T. (1997). Evaluation of groundnut genotypes for heat tolerance. Ann.
Appl. Biol. 131, 481– 489.
Cherry, J. P. (1977). Potential sources of peanut seed proteins and oil in the genus Arachis. J. Agric.
Food Chem. 25, 186–193.
Chiteka, J. A., Gorbet, D. W., Knauft, D. A., Shokes, F. M., and Kucharek, T. A. (1988). Components
of resistance to late leaf spot in peanut. II. Correlation among components and their significance
in breeding for resistance. Peanut Sci. 15, 76–81.
Cho, R. J., Mindrinos, M., Richards, D. R., Sapolsky, R. J., Anderson, M., Drenkard, E., Dewdney, J.,
Reuber, T. L., Stammers, M., Federspiel, N., Theologis, A., Yang, W-H., Hubell, E., Au, M.,
Chung, E. Y., Leshkari, D., Lemieux, B., Dean, C., Lipshutz, R. J., Ausubel, F. M., Davis, R. W.,
and Oefner, P. J. (1999). Genome-wide mapping with biallelic markers in Arabidopsis thaliana.
Nature Genet. 23, 203–207.
Choi, K., Burow, M. D., Church, G., Burow, G., Paterson, A. H., Simpson, C. E., and Starr, J. L.
(1999). Genetics and mechanism of resistance to Meloidogyne arenaria in peanut germplasm.
J. Nematology 31, 283–290.
Coffelt, T. A., and Mozingo, R. W. (1998). Registration of VGP 10 peanut germplasm. Crop Sci. 38,
573 –574.
Coffelt, T. A., and Porter, D. M. (1986). Field screening of reciprocal Chico £ Florigiant peanut
populations for resistance to leafspot in Virginia. Peanut Sci. 13, 57–60.
Coffelt, T. A., Seaton, M. L., and VanScoyoc, S. W. (1989). Reproductive efficiency of 14 Virginia-
type peanut cultivars. Crop Sci. 29, 1217–1220.
Coffelt, T. A., Wynne, J. C., and Monteverde-Penso, E. J. (1993). Genotype £ environment interaction
effects on cultivar development. Oleagineux 48, 1217–1220.
Coffelt, T. A., Phipps, P. M., and Porter, D. M. (1994a). Registration of VGP 9 peanut germplasm.
Crop Sci. 34, 1132–1133.
Coffelt, T. A., Porter, D. M., and Mozingo, R. W. (1994b). Registration of VA 93B peanut. Crop Sci.
34, 1126.
Cole, R. J., Sanders, T. H., Dorner, J. W., and Blankenship, P. D. (1989). Environmental conditions
required to induce preharvest aflatoxin contamination of groundnuts: Summary of six years
research. In “Aflatoxin Contamination of Groundnut: Proc. Int. Workshop” (D. McDonald, V. K
Mehan and S. D. Hall, Eds.), pp. 279 –287. ICRISAT, Patancheru, India.
Cole, R. J., Dorner, J. W., and Holbrook, C. C. (1995). Advances in mycotoxin elimination and
resistance. In “Advances in Peanut Science”. (H. E. Pattee and H. T. Stalker, Eds.), pp. 456–474.
Stillwater, USA.
204 S.L. DWIVEDI ET AL.

Concibido, V. C., Vallee, B. L., Mclaird, P., Pineda, N., Meyer, J., Hummel, L., Yang, J., Wu, K., and
Delannay, X. (2003). Introgression of a quantitative trait locus from Glysine soja into commercial
soybean cultivars. Theor. Appl. Genet. 106, 575– 582.
Cook, D. R. (1999). Medicago truncatula – a model in the making. Special issue: Biotic interactions.
Current-Opinion-in-Plant-Biology 2, 301–304.
Cook, B. G., and Crosthwaite, I. C. (1994). Utilization of Arachis species as forage. In “The
Groundnut Crop: a scientific basis for improvement” (J. Smart, Ed.), pp. 624– 663. Chapman and
Hall, London.
Cook, D. R., VandenBosch, K. A., de Bruijin, F. J., and Huguet, T. (1997). Model legumes get the nod.
Plant Cell 9, 275–281.
Coryell, V. H., Jessen, H., Schupp, J. M., Webb, D., and Keim, P. (1999). Allele-specific hybridization
markers for soybean. Theor. Appl. Genet. 98, 690–696.
Craufurd, P. Q., Wheeler, T. R., Ellis, R. H., Summerfield, R. J., and Williams, J. H. (1999). Effect of
temperature and water deficit on water-use efficiency, carbon isotope discrimination, and specific
leaf area in peanut. Crop Sci. 39, 136–142.
Craufurd, P. Q., Prasad, P. V. V., and Summerfield, J. (2002). Dry matter production and rate of
change of harvest index at high temperature in peanut. Crop Sci. 42, 146–151.
Cregan, P. B., Jarvik, T., Bush, A. L., Shoemaker, R. C., Lark, K. G., Kahler, A. L., Kaya, N.,
VanToai, T. T., Lohnes, D. G., Chung, J., and Specht, J. E. (1999). An integrated gentic linkage
map of the soybean genome. Crop Sci. 39, 1464–1490.
Crouch, J. H. (2001). Molecular marker-assisted breeding: a perspective for small to medium-sized
plant breeding companies. Asia and Pacific Seed Assoc. Tech. Report 30, 1 –14.
Cruickshank, A. L., Wright, G. C., and Nageswara Rao, R. C. (2000). Streeton – An aflatoxin
tolerant peanut cultivar for the Australian peanut industry. Proc. Amer. Peanut Res. Educ. Soc.
32, pp. 27.
Culbreath, A. K., Todd, J. W., and Chamberlain, J. R. (1992). Disease progress of spotted wilt in
peanut cultivars Florunner and Southern Runner. Phytopathology 82, 66–771.
Devos, K. M., and Gale, M. D. (1997). Comparative genetics in the grasses. Plant Mol. Biol. 35,
3–15.
Doehlert, D. S., Wicklow, D. T., and Gardner, H. W. (1993). Evidence implicating the lipoxygenase
pathways in providing resistance to soybean against Aspergillus flavus. Phytopathology 83,
1473–1477.
Dorner, J. W., Cole, R. J., Sanders, T. H., and Blankenship, P. D. (1989). Interrelationship of kernel
water activity, soil temperature, maturity and phytoalexin production in preharvest aflatoxin
contamination of drought-stresses peanuts. Mycopathologia 105, 117–128.
dos Santos, R. C. (1998). EMBRAPA releases BRS 151 Amendoim L 7, large-seeded groundnut
cultivar for the Northeast Region in Brazil. Int. Arachis Newsletter 18, 11 –12.
Drenkard, E., Richter, B. G., Rozen, S., Stutius, L. M., Angell, N. A., Mindrinos, M., Cho, R. J.,
Oefner, P. J., Davis, R. W., and Ausubel, F. M. (2000). A simple procedure for the analysis of
single nucleotide polymorphisms facilitates map-based cloning in Arabidopsis. Plant Physiol.
124, 1483–1492.
Duan, N. X., Tan, Y. J., Jiang, H., and Hu, D. H. (1993). Screening groundnut germplasm for
resistance to bacterial wilt. Oil Crops of China 1, 22– 25.
Duncan, W. G., McCloud, D. E., McGraw, R. L., and Boote, K. J. (1978). Physiological aspects of
peanut yield improvement. Crop Sci. 18, 1015–1020.
Dwivedi, S. L., Thandapani, K., and Nigam, S. N. (1989). Heterosis and combining ability studies and
relationship among fruit and seed characters in peanut. Peanut Sci. 16, 14–20.
Dwivedi, S. L., Nigam, S. N., Jambunathan, R., Sahrawat, K. L., Nagabhushanam, G. V. S., and
Raghunath, K. (1993a). Effects of genotypes and environments on oil content and oil
quality parameters and their correlation in peanut (Arachis hypogaea L.). Peanut Sci. 20,
84–89.
 MOLECULAR BREEDING OF GROUNDNUT 205

Dwivedi, S. L., Reddy, D. V. R., Nigam, S. N., Rao, G. V. R., Wightman, J. A., Amin, P. W.,
Nagabhushanam, G. V. S., Reddy, A. S., Scholberg, E., and Ramraj, V. M. (1993b). Registration
of ICGV 86031 peanut germplasm. Crop Sci. 33, 220.
Dwivedi, S. L., Nigam, S. N., Nagabhushanam, G. V. S., Rao, Y. L. C., and Gibbons, R. W. (1994).
Registration of ICGV 86564 peanut germplasm. Crop Sci. 34, 821.
Dwivedi, S. L., Nagabhushanam, G. V. S., Amin, P. W., Rao, G. V. R., Nigam, S. N., and Wightman,
J. A. (1995a). Registration of four jassid-resistant peanut germplasm lines: ICGV 86252, ICGV
86393, ICGV 86455, and ICGV 86462. Crop Sci. 35, 1716.
Dwivedi, S. L., Nigam, S. N., Reddy, D. V. R., Reddy, A. S., and Rao, G. V. R. (1995b). Progress in
breeding groundnut varieties resistant to peanut bud necrosis virus and its vector. In “Recent
Studies on Peanut Bud Necrosis Disease” (A. A. M. Buiel, J. E. Parlevliet and J. M. Lenne, Eds.),
pp. 35 –40. ICRISAT, Patancheru, India and Agricultural University of Wageningen, The
Netherlands.
Dwivedi, S. L., Nigam, S. N., Rao, Y. L. C., and Nagabhushanam, G. V. S. (1996a). Registration of
ICGV 86325 peanut. Crop Sci. 36, 806.
Dwivedi, S. L., Nigam, S. N., Reddy, D. V. R., Rao, G. V. R., and Reddy, A. S. (1996b). Registration
of ICGV 86388 peanut germplasm. Crop Sci. 36, 1423.
Dwivedi, S. L., Nigam, S. N., Chandra, S., and Ramraj, V. M. (1998a). Combining ability of biomass
and harvest index under short- and long-day conditions in groundnut. Ann. Appl. Biol. 133,
237 –244.
Dwivedi, S. L., Nigam, S. N., and Prasad, M. V. R. (1998b). Induced genetic variation for seed quality
traits in groundnut. Int. Arachis Newsletter 18, 44 –46.
Dwivedi, S. L., Gurtu, S., Chandra, S., Yuejin, W., and Nigam, S. N. (2001). Assessment of genetic
diversity among selected groundnut germplasm. I. RAPD analysis. Plant Breed. 120, 345–349.
FAO. 1999. http://apps.fao.org/page/collections?subset=agriculture.
FAO. 2000. http://apps.fao.org/page/collections?subset=agriculture.
Fatokun, C. A., Menancio-Hautea, D. I., Danesh, D., and Young, N. D. (1993). Evidence of
orthologous seed weight genes in cowpea and mung bean based on RFLP mapping. Genetics 132,
841 –846.
Flavell, R. B., Bennet, M. D., Smith, J. B., and Smith, D. B. (1974). Genome size and the proportion of
repeated nucleotide sequence DNA in plants. Biochem. Genet. 12, 257–269.
Flohr, M-L., Williams, J. H., and Lenz, F. (1990). The effect of photoperiod on the reproductive
development of a photoperiod sensitive groundnut (Arachis hypogaea L.) CV. NC Ac 17090.
Expl. Agric. 26, 397– 406.
Freeman, H. A., Nigam, S. N., Kelley, T. G., and Ntare, B. R. Subrahmanyam, Boughton, D. (1999).
The world groundnut economy: facts, trends, and outlook. ICRISAT, Patancheru, India, pp 52.
Freyre, R., Skroch, P. W., Geffroy, V., Adam-Blondon, A.-F., Shirmohamadali, A., Johnson, W. C.,
Llaca, V., Nodari, R. O., Pereira, P. A., Tsai, S.-M., Tohme, J., Dron, M., Nienhuis, J., Vallejos,
C. E., and Gepts, P. (1998). Towards an integrated linkage map of common bean. 4. Development
of a core linkage map and alignment of RFLP maps. Theor. Appl. Genet. 97, 847–856.
Frisch, M., Bohn, M., and Melchinger, A. E. (1999a). Minimum sample size and optimal positioning
of flanking markers in marker-assisted backcrossing for transfer of a target gene. Crop Sci. 39,
967 –975.
Frisch, M., Bohn, M., and Melchinger, A. E. (1999b). Comparison of selection strategies for marker-
assisted backcrossing of a gene. Crop Sci. 39, 1295– 1301.
Frisch, M., and Melchinger, A. E. (2001a). Marker-assisted backcrossing for simultaneous
introgression of two genes. Crop Sci. 41, 1716–1725.
Frisch, M., and Melchinger, A. E. (2001b). Marker-assisted backcrossing for introgression of a
recessive gene. Crop Sci. 41, 1485–1994.
Gale, M. D., and Davos, K. M. (1998). Comparative genetics in the grasses. Proc. Natl. Acad. Sci. USA
95, 1971–1974.
206 S.L. DWIVEDI ET AL.

Gao, L., Jiang, Z., Long, B., Feng, F., and Deng, F. (1996). Breeding an early maturing and high-
yielding groundnut variety– Nonghua 22 in China. Int. Arachis Newsletter 16, 8–9.
Gardner, M. E. T. B., and Stalker, H. T. (1983). Cytology and leaf spot resistance of section Arachis
amphiploids and their hybrids with Arachis hypogaea. Crop Sci. 23, 1069–1072.
Gardner, H. W., Grove, M. J., and Keller, N. P. (1998). Soybean lipoxygenase is active on nonaqueous
media at low moisture: a constraint to xerophilic fungi and aflatoxins? J. Amer. Oil Chem. Soc.
75, 1801–1808.
Garet, B. (1976). Heterosis et aptitudes a la combinaison chezI’arachide (A. hypogaea L.). Oleagineux
31, 435 –442.
Gedil, M. A., Wye, C., Berry, S., Seger, B., Peleman, J., Jones, R., Leon, A., Slabaugh, M. B., and
Knapp, S. J. (2001). An integrated restriction fragment length polymorphism–amplified fragment
length polymorphism linkage map for cultivated sunflower. Genome 44, 213–221.
Gibori, A., Hillel, J., Cahner, A., and Ashri, A. (1978). A 9 £ 9 diallel analysis in peanuts (A. hypogaea
L.): Flowering time, top weight, pod yield per plant and pod weight. Theor. Appl. Genet. 53,
169–179.
Gimelfarb, A., and Lande, R. (1994). Simulation of marker assisted selection for non-additive traits.
Genetical Res. 63, 39– 47.
Godoy, I. J., and Giandana, E. H. (1992). Groundnut production and research in South America.
Groundnut-A Global Perspective. (S. N. Nigam, Ed.), pp. 77 –85. ICRISAT, Patancheru,
India.
Gorbet, D. W., and Knauft, D. A. (1995). Registration of Andru 93 peanut. Crop Sci. 35, 1507.
Gorbet, D. W., and Knauft, D. A. (1997). Registration of SunOleic 95R peanut. Crop Sci. 37,
1392.
Gorbet, D. W., and Knauft, D. A. (2000). Registration of SunOleic 97R peanut. Crop Sci. 40,
1190–1191.
Gorbet, D. W., Norden, A. J., Shokes, F. M., and Knauft, D. A. (1987). Registration of Southern
Runner peanut. Crop Sci. 27, 817.
Goodrich-Tanrikulu, M., Mahoney, N., and Rodriguez, S. B. (1995). The plant growth regulator
methyl jasmonate inhibits aflatoxin production by Aspergillus flavus. Microbiology 141,
2831–2837.
Grant, D., Cregan, P., and Shoemaker, R. C. (2000). Genome organization in dicots: Genome
duplication in Arabidopsis and synteny between soybean and Arabidopsis. Proc. Nat. Acad. Sci.
USA 97, 4168–4173.
Gregory, M. P., and Gregory, W. C. (1979). Exotic germplasm of Arachis L. interspecific hybrids.
J. Hered. 70, 185–193.
Green, C. C., and Wynne, J. C. (1987). Genetic variability and heritability for resistance to early leaf
spot in four crosses of Virginia-type peanut. Crop Sci. 27, 18–21.
Greenberg, D. C., Williams, J. H., and Ndunguru, J. (1992). Differences in yield determining
processes of groundnut (Arachis hypogaea L.) genotypes in varied drought environments. Ann.
Appl. Biol. 120, 557–566.
Grieshammer, U., and Wynne, J. C. (1990). Isozyme variability in mature seeds of U. S. peanut
cultivars and collections. Peanut Sci. 18, 72 –75.
Grosso, N. R., Nepote, V., and Guzman, C. A. (2000). Chemical composition of some wild peanut
species (Arachis L.) seeds. J. Agric. Food Chem. 48, 806–809.
Guiying, T., Yuying, W., Boshou, L., Dong, L., and Xinming, X. (1995). Release of cultivar Zhonghua
117 in China. Int. Arachis Newsletter 15, 11 –12.
Guok, H. P., Wynne, J. C., and Stalker, H. T. (1986). Recurrent selection within a population from an
interspecific peanut cross. Crop Sci. 26, 249–253.
Hadjichristodoulou, A., Dwivedi, S. L., Wynne, J. C., Nigam, S. N., Alexandrou, G., Theodorides,
Chr., and Mouzouris, M. (1997). Registration of ICGV 88438, ICGV 89214, and ICGV 91098
peanut germplasm. Crop Sci. 37, 1985.
 MOLECULAR BREEDING OF GROUNDNUT 207

Halward, T. M., Stalker, H. T., LaRue, E., and Kochert, G. (1991a). Genetic variation detectable with
molecular markers among unadapted germplasm resources of cultivated peanut and related wild
species. Genome 34, 1013–1020.
Halward, T. M., Wynne, J. C., and Stalker, H. T. (1991b). Recurrent selection progress in a population
derived from an interspecific peanut cross. Euphytica 52, 9–84.
Halward, T., Stalker, T., LaRue, E., and Kochert, G. (1992). Use of single-primer DNA amplifications
in genetic studies of peanut (Arachis hypogaea L.). Plant Mol. Biol. 18, 315–325.
Halward, T., Stalker, H. T., and Kochert, G. (1993). Development of an RFLP linkage map in diploid
peanut species. Theor. Appl. Genet. 87, 379– 384.
Hamid, M. A., Isleib, T. G., Wynne, J. C., and Green, C. C. (1981). Combining ability analysis of
Cercospora leaf spot resistance and agronomic traits in Arachis hypogaea L. Oleagineux 36,
605 –609.
Harch, B. D., Basford, K. E., Delacy, I. H., Lawrence, P. K., and Cruickshank, A. (1995). Patterns of
diversity in fatty acid composition in the Australian groundnut germplasm collection. Genet.
Resour. and Crop Evol. 42, 243– 256.
Harushima, Y., Yano, M., Shomura, A., Sato, M., Shimano, T., Kuboki, Y., Yamamoto, T., Lin, S. Y.,
Antonio, B. A., Parco, A., Kajiya, H., Huang, N., Yamamoto, K., Nagamura, Y., Kurata, N.,
Khush, G. S., and Sasaki, T. (1998). A high-density rice genetic linkage map with 2275 markers
using a single F2 population. Genetics 148, 479– 494.
Hayes, T. R. (1932). Groundnut rosette disease in Gambia. Tropical Agriculture, Trinidad 9, 211–217.
Hayward, A. C. (1964). Characteristics of Pseudomonas solanacearum. J. Appl. Bacteriology 27,
265 –277.
Hayward, A.C., (1990). Diagnosis, distribution, and status of groundnut bacterial wilt. In ACIAR
Proceedings No. 31 (K. J. Middleton, A. C. Hayward, Eds.), pp. 12–17. Genting Highlands,
Malaysia.
Hayward, A. C. (1991). Biology and epidemiology of bacterial wilt caused by Pseudomonas
solanacearum. Ann. Rev. Phytopathology 29, 65–87.
He, G., and Prakash, C. S. (1997). Identification of polymorphic DNA markers in cultivated peanut
(Arachis hypogaea L.). Euphytica 97, 143–149.
He, L. Y., Sequeira, L., and Kelman, A. (1983). Characteristics of strains of Psuedomonas
solanacearum from China. Plant Disease 67, 1357– 1361.
Herendeen, P. S., Crepet, W. L., and Dilcher, D. L. (1992). The fossile history of the leguminosae:
phylogenetic and biogeographic implications. In “Adv. Legume Systematics”. (P. S. Herendeen
and D. L. Dilcher, Eds.), pp. 303–316. Royal Botanic Gardens, Kew, England.
Holbrook, C. C., and Noe, J. P. (1990). Resistance to Meloidogyne arenaria in Arachis spp. and the
implications on development of resistant peanut cultivars. Peanut Sci. 17, 35–38.
Holbrook, C. C., and Noe, J. P. (1992). Resistance to the peanut root-knot nematode (Meloidogyne
arenaria) in Arachis hypogaea. Peanut Sci. 19, 35 –37.
Holbrook, C. C., and Stalker, H. T. (2003). Peanut breeding and genetic resources. Plant Breeding
Reviews 22, 297 –356, print in.
Holbrook, C. C., Noe, J. P., Gorbet, D. W., Stephenson, M. G., and Anderson, W. F. (1996).
Identification and evaluation of additional sources of resistance to peanut root-knot nematode in
Arachis hypogaea L. Peanut Sci. 23, 91–94.
Holbrook, C. C., Noe, J. P., Gorbet, D. W., and Stephenson, M. G. (1998). Evaluation of
peanut breeding lines with resistance to the peanut root-knot nematode. Crop Sci. 38, 260–262.
Holbrook, C. C., Kvien, C. K., Rucker, K. S., Wilson, D. M., Hook, J. E., and Matheron, M. E. (2000).
Preharvest contamination in drought-tolerant and drought-intolerant peanut genotypes. Peanut
Sci. 27, 45 –48.
Hopkins, M. S., Casa, A. M., Wang, T., Michell, S. E., Dean, R. E., Kochert, G. D., and Kresovich, S.
(1999). Discovery and characterization of polymorphic simple sequence repeats (SSRs) in
peanut. Crop Sci. 39, 1243–1247.
208 S.L. DWIVEDI ET AL.

Hospital, F., and Charcosset, A. (1997). Marker-assisted introgression of quantitative trait loci.
Genetics 147, 1469–1485.
Hospital, F., Moreau, L., Lacoudre, F., Charcosset, A., and Gallais, A. (1997). More on the efficiency
of marker-assisted selection. Theor. Appl. Genet. 95, 1181–1189.
Hull, R., and Adams, A. N. (1968). Groundnut rosette and its assister virus. Ann. Appl. Biol. 62,
139–145.
Humphry, M. E., Konduri, V., Lambrides, C. J., Magner, T., McIntyre, C. L., Aitken, E. A. B., and
Liu, C. J. (2002). Development of a mungbean (Vigna radiata) RFLP linkage map and its
comparison with lablab (Lablab purpureus) reveals a high level of colinearity between the two
genomes. Theor. Appl. Genet. 105, 160–166.
ICRISAT. (1994) Medium Term Plan. 1994–1998, volume 1, Patancheru, India.
ICRISAT. (2001) P5 (Biotechnology) Project Annual Report. pp. 43 Patancheru, India.
Isleib, T. G., and Wynne, J. C. (1983). Heterosis in test crosses of 27 exotic peanut cultivars. Crop Sci.
23, 832 –841.
Isleib, T. G., and Wynne, J. C. (1992). Use of plant introductions in peanut improvement. CSSA
Special Publication # 20. In “Uses of Plant Introductions in Cultivar Development Part 2”. (H. L.
Sands, L. E. Sands, P. S. Baenziger and S. H. Mickelson, Eds.), pp. 75–116, Medison, USA.
Isleib, T. G., Wynne, J. C., and Rawlings, J. O. (1978). Estimates of epistasis for diverse peanut
cultivars. Peanut Sci. 5, 106–108.
Isleib, T. G., Wynne, J. C., and Nigam, S. N. (1994). Groundnut breeding. In “The Groundnut Crop: A
Scientific Basis for Improvement”. (J. Smart, Ed.), pp. 552–623. Chapman and Hall, London.
Isleib, T. G., Rice, P. W., Mozingo, R. W., Mozingo II, R. W., and Pattee, H. E. (1999). Registration of
Gregory peanut. Crop Sci. 39, 1526.
Jain, A. K., Basha, S. M., Holbrook, C. C., (2001). Identification of drought-responsive transcripts in
peanut (Arachis hypogaea L.). Electronic J. Biotechnology 4. http://www.ejb.org/content/vol14/
issue2/full/2/index.html.
Jambunathan, R., Singh, A. K., Gurtu, S., and Raghunath, K. (1993). Amino acid composition, fatty
acid composition, and levels of protease inhibitors in seeds of wild Arachis species. Oleagineux
48, 415 –419.
James, S. L. H., and Young, C. T. (1983). Comparison of fatty acid content of imported peanut.
J. Amer. Oil Chem. Soc. 60, 945 –947.
Jayalakshmi, V., Rajareddy, C., Reddy, P. V., and Nageswara Rao, R. C. (1999). Genetic analysis of
carbon isotope discrimination and specific leaf area in groundnut (A. hypogaea L.). J. Oilseeds
Res. 16, 1 –5.
Jogloy, S., Wynne, J. C., and Beute, M. K. (1987). Inheritance of late leaf spot resistance and
agronomic traits in peanut. Peanut Sci. 14, 86–90.
Jung, S., Swift, D., Sengoku, E., Patel, M., Teule, F., Powell, G., Moore, K., and Abbott, A. (2000a).
The high oleate trait in the cultivated peanut (Arachis hypogaea L.). I. Isolation and
characterization of two genes encoding microsomal oleoyl-PC desaturases. Mole. Gen. Genet.
263, 796 –805.
Jung, S., Powell, G., Moore, K., and Abbott, A. (2000b). The high oleate trait in the cultivated peanut
(Arachis hypogaea L.).II. Molecular basis and genetics of the trait. Mol. Gen. Genet. 263,
806–811.
Kaga, A., Ishii, T., Tsukimoto, K., Tokoro, E., and Kamijima, O. (2000). Comparative molecular
mapping in Ceratotropis species using an interspecific cross between azuki bean (Vigna
angularis) and rice bean (V. umbellata). Theor. Appl. Genet. 100, 207 –213.
Kalekar, A. R., Patil, B. C., and Deokar, A. B. (1984). Inheritance of resistance to rust in groundnut.
Madras Agric. J. 71, 125 –126.
Kanazin, V., Talbert, H., See, D., DeCamp, P., Nevo, E., and Blake, T. (2002). Discovery and
assay of single-nucleotide polymorphisms in barley (Hordeum vulgare). Plant Mol. Biol. 48,
529–537.
 MOLECULAR BREEDING OF GROUNDNUT 209

Keim, P., Schupp, J. M., Travis, S. E., Clayton, K., Zhu, T., Shi, L., Ferreira, A., and Webb, D. M.
(1997). A high-density soybean genetic map based on AFLP markers. Crop Sci. 37, 537–543.
Kerridge, P.C., Hardy, B., (eds.). 1994. Biology and agronomy of forage Arachis. pp. 206. CIAT, Cali,
Colombia. CIAT Publication No. 240.
Ketring, D. L. (1984). Root diversity among peanut genotypes. Crop Sci. 24, 229–232.
Kirby, J. S., Melouk, H. A., Stevens, Jr., T.E, Banks, D. J., Sholar, J. R., Damicone, J. P., and Jackson,
K. E. (1998). Registration of Southwest Runner peanut. Crop Sci. 38, 545–546.
Knauft, D.A. (1987). Inheritance of rust resistance in groundnut. In “Groundnut Rust Disease: Proc.
Discussion Group Meeting” (D. McDonald, P. Subrahmanyam, J. A. Wightman, J. V. Mertin,
eds.), pp.183–187. ICRISAT, Patancheru, India.
Knauft, D. A., and Wynne, J. C. (1995). Peanut Breeding and Genetics. Adv. Agronomy 55, 393–445.
Kochert, G., Halward, T., Branch, W. D., and Simpson, C. E. (1991). RFLP variability in peanut
(Arachis hypogaea L.) cultivars and wild species. Theor. Appl. Genet. 81, 565 –570.
Kornegay, J. L., Beute, M. K., and Wynne, J. C. (1980). Inheritance of resistance to Cercospora
arachidicola and Cercosporodium personatum in six Virginia-type peanut lines. Peanut Sci. 7,
4 –9.
Kota, R., Wolf, M., Michalek, W., and Graner, A. (2001a). Application of denaturing high-
performance liquid chromatography for mapping of single nucleotide polymrphisms in barley
(Hordeum vulgare L.). Genome 44, 523 –528.
Kota, R., Varshney, R. K., Thiel, T., Dehmer, K. J., and Graner, A. (2001b). Generation and
comparison of EST-derived SSRs and SNPs in barley (Hordeum vulgare L.). Hereditas 135,
145 –151.
Kowalski, S. P., Lan, T-H., Fieldmann, K. A., and Paterson, A. H. (1994). Comparative mapping of
Arabidopsis thaliana and Brassica oleracea chromosomes reveals islands of conserved
organization. Genetics 138, 499 –510.
Krapovikas, A., and Gregory, W. C. (1994). Taxonomy of genus Arachis (Legumonisae). Bonplandia
8, 1–186.
Kruglyak, L. (1997). The use of a genetic map of biallelic markers in linkage studies. Nature Genet.
17, 21–24.
Ku, Hsin-Mei, Vision, T., Liu, J., and Tanksley, S. D. (2000). Comparing sequenced segments of the
tomato and Arabidopsis genomes: Large-scale duplication followed by selective gene loss creates
a network of synteny. Proc. Nat. Acad. Sci., USA 97, 9121–9126.
Ku, Hsin-Mei, Liu, J., Doganlar, S., and Tanksley, S. D (2001). Exploitation of Arabidopsis-tomato
synteny to construct a high-resolution map of the ovate-containing region in tomato chromosome
2. Genome 44, 470–475.
Kwok, P. Y., and Gu, Z. (1999). SNP libraries: why and how are we building them? Molecular
Medicine Today 12, 538–543.
Kwok, P. Y., Deng, Q., Zakeri, H., Taylor, S. L., and Nickerson, D. A. (1996). Increasing the
information content of STS-based genome maps: identifying polymorphisms in the mapped
STSs. Genomics 31, 123–126.
Lande, R., and Thompson, R. (1990). Efficiency of marker-assisted selection in the improvement of
quantitative traits. Genetics 124, 743 –756.
Lander, E. S., and Botstein, D. (1989). Mapping Mendelian factors underlying quantitative traits using
RFLP linkage maps. Genetics 121, 185 –199.
Lanham, P. G., Fennell, S., Moss, J. P., and Powell, W. (1992). Detection of polymorphic loci in
Arachis germplasm using random amplified polymorphic DNAs. Genome 35, 885 –889.
Larkin, P. D., and Park, W. D. (1999). Transcript accumulation and utilization of alternate and non-
consesus splice sites in rice granule-bound starch synthase are temperature-sensitive and
controlled by a single nucleotide polymorphism. Plant Mol. Biol. 40, 719– 727.
Layrisse, A., Wynne, J. C., and Isleib, T. G. (1980). Combining ability for yield, protein and oil of
peanut lines from South American Centers of diversity. Euphytica 29, 561–570.
210 S.L. DWIVEDI ET AL.

Lee, S. H., Bailey, M. A., Mian, M. A. R., Shipe, E. R., Ashley, D. A., Parrott, W. A., Hussey, R. S.,
and Boerma, H. R. (1996a). Identification of quantitative trait loci for plant height, lodging,
and maturity in a soybean population segregating for growth habit. Theor. Appl. Genet. 92,
516–523.
Lee, S. H., Bailey, M. A., Mian, M. A. R., Carter, Jr. T-E., Ashley, D.A., Hussey, R.S., Parrott, W.A.,
and Boerma, H.R., (1996b) Molecular markers associated with soybean plant height, lodging, and
maturity across locations. Crop Sci. 36, 728 –735.
Lee, J. M., Grant, D., Vallejos, C. E., and Shoemaker, R. S. (2001). Genome organization in dicots: II.
Arabidopsis as a ‘bridging species’ to resolve genome evolution events among legumes. Theor.
Appl. Genet. 103, 765 –773.
Liang, X. (1998). Present status of groundnut bacterial wilt in South China. In Groundnut Bacterial
Wilt: Proc. Fourth Working Group Meeting (S. Pande, L. Boshou, N. X. Hong, C. Johnson, and
C. L. L. Gowda, Eds.), pp. 23–26. VASI Hanoi, Vietnam, and ICRISAT, Patancheru, India.
Liang, X., Shaoxing, L., Yicong, L., and Guiyuan, Z. (1999a). Release of cultivars Yue You 5 and Yue
You 79 in Guangdong, China. Int. Arachis Newsletter 19, 15.
Liang, X., Yicong, L., Shaoxiong, L., and Guiyuan, Z. (1999b). Yue You 223: A high-yielding
Chinese Cultivar with good resistance to rust. Int. Arachis Newsletter 19, 16.
Liao, B. S., Li, W., and Sun, D. (1986). Inheritance of resistance to Pseudomonas solanacearum E. F.
Smith in Arachis hypogaea L. Oil Crops of China 3, 1–8.
Liao, B. S., Wang, L. Y., Xia, X., Tang, G., Tan, Y., and Sun, D. (1990). Genetics and breeding aspects
of resistance to bacterial wilt in groundnut. In “Proc. ACIAR/ICRISAT Collaborative Research
Planning Meeting” (K. J. Middleton and A. C. Hayward, Eds.), pp. 39–43. Genting Highlands,
Malayasia.
Liao, B. S., Jiang, W. R., Tan, Y. J., Rao, R. H., and Tang, G. Y. (1992). A study on characteristics
of nitrogen fixation in some groundnut lines resistant to bacterial wilt. Oil Crops of China 4,
34–37.
Liao, B., Naixiong, D., Yujun, T., and Yong, W. (1998). Status of groundnut bacterial wilt in Central
China. In “Groundnut Bacterial Wilt: Proc. Fourth Working Group Meeting” (S. Pande,
L. Boshou, N. X.) India.
Lopez, Y., Nadef, H. L., Smith, O. D., Connell, J. P., Reddy, A. S., and Fritz, A. K. (2000). Isolation
and characterization of the D12-fatty acid desaturase in peanut (Arachis hypogaea L.) and search
for polymorphisms for the high oleat trait in Spanish market-type lines. Theor. Appl. Genet. 10,
1131–1138.
Lynch, R. E. (1990). Resistance in peanut to major arthopod pests. Florida Entomologist 73, 422 –445.
Lynch, R. E., and Mack, T. P. (1995). Biological and biotechnical advances for insect management in
peanut. In “Advances in Peanut Science” (H. E. Pattee and H. T Pattee, Eds.), pp. 95– 159.
Stillwater, OK, USA.
Machmud, M., and Rais, S.A. (1994). Status of groundnut bacterial wilt research in Indonesia. In
“Groundnut Bacterial Wilt in Asia: Proc. Third Working Group Meeting” (V. K. Mehan and
D. McDonald, Eds.), pp. 115–119. OCRI, China, and ICRISAT Patancheru, India.
Makne, V. G., and Bhale, N. L. (1987). Combining ability analysis for yield, protein and oil in
groundnut. Euphytica 40, 169 –172.
Mallikarjuna, N. (2002). Gene introgression from Arachis glabrata into A. hypogaea, A. duranensis,
and A. digoi. Euphytica 124, 99 –104.
Mallikarjuna, N., and Bramel, P. (2001). Crossability in genus Arachis L. Amer. Peanut Res. Educ.
Soc. 33, 57.
Mauboussin, J. C., Laurent, P., and Delafond, G. (1970). Les varieties d arachides recommandees au
Senegal et leur emploi. Cah. Agric. Prat. Pays Chauds 2, 63–89.
Maughan, P. J., Saghai Maroof. M.A., and Buss, G.R. (1996). Molecular-marker analysis of seed-
weight: genomic location, gene action, and evidence for orthologous evolution among three
legume species. Theor. Appl. Genet. 93, 574–579.
 MOLECULAR BREEDING OF GROUNDNUT 211

Mayer, K., Murphy, G., Tarchini, R., Wambutt, R., Volckaert, G., Pohl, T., Dusterhoft, A., Stiekema,
W., Entian, K-D., Terryn, N., Lemcke, K., Haase, D., Hall, C. R., Dodeweerd, A-M van, Tingey, S.
V., Mewes, H-W., Bevan, M.W., and Bancroft, I. (2001). Conservation of microstructure between
a sequenced region of the genome of rice and multiple segments of the genome of Arabidopsis
thaliana. Genome Res. 10.1101/gr.161701 (http://www.genome.org/cgi/doi/10.1101/gr.161701).
McCouch, S. (1998). Toward a plant genomics initiative: Thoughts on the value of cross-species and
cross-genera comparisons in the grasses. Proc. Natl. Acad. Sci., USA 95, 1983–1985.
Mehan, V. K. (1989). Screening groundnut for resistance to seed invasion by Aspergillus flavus
and aflatoxin production. In “Aflatoxin Contamination of Groundnut: Proc. Int. Workshop”
(D. McDonald and V. K. Mehan, Eds.), pp. 323–334. ICRISAT, Patancheru, India.
Mehan, V. K., McDonald, D., and Ramakrishna, N. (1986). Varietal resistance in peanut to aflatoxin
production. Peanut Sci. 13, 7– 10.
Mehan, V. K., McDonald, D., Haravu, L. J., and Jayanthi, S. (1991). The groundnut aflatoxin problem:
review and literature data base. ICRISAT, Patancheru, India. pp 369.
Mehan, V. K., McDonald, D., Singh, A. K., and Moss, J. P. (1992). Aflatoxin production in seeds of
wild Arachis species. Oleagineux 47, 87–89.
Mehan, V. K., Reddy, P. M., Rao, K. V., and McDonald, D. (1994). Components of rust resistance in
peanut genotypes. Phytopathology 84, 1421–1426.
Meksem, K., Ruben, E., Hyten, D., Triwitayakorn, K., and Lightfoot, D. A. (2001). Conversion of
AFLP bands into high-throughput DNA markers. Mole. Genet. Genom. 265, 207 –214.
Melotto, M., and Kelly, J. D. (2001). Fine mapping of the C0-4 locus of common bean reveals a
resistance gene candidate, COK-4, that encodes for a protein kinase. Theor. Appl. Genet. 103,
508 –517.
Menancio-Hautea, D., Fatokun, C. A., Kumar, L., Danesh, D., and Young, N. D. (1993). Comparative
genome analysis of mungbean (Vigna radiata L. Wilczek) and cowpea (V. unguiculata
L. Walpers) using RFLP mapping data. Theor. Appl. Genet. 86, 797–810.
Menendez, C. M., Hall, A. E., and Gepts, P. (1997). A genetic linkage map of cowpea (Vigna
unguiculata) developed from a cross between two inbred, domesticated lines. Theor. Appl. Genet.
95, 1210–1217.
Mercer, L. C., Wynne, J. C., and Young, C. T. (1990). Inheritance of fatty acid content in peanut oil.
Peanut Sci. 17, 17–21.
Michelmore, R. W., Paran, I., and Kesseli, R. V. (1991). Identification of markers linked to disease-
resistance genes by bulked segregant analysis: A rapid method to detect markers in specific
genomic regions by using segregating populations. Proc. Nat. Acad. Sci. USA 88, 9828–9832.
Minja, E. M., van der Merwe, P. J. A., Kimmins, F. M., and Subrahmanyam, P. (1999). Screening of
groundnut lines for resistance to Aphids. Aphis craccivora Koch. Int. Arachis Newsletter 19,
21 –23.
Mixon, A. C. (1976). Peanut breeding strategy to minimize aflatoxin contamination. J. Amer. Peanut
Res. Educ. Assoc. 8, 54–58.
Mixon, A. C. (1986). Reducing Aspergillus species infection of peanut seed using resistant genotypes.
J. Environ. Quality 15, 101–103.
Moss, J. P. (1985). Breeding strategies for utilization of wild species of Arachis in groundnut
improvement. In “Proc. Int. Workshop on Cytogenetics of Arachis” (J. P. Moss and S. D. Feakin,
Eds.), pp. 93–99. ICRISAT, Patancheru, India.
Moss, J. P., Singh, A. K., Reddy, L. J., Nigam, S. N., Subrahmanyam, P., McDonald, D., and Reddy,
A. G. S. (1997). Registration of ICGV 87165 peanut germplasm with multiple resistance. Crop
Sci. 37, 1028.
Moss, J. P., Singh, A. K., Nigam, S. N., Hildebrand, G. L., Govinden, N., Ismael, F. M.,
Subrahmanyam, P., and Reddy, L. J. (1998). Registration of ICGV-SM 86715. Crop Sci. 38, 572.
Mozingo, R. W., Coffelt, T. A., and Wynne, J. C. (1987). Genetic improvement in large-seeded
Virginia-type peanut cultivars since 1944. Crop Sci. 27, 228 –231.
212 S.L. DWIVEDI ET AL.

Mozingo, R. W., Wynne, J. C., Porter, D. M., Coffelt, T. A., and Isleib, T. G. (1994). Registration of
VA-C 92R. Crop Sci. 34, 539 –540.
Mozingo, R. W., Coffelt, T. A., and Isleib, T. G. (2000). Registration of VA 98R peanut. Crop Sci. 40,
1202–1203.
Moore, K. M., and Knauft, D. A. (1989). Inheritance of high oleic acid in peanut. J. Heredity 80,
252–253.
Muehlbauer, G. J., Specht, J. E., Thomas-Compton, M. A., Staswick, P. E., and Bernard, R. L. (1988).
Near-isogenic lines–A potential resource in the integration of conventional and molecular
marker linkage maps. Crop Sci. 28, 729–735.
Murant, A., and Kumar, I. (1990). Different variants of satellite RNA of groundnut rosette virus are
responsible for the chlorotic and green forms of groundnut rosette disease. Ann. Appl. Biol. 117,
85–92.
Murant, A., Rajeshwari, R., Robinson, D., and Raschke, J. (1988). A satellite RNA of groundnut
rosette virus that is largely responsible for symptoms of groundnut rosette disease. J. General
Biol. 69, 1479–1486.
Murthy, T. G. K., and Reddy, P. S. (1993). Cytogenetics and genetics of groundnuts. Intercept Ltd,
Andover, England.
Nageswara Rao, R. C., and Nigam, S. N. (2003). Genetic options for drought management in
groundnut. In “Management of Agricultural Drought: Agronomic and Genetic Options” (N. P.
Saxena, Ed.), pp. 123–141. Science Publisher Inc., New Hampshire, USA.
Nageswara Rao, R. C., and Wright, G. C. (1994). Stability of relationship between specific leaf area
and carbon isotope discrimination across environments in peanut. Crop Sci. 34, 98 –103.
Nageswara Rao, R. C., Williams, J. H., and Singh, M. (1989). Genotypic sensitivity to drought and
yield potential of peanut. Agron. J. 81, 887– 893.
Nageswara Rao, R. C, Reddy, L. J., and Nigam, S. N. (1992). Influence of soil type on the adaptation
of groundnut genotypes. In “Proc. of the Fifth Regional Groundnut Workshop for Southern
Africa” (R. C. Nageswara Rao and P. Subrahmanyam, Eds.), pp. 15–18. ICRISAT, Patancheru,
India.
Nageswara Rao, R. C., Williams, J. H., Wadia, K. D. R., Hubick, K. T., and Farquhar, G. D. (1993).
Crop growth, water-use efficiency, and carbon isotope discrimination in groundnut
(Arachis hypogaea L.) genotypes under end-of-season drought conditions. Ann. Appl. Biol.
122, 357 –367.
Naidu, R., Robinson, D., and Kimmins, F. (1998). Detection of each of the causal agents of groundnut
rosette disease in plants and vector aphids by RT-PCR. J. Virol. Methods 76, 9–18.
Narvel, J. M., Fehr, W. R., Chu, W-C., Grant, D., and Shoemaker, R. C. (2000). Simple sequence
repeat diversity among soybean plant introductions and elite genotypes. Crop Sci. 40,
1452–1458.
Nelson, S. C., Simpson, C. E., and Starr, J. L. (1989). Resistance to Meloidogyne arenaria in Arachis
spp. Germplasm. Supplement to the J. Nematology 21, 654–660.
Nevill, D. J. (1981). Components of resistance to Cercospora arachidicola and Cercosporodium
personatum in groundnut. Ann. Appl. Biol. 99, 77 –86.
Nevill, D. J. (1982). Inheritance of resistance to Cercosporidium personatum in groundnuts: a genetic
model and its implications for selection. Oleagineux 37, 355–362.
Nigam, S. N. (2000). Some strategic issues in breeding for high and stable yield in groundnut in India.
J. Oilseeds Res. 17, 1–10.
Nigam, S. N., and Bock, K. R. (1990). Inheritance of resistance to groundnut rosette virus in groundnut
(Arachis hypogaea L.). Ann. Appl. Biol. 117, 553– 560.
Nigam, S. N., Dwivedi, S. L., Rao, Y. L. C., and Gibbons, R. W. (1990a). Registration of ICGV 87128.
Crop Sci. 30, 959.
Nigam, S. N., Dwivedi, S. L., Rao, Y. L. C., and Gibbons, R. W. (1990b). Registration of ICGS 11.
Crop Sci. 30, 960.
 MOLECULAR BREEDING OF GROUNDNUT 213

Nigam, S. N., Dwivedi, S. L., Rao, Y. L. C., and Gibbons, R. W. (1991a). Registration of ICGV 87141
peanut. Crop Sci. 31, 1096.
Nigam, S. N., Dwivedi, S. L., Rao, Y. L. C., and Gibbons, R. W. (1991b). Registration of ICGS 1
peanut cultivar. Crop Sci. 31, 1382–1383.
Nigam, S. N., Dwivedi, S. L., Rao, Y. L. C., and Gibbons, R. W. (1992a). Registration of ICGV 87187
peanut. Crop Sci. 32, 278–279.
Nigam, S. N., Reddy, L. J., Subrahmanyam, P., Reddy, A. G. S., McDonald, D., and Gibbons, R. W.
(1992b). Registration of ICGV 87157, an elite peanut germplasm with multiple resistance to
diseases. Crop Sci. 32, 837.
Nigam, S. N., Dwivedi, S. L., Rao, Y. L. C., and Gibbons, R. W. (1992c). Registration of ICGV 87121
peanut germplasm. Crop Sci. 32, 1078–1079.
Nigam, S. N., Marfo, K. O., Assibi, M. A., Dwivedi, S. L., Rao, Y. L. C., and Gibbons, R. W. (1993).
Registration of Sinkarzei peanut. Crop Sci. 33, 212.
Nigam, S. N., Dwivedi, S. L., and Rao, Y. L. C. (1994). Registration of ICGS 35 peanut germplasm.
Crop Sci. 34, 1423–1424.
Nigam, S. N., Nageswara Rao, R.C., Wynne, J.C., Williams, J.H., Fitzner, M., and Nagabhushanam,
G.V.S. (1994b). Effect and interaction of temperature and photoperiod on growth and partitioning
in three groundnut (A. hypogaea L.) genotypes. Ann. Appl. Biol. 125, 541–552.
Nigam, S. N., Hildebrand, G. L., and Bock, K. R. (1995). Registration of ICGV-SM 83708 peanut
germplasm. Crop Sci. 35, 1718.
Nigam, S. N., Rao, M. J. V., Upadhyaya, H. D., Rao, Y. L. C., and Reddy, N. S. (1995). Registration of
an early maturing peanut germplasm ICGV 86015. Crop Sci. 35, 1718–1719.
Nigam, S. N., Dwivedi, S. L., Ramraj, V. M., and Chandra, S. (1997). Combining ability of response to
photoperiod in peanut. Crop Sci. 37, 1159–1162.
Nigam, S. N., Hildebrand, G. L., Syamasonta, B., Bock, K. R., Subrahmanyam, P., and Reddy, L. J.
(1998a). Registration of ICGV-SM 83005 peanut germplasm. Crop Sci. 38, 571.
Nigam, S. N., Hildebrand, G. L., Bock, K. R., Ismael, F. M., Govinden, N., Subrahmanyam, P., and
Reddy, L. J. (1998b). Registration of ICGV-SM 85048 peanut germplasm. Crop Sci. 38,
572 –573.
Nigam, S. N., Nageswara Rao, R.C., and Wynne, J.C. (1998c). Effects of temperature and photoperiod
on vegetative and reproductive growth of groundnut (Arachis hypogaea L.). J. Agron. Crop Sci.
181, 117–124.
Nigam, S. N., Upadhyaya, H. D., Chandra, S., Nageswara Rao, R.C., Wright, G.C., and Reddy, A.G.S.
(2001). Gene effects for specific leaf area and harvest index in three crosses of groundnut (Arachis
hypogaea). Ann. Appl. Biol. 139, 301–306.
Norden, A. J., Gorbet, D. W., Knauft, D. A., and Young, C. T. (1987). Variability in oil quality among
peanut genotypes in the Florida breeding program. Peanut Sci. 14, 7–11.
Ntare, B. R., and Williams, J. H. (1998). Heritability of components of a simple physiological model
for yield in groundnut under semi arid rainfed conditions. Field Crops Res. 58, 25–33.
Oliver, M. J., Wood, A. J., and O’Mahony, P. (1998). To dryness and beyond-preparation for the dried
state and rehydration in vegetative desiccation-tolerant plants. Plant Growth Regulation 24,
193 –201.
Olorunju, P. E., Kuhn, C. W., and Demski, J. W. (1992). Inheritance of resistance in peanut to mixed
infections of groundnut rosette virus (GRV) and groundnut rosette assistor virus and a single
infection of GRV. Plant Disease 76, 95 –100.
Olorunju, P. E., Ntare, B. R., Pande, S., and Reddy, S. V. (2001). Additional sources of resistance to
groundnut rosette disease in groundnut germplasm and breeding lines. Ann. Appl. Biol. 139,
259 –268.
Orf, J. H., Chase, K., Jarvik, T., Mansur, L. M., Cregan, P. B., Adler, F. R., and Lark, K. G. (1999).
Genetics of soybean agronomic traits: I. Comparison of three related recombinant inbred
populations. Crop Sci. 39, 1642– 1651.
214 S.L. DWIVEDI ET AL.

Padagham, D. E., Kimmins, F. M., and Rao, G. V. R. (1990). Resistance in groundnut (Arachis
hypogaea L.) to Aphis craccivora (Koch). Ann. Appl. Biol. 131, 481–489.
Paik-Ro, O. G., Smith, R. L., and Knauft, D. A. (1992). Restriction fragment length polymorphism
evaluation of six peanut species within the Arachis section. Theor. Appl. Genet. 84,
201–208.
Pande, S., and Rao, J. N. (2001). Resistance of wild Arachis species to late leaf spot and rust in
greenhouse trials. Plant Disease 85, 851 –855.
Pande, S., Boshou, L., Hong, N. X., Johnson, C., and Gowda, C. L. L. (eds.) (1998). Groundnut
Bacterial Wilt in Asia. Proc. of the Fourth Working Group Meeting, pp. 87. VASI Hanoi,
Vietnam, and ICRISAT, Patancheru, India (Summaries in English).
Paramasivam, K., Jayasekhar, M., Rajasekharan, R., and Veerabadhiran, P. (1990). Inheritance of rust
resistance in groundnut (A. hypogaea L.). Madras Agric. J. 77, 50–52.
Parker, R. C., Wynne, J. C., and Emery, D. A. (1970). Combining ability estimates in Arachis
hypogaea L. I. Seedling response in controlled environment. Crop Sci. 10, 429 –432.
Paterson, A. H., Lander, E. S., Hewitt, J. D., Peterson, S., Lincoln, S. E., and Tanksley, S. D. (1988).
Resolution of quantitative traits into Mendelian factors, using a complete linkage map of
restriction fragment length polymorphism. Nature 335, 721 –726.
Paterson, A. H., Lan, T-H., Reischmann, K. P., Chang, C., Lin, Y-R., Liu, S-C., Burow, M. D.,
Kowalski, S. P., Katsar, C. S., DelMonte, T. A., Feldmann, K. A., Schertz, K. F., and Wendel, J. F.
(1996). Toward a unified genetic map of higher plants, transcending the monocot-dicot
divergence. Nature Genomics 14, 380– 382.
Pattee, H. E., Isleib, T. G., and Giesbrecht, F. G. (1998). Variation in intensity of sweet and bitter
sensory attributes across peanut genotypes. Peanut Sci. 25, 63–69.
Pattee, H. E., Isleib, T. G., Giesbrecht, F. G., and McFeeters, R. F. (2000). Investigations into
genotypic variations of peanut carbohydrates. J. Agric. Food Chem. 48, 750– 756.
Pettit, R. E., Azaizeh, H. A., Taber, R. A., Szerszen, J. B., and Smith, O. D. (1989). Screening
groundnut cultuivars for resistance to Aspergillus flavus, Aspergillus parasiticus, and
aflatoxin contamination. In “Aflatoxin Contamination of Groundnuts: Proc. Int. Workshop”
(D. McDonald and V. K. Mehan, Eds.), pp. 291–303. ICRISAT, Patancheru, India.
Polzin, K. M., Lohnes, D. G., Nickell, C. D., and Shoemaker, R. C. (1994). Integration of Rps2, Rmd,
and Rj2 into linkage group J of the soybean molecular map. J. Heredity 85, 300– 303.
Prasad, P. V. V., Craufurd, P. Q., and Summerfield, R. J. (1999). Sensitivity of peanut to timing of heat
stress during reproductive development. Crop Sci. 39, 1352–1357.
Quiros, C. F., Grellet, F., Sadowski, J., Suzuki, T., Li, G., and Wroblewski, T. (2001). Arabidopsis and
Brassica comparative genomics: sequence, structure and gene content in the AB11-Rps2-Ck1
chromosomal segment and related regions. Genetics 157, 1321–1330.
Rajgopal, K., Chandran, K., Bhagat, N. R., and Bhalodia, P. K. (1997). Morphological
characterization of Valencia and Virginia bunch peanut (Arachis hypogaea L.) germplasm.
Plant Genet. Resour. Newsletter 109, 27–29.
Rao, G. V. R., and Wightman, J. A. (1999). Status of the integrated management of groundnut pests in
India. In “IPM System in Agriculture 5” (R. K. Upadhyay, K. G. Mukerji and R. L. Rajak, Eds.),
pp. 435 –459. Aditya Books, New Delhi, India.
Rao, M. J. V., Upadhyaya, H. D., Mehan, V. K., Nigam, S. N., and Reddy, N. S. (1995). Registration
of peanut germplasm ICGV 88145 and ICGV 89104 resistant to seed infection by Aspergillus
flavus. Crop Sci. 35, 1717.
Ratna, A. S., Rao, A. S., Reddy, A. S., Nolt, B. L., Reddy, D. V. R., Vijayalakshmi, M., and
McDonald, D. (1991). Studies on transmission of Indian peanut clump virus disease by Polymyxa
graminis. Ann. Appl. Biol. 118, 71 –78.
Raven, P. H., and Polhill, R. M. (1981). Biogeography of the legumes. In “Advances in Legume
Systematics” (R. M. Polhill and P. H. Raven, Eds.), pp. 27–34. Crown, Royal Botanic Gardens,
Kew, England.
 MOLECULAR BREEDING OF GROUNDNUT 215

Reddy, D. V. R., Rajeshwari, R., Iizuka, N., Lesemann, D. E., Nolt, B. L., and Gogo, T. (1983). The
occurrence of Indian peanut clump, a soil-born virus disease of groundnut (Arachis hypogaea L.)
in India. Ann. Appl. Biol. 102, 305–310.
Reddy, L. J., Nigam, S. N., Dwivedi, S. L., and Gibbons, R. W. (1987). Breeding groundnut cultivars
resistant to rust (Puccinina arachidis Speg.). In “Groundnut Rust Disease: Proc. Discussion
Group Meeting” (D. McDonald, P. Subrahmanyam, J. A. Wightman and J. V. Mertin, Eds.),
pp. 17–25. ICRISAT, Patancheru, India.
Reddy, D. V. R., Nolt, B. L., Hobbs, H. A., Reddy, A. S., Rajeshwari, R., Rao, A. S., Reddy, D. D. R.,
and McDonald, D. (1988). Clump virus in India: isolates, host range, transmission and
management. In “Viruses with Fungal Vectors” (J. I. Cooper and M. J. C. Asher, Eds.),
pp. 239–246. The Assoc. Appl. Biol. Wellsbourne, UK.
Reddy, D. V. R., Wightman, J. A., Beshear, R. J., Highland, B., Black, M., Sreenivasulu, M., Dwivedi,
S. L., Demski, J. W., McDonald, D., Smith Jr., J. W., and Smith, D. H. (1991). Bud necrosis: a
disease of groundnut caused by tomato spotted wilt virus. Information Bulletin No. 31 pp. 20
ICRISAT, Patancheru, India.
Reddy, L. J., Nigam, S. N., Subrahmanyam, P., Reddy, A. G. S., McDonald, D., Gibbons, R. W., and
Pentaiah, V. (1992). Registration of ICGV 87160 peanut. Crop Sci. 32, 1075.
Reddy, L. J., Nigam, S. N., Subrahmanyam, P., Reddy, A. G. S., McDonald, D., Gibbons, R. W., and
Pentaiah, V. (1993). Registration of ICGV 86590 peanut cultivar. Crop Sci. 33, 357–358.
Reddy, D. V. R., Buiel, A. A. M., Satyanarayana, T., Dwivedi, S. L., Reddy, A. S., Ratna, A. S.,
Lakshmi, K. V., Rao, G. V. R., Naidu, R. A., and Wightman, J. A. (1995). In “Proc. of a Meeting
on Recent Studies on Peanut Bud Necrosis Disease” (A. A. M. Buiel, J. E. Parlevliet and J. M.
Lenne, Eds.), pp. 3– 7. ICRISAT/Agricultural University of Wageningen, Patancheru, India/
The Netherlands.
Reddy, L. J., Nigam, S. N., Moss, J. P., Singh, A. K., Subrahmanyam, P., McDonald, D., and Reddy,
A. G. S. (1996). Registration of ICGV 86699 peanut germplasm line with multiple disease and
insect resistance. Crop Sci. 36, 821.
Reddy, C. D. R., Reddy, P. N., Reddy, M. V., Babu, P. R., Babu, A. M., and Reddy, M. L. (1998). Kadiri
4: A high-yielding, short duration groundnut cultivar for the postrainy and summer seasons: A
boon to Rayalaseema farmers of Andhra Pradesh, India. Int. Arachis Newsletter 18, 20–21.
Reddy, A. S., Reddy, L. J., Mallikarjuna, N., Abdurahman, M. D., Reddy, Y. V., Bramel, P. J., and
Reddy, D. V. R. (2000a). Identification of resistance to peanut bud necrosis virus (PBNV) in wild
Arachis germplasm. Ann. Appl. Biol. 137, 135–139.
Reddy, L. J., Nigam, S. N., Subrahmanyam, P., Ismael, F. M., Govinden, N., and van der Merwe,
P. J. A. (2000b). Registration of groundnut cultivar Venus (ICGV 87853). Int. Arachis Newsletter
20, 29–31.
Reddy, L. J., Nigam, S. N., Subrahmanyam, P., Ismael, F. M., Govinden, N., and van der Merwe,
P. J. A. (2001a). Registration of groundnut cultivar Sylvia (ICGV 93207). Int. Arachis Newsletter
21, 20–22.
Reddy, L. J., Nigam, S. N., Nageswara Rao, R. C., and Reddy, N.S. (2001b). Registration of ICGV
87354 peanut germplasm with drought tolerance and rust resistance. Crop Sci. 41, 274–275.
Rucker, K. S., Kvien, C. K., Holbrook, C. C., and Wood, J. E. (1995). Identification of peanut
genotypes with improved drought avoidence traits. Peanut Sci. 22, 14–18.
Ruiz, T. M., Ramos-Santana, R., and Sotomayor-Rios, A. (2000). Dry matter yield of rhizoma
perennial peanut (Arachis glabrata) harvested at six, nine and 12 weeks at two semiarid sites.
J. Agric. Univ. P.R. 84, 115 –131.
Ryder, C. D., Smith, L. B., Teakle, G. R., and King, G. J. (2001). Contrasting genome organization:
two regions of the Brassica oleracea genome conserved with collinear regions of the Arabidopsis
thaliana genome. Genome 44, 808–817.
Sandhu, B. S., and Khera, A. S. (1976). The role of epistasis in the inheritance of yield and its
components in groundnut. Crop Improv. 3, 9–17.
216 S.L. DWIVEDI ET AL.

Sangha, A. S., and Labana, K. S. (1982). Diallel analysis in groundnut (Arachis hypogaea L.). Part 2.
Pod number and pod yield. Theor. Appl. Genet. 64, 59–63.
Santra, D. K., Tekeoglu, M., Ratnaparkhe, M., Kaiser, W. J., and Muehlbauer, F. J. (2000).
Identification and mapping of QTL conferring resistance to Ascochyta blight in chickpea. Crop
Sci. 40, 606–1612.
Savage, G. P., and Keenan, J. I. (1994). The composition and nutritive value of groundnut kernels.
In “The Groundnut Crop: A Scientific Basis for Improvement” (J. Smart, Ed.), pp. 173– 213.
Chapman and Hall, London.
Schearer, C., Wright, G. C., Krosch, S., Tatnell, J., and Kyei, A. (1999). Effect of temperature on
growth and aflatoxin production by non-toxigenic (strain 4351) and toxigenic (strain 2754).
Aspargillus flavus. Food Legumes Newsletter 29, 4–7.
Schmitz, J., Franzen, R., Nguyen, T. H., Garcia-Maroto, F., Pozzi, C., Salamini, F., and Rohde, W.
(2000). Cloning, mapping and expression analysis of barley MADS-box genes. Plant Mol. Biol.
42, 899 –913.
Sebolt, A. M., Shoemaker, R. C., and Diers, B. W. (2000). Analysis of a quantitative trait locus
allele from wild soybean that increases seed protein concentration in soybean. Crop Sci. 40,
1438–1444.
Sen, S., and Churchill, G. A. (2001). A statistical framework for quantitative trait mapping. Genetics
159, 371 –387.
Shanlin, Y., Yuliang, C., Shuyuan, G., and Ping, M. (2000). A new high-yielding low oil content
groundnut variety. Int. Arachis Newsletter 20, 32–33.
Sharief, Y., Rawlings, J. O., and Gregory, W. C. (1978). Estimates of leaf spot resistance in three
interspecific hybrids of Arachis. Euphytica 27, 741–751.
Sharma, K. K., and Lavanya, M. (2002). Recent developments in transgenics for abiotic stress in
legumes of the semi-arid tropics. JIRCAS Working Report No. 23, 61–73.
Sharma, S. B., Waliar, F., and Ndunguru, B. J. (1991). Plant-parasitic nematodes associated with
groundnut in parts of Benin. Burkina Faso, and Nigeria. Oleagineux 46, 119 –120.
Sharma, S. B., Waliar, F., Subrahmanyam, P., and Ndunguru, B. J. (1992). Role of Scutellonema
clathricaudatum in etiology of groundnut growth variability in Niger. Plant and Soil 143,
133–139.
Sharma, S. B., Ansari, M. A., Varaprasad, K. S., Singh, A. K., and Reddy, L. J. (1999). Resistance to
Meloidogyne javanica in wild Arachis species. Genet. Resour. Crop Evol. 46, 557 –568.
Sharma, H. C., Sharma, K. K., Seetharama, N., and Ortiz, R. (2000). Prospects for using transgenic
resistance to insect pests in crop improvement. Electron. J. Biotechnol. 3, http://www. ejb.org/
content/vol 3/issue 2/full/20.
Shionzaki, K., and Yamaguchi-Shinozaki. (1996). Molecular responses to drought and cold stress.
Current-Opinion-in-Biotechnology 7, 161 –167.
Shorter, R., and Hammons, R. O. (1985). Pattern analysis of genotype adaptation and genotype £
environment interactions in the uniform peanut performance tests. Peanut Sci. 12, 35–41.
Shuren, G., Caibin, W., Shanlin, Y., Jimin, Z., Xueqing, Z., Shubo, W., Shouxiang, T., Chuantang,
W., and Ruorui, Q. (1996). Present situation and prospects for groundnut production in China.
In “Achieving High Groundnut Yields: Proc. Int. Workshop” (C. L. L Gowda, S. N. Nigam, C.
Johansen and C. Renard, Eds.), pp. 17–26. Laixi City, Shandong, China (In English.
Summaries in English and Chinese) and ICRISAT, Patancheru, India.
Simon, C. J., and Muehlbauer, F. J. (1997). Construction of a chickpea linkage map and its comparison
with maps of pea and lentil. J. Heredity 88, 115–119.
Simpson, C. E. (2001). Use of wild Arachis species/introgression of genes into A. hypogaea L. Peanut
Sci. 28, 114–116.
Simpson, C. E., and Starr, J. L. (2001). Registration of Coan peanut. Crop Sci. 41, 918.
Simpson, C. E., Nelson, S. C., Starr, J. L., Woodard, K. E., and Smith, O. D. (1993). Registration of
TxAG-6 and TxAG-7 peanut germplasm lines. Crop Sci. 33, 1418.
 MOLECULAR BREEDING OF GROUNDNUT 217

Simpson, C. E., Smith, O. D., and Melouk, H. A. (2000). Registration of Tamrun 98 peanut. Crop Sci.
40, 859.
Singh, A. K. (1986a). Utilization of wild relatives in genetic improvement of Arachis hypogaea L.
Part 7. Autotetraploid production and prospects in interspecific breeding. Theor. Appl. Genet. 72,
164 –169.
Singh, A. K. (1986b). Utilization of wild relatives in genetic improvement of Arachis hypogaea L.
Part 8. Synthetic amphiploids and their importance in interspecific breeding. Theor. Appl. Genet.
72, 433–439.
Singh, A. K., and Nigam, S. N. (1997). Groundnut. In “Biodiversity in Trust” (D. Fuccillo, L. Sears
and P. Stapleton, Eds.), pp. 114–127. Cambridge University Press, UK.
Singh, A. K., Subrahmanyam, P., and Moss, J. P. (1984). The dominant nature of resistance to
Puccinia arachidis in certain wild Arachis species. Oleagineux 39, 535 –538.
Singh, A. K., Stalker, H. T., and Moss, J. P. (1991). Cytogenetics and use of alien genetic variation in
groundnut improvement. In “Chromosome Engineering in Plants: Genetics, Breeding, Evolution,
Part B”. (T. Tsuchiya and P. K. Gupta, Eds.), pp. 65 –77. Elsevier Science Publishers,
Amsterdam, The Netherlands.
Singh, K. P., Raina, S. N., and Singh, A. K. (1996). Variation in chromosomal DNA associated with
the evolution of Arachis species. Genome 39, 890 –897.
Singh, A. K., Mehan, V. K., and Nigam, S. N. (1997). Sources of resistance to groundnut fungal and
bacterial wilt diseases: an update and appraisal. Information Bulletin No. 50; 48 ICRISAT,
Patancheru, India.
Singh, A. K., Smartt, J., Simpson, C. E., and Raina, S. N. (1998). Genetic variation vis-à-vis
molecular polymorphism in groundnut. Arachis hypogaea L. Genet. Resour. and Crop Evol. 45,
119 –126.
Skriver, K., and Mundy, J. (1990). Gene expression in response to abscisic acid and osmotic stress.
Plant Cell 2, 503 –512.
Smart, J. (1961). The disease of groundnut in Northern Rhodesia. Emp. J. Expt. Agric. 29, 79–87.
Smart, J. (1978). Makulu Red–A ‘green revolution’ groundnut variety? Euphytica 27, 605–608.
Smirhoff, N., and Colombe, S. V. (1989). Drought influences the activity of enzymes of the chloroplast
hydrogen peroxide scavenging system. J. Expl. Bot. 39, 1097–1180.
Stalker, H. T., and Beute, M. K. (1993). Registration of four leafspot resistant peanut germplasm lines.
Crop Sci. 33, 1117.
Stalker, H. T., and Campbell, W. V. (1983). Resistance of wild species of peanut to an insect complex.
Peanut Sci. 10, 30–33.
Stalker, H. T., and Lynch, R. E. (2002). Registration of four insect-resistant peanut germplasm lines.
Crop Sci. 42, 312– 313.
Stalker, H. T., and Moss, J. P. (1987). Speciation, cytogenetics, and utilization of Arachis species. Adv.
Agron. 41, 1 –40.
Stalker, H. T., and Simpson, C. E. (1995). Germplasm resources in Arachis. In “Advanced Peanut
Science” (H. E. Pattee and H. T Stalker, Eds.), pp. 14– 53. Stillwater, USA.
Stalker, H. T., Young, C. T., and Jones, T. M. (1989). A survey of the fatty acids of peanut species.
Oleagineux 44, 419 –424.
Stalker, H. T., Beute, M. K., Shew, B. B., and Barker, K. R. (2002a). Registration of two root-knot
nematode resistant peanut germplasm lines. Crop Sci. 42, 312 –313.
Stalker, H. T., Beute, M. K., Shew, B. B., and Isleib, T. G. (2002b). Registration of five leaf spot
resistant peanut germplasm lines. Crop Sci. 42, 313–314.
Subrahmanyam, P., van Wyk, P. S., Kisyombe, C. T., Cole, D. L., Hildebrand, G. L., Chiyembekeza,
A. J., and van der Merwe, P. J. A. (1977). Diseases of groundnut in the Southern African
Development Community Region and their management. Int. J. Pest Management 43, 261–273.
Subrahmanyam, P., McDonald, D., Gibbons, R. W., and Rao, P. V. S. (1983a). Components of
resistance to Puccinia arachidis in peanut. Phytopathology 73, 253–256.
218 S.L. DWIVEDI ET AL.

Subrahmanyam, P., Moss, J. P., and Rao, V. R. (1983b). Resistance to peanut rust in wild Arachis
species. Plant Disease 67, 209–212.
Subrahmanyam, P., Williams, J. H., McDonald, D., and Gibbons, R. W. (1984). The influence of foliar
diseases and their control by selective fungicides on a range of groundnut genotypes. Ann. Appl.
Biol. 104, 467–476.
Subrahmanyam, P., Ghanekar, A. M., Nolt, B. L., Reddy, D. V. R., and McDonald, D. (1985a).
Resistance to groundnut diseases in wild Arachis species. In “Proc. Int. Workshop on
Cytogenetics of Arachis” (J. P. Moss and S. D. Feakin, Eds.), pp. 49–55. ICRISAT, Patancheru,
India.
Subrahmanyam, P., Moss, J. P., McDonald, D., Rao, P. V. S., and Rao, V. R. (1985b). Resistance to
leaf spot caused by Cercosporidium personatum in wild Arachis species. Plant Disease 69,
951–954.
Subrahmanyam, P., Reddy, L. J., Gibbons, R. W., and McDonald, D. (1985). Peanut rust: A major
threat to peanut production in the semi-arid tropics. Plant Dis. 69, 813 –819.
Subrahmanyam, P., Greenberg, D. C., Savary, S., and Bosc, J. P. (1991). Diseases of groundnut in
West Africa and their management: research priorities and strategies. Trop. Pest Management 37,
259–269.
Subrahmanyam, P., Hildebrand, G. L., Naidu, R. A., Reddy, L. J., and Singh, A. K. (1998). Sources of
resistance to groundnut rosette disease in global groundnut germplasm. Ann. Appl. Biol. 132,
473–485.
Subrahmanyam, P., Naidu, R. A., Reddy, L. J., Kumar, P. L., and Ferguson, M. E. (2001). Resistance
to groundnut rosette disease in wild Arachis species. Ann. Appl. Biol. 139, 45–50.
Subramanian, V., Gurtu, S., Nageswara Rao. R. C., and Nigam, S. N. (2000). Identification of DNA
polymorphism in cultivated groundnut using random amplified polymorphic DNA (RAPD)
assay. Genome 43, 656–660.
Swe, S. T., and Branch, W. D. (1986). Estimates of combining ability and heterosis among peanut
cultivars. Peanut Sci. 13, 70–74.
Tabaeizadeh, Z. (1998). Drought-induced responses in plant cells. Int. Rev. Cytology 182, 193– 247.
Tai, P. Y. P., and Hammons, R. O. (1978). Genotype-environment interaction effects in peanut variety
evaluation. Peanut Sci. 5, 72–74.
Talwar, H. S., Takeda, H., Yashima, S., and Senboku, T. (1999). Growth and photosynthetic responses
of groundnut genotypes to high temperature. Crop Sci. 39, 460– 466.
Tanksley, S. D., and McCouch, S. R. (1997). Seed banks and molecular maps: Unlocking genetic
potential from the wild. Science 277, 1063–1066.
Tanksley, S. D., and Nelson, J. C. (1996). Advance backcross QTL analysis: a method for the
simultaneous discovery and transfer of valuable QTL from unadapted germplasm to elite
breeding lines. Theor. Appl. Genet. 92, 191– 203.
Tanksley, S. D., Young, N. D., Paterson, A. H., and Bonierbale, M. W. (1989). RFLP mapping in plant
breeding: new tools for an old science. BioTechnology 7, 257–264.
Tanksley, S. D., Ganal, M. W., Prince, J. P., de Vicente, M. C., Bonierbale, M. W., Broun, P., Fulton,
T. M., Giovannoni, J. J., Grandillo, S., Martin, G. B., Messeguer, R., Miller, J. C., Miller, L.,
Paterson, A. H., Pineda, O., Roder, M. S., Wing, R. A., Wu, W., and Young, N. D. (1992). High
density molecular linkage maps of the tomato and potato genomes. Genetics 132, 1141–1160.
Temsch, E. M., and Greilhuber, J. (2000). Genome size variation in Arachis hypogaea and A.
monticola re-evaluated. Genome 43, 440–451.
Tenaillon, M. I., Sawkins, M. C., Long, A. D., Gaut, R. L., Doebley, J. F., and Gaut, B. S. (2001).
Patterns of DNA sequence polymorphism along chromosome 1 of maize (Zea Mays ssp. Mays
L.). Proc. Nat. Acad. Sci. USA 98, 9161–9166.
Thakur, R. P., Rao, V. P., Reddy, S. V., and Ferguson, M. (2000). Evaluation of wild Arachis
germplasm accessions for in vitro seed colonization and aflatoxin production by A. flavus. Int.
Arachis Newsletter 20, 44–46.
 MOLECULAR BREEDING OF GROUNDNUT 219

The International SNP Map Working Group (2001). A map of human genome sequence variation
containing 1.42 million single nucleotide polymrphisms. Nature 409, 928–933.
Thouvenel, J. C., Fauquet, C., Fargette, D., and Fishpool, L. D. C. (1988). Peanut clump virus in West
Africa. In “Viruses with Fungal Vectors” (J. I. Cooper and M. J. C. Asher, Eds.), pp. 247–254.
The Association of Applied Biology, Wellesbourne, UK.
Timper, P., Holbrook, C. C., and Xue, H. Q. (2000). Expression of nematode resistance in plant
introduction of Arachis hypogaea. Peanut Sci. 27, 78– 82.
Tiwari, S. P., Ghewande, M. P., and Mishra, D. P. (1984). Inheritance of resistance to rust and late leaf
spot in groundnut (A. hypogaea). J. Cytol. and Genet. 19, 97 –101.
Upadhyaya, H. D., and Nigam, S. N. (1994). Inheritance of two components of early maturity in
groundnut (Arachis hypogaea L.). Euphytica 78, 59–67.
Upadhyaya, H. D., and Nigam, S. N. (1998). Epistasis for vegetative and reproductive traits in peanut.
Crop Sci. 38, 44– 49.
Upadhyaya, H. D., and Nigam, S. N. (1999a). Inheritance of fresh seed dormancy in peanut. Crop Sci.
39, 98–101.
Upadhyaya, H. D., and Nigam, S. N. (1999b). Detection of epistasis for protein and oil contents and oil
quality parameters in peanut. Crop Sci. 39, 115–118.
Upadhyaya, H. D., Nigam, S. N., Rao, M. J. V., Reddy, A. G. S., Yellaiah, N., and Reddy, N. S.
(1997a). Registration of five Spanish peanut germplasm lines with fresh seed dormancy. Crop Sci.
37, 1027.
Upadhyaya, H. D., Nigam, S. N., Rao, M. J. V., Reddy, A. G. S., Yellaiah, N., and Reddy, N. S.
(1997b). Registration of ICGV 86143 peanut germplasm. Crop Sci. 37, 1986.
Upadhyaya, H. D., Nigam, S. N., Rao, M. J. V., Reddy, A. G. S., Yellaiah, N., and Reddy, N. S. (1998).
Registration of early-maturing peanut germplasm ICGV 92196, ICGV 92206, ICGV 92234, and
ICGV 92243. Crop Sci. 38, 900–901.
Upadhyaya, H. D., Nigam, S. N., Reddy, A. G. S., and Yellaiah, N. (2001a). Registration of early-
maturing fresh seed dormant peanut germplasm ICGV 93470. Crop Sci. 41, 597–598.
Upadhyaya, H. D., Nigam, S. N., Pande, S., Reddy, A. G. S., and Yellaiah, N. (2001b). Registration of
early maturing, moderately resistant to rust peanut germplasm. Crop Sci. 41, 598– 599.
Upadhyaya, H. D., Nigam, S. N., Mehan, V. K., Reddy, A. G. S., and Yellaiah, N. (2001c).
Registration of Aspergillus flavus seed infection resistant peanut germplasm ICGV# 91278,
91283, and 91284. Crop Sci. 41, 599–600.
Upadhyaya, H. D., Nigam, S. N., and Singh, S. (2001d). Evaluation of groundnut core collection to
identify sources of tolerance to low temperature at germination. Indian J. Plant Genet. Resour.
14, 165–167.
Upadhyaya, H. D., Ferguson, M. E., and Bramel, P. J. (2001e). Status of Arachis germplasm collection
at ICRISAT. Peanut Sci. 28, 89 –96.
Upadhyaya, H. D., Ortiz, R., Bramel, P. J., and Singh, S. (2003). Development of a groundnut core
collection using taxonomical, geographical, and morphological descriptors. Genetic Resour.
Crop Evol. 50, 139 –148.
Utomo, S.D., Anderson, W.F., Wynne, J.C., Beute, M.K., Hagler Jr., W.M., and Payne, G.A. (1990).
Estimates of heritability and correlation among three mechanisms of resistance to Aspergillus
parasiticus in peanut. Proc. Amer. Peanut Res. Educ. Soc. 22, pp. 26 (Abstract).
van der Merwe, P. J. A., Subrahmanyam, P., Hildebrand, G. L., Reddy, L. J., Nigam, S. N.,
Chiyembekeza, A. J., Busolo-Bulafu, C. M., and Kapewa, T. (2001). Registration of groundnut
cultivar ICGV-SM 90704 with resistance to groundnut rosette. Int. Arachis Newsletter 21, 19–20.
van Dodeweerd, A-M., Caroline, C. R., Bent, E. G., Johnson, S. J., Bevan, M. W., and Bancroft, I.
(1999). Identification and analysis of homoeologous segments of the genomes of rice and
Arabidopsis thaliana. Genome 42, 887–892.
Varman, P. V., Ravendran, T. S., and Ganapathy, T. (1991). Genetic analysis of rust resistance in
groundnut Arachis hypogaea L. J. Oilseeds Res. 8, 35–39.
220 S.L. DWIVEDI ET AL.

Varman, P. V., Joel, A. J., Mylswami, V., Nagarajan, P., Raveendran, T. S., Sridharan, C. S.,
Dwivedi, S. L., Nigam, S. N., and Rao, G. V. R. (1998). Registration of ALR 2 peanut. Crop
Sci. 38, 1716.
Vindhiyavarman, P., and Mohammed, S. E. N. (2001). Release of foliar disease resistant groundnut
cultivar VRI Gn 5 in Tamil Nadu. India. Int. Arachis Newsletter 21, 16 –17.
Visscher, P. M., Haley, C. S., and Thompson, R. (1996). Marker-assisted introgression in backcross
breeding programs. Genetics 144, 1923–1932.
Waliyar, F., (1991). Evaluation of yield losses due to groundnut leaf diseases in West Africa. In
“Summary Proc. of the Second ICRISAT Regional Groundnut Meeting for West Africa”
(B. J. Nduguru, F. Waliar, and B. R. Ntare, Eds.), pp. 32–33. ICRISAT Sahelian Centre, Niamey,
Niger.
Waliyar, F., McDonald, D., Rao, P. V. S., and Reddy, P. M. (1993). Components of resistance to an
Indian source of Cercospora arachidicola in selected peanut lines. Peanut Sci. 20, 93 –96.
Wang, Y. Y., Wang, C. H., and Xia, X. M. (1985). A preliminary study on inheritance of resistance to
bacterial wilt in peanut. Oil Crops of China 4, 15–17.
Weiss, E.A., (1983). Oilseed Crops. Longman, London, pp. 660.
Weeden, N. F., Muehlbauer, F. J., and Ladizinsky, G. (1992). Extensive conservation of linkage
relationships between pea and lentil genetic maps. J. Hered. 83, 123 –129.
Wheeler, T. R. W., Chatzialioglou, A., Craufurd, P. Q., Ellis, R. H., and Summerfield, R. J. (1997). Dry
matter partitioning in groundnut exposed to high temperature stress. Crop Sci. 37, 1507–1513.
Wicking, C., and Williamson, B. (1991). From linked marker to gene. Trends Genet. 7, 288–293.
Wightman, J. A., and Rao, G. V. R. (1994). Groundnut pests. In “The Groundnut Crop: A Scientific
Basis for Improvement”. (J. Smart, Ed.), pp. 395–479. Chapman and Hall, London.
Wightman, J. A., Dick, K. M., Rao, G. V. R., Shanower, T. G., and Gold, C. G. (1990). Pests of
groundnut in the Semi-Arid Tropics. In “Insect Pests of Food Legumes” (R. K. Singh, Ed.), pp.
243–322. Wiley, London.
Wightman, J. A., Rao, G. V. R., and Lakshmi, K. V. (1995). Thrips palmi, general pest and vector of
some tospoviruses in Asia. In “Recent Studies in Peanut Bud Necrosis Disease: Proc. of a
Meeting” (A. A. M. Buiel, J. E. Parlevliet, J. M. Lenne, Eds.), pp. 11–15. ICRISAT Patancheru,
India and Agricultural University of Wageningen, The Netherlands.
Wilcut, J. W., York, A. C., Grichar, W. J., and Wehtje, G. R. (1995). The biology and management of
weeds in peanut (Arachis hypogaea L.). In “Advances in Peanut Science” (H. E. Pattee, H. T.
Stalker, Eds.), pp. 207 –244. Stillwater, USA.
Wissuwa, M., and Ae, N. (2001). Genotypic differences in the presence of hairs on roots and
gynophores of peanuts (Arachis hypogaea L.) and their significance for phosphorus uptake.
J. Expl. Bot. 52, 1703–1710.
Wright, G. C., Nageswara Rao, R. C., and Farquhar, G. D. (1994). Water-use efficiency and carbon
isotope discrimination in peanut under water deficit conditions. Crop Sci. 34, 92–97.
Wynne, J. C., and Gregory, W. C. (1981). Peanut breeding. Adv. Agron. 34, 39– 71.
Wynne, J. C., and Isleib, T. G. (1978). Cultivar £ environment interaction in peanut yield test. Peanut
Sci. 5, 102–105.
Wynne, J. C., Emery, D. A., and Rice, P. N. (1970). Combining ability estimates in Arachis hypogaea
L. II. Field performance of F1 hybrids. Crop Sci. 10, 713 –715.
Wynne, J. C., Rawlings, J. O., and Emery, D. A. (1975). Combining ability estimates in Arachis
hypogaea. L., III. F2 generation of intra- and inter-subspecific crosses. Peanut Sci. 2, 50 –54.
Wynne, J. C., Beute, M. K., and Nigam, S. N. (1991). Breeding for disease resistance in peanut
(Arachis hypogaea L.). Annu. Rev. Phytopathol. 29, 279–303.
Xiao, J., Grandillo, S., Ahn, S. N., McCouch, S. R., Tanksley, S. D., Li, J., and Yuan, L. (1996). Genes
from wild rice improve yield. Nature 384, 223 –224.
Yanhao, S., and Caibin, W. (1990). Factors contributing to high yields of groundnut in Shandong,
China. Int. Arachis Newsletter 8, 7– 9.
 MOLECULAR BREEDING OF GROUNDNUT 221

Youn-Sup, O., Young-Keun, C., Jong-chul, K., Jong-Tae, K., Myung-Kyu, O., Jung-Gon, K., Ki-Hun,
P., Young-Sun, J., Moon-Soo, P., Su-Yeon, C., Dwivedi, S. L., and Nigam, S. N. (2000).
Registration of Jeokwangtangkong peanut. Crop Sci. 40, 292.
Young, C. T., and Hammons, R. O. (1973). Variation in the protein levels of a wide range of peanut
genotypes (Arachis hypogaea L.). Oleaginuex 28, 293–297.
Young, N.D., Weeden, N.F., and Kochert, G. (1996). Genome mapping in legumes (Fam. Fabaceae).
In “Genome Mapping in Plants” (A. H. Paterson, Ed.), pp. 211–227. Landes Co, USA.
Zeringue, H., Brown, R., Neucere, J., and Cleveland, T. (1996). Relationship between C6-C12 alkanal
and alkenal volatile contents and resistance of maize genotypes to Aspergillus flavus and aflatoxin
production. J. Agric. Food Chem. 44, 403–407.
Zhengchao, L., and Qingshu, Q. (2000). Huayu 16: A new high-yielding, improved quality groundnut
cultivar with wide adaptability for Northern China. Int. Arachis Newsletter 20, 31–32.
Zhengchao, L., Qingshu, Q., Chuantang, W., Huarong, M., and Fuyu, S. (1997). Luhua 15: A high-
yielding, small-seeded groundnut cultivar with improved seed quality. Int. Arachis Newsletter 17,
29 –30.
 AN ECOLOGICAL APPROACH TO
STRENGTHEN WEED
MANAGEMENT IN THE SEMIARID
GREAT PLAINS
Randy Anderson
USDA—Agricultural Research Service, 2923 Medary Ave.,
Brookings, SD 57006, USA

I. Introduction
A. Changes in Cropping Practices
B. Concerns with Herbicide Resistance and Input Costs
II. Ecologically Based Weed Management
A. Reducing Weed Density
B. Enhancing Crop Competitiveness
C. Opportunity to Integrate Ecological Processes with Weed
Management
III. Reducing Weed Density
A. Rotations Comprised of Crops with Different Life Cycles
B. Tillage Minimizes Rotation Effect on Weed Density
IV. Enhancing Crop Competitiveness
A. Winter Wheat
B. Corn
C. Sunflower
D. Proso Millet
V. Summary
A. Rotations Based on a Cycle-of-Four Aid Weed Management
but also Accrue Other Benefits
B. Semiarid Paradox with Tillage and Herbicides
References

Cropping systems in the semiarid Great Plains are rapidly changing.

Previously, winter wheat – fallow was the prevalent system; now, because of
no-till practices, producers are diversifying their rotations to include
alternative crops. Yet, weed management is often ineffective because of
herbicide-resistant weeds and low profit margins. A possible solution is
ecologically based weed management, where cropping systems are designed
to lower weed community densities and improve crop competitiveness to
weeds. Both winter and summer annual crops are grown in the Great Plains,
which can help manage weeds because growth periods differ considerably
between crop types. Designing rotations to include two winter/spring crops

33
Advances in Agronomy, Volume 80
Copyright q 2003 by Academic Press. All rights of reproduction in any form reserved
0065-2113/03$35.00
34 R. ANDERSON

followed by two summer annual crops can reduce weed community density
12-fold compared to less diverse rotations. This rotation design favors natural
weed seed decline in soil yet avoids proliferation of weed densities in crops
with similar life cycles. However, tillage minimizes the effect of rotation
design on weed community. A second component of the ecological approach
is to strengthen crop competitiveness with cultural practices. Combining
three practices together improved crop competitiveness three- to five-fold.
With some crops, such as proso millet, cultural systems are so effective that
herbicides may not be needed for in-crop weed control. Designing rotations
based on a cycle-of-four with winter and summer annual crops also accrues
ancillary benefits such as increased yields and economic returns as well as
improved resource-use-efficiency. Furthermore, the ecologically based
approach will enable producers to ameliorate the negative impacts of
herbicide-resistant weeds and rising input costs. q 2003 Academic Press.

I. INTRODUCTION

The Central Great Plains, an area in eastern Colorado and Wyoming, and
western Kansas and Nebraska, is part of the shortgrass prairies of the USA. In
spite of its semiarid climate, dryland cropping with winter wheat (Triticum
aestivum L.) is a major component of the region’s agriculture. Water relations are
the most critical aspect of crop production, as precipitation ranges from 250 to
450 mm per year.
When the region was settled in the late 1800s, settlers followed cropping
practices common at their place of origin, most often being the humid east.
Initially, winter wheat, corn (Zea mays L.), and sorghum [Sorghum bicolor (L.)
Moench] were grown in various rotations (Greb et al., 1974). Droughts of the
1930s and 1950s, however, led producers to adopt the winter wheat – fallow
rotation to compensate for the region’s erratic precipitation. During fallow,
neither crops nor weeds are allowed to grow, as the goal of fallow is to store
precipitation in soil. Soil water gained during fallow improves growth of the
following winter wheat, subsequently reducing yield variability and crop loss due
to drought stress.

A. CHANGES IN CROPPING PRACTICES

Producers in the Central Great Plains are changing their crop rotations from
winter wheat – fallow because of two factors. First, herbicides allow producers
to eliminate tillage for weed control during fallow; with no-till systems,
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 35

producers preserve crop residue on the soil surface, which improves

precipitation storage in soil during non-crop periods. Extra water in soil allows
producers to grow more crops in succession before fallow is needed (Peterson
et al., 1996). Producers now are growing summer annual crops such as corn,
sorghum, proso millet (Panicum miliaceum L.), and sunflower (Helianthus
annuus L.) in sequence with winter wheat and fallow (Peterson et al., 1993a;
Anderson et al., 1999). Because no-till systems have been so effective,
producers are seeking to eliminate fallow for both economic and environmental
reasons. Rotations with continuous cropping, such as winter wheat – proso
millet and winter wheat –corn – proso millet, have been successful on favorable
soils.
A second factor encouraging changes in rotations is proliferation of winter
annual grasses such as jointed goatgrass (Aegilops cylindrica Host), feral rye
(Secale cereale L.) and downy brome (Bromus tectorum L.) (Wicks, 1984;
Donald and Ogg, 1991). These annual grasses are well-adapted to winter wheat –
fallow because their life cycles are similar to winter wheat. Furthermore,
herbicide options to control these grasses in winter wheat are limited, and in most
situations, ineffective. Thus, producers started growing summer annual crops
after winter wheat to help manage these weeds.

B. CONCERNS WITH HERBICIDE RESISTANCE AND INPUT COSTS

With no-till systems, weeds are controlled with herbicides instead of tillage.
A startling consequence of herbicide use, however, is the appearance of resistant
weeds. For many years, atrazine effectively controlled weeds during fallow as
well as in corn. But now, kochia [Kochia scoparia (L.) Schrad.], green foxtail
[Setaria viridis (L.) Beauv.], and downy brome are resistant to atrazine (Holt
and LeBaron, 1990). A second class of herbicides, whose mode of action is
inhibiting the acetolactate synthase (ALS) enzyme, was registered for use in
winter wheat during the early 1980s. Yet, within only five years, species such as
kochia, prickly lettuce (Lactuca serriola L.), and Russian thistle (Salsola iberica
Sennen & Pau) developed resistance to this new class of herbicides (Lyon et al.,
1996).
Glyphosate is used extensively to control weeds during non-crop periods
and fallow. With glyphosate, scientists hypothesized that weed resistance would
be rare because of its unique mode of action (Kishmore et al., 1992). Yet, world
wide, four weed species are now resistant to glyphosate (Heap, 2002), including
horseweed [Conyza canadensis (L.) Cronq.], which is common in the Central
Great Plains. A further concern with glyphosate is increased tolerance in
weeds. Several species in the Central Great Plains, such as toothed spurge
(Euphorbia serrata), tumble windmillgrass (Chloris verticillata Nutt.), common
eveningprimrose (Oenothera biennis L.), and wild buckwheat (Polygonum
36 R. ANDERSON

convolvulus L.), require double the normal-use rates of glyphosate for control.1
A similar situation exists in Australia. After 15 years of no-till winter wheat
systems, 17 weed species are tolerant of glyphosate at normal use rates (Wicks
et al., 2000).
A further disturbing trend is multiple resistance; weeds resistant to herbicides
with one mode of action can become resistant to herbicides with other modes of
action, even if these herbicides have never been applied to the species (Powles
et al., 1997). Cause of multiple resistance is unknown, but selection pressure can
result in weeds with different metabolism, thus disrupting activity of other
metabolic-based herbicides on that species. An example of multiple resistance is
rigid ryegrass (Lolium rigidum) in Australia, which is resistant to herbicides of
most modes of action used in that country, including glyphosate (Powles et al.,
1998).
A second consequence of new cropping systems in the Central Great Plains is
that production costs have increased considerably. Producers still lose
considerable crop yield due to weeds, yet herbicide resistance is forcing
producers to use more expensive management tactics (Beckie et al., 1999).
Furthermore, fertility needs are higher because the nitrogen-supplying aspect of
fallow is minimized with more intensive cropping (Westfall et al., 1996). With
prices for crop commodities remaining relatively static or even decreasing, rising
input costs are straining economic viability of semiarid crop production;
consequently, producers are seeking to reduce inputs associated with weed
management.

II. ECOLOGICALLY BASED WEED MANAGEMENT

A possible solution to the above concerns is integrating weed management

with ecological processes. With ecologically based management, emphasis is
placed on (1) lowering weed community density, and (2) increasing crop
competitiveness with weeds (Pedigo, 1995). Diversifying rotations with crops of
different life cycles is one tactic for lowering weed community density, whereas
any cultural practice that helps crops access resources earlier than weeds
improves a crop’s competitiveness. With the ecologically based approach,
herbicides supplement cultural practices rather than serve as the sole tactic for
weed control. Thus, herbicides become an option rather than a requirement for
cropping success.

1
Gail Wicks, personal communications. 2002. North Platte NE.
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 37

A. REDUCING WEED DENSITY

Prominent weeds in Central Great Plains cropping systems are annual species,
whose population growth depends on seed production. Annual weeds associate
with crops that have similar life cycles because of greater opportunities for weed
seedlings to emerge, establish, and produce seeds (Froud-Williams, 1988). Weed
density increases substantially if rotations are comprised of crops with similar life
cycles (Striebig, 1979), as demonstrated in the Central Great Plains with winter
annual grasses infesting winter wheat.
Rotating winter and summer annual crops has the greatest impact on weed
community dynamics (Thomas and Frick, 1993). Different planting and harvest
dates among these crops provide opportunities for producers to prevent either
plant establishment or seed production. Diversifying a rotation with crops of
different life cycles was a key weed management tactic before herbicides were
developed (Leighty, 1938). This rotational strategy can be particularly valuable
in the Central Great Plains, as both winter and summer annual crops are
commonly grown and are economically viable.

B. ENHANCING CROP COMPETITIVENESS

A second component of ecologically based management is improving crop

competitiveness with weeds. Crop interaction with weeds varies with crop, weed
species, and environmental conditions, but consistently, any plant, whether a crop
or a weed, that captures resources first gains a competitive advantage over
neighboring plants (Mortimer, 1984). For example, some cultivars of winter
wheat are more tolerant of downy brome interference (Challaiah et al., 1986).
Tolerant cultivars are usually taller and intercept more solar radiation; less light
within the crop canopy reduces downy brome growth.
A second example is planting winter wheat at higher seeding rates with nar-
rower row spacing, which leads to earlier canopy closure. This strategy improves
winter wheat’s competitiveness with cheatgrass (Bromus secalinus L.) 10 – 25%
compared to conventional practices, reducing yield loss as well as cheatgrass seed
production (Koscelny et al., 1991). Another strategy favorable for crops is
fertilizer placement. Mesbah and Miller (1999) found that placing nitrogen (N)
fertilizer in a band below the crop seed reduces jointed goatgrass biomass in winter
wheat 15 – 20% because winter wheat accesses N earlier than jointed goatgrass.

C. OPPORTUNITY TO INTEGRATE ECOLOGICAL

PROCESSES WITH WEED MANAGEMENT

Because cropping systems are rapidly changing in the Central Great Plains,
an opportunity exists for scientists and producers to design cropping systems that
38 R. ANDERSON

integrate crop diversity and sequencing with our greater understanding of

weed ecology. This review examines weed dynamics as affected by rotations and
cultural practices in the Central Great Plains. Our goal is to relate research trends
to ecological processes and characteristics of the region’s weed community, then
suggest guidelines for integrating these processes with semiarid cropping
systems. With this approach, producers may be able to adequately manage
weeds in their production systems, yet reduce reliance on herbicides, minimize
development of herbicide resistance, and lower input costs for controlling
weeds.

III. REDUCING WEED DENSITY

A. ROTATIONS COMPRISED OF CROPS WITH DIFFERENT

LIFE CYCLES

As mentioned previously, rotating crops with different life cycles can disrupt
population growth of weeds (Froud-Williams, 1988; Thomas and Frick, 1993).
However, we have also noticed that duration of cropping within a life cycle
category, i.e., winter or summer annual life cycles, affects weed density.
We initially observed this trend in a cropping system study at Akron, CO,
where after several years, weed levels differed in proso millet among various
rotations (Anderson 1998a). To quantify this difference, we measured the
weed biomass in proso millet of four rotations: continuous proso millet (M –M),
winter wheat –proso millet – fallow (W – M– F), winter wheat – corn – proso millet
(W – C – M), and winter wheat – corn – proso millet – fallow (W – C – M – F).
Herbicides used in proso millet controlled only broadleaf weeds, which left
green foxtail and longspine sandbur [Cenchrus longispinus (Hack.) Fern.] as the
predominant species.
Biomass of these grasses was approximately 180 g/m2 in continuous proso
millet (Fig. 1). In contrast, weed biomass was six-fold less in W –M – F, a rotation
with only one summer crop in three years. Weed biomass in W – C –M was one
half of continuous proso millet (M –M), whereas weed biomass was three-fold
less in W – C –M –F compared to W – C –M. An intriguing finding was that weed
biomass did not differ between W –M – F and W – C – M– F, even though W –C –
M –F had two summer crops in a row. One similarity between W –M –F and W –
C – M– F is the two-year interval without summer crops (winter wheat and
fallow). During this interval, grasses were effectively controlled in wheat and
fallow, thus preventing seed production. Grass control in corn was approximately
90%, which allowed isolated plants to produce seeds during the growing season,
yet the two-year interval still reduced grass biomass in proso millet of W –C – M–
F similarly to W –M –F.
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 39

Figure 1 Biomass of green foxtail and longspine sandbur in proso millet as influenced by four
rotations at Akron, CO. Study initiated in 1990, with weed biomass measured in 1997 and 1998. Data
averaged across years; bars with the same letter are not significantly different based on Fisher’s LSD
(0.05). Abbreviations: W, winter wheat; M, proso millet; C, corn; and F, fallow. (Adapted from
Anderson, 1998a).

A two-year interval also reduces density of winter annual or cool season

weeds. Density of both jointed goatgrass and downy brome was reduced more
than 90% with rotations such as winter wheat –sunflower – fallow or winter
wheat – proso millet –fallow, compared to winter wheat – fallow (Lyon and
Baltensperger, 1995). Winter annual grasses were easily controlled during the
summer annual crop and fallow to prevent seed production. A second example is
fairy candelabra (Androsace occidentalis Lunell), a native cool-season species
that proliferates in no-till spring wheat –fallow in the Northern Great Plains. A
rotation comprised of spring wheat – winter wheat – sunflower eliminated fairy
candelabra (Anderson et al., 1998). Fairy candelabra is easily controlled before
sunflower is planted, whereas winter wheat growth in the spring is more
competitive than spring wheat and prevents fairy candelabra seedlings from
establishing. The two-year interval out of spring wheat prevented fairy
candelabra proliferation in this rotation.
These data trends led us to hypothesize that weed community density would
be less if rotations were designed to include two-year intervals within each life
cycle of winter or summer annual crops. Our hypothesis was supported by data
from a crop rotation study at Pierre, SD (R. L. Anderson and D. L. Beck,
unpublished data, 2002). In this study, we measured weed density in four
rotations: winter wheat – fallow (W – F), winter wheat – chickpea (Cicer
arietinum L.) (W – CP), winter wheat –corn –chickpea (W –C – CP), and winter
wheat – corn – soybean (Glycine max Merrill) – dry pea (Pisum sativum L.)
(W – C –SB – Pea). Rotations were comprised of a range of winter and summer
40 R. ANDERSON

annual crops, with winter wheat and dry pea being winter/spring annual species
whereas corn, soybean, and chickpea are summer annual crops. Weeds were
controlled within and between crops with herbicides and management practices
commonly used by producers in the region. Weed density means represent an
average across all crops in a rotation.
Weed density in W –F was 31 plants/m2, with downy brome being the main
weed species (Fig. 2). When chickpea was grown with winter wheat (W – CP),
weed density in the rotation increased to 60 plants/m2, and included summer
annual weeds such as green foxtail, witchgrass (Panicum capillare L.), and
redroot pigweed (Amaranthus retroflexus L.) as well as downy brome in the weed
community. When a second summer annual crop, corn, was added to the rotation
(W – C – CP), downy brome was eliminated in the rotation’s weed community, but
summer annual weeds remained at a density of 25 plants/m2. In the four-year
rotation of W –C – SB – Pea, where two winter/spring crops (winter wheat and dry
pea) were followed by two summer annual crops (corn and soybean), weed
density was only five plants/m2. Weed density in the four-year rotation was 12-
fold less compared to W – CP and five-fold less compared to W –C – CP. Two
years in summer annual crops reduced winter annual weed density whereas two
winter/spring season crops minimized warm season species. This interaction
between rotations and weed dynamics may be partially attributed to two
ecological processes: weed seed decline in soil and rate of population growth of a
weed species.

Figure 2 Weed density in four rotations of a cropping systems study, Pierre, SD. Study initiated
in 1990; weed community assessed in 2001. Treatment means represent weed density averaged across
all crops within each rotation. Bars with the same letter are not significantly different based on Fisher’s
LSD (0.05). Abbreviations: W, winter wheat; CP, chickpea, C, corn; SB, soybean; and Pea, dry pea.
(R. L. Anderson and D. L. Beck, unpublished data, 2002).
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 41

1. Seed Bank Dynamics

Shedding of seeds by weeds leads to a reservoir of weed seeds in soil, known

as the seed bank, and is the main source of weed infestation in future crops
(Roberts, 1981). Seeds in soil either germinate, die of natural causes, or are
consumed by microorganisms such as fungi and nematodes and predators such as
field mice or birds (Harper, 1977). Decline of seed density in soil follows a
typical trend, with rapid loss of live seeds in the first two years after shedding
(Roberts, 1981; Egley and Williams, 1990).
With green foxtail and longspine sandbur, less than 10% of seeds are alive
after two years in soil, a four-fold difference compared to seed density after one
year (Fig. 3a). Similar trends occur with winter annual grasses, as jointed
goatgrass and downy brome seed density also declines to less than 10% after two
years in the seed bank (Fig. 3b). Seed densities of all species were at least three-
fold less after two years compared to one year in soil.
Rapid loss of viable seeds in soil has been found with numerous weed species
(Roberts, 1981; Froud-Williams et al., 1984; Egley and Williams, 1990).
Designing rotations to include two-year intervals with crops of similar life cycles
favors natural decline of weed seeds with dissimilar life cycles in soil; with most
weed species, potential density can be reduced 90%.

2. Natural Rate of Increase: Green Foxtail as an Example

A second factor influencing population dynamics of weeds in rotations is the

intrinsic rate of increase of a species, which quantifies the rate of population
change in an ecosystem. This concept reflects survival and seed production of
plants within constraints of a cropping system.
Our interest in this factor was stimulated by producer experiences with weeds
in corn. Seeking to eliminate jointed goatgrass in winter wheat, producers grew
corn three years in a row to reduce density of jointed goatgrass seeds in soil.
However, this strategy led to a dramatic proliferation of summer annual grasses
in corn.
To understand this trend, we quantified the following parameters that
influence population growth of green foxtail (Anderson, 1998a). First, green
foxtail growing in corn produces approximately 2300 seeds per plant, and
secondly, 8% of green foxtail seeds in soil emerge each year. We also noted that
weed control in corn generally eliminates 90% of green foxtail plants.
Based on these parameters, one green foxtail plant establishing during the
first year of corn will lead to 18 plants the second year of corn and 324 plants in
the third year, a rate of increase of 18-fold per year. Exponential growth by
green foxtail led to severe infestation after three years of continuous corn.
We have also observed a similar rate of increase for downy brome infesting
42 R. ANDERSON

Figure 3 Persistence of seed viability of green foxtail and longspine sandbur (a), and downy
brome and jointed goatgrass (b), when seed remains within the top 5 cm of soil. (Adapted from
Dawson and Burns, 1975; Donald and Zimdahl, 1987; Donald, 1991; and Anderson, 1998a).

winter wheat.2 These trends with green foxtail and downy brome indicate that
growing crops with similar life cycles more than two years in a row favors
exponential growth in weed density. If rotations are comprised only of crops
with similar life cycles, producers will have to compensate for the natural rate
of increase in population with herbicides or other weed control tactics.

2
Stymiest, C. E. and R. L. Anderson. 2001. p. 59, Annual Research Report: West River Agricultural
Center Crops and Soils Research. 1905 Plaza Boulevard, Rapid City, SD.
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 43

B. TILLAGE MINIMIZES ROTATION EFFECT ON WEED DENSITY

Another rotation study located at Wall, SD, also demonstrated the impact of
arranging crops by life cycle. Weed density among rotations was lowest in four-
year rotations comprised of two winter/spring crops followed by two summer
annual crops.3 However, we were surprised at the magnitude of difference in
weed density between rotations in this study and the study at Pierre, SD. Weed
density differed only three-fold between four-crop and two-crop rotations at Wall
(data not shown), contrasting with the 12-fold difference at the Pierre study
(Fig. 2). Similar crops were compared, and the predominant weed species at both
sites were downy brome and green foxtail.
A key difference between the two studies was tillage. At Wall, tillage with the
sweep plow4 incorporated herbicides and fertilizer, with at least one tillage
operation occurring each year; the study at Pierre was no-till in all years. The
difference of rotation impact on weed density between locations suggests that
tillage may affect long-term weed dynamics.
Tillage usually stimulates a flush of seedlings by placing weed seeds in more
favorable sites in soil for germination (Roberts, 1981). Scientists have speculated
that tillage, by stimulating germination, may reduce weed seed density in soil.
However, burial of weed seeds in soil by tillage also prolongs their survival
because soil protects seeds from environmental extremes (Froud-Williams et al.,
1984; Egley, 1986). For example, persistence of green foxtail seeds after two
years was greater than 50% when seeds were buried 10 cm, contrasting with less
than 10% of seeds surviving when they remained on the soil surface (Fig. 4).
Even when seeds were buried less than 5 cm in soil, the survival rate was still
two-fold greater after two years compared to seeds on the soil surface.
Egley and Williams (1990), comparing seed longevity of several weed species
between tillage and no-till across five years, also found that tillage did not
enhance seed loss in soil. Loss of seed viability was greatest if seeds remained on
the soil surface, especially during the first year. Sagar and Mortimer (1976)
reported similar results with wild oat (Avena fatua L.); loss of viable seed over
winter was five times greater when seeds remained on the soil surface compared
to burial 5 cm deep.
Mohler (1993), reviewing the literature, reported that long-term seed
survival of most weed species is greater if seeds are buried in soil compared
to lying on the soil surface. Jointed goatgrass was noted as one for the few
exceptions to this trend. Donald (1991) compared survival of jointed
goatgrass seeds left on the soil surface to burial 5 cm deep; after 1.5

3
Stymiest, C. E. and R. L. Anderson. 2001. p. 59, Annual Research Report: West River Agricultural
Center Crops and Soils Research. 1905 Plaza Boulevard, Rapid City, SD.
4
A sweep plow is comprised of V-shaped blades that sever weed roots with minimum soil
disturbance, tilling to a depth of 5 to 8 cm.
44 R. ANDERSON

Figure 4 Impact of seed depth in soil on longevity of green foxtail. Study conducted in the
semiarid prairies of Canada. (Adapted from Banting et al., 1973; Thomas et al., 1986).

years, 4% of seeds were alive if buried, whereas 7% of seeds survived on the

soil surface. Difference between jointed goatgrass and other weed species
may be attributed to the dispersal unit of jointed goatgrass, where seeds
enclosed within the spikelet are protected from environmental conditions on
the soil surface.
Because of this protective dispersal unit, we wondered if tilling to stimulate
germination would be favorable for seed bank depletion of jointed goatgrass, in
contrast with other weed species. Therefore, we compared seed survival of
jointed goatgrass spikelets buried in soil with a sweep plow to jointed
goatgrass spikelets left on the soil surface across five years (Anderson, 1998b).
During the first year, tillage increased seedling emergence two-fold compared
to seeds remaining on the soil surface. Yet, seedling emergence did not differ
between tillage and no-till during the next four years; tillage did not influence
long-term seed bank dynamics. Apparently, loss of jointed goatgrass seed
viability over winter compensated for increased germination by tillage during
the first year.
With green foxtail and downy brome, the studies at Wall and Pierre suggest
that tillage lessens impact of rotation design on population dynamics of weeds.
Yet, even with tillage, rotations comprised of two summer annual crops followed
by two winter annual crops (fallow, if used, would fit in either life cycle category)
lower weed community density. This rotation design accentuates the ecological
process of seed decline in soil yet avoids the exponential population growth of
weed species that occurs with several years of crops with similar life cycles.
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 45

IV. ENHANCING CROP COMPETITIVENESS

A second component of ecologically based weed management is improving a

crop’s ability to compete with and tolerate weeds. Cultural practices are devised
that help the crop capture resources such as nutrients or solar radiation before
weeds. To improve competitiveness of winter wheat, corn, sunflower, and proso
millet, we evaluated cultural practices such as N fertilizer placement, narrow row
spacing, higher plant densities, or delayed planting, either alone or in
combinations, to determine their impact on weed growth and crop yields.

A. WINTER WHEAT

Winter annual grasses are difficult to control in winter wheat because

herbicides are not available for in-crop control and winter wheat –fallow favors
population growth of these grasses. Because of these limitations, we focused our
research on strengthening the winter wheat canopy. A series of cultural systems,
comprised of various cultural practices such as banding N fertilizer with the seed,
higher seeding rates, and tall cultivars, were evaluated to determine if seed
production of winter annual grasses could be reduced (Anderson, 1997). We
compared these cultural systems to the common production system used by
producers [seeding rate of 45 kg/ha, semidwarf cultivars, and N fertilizer-applied
broadcast].
Seed production of both feral rye and jointed goatgrass were reduced 40 – 45%
by the most effective cultural system: seeding rate of 67 kg/ha, tall cultivar, and N
banded with the seed at planting (Fig. 5). When only one or two cultural practices
were used, seed production was reduced less than 20%; combining three practices
led to a synergistic impact on reducing weed growth. Similar results occurred
with downy brome in another study. Seed production of downy brome was
reduced more than 40% by strengthening winter wheat’s canopy with cultural
practices (Anderson, 1996).
Producers may accrue additional benefits for weed management with this
strategy, as strengthening the canopy of winter wheat can reduce weeds in
following crops. Wicks et al. (1986) observed that summer annual weed
density in corn varied among cultivars of the previous winter wheat crop.
They attributed this effect to two factors: more competitive canopies and
higher crop residue levels after harvest. Any cultural practice that improved
resource capture by winter wheat, such as increased seeding rate, narrow row
spacing, or nitrogen and phosphorus placement, reduced weed density and
seed production (Vander Vorst et al., 1983; Valenti and Wicks, 1992). Less
seed production by weeds in winter wheat led to lower weed density in corn
or sorghum.
46 R. ANDERSON

Figure 5 Seed production of jointed goatgrass and feral rye as affected by cultural practices in
winter wheat, Akron CO. Treatment means were compared to the conventional system of winter wheat
production: semidwarf variety, normal seeding rate, and N applied broadcast. Data averaged across
three years; bars with the same letter are not significantly different based on Fisher’s LSD (0.05).
Means for N placement did not differ from the conventional system. (Adapted from Anderson, 1997).

The second component involved with winter wheat’s suppression of future

weed density was residue production. Crutchfield et al. (1986) found that
suppression of weed emergence and establishment was related to crop residue
quantity on the soil surface. Weed density was reduced approximately 30% by
3500 kg/ha of crop residue and 60% by residue levels of 8000 kg/ha.
Producers can improve management of winter annual grasses in winter wheat
with cultural practices that help wheat capture resources before weeds. In
addition, this strategy may provide an ancillary benefit for weed management by
reducing the density of summer annual weeds in following crops such as corn or
sunflower.

B. CORN

Corn hectarage is rapidly increasing because of its favorable performance

when planted after winter wheat. But, corn is not competitive with weeds because
it is grown in wide rows (76 cm) and planted at low populations (30,000–
38,000 plants/ha). Consequently, current production practices rely on high
herbicide inputs to manage weeds. To improve weed management in corn, we
examined two approaches: (1) management tactics in winter wheat and during the
interval between wheat harvest and corn planting for impact on weed density in
corn, and (2) cultural strategies in corn to improve its competitiveness with weeds.
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 47

In our study examining cultural systems in winter wheat (Anderson, 1997), we

observed that the quantity of crop residue remaining after harvest varied among
cultural systems. Winter wheat grown with conventional practices left
approximately 4000 –4500 kg/ha of crop residue on the soil surface. In contrast,
winter wheat produced 6000– 6500 kg/ha of crop residue with a cultural system
comprised of a higher seeding rate, a taller cultivar, and N banding with the seed.
We speculated that weed suppression would be greater with higher residue levels
produced by cultural systems in winter wheat, based on the results of Crutchfield
et al. (1986) that related weed suppression to residue quantity.
A second area of interest with crop residue involved tillage. Producers may
replace herbicides with tillage as a management strategy to minimize herbicide
resistance. Because tillage increases weed emergence (Roberts, 1981), we
wondered if extra crop residue produced with cultural systems in winter wheat
could eliminate tillage-enhanced emergence of weeds. Therefore, we compared a
cultural system designed to strengthen winter wheat’s canopy to the conventional
system of winter wheat production for impact on weed density in corn planted the
next year. The cultural system was comprised of higher seeding rate, taller
cultivar, and N placed with the seed. We also compared tillage and no-till during
the interval after wheat harvest and before corn planting with both systems of
winter wheat production. We tilled with the sweep plow because it buries only
10% of crop residue on the soil surface with each operation, contrasting with 60 –
80% burial by a tandem disk harrow (Good and Smika, 1978). Plots were tilled
twice with the sweep plow before planting corn, whereas in no-till plots,
herbicides controlled weeds between winter wheat harvest and corn planting.
With no-till, the cultural system in winter wheat, with its high residue quantity,
reduced weed emergence 40% in corn compared to the conventional system
(Anderson, 1999a). However, our hypothesis that extra crop residue would mask
the effect of tillage on weed emergence was proven false. Not only did tillage
eliminate the residue effect on weed emergence, but surprisingly, tillage
increased weed density 30% compared to the no-till conventional winter wheat
with less residue. Burial of residue and weed seeds by tillage apparently altered
the weed seed-soil interaction such that weed emergence was increased
regardless of residue quantity on the soil surface.
Our study demonstrates that weed density in corn can be reduced with cultural
practices in winter wheat, whereas tillage eliminates this benefit. Because tillage
increases weed density, producers may have to increase inputs to control weeds in
corn with tilled systems. A further concern is that corn yields less in tilled
systems. An earlier study showed that tilling with the sweep plow before planting
reduced corn yields 30– 40% compared to a no-till system; yield loss was
attributed to less effective use of precipitation (Anderson, 1990). Using tillage as
a management tactic for herbicide resistance may not be favorable for semiarid
corn production, as tillage disrupts the ecological benefits of crop residue and no-
till in relation to weed dynamics.
48 R. ANDERSON

Our second goal with weed management in corn was to improve its
competitiveness with weeds. Common practices for growing corn in the Great
Plains include a target population of 37,000 plants/ha, row spacing of 76 cm, and
N fertilizer applied broadcast. To strengthen corn, we evaluated three cultural
practices: (1) banding N near the seed, (2) higher corn density (47,000 plants/ha),
and (3) narrow row spacing (38 cm) (Anderson, 2000b). All possible
combinations of these cultural practices were evaluated, with the study
established in a no-till system. The conventional system comprised of common
practices was included for comparison, and treatments were split into weed-free
and weed-infested subplots.
Grass weed (Setaria spp.) biomass was reduced 60% when three cultural
practices were combined in one system. In contrast, weed biomass was reduced
10% or less if only one practice was used, whereas combining any two practices
reduced biomass only 10% – 25%. As found with winter wheat, synergism among
cultural practices occurred; impact on weed biomass was six-fold greater when
three cultural practices were combined compared to any one practice alone.
Corn yield did not differ among treatments in weed-free conditions, but
cultural practices improved corn tolerance to weeds. Yield loss due to grass
interference was 43% with the conventional system, a three-fold difference
compared to 13% yield loss with the system comprised of three cultural practices
(Fig. 6). If only one or two cultural practices were used, yield loss due to weed
interference was still reduced, but not to the extent of combining all cultural
practices together.
Producers have a multitude of options to favor corn over weeds. First,
strengthening the canopies of both winter wheat and corn reduces weed density
and biomass in corn. Secondly, no-till systems improve weed control in contrast
to tilling during the interval between winter wheat harvest and corn planting.
Tillage favors weed emergence and increases weed density in corn. A further
consequence of tillage is lower corn yields.

C. SUNFLOWER

Sunflower hectarage is increasing in the Central Great Plains, and it is

commonly grown in a winter wheat –corn –sunflower – fallow rotation. However,
sunflower, a summer annual crop, has a life cycle similar to corn, thus similar
weeds infest both crops. Furthermore, like corn, sunflower is grown in wide rows;
consequently, sunflower is not competitive with weeds.
Yet, growing sunflower after corn still offers producers an ecological benefit
for weed management. Sunflower is normally planted in early June, 3– 4 weeks
later than corn. Planting sunflower later enables producers to reduce potential
weed density, as weed emergence declines during the summer. This trend was
shown in a long-term study that monitored weed emergence between April 1 and
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 49

Figure 6 Corn grain yield in weed-free and weed-infested conditions as affected by cultural
practice combinations. Conventional system was 37,000 plants/ha at a row spacing of 76 cm, with N
fertilizer broadcast at planting. Cultural practices were banding N near the seed, increasing crop
density to 47,000 plants/ha, and reducing row spacing to 38 cm. Data averaged across three years; bars
with the same letter are not significantly different based on Fisher’s LSD (0.05). (Adapted from
Anderson, 2000b).

September 1 (Anderson, 1994a). Averaged across seven years, weed emergence

showed two peaks, the first between April 25 and May 9, and the second between
May 23 and June 6 (Fig. 7). The first peak was comprised of cool season species,
whereas warm season species were dominant in the second peak. Emergence
after June 1 was only 30% of the total seasonal emergence; thus, planting
sunflower later reduces potential weed emergence in the crop 35 – 40% compared
to corn. Diversifying crops within a life cycle category, i.e., summer annual
crops, will moderate the tendency of weeds to proliferate when two crops of
similar life cycles are grown in sequence.
To further improve weed management, we evaluated cultural practices to
strengthen sunflower competitiveness (Anderson, 1999b; Tanaka and Anderson,
2000). Common production practices for sunflower in this region include plant
populations of 35,000 –39,000 plants/ha planted in rows spaced 76 cm apart, with
N fertilizer applied broadcast (referred to as the conventional system). To
improve canopy development, we grew sunflower in narrow rows (50 cm wide),
increased plant population (47,000 plants/ha), and banded N fertilizer adjacent to
the seed row (referred to as the cultural system). We compared these systems at
two planting dates of early June and two weeks later. Treatments were split into
weed-free and weed-infested subplots.
50 R. ANDERSON

Figure 7 Seedling emergence pattern of a weed community in the Central Great Plains. Data were
collected from two tillage treatments, no-till and tillage with a sweep plow, and were averaged across
seven years. The dotted line represents one standard deviation. (Adapted from Anderson, 1994a).

With early planting, weed biomass was reduced approximately 65% by the
cultural system, compared to the conventional system. When planting was
delayed two weeks, the cultural system reduced weed biomass 85% whereas
weed biomass was 50% less with the conventional system, compared to the
conventional system planted early. Later planting provided an additional two
weeks to control weeds.
The cultural system eliminated seed yield loss due to weed interference at both
planting dates (Fig. 8). In contrast, with the conventional system, weeds reduced
yield 24% at the early planting date but did not affect yield when planting was
delayed. Sunflower usually yields less when planted late; comparing the
conventional systems in weed-free conditions, later planting reduced yield 17%.
Surprisingly, delayed planting did not reduce yield with the cultural system. We
speculate that the cultural system improved growth efficiency of sunflower, thus
minimizing the detrimental effect of late planting.

D. PROSO MILLET

Proso millet, a summer annual grass, was historically grown as a catch crop to
replace hail-damaged wheat or in a winter wheat – proso millet – fallow rotation.
In recent years, proso millet is being included in a multitude of planned rotations,
generally following winter wheat, corn, or sunflower. As with sunflower, proso
millet is usually planted in early June, with optimum planting dates in no-till
systems being June 1 to June 10 (Anderson, 1994b).
Proso millet is competitive with weeds because it is planted in narrow rows
(20 –30 cm wide) and at high plant populations (2 million seeds/ha). In a study
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 51

Figure 8 Seed yield of sunflower grown in two production systems. The conventional system
(denoted by Conv.) was comprised of rows spaced 76 cm apart, crop population of 39,000 plants/ha,
and N fertilizer applied broadcast. The cultural system included row spacing of 50 cm, crop
population of 47,000 plants/ha, and N fertilizer applied in a band by the seed at planting. Planting
dates differed by two weeks. Data averaged across two years; bars with the same letter are not
significantly different based on Fisher’s LSD (0.05). (Adapted from Anderson, 1999b; Tanaka and
Anderson, 2000).

where weeds were allowed to grow for seven weeks after planting, weed biomass
in proso millet was three-fold less than in sunflower and 10-fold less than in corn
(Anderson, 1999a). Because proso millet is more competitive than corn and
sunflower, we hypothesized that production systems could be designed that
eliminate need for herbicides in proso millet. Not only would this tactic help
producers reduce input costs, but it may avoid herbicide injury to proso millet.
Lyon and Miller (1999) found that, in some years, herbicides used in proso millet
reduced yield 20% because of injury.
Therefore, we compared two systems of proso millet production, the
conventional system used by producers versus a cultural system designed to
counter ecological trends of weeds (Anderson, 2000a). The cultural system
strengthened proso millet’s canopy by increasing the seeding rate 50% and
banding N fertilizer near the seed; also, planting was delayed two weeks
compared to the conventional system. With the conventional system, N was
applied broadcast whereas proso millet was planted in early June.
We also evaluated two tillage treatments, no-till and tillage with the sweep
plow, on weed dynamics in each production system. The weed community was
primarily redroot pigweed (Amaranthus retroflexus L.). Each plot was split into
weed-free and weed-infested subplots.
52 R. ANDERSON

The cultural system eliminated yield loss due to redroot pigweed interference.
With tillage, weeds reduced grain yield 29% in the conventional system, but yield
loss was only 2% with the cultural system (Table I). In no-till, yield loss
decreased from 4% in the conventional system to 0% with the cultural system.
Yield differences reflect production system impact on redroot pigweed density
and biomass. With tillage, only 8 plants/m2 infested the cultural system
whereas 42 plants/m2 established in the conventional system, a five-fold increase
(Table I). Biomass was reduced nine-fold by the cultural system. Similar trends
occurred in no-till; redroot pigweed density and biomass were several-fold less
with the cultural system.
Grain yield in weed-free conditions did not differ between systems within a
tillage treatment. But, yield was 25% greater in no-till, which was attributed to
improved water relations.
Producers are concerned that without herbicides, seed production of isolated
weeds in the crop will rapidly increase weed density in future years. Weeds
produce a considerable number of seeds per plant (Anderson, 1998a). Yet,
research has shown that weed density in rotations remains relatively stable if
control efficacy of either weed seed (Firbank and Watkinson, 1986; Cousens et al.,
1987) or biomass production (Bosnic and Swanton, 1997) approaches 85 –90%.
In our study, biomass of redroot pigweed was reduced almost 90% with the
cultural system in both tillage treatments (Table I). Furthermore, seed production
per square meter by redroot pigweed was reduced 10- to 16-fold by the cultural

Table I
Effect of Production Systems on Redroot Pigweed Growth and Interference in Proso Millet.
Data Averaged Across Three Years. Tillage was Performed with a Sweep Plow. (Adapted from
Anderson, 2000a)

Redroot pigweed

Production Density Biomass Seed production Weed-infested proso

system (plants/m2) (g/m2) (seeds/m2) millet yield loss (%)

Tilled
Conventionala 42 ab 475 a 88,400 a 29 a
Cultural 8b 56 b 8,100 b 2 bc
No-till
Conventional 6b 72 b 9,700 b 4b
Cultural 1c 7c 600 c 0c
a
Conventional system was comprised of a seeding rate of 11 kg/ha, N broadcast, and early planting;
cultural system was proso millet planted at 17 kg/ha, N banded near the seed, and planting delayed two
weeks.
b
Means within columns followed by the same letter do not differ as determined with the Fisher’s LSD
(0.05).
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 53

system. With no-till, one plant in the cultural system produced 600 seeds; in
contrast, six plants in the conventional system produced 9700 seeds, or
approximately 1600 seeds/plant (Table I). A similar trend occur between systems
in the tilled treatment; seed production per square meter was reduced 10-fold with
the cultural system.
Our data suggest that ecologically based weed management will not lead to
population growth of weeds, as seed production is severely reduced with more
competitive proso millet. We suggest that a similar trend with weed seed
production may occur with cultural systems in sunflower, as weed biomass can be
reduced 85% (Anderson, 1999b; Tanaka and Anderson, 2000). Seed production
by weeds is related to its biomass; generally, larger plants produce more seeds
(Lindquist et al., 1995; Smith et al., 1999). With both proso millet and sunflower,
cultural systems may enable producers to eliminate herbicide use, consequently
reducing input costs as well as minimizing selection pressure for resistant weed
species.
A startling trend with our study was impact of tillage on weed dynamics.
Redroot pigweed density and biomass was at least six-fold greater with tillage
compared to no-till (Table I). Tilling with the sweep plow apparently placed
redroot pigweed seed in more favorable germination sites in soil. A second factor
of tillage was that proso millet yielded less with tillage. These trends were similar
to results found with corn where tillage increased weed density but decreased
grain yield.

V. SUMMARY

A. ROTATIONS BASED ON A CYCLE-OF- FOUR AID WEED

MANAGEMENT BUT ALSO ACCRUE OTHER BENEFITS

Cropping systems are changing in the Great Plains, with producers exploring a
multitude of rotations. Designing rotations in a cycle-of-four, with two-year
intervals of summer and winter annual crops (fallow serving in either life-cycle
category), will enable producers to integrate ecological principles related to weed
dynamics with crop choices and sequencing. The ecological approach improves
weed management by reducing weed community density and favoring the crop in
capturing resources, thus reducing need for herbicides. This approach will be
further improved if producers choose crops within a life-cycle category that have
different planting dates, such as corn and sunflower. A range of planting dates
provides an opportunity to disrupt population growth of weeds associated with a
life cycle.
Producers in the Central Great Plains who have adopted the cycle-of-four
rotation have found that their weed control costs declined 50% compared to
54 R. ANDERSON

rotations such as winter wheat – fallow or winter wheat – corn –proso millet.
Furthermore, this rotational design provides additional benefits that favor
agronomic and economic success in this semiarid region.

1. Crop Yield Benefits

Diversifying crops in rotations can improve crop yield, a response known as

the rotation effect. Cause of this effect has not been identified, but it most likely
represents a multitude of factors (Wright 1990). One contributing factor is plant
disease. Diversifying crops in a rotation reduces the frequency a crop is grown;
plant diseases proliferate if a crop is grown too frequently. Sunflower is
especially responsive to cropping frequency, yielding more if grown only once
every four years or less frequently (Anderson et al., 1999). Similarly, winter
wheat and corn yields are highest when grown only once every three years (Cook
and Veseth, 1991; Porter et al., 1997; Anderson, 1998a).

2. Improved Land Productivity and Economics

Land productivity in the Central Great Plains can be almost doubled with
rotations such as winter wheat –corn –proso millet –fallow, or winter wheat –
corn – sunflower – fallow compared to winter wheat –fallow (Anderson et al.,
1999). Not only is land yield improved, but net returns are increased. For
example, winter wheat – corn – proso millet –yielded 25% more net return than
winter wheat –fallow (Peterson et al., 1993b). Furthermore, diverse rotations
reduced financial risks for producers (Dhyuvetter et al., 1996). Similar trends
occur in the semiarid prairies of Canada; rotations comprised of four different
crops yielded more net return with less variability than rotations with fewer crops
(Zentner et al., 2002).

3. Increased Resource-Use-Efficiency

Winter wheat– fallow generally converts 40– 45% of precipitation into grain;
the remaining precipitation is lost either by evaporation or runoff (Farahani et al.,
1998). In contrast, diverse rotations such as winter wheat – corn –sunflower –
fallow or winter wheat – corn – proso millet – fallow convert 60 – 70% of
precipitation into grain (Peterson et al., 1996; Anderson 1998a). With continuous
cropping, precipitation use approaches 75% (Farahani et al., 1998). Thus, more
intensive rotations can improve a cropping system’s water-use-efficiency almost
two-fold.
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 55

Also, this design, by promoting crop diversity, can help producers in adjusting
to precipitation extremes prevalent in semiarid regions because crop diversity
improves water-use-efficiency of individual crops. With corn, Drury and Tan
(1995) found that not only was grain yield increased, but yield variability was
reduced two-fold in diverse rotations compared to continuous corn. They
attributed this response to improved root growth. Sahs and Lesoing (1985)
reported similar results: diverse rotations improved water-use-efficiency and
yield of corn in dry years.
More intensive rotations also improve nutrient-use-efficiency. Bowman and
Halvorson (1997) found that phosphorus-use-efficiency was greater in
cropping systems where fallow was less frequent. Phosphorus (P) is more
available for plant uptake if P remains in the organic phase, either in plant
biomass or with organic matter. During fallow, chemical reactions in soil
convert P into inorganic forms, whereas yearly contributions of plant biomass
in intensive cropping favor the organic phase of P.
An ancillary benefit of lower weed community density is improved herbicide
performance; both soil- and foliar-applied herbicides are more effective at lower
weed density (Winkle et al., 1981; Dieleman et al., 1999). A further benefit with
the ecologically based approach is that some crops can be grown without
herbicides. This will benefit producers from an efficiency perspective, as weed
infestations are variable in fields and often include areas with no weeds, whereas
crops can be injured by herbicides (Lyon and Miller, 1999). The ecologically
based approach may avoid unnecessary herbicide use and possible yield loss due
to herbicide injury.

4. Management of Herbicide Resistance

Crop diversity in rotations provides more options for producers to rotate

herbicides, thus reducing the frequency that herbicides with one mode of
action are used. Impact of this strategy has been demonstrated with wild oat
(Avena fatua L.) that has developed resistance to herbicides that inhibit the
ACCase enzyme. Cavan et al. (2001), using a simulation model based on
wild oat genetics and control efficacy, estimated that resistant biotypes of
wild oat would appear in 13 years if an ACCase-inhibiting herbicide was
used every year for control (Fig. 9). In contrast, if this class of herbicide was
applied just once every three years, resistance would not appear until after
126 years, a 10-fold difference. Resistant wild oat would appear in 26 years
if the herbicide was applied every other year. Because of the relationship
between resistance and frequency of use, producers are modifying their
cropping systems to ensure that these herbicides are applied only once every
three years in the same field (Morrison and Bourgeois, 1995).
56 R. ANDERSON

Figure 9 Time for resistance in wild oat to appear, as affected by frequency of herbicide
application across years. Simulation model based on agronomic and genetic characteristics of
resistance in wild oat. Terms on the x axis refer to applications across years, i.e., 1 in 2 represents 1
herbicide application in 2 years. (Adapted from Cavan et al., 2001).

The ecologically based approach further supports resistance management by

eliminating herbicide use in some crops. Consequently, selection pressure by
herbicides is not imposed on the weed community in that crop, which will limit
population growth of resistant weeds (Holt and LeBaron, 1990).

5. Opportunity for Great Plains Producers

Diverse rotations, especially if based on the cycle-of-four design, enable

producers to accrue a multitude of benefits, both for weed management and crop
production efficiency. A perplexing aspect of rotations, however, is that
producers often fail to accrue its benefits. An example of this anomaly was
reported by Bourgeois et al. (1997), who surveyed producers in Manitoba,
Canada, about strategies to manage herbicide resistance. They found that
producers recognized diverse rotations as a key strategy in managing herbicide
resistance; yet, surprisingly, rotations were also viewed as the major impediment
to adopting management strategies. Producers preferred the convenience and
habit of their current rotations, which, in some cases, precluded options for
rotating herbicides. Their approach to resistance management with rotations was
curative, not preventive. In other words, producers changed their rotations after
resistant weeds appeared in their fields. Powles et al. (1997) found a similar
perception among Australia producers, where changes in rotations were driven
more by production obstacles than ecological considerations.
Because producers in the Central Great Plains are actively seeking new
rotations, an opportunity exists for the agricultural community of this region to
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 57

develop production systems guided by ecological principles. With appropriate

crop sequencing, producers can integrate ecological processes with other
considerations involved in rotation design. The numerous benefits of diverse
rotations will help producers not only improve weed management, but also
strengthen the region economically.

B. SEMIARID PARADOX WITH TILLAGE AND HERBICIDES

No-till systems have changed cropping practices in the Central Great Plains
because of its beneficial impact on water relations (Peterson et al., 1996) and soil
health (Bowman et al., 1999). Peterson et al., (1993) suggested that no-till
systems have initiated a spiral of soil regeneration in this region, where
interactions among more favorable water relations, residue production, and crop
yield are continually improving soil health and subsequently, future crop
performance.
However, weed resistance is threatening the no-till system in the Central Great
Plains. Currently, resistant weeds exist for most herbicides used in the region
(Lyon et al., 1996). A second concern is that no-till systems enhance
development of weed resistance because weed seeds do not survive as long in
the seed bank as with tilled systems. This characteristic of no-till minimizes the
buffering effect of the seed bank on population dynamics (Gressel, 1993; Cavan
et al., 2000). A third concern with no-till rotations is that extensive intervals
occur when crops are not grown, yet weeds can emerge in any month of the year.
Producers, guided by economic, control efficacy, and crop flexibility reasons, rely
extensively on glyphosate to control weeds during these non-crop intervals. Yet,
even with diverse rotations, glyphosate is applied several times a year and
producers are now faced with glyphosate-tolerant weeds.5
One suggested strategy for managing herbicide resistance is tillage (Retzinger
and Mallory-Smith, 1997). Producers in this region are considering tillage again
for two reasons. First, input costs escalate rapidly with herbicide alternatives to
glyphosate, whereas tillage is perceived as economically favorable. Second,
tillage was the standard practice to control weeds before no-till, therefore
producers have a historical precedence to till.
Yet, tillage is detrimental to semiarid crop production. Because tillage
reduces efficiency of water use, crop yield is less in tilled systems (Anderson,
1990; 2000a). Tillage also disrupts the spiral of soil regeneration associated
with no-till systems (Peterson et al., 1993). Furthermore, tillage increases
weed density in crops thus minimizing impact of rotations on weed community
density.

5
Gail Wicks, personal communications. 2002. North Platte NE.
58 R. ANDERSON

Producers are struggling with this paradox: Tools (herbicides) needed for
more intensive cropping are being rendered ineffective by resistant weeds, yet
intensive cropping is less feasible in tilled systems. For weed management in
crops, the ecologically based approach expands producer options to control
weeds and avoid herbicide resistance. However, producers need additional
options to reduce use of glyphosate during fallow, yet still avoid tillage.
Because crops can be effective in suppressing weeds, one possibility is green
fallow, where a crop is grown only for vegetative growth before being killed. For
example, sweetclover (Melilotus officinalis Lam.) grown over winter reduced
weed density 75– 97% during fallow, compared to conventional fallow (Black-
shaw et al., 2001). But in the Central Great Plains, crop growth required for that
level of weed suppression reduced winter wheat yield because of excessive water
use (Schlegel and Havlin, 1997). Yet, Tanaka et al. (1997) found that wheat yields
were not affected if green fallow was maintained for only six to eight weeks.
If green fallow could suppress weeds for six weeks, producers may be able to
vary timing of selection pressure by glyphosate on the weed community. The
interval of green fallow suppression could be related to emergence periods of
selected weeds that are developing either tolerance or resistance. If successful,
green fallow would be another tool for producers to control weeds during fallow.
Producers also would accrue ancillary benefits with green fallow, such as
increased nutrient cycling, improved grain yields, and soil protection from
erosion (Biederbeck and Bouman, 1994).
Herbicide resistance is now a major concern in semiarid production. A key
issue facing producers and scientists is devising production systems that do not
favor herbicide-resistant weeds. Otherwise, producers may return to tillage-based
systems that are not as effective as no-till from an economic, agronomic, or
environmental perspective.

REFERENCES

Anderson, R. L. (1990). Tolerance of safflower, corn, and proso millet to clomazone. Weed Technol. 4,
606–611.
Anderson, R. L. (1994a). Characterizing weed community seedling emergence for a semiarid site in
Colorado. Weed Technol. 8, 245–249.
Anderson, R. L. (1994b). Planting date effect on no-till proso millet. J. Prod. Agric. 7, 454–458.
Anderson, R. L. (1996). Enhancing winter wheat tolerance to downy brome. West. Soc. Weed Sci. Res.
Rpt. p. 104.
Anderson, R. L. (1997). Cultural systems can reduce reproductive potential of winter annual grasses.
Weed Technol. 11, 608–613.
Anderson, R. L. (1998a). Designing rotations for a semiarid region. pp. 4–15. In Proc. 10th Annual
Meeting, Colorado Conservation Tillage Association, Sterling, CO. Colorado Conservation
Tillage Association, 40335 County Road GG, Akron CO.
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 59

Anderson, R. L. (1998b). Seedling emergence of winter annual grasses as affected by limited tillage
and crop canopy. Weed Technol. 12, 262–267.
Anderson, R. L., Tanaka, D. L., Black, A. L., and Schweizer, E. E. (1998). Weed community and
species response to crop rotation, tillage, and nitrogen fertility. Weed Technol. 12, 531 –536.
Anderson, R. L. (1999a). Cultural strategies reduce weed densities in summer annual crops. Weed
Technol. 13, 314–319.
Anderson, R.L. (1999b). Improving weed control in corn and sunflowers with narrow rows. pp. 88-97.
In Proc. 11th Annual Meeting, Colorado Conservation Tillage Association, Sterling, CO. Colo.
Conserv. Till. Assoc., 40335 County Road GG, Akron CO.
Anderson, R. L. (2000a). A cultural systems approach eliminates the need for herbicides in semiarid
proso millet. Weed Technol. 14, 602–607.
Anderson, R. L. (2000b). Cultural systems to aid weed management in semiarid corn (Zea mays).
Weed Technol. 14, 630 –634.
Anderson, R. L., Bowman, R. A., Nielsen, D. C., Vigil, M. F., Aiken, R. M., and Benjamin, J. G.
(1999). Alternative crop rotations for the Central Great Plains. J. Prod. Agric. 12, 95– 99.
Banting, J. D., Molberg, E. S., and Gephardt, J. P. (1973). Seasonal emergence and persistance of
green foxtail. Can. J. Plant Sci. 53, 369 –376.
Beckie, H. J., Thomas, A. G., Legere, A., Kelner, D. J., van Acker, R. C., and Meers, S. (1999). Nature,
occurrence, and cost of herbicide-resistant wild oat (Avena fatua) in small-grain production area.
Weed Technol. 13, 612 –625.
Biederbeck, V. O., and Bouman, O. T. (1994). Water use by annual green manure legumes in dryland
cropping systems. Agron. J. 86, 543– 549.
Blackshaw, R. E., Moyer, J. R., Doran, R. C., and Boswell, A. L. (2001). Yellow sweetclover, green
manure, and its residues effectively suppress weeds during fallow. Weed Sci. 49, 406 –413.
Bosnic, A. C., and Swanton, C. J. (1997). Economic decision rules for postemergence herbicide
control of barnyardgrass (Echinochloa crus-galli) in corn (Zea mays). Weed Sci. 45, 557 –563.
Bourgeois, L., Morrison, I. N., and Kelner, D. (1997). Field and producer survey of ACCase resistant
wild oat in Manitoba. Can. J. Plant Sci. 77, 709–715.
Bowman, R. A., and Halvorson, A. D. (1997). Crop rotation and tillage effects on phosphorus
distribution in the Central Great Plains. Soil Sci. Soc. Am. J. 61, 1418–1422.
Bowman, R. A., Vigil, M. F., Nielsen, D. C., and Anderson, R. L. (1999). Soil organic matter changes
in intensively cropped dryland systems. Soil Sci. Soc. Am. J. 63, 186–191.
Cavan, G., Cussans, J., and Moss, S. (2001). Managing the risks of herbicide resistance in wild oat.
Weed Sci. 49, 236–240.
Challaiah, Burnside, O. C., Wicks, G. A., and Johnson, V. A. (1986). Competition between winter wheat
(Triticum aestivum) cultivars and downy brome (Bromus tectorum). Weed Sci. 34, 689 –693.
Cook, R. J., and Veseth, R. J. (1991). In “Wheat health management”, 152p. American
Phytopathological Society Press, St. Paul MN.
Cousens, R., Moss, S. R., Cussans, G. W., and Wilson, B. J. (1987). Modeling weed populations in
cereals. Reviews of Weed Science 3, 93–112. Weed Science Society of America, Lawrence, KS.
Crutchfield, D. A., Wicks, G. A., and Burnside, O. C. (1986). Effect of winter wheat (Triticum
aestivum) straw mulch level on weed control. Weed Sci. 34, 110–114.
Dawson, J. H., and Burns, V. F. (1975). Longevity of barnyardgrass, green foxtail, and yellow foxtail
seeds in soil. Weed Sci. 23, 437–440.
Dieleman, J. A., Mortensen, D. A., and Martin, A. R. (1999). Influence of velvetleaf (Abutilon
theophrasti) and common sunflower (Helianthus annuus) density variation on weed management
outcomes. Weed Sci. 47, 81–87.
Dhuyvetter, K. C., Thompson, C. R., Norwood, C. A., and Halvorson, A. D. (1996). Economics of
dryland cropping systems in the Great Plains: a review. J. Prod. Agric. 9, 212 –216.
Donald, W. W. (1991). Seed survival, germination ability, and emergence of jointed goatgrass
(Aegilops cylindrica). Weed Sci. 39, 210–216.
60 R. ANDERSON

Donald, W. W., and Ogg, A. G. (1991). Biology and control of jointed goatgrass (Aegilops cylindrica),
a review. Weed Technol. 5, 3 –17.
Donald, W. W., and Zimdahl, R. L. (1987). Persistence, germinability, and distribution of jointed
goatgrass (Aegilops cylindrica) seed in soil. Weed Sci. 35, 149 –154.
Drury, C. F., and Tan, C. S. (1995). Long-term (35 years) effects of fertilization, rotation, and weather
on corn yields. Can. J. Plant Sci. 75, 355 –362.
Egley, G. H. (1986). Stimulation of weed seed germination in soil. Reviews of Weed Science 2, 67 –89.
Weed Science Society of America, Lawrence, KS.
Egley, G. H., and Williams, R. D. (1990). Decline of weed seeds and seedling emergence over five
years as affected by soil disturbance. Weed Sci. 38, 504– 510.
Farahani, H. J., Peterson, G. A., and Westfall, D. G. (1998). Dryland cropping intensification:
a fundamental solution to efficient use of precipitation. In “Advances in Agronomy”. 64,
pp. 197 –223. Academic Press, New York.
Firbank, L. G., and Watkinson, A. R. (1986). Modelling the population dynamics of an arable weed
and its effects upon crop yield. J. Appl. Ecol. 23, 147–159.
Froud-Williams, R. J. (1988). Changes in weed flora with different tillage and agronomic management
systems. In “Weed management in agroecosystems: ecological approaches”. (M. A. Altieri and
M. Liebman, Eds.), pp. 213–236. CRC Press Inc., Boca Raton, FL.
Froud-Williams, R. J., Chancellor, R. J., and Drennan, D. S. H. (1984). The effects of seed burial and
soil disturbance on emergence and survival of arable weeds in relation to minimal cultivation.
J. Appl. Ecol. 21, 629–641.
Good, L. G., and Smika, D. E. (1978). Chemical fallow for soil and water conservation in the Great
Plains. J. Soil Water Conserv. 33, 89–90.
Greb, B. W., Smika, D. E., Woodruff, N. P., and Whitfield, C. J. (1974). In “Summer fallow in the
western United States”. USDA-ARS Conserv. Res. Report no. 17., pp. 51 –85.
Gressel, J. (1993). Management of herbicide-resistant weeds in crop production. In “International
Crop Science”. (D. R. Buxton, R. Shibles, R. A. Fosberg, B. L. Blad, K. H. Asay, G. M. Paulsen
and R. F. Wilson, Eds.), 1, pp. 121 –127. American Society of Agronomy, Madison WI.
Harper, J. L. (1977). Population biology of plants. In “The seed bank”, pp. 83–110. Academic Press,
New York.
Heap, I. (2002). The International Survey of Herbicide Resistant Weeds. Online. Internet. June 1,
2002. Available www.weedscience.com.
Holt, J. S., and LeBaron, H. M. (1990). Significance and distribution of herbicide resistance. Weed
Technol. 4, 141– 149.
Kishmore, G. M., Padgette, S. R., and Fraley, R. T. (1992). History of herbicide-tolerant crops:
methods of development and current state of the art—emphasis on glyphosate tolerance. Weed
Technol. 6, 626– 634.
Koscelny, J. A., Peeper, T. F., Solie, J. B., and Solomon, S. G., Jr (1991). Seeding date, seeding rate,
and row spacing affect wheat (Triticum aestivum) and cheat (Bromus secalinus). Weed Technol.
5, 707– 712.
Leighty, C. E. (1938). Crop Rotation. In “U.S. Department of Agriculture Yearbook of Agriculture:
Soils and Men”, pp. 406–430. U.S. Government. Printing Office, Washington, DC.
Lindquist, J. L., Maxwell, B. D., Buhler, D. D., and Gunsolus, J. L. (1995). Velvetleaf (Abutilon
theophrasti) recruitment, survival, seed production, and interference in soybean (Glycine max).
Weed Sci. 43, 226 –232.
Lyon, D. J., and Baltensperger, D. D. (1995). Cropping systems control winter annual grass weeds in
winter wheat. J. Prod. Agric. 8, 535–539.
Lyon, D. J., Miller, S. D., and Wicks, G. A. (1996). The future of herbicides in weed control systems of
the Great Plains. J. Prod. Agric. 9, 209–215.
Lyon, D. J., and Miller, S. D. (1999). Herbicide injury in proso and foxtail millets. Proc., Western Soc.
Weed Sci. Annual Meeting 52, 24.
 ECOLOGICAL WEED MANAGEMENT IN THE GREAT PLAINS 61

Mesbah, A. O., and Miller, S. D. (1999). Fertilizer placement affects joined goatgrass (Aegilops
cylindrica) competition in winter wheat (Triticum aestivum). Weed Technol. 13, 374– 377.
Mohler, C. L. (1993). A model of the effects of tillage on emergence of weed seedlings. Ecol. Applic.
3, 53–73.
Morrison, I. N., and Bourgeois, L. (1995). Approaches to managing ACCase inhibitor resistance in
wild oat on the Canadian prairies. In “Proceedings of Brighton Crop Protection Conference”.
Brighton UK, 2, 567–576.
Mortimer, A. M. (1984). Perspective on plant population ecology. In “Population ecology and weed
science”. (R. Dirzo and J. Sarukhan, Eds.), pp. 363–388. Sinauer Associates Inc., Sunderland,
MS.
Pedigo, L. P. (1995). Closing the gap between IPM theory and practice. J. Agric. Entomol. 12, 171–181.
Peterson, G. A., Westfall, D. G., and Cole, C. V. (1993a). Agroecosystem approach to soil and crop
management research. Soil Sci. Soc. Am. J. 57, 1354–1360.
Peterson, G. A., Westfall, D. G., Toman, N. E., and Anderson, R. L. (1993b). Sustainable dryland
cropping systems: Economic analysis. In “Agric. Exp. Stn. Tech. Bull. TB93-3”. Colo. St. Univ.
Peterson, G. A., Schlegel, A. J., Tanaka, D. L., and Jones, O. R. (1996). Precipitation use efficiency as
affected by cropping and tillage systems. J. Prod. Agric. 9, 180–186.
Porter, P. M., Crookston, R. K., Ford, J. H., Huggins, D. R., and Lueschen, W. E. (1997). Interrupting
yield depression in monoculture corn: comparative effectiveness of grasses and dicots. Agron. J.
89, 247–250.
Powles, S. B., Preston, C., Bryan, I. B., and Justin, A. R. (1997). Herbicide resistance: impact and
management. In “Advances in Agronomy”. 58, pp. 57-93. Academic Press, Inc. New York.
Powles, S. B., Lorraine-Colwill, D. F., Dellow, J. J., and Preston, C. (1998). Evolved resistance to
glyphosate in rigid ryegrass (Lolium rigidum) in Australia. Weed Sci. 46, 604 –607.
Retzinger, E. J., Jr, and Mallory-Smith, C. A. (1997). Classification of herbicides by site of action for
weed resistance management strategies. Weed Technol. 11, 384–393.
Roberts, H. A. (1981). Seed banks in soils. Adv. Appl. Biol. 6, 1–55.
Sagar, G. R., and Mortimer, A. M. (1976). An approach to the study of the population dynamics of
plants with special reference to weeds. Adv. Appl. Biol. 1, 1 –47.
Sahs, W. W., and Lesoing, G. (1985). Crop rotations and manure versus agricultural chemicals in
dryland grain production. J. Soil & Water Conserv. 40, 511 –516.
Schlegel, A. J., and Havlin, J. L. (1997). Green fallow for the Central Great Plains. Agron. J. 89,
762 –767.
Smith, G. L., Freckleton, R. P., Firbank, L. G., and Watkinson, A. R. (1999). The population dynamics
of Anisantha sterilis in winter wheat: comparative demography and the role of management.
J. Appl. Ecol. 36, 455 –471.
Streibig, J. C. (1979). Numerical methods illustrating the phytosociology of crops in relation to weed
flora. J. Appl. Ecol. 16, 577–587.
Tanaka, D.L., Anderson, R.L., (2000). Integrated approach to weed control in sunflowers. pp. 54–61,
in Proc., 22nd National Sunflower Association Research Workshop, Fargo, ND. Nat. Sunflower
Assoc., 4023 State Street, Bismarck ND.
Tanaka, D. L., Bauer, A., and Black, A. L. (1997). Annual legume cover crops in spring wheat–fallow
systems. J. Prod. Agric. 10, 251 –255.
Thomas, A. G., Banting, J. D., and Bowes, G. (1986). Longevity of green foxtail seeds in a Canadian
prairie soil. Can. J. Plant Sci. 66, 189–192.
Thomas, A. G., and Frick, B. L. (1993). Influence of tillage systems on weed abundance in
southwestern Ontario. Weed Technol. 7, 699 –705.
Valenti, S. A., and Wicks, G. A. (1992). Influence of nitrogen rates and wheat (Triticum aestivum)
cultivars on weed control. Weed Sci. 40, 115–121.
Vander Vorst, P. B., Wicks, G. A., and Burnside, O. C. (1983). Weed control in a winter wheat– corn–
ecofarming rotation. Agron. J. 75, 507–511.
62 R. ANDERSON

Westfall, D. G., Havlin, J. L., Hergert, G. W., and Raun, W. R. (1996). Nitrogen management in
dryland cropping systems. J. Prod. Agric. 9, 192–199.
Wicks, G. A. (1984). Integrated systems for control and management of downy brome (Bromus
tectorum) in cropland. Weed Sci., 32(Suppl. 1), 26–31.
Wicks, G. A., Ramsel, R. E., Nordquist, P. T., Schmidt, J. W., and Challaiah. (1986). Impact of wheat
cultivars on establishment and suppression of summer annual weeds. Agron. J. 78, 59–62.
Wicks, G. A., Felton, W. L., Murison, R. D., and Martin, R. J. (2000). Changes in fallow weed species
in continuous wheat in northern New South Wales, 1981-1990. Australian J. Exp. Agric. 40,
831–842.
Winkle, M. E., Leavitt, J. R. C., and Burnside, O. C. (1981). Effects of weed density on herbicide
absorption and bioactivity. Weed Sci. 29, 405–409.
Wright, A. T. (1990). Yield effect of pulses on subsequent cereal crops in the northern plains. Can.
J. Plant Sci. 70, 1023–1032.
Zentner, R. P., Wall, D. D., Nagy, C. N., et al. (2002). Economics of crop diversification and soil
tillage opportunities in the Canadian prairies. Agron. J. 94, 216–230.
 SCALE-DEPENDENT DISPERSION IN
SOILS: AN OVERVIEW
L. Zhou and H. M. Selim
Agronomy Department
Louisiana State University Agricultural Center
Baton Rouge, Louisiana 70803 USA

I. Introduction
A. Justification
B. Laboratory Studies
C. Field Tests
D. Interpretations
E. Outline
II. Stochastic Approaches
A. Macrodispersivity Theory
B. Fractal Models
C. Transfer Function Models
III. Analytical Solutions of CDE with Scale-dependent Dispersivity
A. Solutions of CDE with Time-dependent Dispersivity
B. Solutions of CDE with Distance-dependent Dispersivity
IV. Numerical Approaches
V. Fractional Order Advection – dispersion Equation (FADE)
A. Derivation of FADE
B. Solutions
C. Applications
VI. Concluding Remarks
References

I. INTRODUCTION

Accurate description of the transport of contaminants in porous media is very

important to the evaluation and remediation of contamination in soils and
aquifers. The advection – dispersion equation (ADE) is the most widely used
governing equation in describing solute transport in hydrogeologic systems.
Advection represents the transport of chemicals with the bulk fluid movement or
mass flow. The term dispersion in ADE is a simplification for hydrodynamic
223
Advances in Agronomy, Volume 80
Copyright q 2003 by Academic Press. All rights of reproduction in any form reserved
0065-2113/03$35.00
224 L. ZHOU AND H. M. SELIM

dispersion. The hydrodynamic dispersion includes the processes of mechanical

dispersion and molecular diffusion. The mechanical dispersion is caused by
variation in fluid velocity in the pore space of porous media, and diffusion is
caused by molecules’ random motion. Dispersivity is commonly used to
characterize dispersion behavior of solutes in porous media.

A. JUSTIFICATION

Early studies led to the belief that a single value of the dispersivity parameter
(a) for an entire medium is sufficient to characterize the spreading processes of
tracer solutes in porous media (Bear, 1972). However, numerous studies have
shown that the dispersivity measured in the laboratory often fails to give adequate
description of transport behavior at the field scale. Dispersivities estimated from
field observations are often much larger than those measured in the laboratory for
the same type of porous material (Pickens and Grisak, 1981a). On the other hand,
numerous studies suggest that the dispersivity depends on the mean travel distance
and/or the scale of the observations (Peaudecerf and Sauty, 1978; Sudicky and
Cherry, 1979; Pickens and Grisak, 1981a). These results suggest that a unique
dispersivity for the entire medium is not adequate to describe solute dispersion in
hydrogeologic systems at the field scale. The scale dependency of the dispersion
is an added complexity in characterizing solute transport in the field.
Scale dependency or scale effect of solute dispersion in porous media has been
recognized for more than four decades. Pickens and Grisak (1981a) gave a
detailed review of scale effects based on field dispersion investigations. They
summarized results from computer simulation, laboratory studies, and field tracer
tests. Their values of longitudinal dispersivities obtained from computer
modeling ranged from 12 to 61 m. There existed some trend that larger
dispersivities were associated with larger contamination zones. The longitudinal
dispersivities obtained from the analysis of laboratory breakthrough data on
repacked granular materials were of the order of 0.01 – 1 cm. In contrast to
computer simulations and laboratory data, the longitudinal dispersivities
obtained from analysis of various types of field tracer tests ranged between
0.012 and 15.2 m (see Table 2 in Pickens and Grisak, 1981a). Other scholars also
observed the scale effect in solute dispersion. Fried (1972) reported longitudinal
dispersivities from several sites. Estimated values of dispersivities were
0.1 – 0.6 m for the local scale (aquifer stratum), 5 – 11 m for the global scale
(aquifer thickness), and 12.2 m for the regional scale (several kilometers). Later,
Fried (1975) redefined these scales in terms of the mean travel distance of a tracer
solute or a contaminant. Mean travel distance of 2 –4 m was assigned for the local
scale, 4– 20 m for the global scale 1, 20– 100 m for the global scale 2, and greater
than 100 m (usually several kilometers) for the regional scale. Later, Gelhar et al.
(1992) presented a critical review of dispersivity observations from 59 different
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 225

field sites and found that longitudinal dispersivities ranged from 1022 to 104 m
for scales of observation from 1021 to 105 m. At a given scale, the longitudinal
dispersivity values were found to range over 2– 3 orders of magnitude (see Fig. 1
in Gelhar et al., 1992). Overall, the data indicated a trend of a systematic increase
of the longitudinal dispersivity with observation scale.

B. LABORATORY STUDIES

A number of case studies have also been conducted in the laboratory to

investigate the scale effect on dispersivity for various observation scales. Khan
and Jury (1990) conducted a series of tracer transport experiments on undisturbed
and repacked soil columns having three different lengths under three different
flow rates. The shorter columns were obtained by cutting longer ones in halves.
They found that the dispersivity significantly increased with increasing column
length for the undisturbed soil columns at the higher flux rates. However, the
dispersivity was length-independent for the repacked columns. Porro et al. (1993)
investigated solute transport through two large soil columns. One column was
uniformly filled with loamy fine sand and the other filled with alternating 20 cm-
thick layers of loamy fine sand and silty clay loam. Various tracers such as
tritiated water, bromide and chloride were applied under steady flow conditions
to the tops of soil columns. Their results displayed no clear relationship between
the dispersion coefficients and depth for any of the tracers for either column. In a
different column study, Zhang et al. (1994) packed a 12.5 m soil column in

Figure 1 Breakthrough curves at the point x=A1 ¼ 12 for the classical convection– dispersion
model (solid line) and the extended model with Ct ¼ C rsj 0 ð2pÞ1=2 =m (dashed line). After Gelhar
et al. (1979). Copyright q 1979 by the American Geophysical Union. Reproduced with permission of
AGU.
226 L. ZHOU AND H. M. SELIM

different ways, i.e., homogeneous and heterogeneous packings. The observed

dispersivity increased with distance for both homogeneous and heterogeneous
columns.
In an effort to find laboratory evidence of the scale effect in dispersion,
Silliman and Simpson (1987) conducted transport experiments in a sandbox. The
sandbox was filled in four different arrangements: uniform sand, layered packing,
layered-block packing, and uniform heterogeneity. For uniform sand packing,
20-mesh sand (d50 ¼ 0:70 mm) was packed uniformly. In layered packing, a
horizontal layer of 90-mesh sand (d50 ¼ 0:16 mm) was embedded in the 20-mesh
sand. In layered-block packing, “blocks” of 90-mesh sand were uniformly
distributed in a horizontal layer occupying one-third of the vertical height of the
sand. In uniform heterogeneity packing, blocks of 90-mesh sand were uniformly
distributed throughout the sandbox. No scale effect was found for the uniform
sand packing. Analysis of the breakthrough data from the layered packing
suggested a three-stage transport process. Breakthrough data from the layered-
block packing showed a three-stage process, too. However, for layered-block
packing, the duration of the central segment of breakthrough curves showed
dependence on the distance from source indicating scale effects for this packing.
Analysis of the breakthrough curves obtained in one run for the uniform
heterogeneity packing showed that the dispersivity increased with distance. In a
two-dimensional sandbox study, Burns (1996) estimated longitudinal dispersiv-
ities at different distances. He found that the dispersivity increased slightly with
distance. To verify analytical solutions for a one-dimensional scale-dependent
dispersion model, Pang and Hunt (2001) conducted transport experiments in an
8 m-long homogeneous pea-gravel column. The apparent dispersivities obtained
at different distances showed a clear increase with distance. Other laboratory
studies are available in the literature in support of the general finding that the
dispersivity is found to be scale dependent in some studies but not in others.

C. FIELD TESTS

Though it is more difficult to monitor solute transport at the field scale than at
laboratory scale, numerous field transport experiments have also been conducted
to study the scale-dependent dispersion. For example, based on analyses of a
series of field tests, Sauty (1980) found that dispersivity at first increased with
distance until a characteristic value was reached, which then remained constant.
Butters (1987) conducted field transport experiments in an unsaturated soil to
study the transport of bromide. Based on the analyses of the data, Butters and Jury
(1989) claimed that a nearly linear scale effect in dispersivity to a depth of at least
14.8 m was displayed. Leland and Hillel (1982) conducted a field study of solute
dispersion in an unconfined aquifer. They installed two lines of samplers. The
results from both sampler lines were inconsistent. One line showed dispersivity
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 227

increasing with distance up to the fourth sampler in the line, while the other line
showed dispersivity tending to decrease with distance from the injection plane.
To investigate scale-dependent dispersivity, Taylor and Howard (1987)
conducted a 40 m natural gradient tracer test in an unconfined sand aquifer. The
studied area was covered by an extensive array of monitoring wells. The obtained
dispersivity values were found to be randomly distributed at each monitoring
line. Overall, the field data did not indicate a clear trend of increase in dispersivity
with distance from the injection well. Recently, Himmelsbach et al. (1998)
studied solute transport processes in a highly permeable fault zone of Lindau
fractured rock test site in Germany. For a distance of 11.2 m, they obtained
dispersivities ranging between 0.17 and 0.3 m. However, the dispersivity based
on a different experimental run was 2.1 m for a distance of 21.4 m. They did not
expect that a sudden increase in dispersivity could be explained by the increase of
distance from 11.2 to 21.4 m. In summary, results from these field studies
indicate inconsistent relationship between the dispersivity and distance.
In addition to the dependence of dispersivity on distance from source of a
solute or contaminant, the dispersivity is often found to be dependent on the time
after the release of solutes. Dieulin (1980) described the results of a tracer test
conducted in a heterogeneous alluvial aquifer and found it necessary to divide the
breakthrough curves measured from each observation well into a series of
segments. Each portion was then simulated individually and a different
dispersivity value determined. The results led to a conclusion that the dispersivity
increased with time (Dieulin, 1980; cited in Sudicky et al., 1983). A similar
conclusion was reached by Freyberg (1986). In an analysis of solute transport data
at the Borden site, Freyberg (1986) found a non-linear increase of the variance of
travel distance with time, which indicated a time-dependent dispersivity. He also
found that the dispersivity reached an asymptotic value of 0.49 m.

D. INTERPRETATIONS

Several efforts have been made to account for the dependence of dispersivity
on time or distance. Most of the theories rely on the heterogeneous nature of the
porous media. Stochastic theories (Gelhar et al., 1979; Dagan, 1982, 1984;
Neuman et al., 1987 among others) relate macrodispersivity with the spatial
correlation structure of the hydraulic conductivity of the porous media.
According to stochastic models, macrodispersivity is a function of time initially
and will reach asymptotic constant value after all scales of variability have been
experienced. Wheatcraft and Tyler (1988) proposed a fractal model to explain the
scale effect of dispersivity. Their model showed that dispersivity increased
proportionally with mean travel distance to the power of D 2 1; where D is
the fractal dimension of streamtubes. Other researchers argued that some part
of the scale effect might be explained as an artifact of the models used (Domenico
228 L. ZHOU AND H. M. SELIM

and Robbins, 1984). Based on synthetic data, Domenico and Robbins (1984)
showed that scaling-up of dispersivity would occur whenever a ðn 2 1Þ-
dimensional model was calibrated or otherwise employed to describe an n-
dimensional system. Similarly, Pickens and Grisak (1981a) have attributed the
scaling-up of apparent dispersion coefficients to the use of a one-dimensional
model to calibrate a two-dimensional flow field in which the convective velocity
varied with distance along the radial direction. Serrano (1992) attempted to
describe field scale solute transport parameters in terms of regional hydrologic
and aquifer hydraulic properties. An equation of dispersion in one- and two-
dimensional homogeneous and heterogeneous aquifers with scale-dependent
parameters given as a function of natural recharge rate from rainfall, aquifer
transmissivity, hydraulic gradient, aquifer thickness and aquifer soil porosity
were derived. His results showed that the recharge rate substantially affected the
contaminant distribution and might partially explain the scale dependence of
dispersion parameters. The reasons for observed scale effects on dispersion of
solute in geologic systems remain unresolved, however.

E. OUTLINE

In this contribution, we present a comprehensive analysis of recent advances

on studies focusing on scale-dependent dispersion in porous media. First, we
focus on stochastic approaches of time-dependent dispersivity models.
Application of fractal geometry and development in transfer function theory
will also be included in this section. Second, we discuss analytical and numerical
solutions for the convection –dispersion equation (CDE) with either time-
dependent or distance-dependent dispersivities. Third, a new approach, fractional
convection –dispersion equation, which represents anomalous dispersion in
heterogeneous media, is presented. Finally, several issues in current studies on
scale-dependent dispersion are addressed.

II. STOCHASTIC APPROACHES

A. MACRODISPERSIVITY THEORY

Stochastic methods have been employed to examine macroscopic flow and

transport processes in recent years (Gelhar et al., 1979; Dagan, 1982; Neuman
et al., 1987). The basic idea of stochastic methods is to describe solute transport
in terms of the random trajectory of a solute particle moving through a medium
with a random velocity field. The local dispersion is often assumed to be small.
The velocity field UðxÞ; which does not depend on the solute concentration, is
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 229

assumed to be a stochastic process, stationary in space and time. The velocity is

further split into two components as
UðxÞ ¼ u
þ uf ðxÞ ð1Þ
where u
is the mean velocity and uf the zero mean velocity fluctuation. The
fluctuation uf is assumed to be a Gaussian process with covariance tensor
suf ðx; yÞ U kuf ðxÞuf ðyÞl ð2Þ
The trajectory XðtÞ of a particle that starts at (x ¼ 0; t ¼ 0) is a realization of the
stochastic process XðtÞ; which is defined by
ðt
XðtÞ ¼ UðXðt0 ÞÞdt0 þ Xd ðtÞ ð3Þ
0

where Xd ðtÞ is a Gaussian process with zero mean and a covariance 2dd t that
accounts for local dispersion. Substituting Eq. (1) into Eq. (3) gives
ðt

þ uf ðXðt0 ÞÞdt0 þ Xd ðtÞ
XðtÞ ¼ ut ð4Þ
0

This integral equation is very difficult to solve because uf depends on the

trajectory XðtÞ (Jury and Roth, 1990). Dagan (1984) used a first order
approximation by replacing XðtÞ by its expectation kXðtÞl ¼ ut:

With the same
approximation, Eq. (4) can be simplified to
ðt
0
XðtÞ ¼ ut

þ uf ðutÞdt

þ Xd ðtÞ ð5Þ
0

The trajectory XðtÞ is a Gaussian process. The covariance tensor of XðtÞ may be
written as
ðt ðt
sX; jk ¼ suf; jk ðut
2 Þdt1 dt2 þ 2dd; jk t

1 ; ut ð6Þ
0 0

where j; k ¼ x; y; z: According to Eq. (6), the covariance tensor of the

process XðtÞ can be calculated in terms of the covariance of the velocity
fluctuations. Based on the covariance tensor, the macrodispersion tensor can
be defined as
1 dsX; jk
Djk U ð7Þ
2 dt
The macrodispersion tensor can be calculated from hydraulic conductivity
tensor KðxÞ: Because hydraulic conductivity often follows a lognormal
distribution, we define another tensor Y as
Y U lnðK=K0 Þ ð8Þ
where K0 is a characteristic value of lKðxÞl (Dagan, 1984 among others).
230 L. ZHOU AND H. M. SELIM

Dagan (1984) obtained an analytic expression for the dispersion tensor D

in two and three dimensions for an isotropic exponential covariance function
of Y
 
jxj
sY ðxÞ ¼ s2Y exp 2 ð9Þ
lY
where s2Y is the variance and lY the integral scale of Y: The covariance
tensor (Eq. (6)) of the particle positions calculated by Dagan (1984) reads
   
8 4 8 8 1
sX; xx ¼ 2s2Y l2Y t 2 þ 2 3 þ 2 1 þ expð2tÞ þ 2dL t ð10Þ
3 t t t t
where t U u
t=lY ; and dL is the local longitudinal dispersion coefficient. Here
we list only the principal component along the flow direction. Accordingly,
the longitudinal component of the macrodispersion tensor of Eq. (7) is
given by
   
2 4 24 24 24 8
Dxx ¼ sY lY u
1 2 2 þ 4 2 þ 3 þ 2 expð2tÞ þ dL ð11Þ
t t t4 t t
Jury and Roth (1990) inspected the approximation of Dxx for different t: For
t q 1; Eq. (11) can be approximated by
Dxx < dL þ s2Y lY u
ð12Þ
From Eq. (12), we recognize that the longitudinal macrodispersion coefficient
Dxx approaches a constant value in an isotropic stochastic continuum at very
large time. For t , 1; Eq. (11) can be approximated by
8 2 2
Dxx < dL þ s u
t ð13Þ
15 Y
From Eq. (13), we recognize that Dxx increase linearly with time if the mean
travel distance is less than the integral scale of Y: For very small time
ðt p 1Þ; the second term in Eq. (13), which accounts for dispersion induced
by velocity fluctuation, could be ignored. Russo (1998) applied the same
approach to the transport process in the vadose zone. The expression for
macrodispersion in the vadose zone is more complicated and also a function
of soil moisture content.
Gelhar et al. (1979) analyzed the development of the dispersion process for the
case of a perfectly stratified aquifer. Their analysis shows that the dispersivity
initially grows with displacement distance and that departures from the classical
Gaussian distribution are very significant during this development process
(Fig. 1). Later on, Gelhar and Axness (1983) conducted three-dimensional
stochastic analysis of macrodispersion in aquifers. Their analysis yields a finite
asymptotic longitudinal dispersivity using simple monotonic covariance
functions. Neuman et al. (1987) developed a stochastic theory of field-scale
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 231

Fickian dispersion in anisotropic porous media. Their theory also shows that the
field-scale longitudinal dispersivity reaches a non-zero constant value asympto-
tically. Matheron and De Marsily (1980) showed that the transport of solute could
not, in general, be represented by the usual convection– diffusion equation, even
for large time. They also showed, however, that when the flow is not exactly
parallel to the stratification, diffusive behavior is much more likely to appear.
Smith and Schwartz (1980) conducted a stochastic analysis of macroscopic
dispersion and their results suggested that the diffusion model for macroscopic
dispersion might be inadequate to describe mass transport in geologic units.
Mercado (1967) studied the spreading pattern of injected water in a permeability-
stratified aquifer. Interestingly, he concluded that the variance of travel distance
would grow proportionally to the square of the mean travel distance in aquifers.
In turn, the dispersivity increased linearly with mean travel distance or time. For
solute transport in anisotropic heterogeneous aquifers, Dagan (1988) found that
the macrodispersion coefficients were functions of time. He also concluded that
the suggested formulae were valid for large Peclet numbers, as encountered in
most groundwater applications and for average flow parallel to the plane of
isotropy. Dagan (1988) also suggested that the travel distance required to reach
asymptotic constant value is of the order of tens of horizontal log conductivity
integral scales.

B. FRACTAL MODELS

Classical Brownian motion is often used to describe solute transport in porous

media. The characteristics of Brownian motion lies in that it predicts a linear
increase of variance of travel distance with time. However, non-linear increase of
variance of travel distance has been observed in large field transport experiments.
One possible idea to study this anomalous transport is to extend traditional
Brownian motion to fractional Brownian motion (fBm). Many researchers
(Hewett, 1986; Philip, 1986; Arya et al., 1988; Ababou and Gelhar, 1990;
Neuman, 1990) have developed several different fractal models based on fBm.
These fractal models were compared by Tyler and Wheatcraft (1992). Sahimi
(1993) also presented a discussion about differences among the fractal models in
terms of Hurst coefficient. Ndumu and Addison (2001) gave a review on fBm and
fractional Gaussian noise (fGn) in subsurface hydrology.
The fBm was introduced by Mandelbrot and van Ness (1968) as scaling limits
of random walks or Brownian motions with power-law transition probability,
which constitute random fractal functions described by the index H ð0 , H , 1Þ;
the Hurst exponent. The value of H defines the intermittency or auto-correlation
property of the fBm function. One-dimensional fBm with index H that goes
through the origin (i.e., BH ð0Þ ¼ 0 at t ¼ 0) can be derived by substracting
the value of the random function at t ¼ 0 from the original notation of fBm given
232 L. ZHOU AND H. M. SELIM

in Feder (1988). The resulting expression for fBm can be written as

ð0
1
BH ðtÞ 2 BH ð0Þ ¼ ½ðt 2 tÞH21=2 2 ð2tÞH21=2 dWðtÞ
GðH þ 1=2Þ 21
ðt  ð14Þ
H21=2
þ ðt 2 tÞ dWðtÞ
0

where dWðtÞ stands for a Gaussian random function with zero mean and unit
variance; and GðH þ 1=2Þ is gamma function. The arguments t and t in Eq. (14)
represent time with t . t:
Increments of BH ðtÞ defined by WH ðtÞ are given by

WH ðtk21 Þ ¼ BH ðtk Þ 2 BH ðtk21 Þ ð15Þ

and are termed fractional white noise or fGn. WH ðtÞ becomes ordinary white noise
when H ¼ 1=2: It should be noted that WH ðtÞ is a discrete stationary process with
zero mean and variance that scales as , ðtk 2 tk21 Þ2H : The important property of
fBm increments (for our purposes) is its correlation. The autocorrelation function
CH ðtÞ of fractional white noise is defined by (Molz et al., 1997)

CðtÞ ¼ s2 h2 ½Hð2H 2 1Þjtj2H22 ð16Þ

where t is separation or lag distance, h the fixed increment, and s2 is the variance
for increment h ¼ 1: Apparently, H is imposed to vary between 0 and 1. For
H . 1; 2H 2 2 . 0 and the autocorrelation of fractional white noise grows
without bound with t: Negative H yields a strictly positive correlation that drops
off too rapidly with t: For H ¼ 1=2; CH ðtÞ ¼ 0 for ltl . 0: For 0 , H , 1=2;
CH ðtÞ , 0; and for 1=2 , H , 1; CH ðtÞ . 0: Only if H ¼ 1=2 is the correlation
between the increments of fBm zero and independent of lag distance t: Hurst
(1951, 1957) discovered and studied different types of processes exhibited by the
fBm in terms of H: The expression 0 , H , 1=2 defines zones of negative
correlation (a negative increment tends to be followed by a positive increment),
and Eq. (14) models subdiffusive, or antipersistent behavior. The value H ¼ 1=2
defines zero correlation. This is the special case of Gaussian noise, increments of
Brownian motion, or random walk; thus, Eq. (14) models Fickian diffusion. The
expression 1=2 , H , 1 defines the zone of positive correlation (a negative
increment tends to be followed by another negative one) and models
superdiffusion, or persistent behavior. Ndumu and Addison (2001) illustrated
the comparative diffusion as a function of time t of independent tracer particles
undergoing fBm diffusion for different Hurst coefficients H: They found that
higher values of the Hurst exponent H result in greater diffusion because of the
increased persistence of BH ðtÞ as H increases (Fig. 2; note the different scalings in
the y-axis for the same time t of Fig. 2a –c). For our purpose, we are interested
only in superdiffusion, i.e., 1=2 , H , 1: The box counting fractal dimension of
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 233

Figure 2 FBM spreading of 10 tracer particles released at origin at time t ¼ 0 with

BH ðt ¼ 0Þ ¼ 0 for 0 , H , 1: From Ndumu and Addison (2001). Copyright q 2001 by American
Society of Civil Engineering. Reproduced with permission of ASCE.
234 L. ZHOU AND H. M. SELIM

a fBm trajectory is Df ¼ 1=H: Hewett (1986) gave a different relationship

between fractal dimension and Hurst exponent as H ¼ 2 2 Df :
The spectral properties of fBm/fGn are also important in the study of
anomalous transport, especially for the synthetic generation of these processes.
The Fourier –Stieltjes integral is used for representing the ensemble that
constitutes a stationary stochastic process. Detailed discussion on spectral
representation and spectrum of fBm/fGn can be found in Yaglom (1987) and
Molz et al. (1997) among others. The spectrum relations of fBm and fGn are

1
fm ðvÞ / ð17Þ
v2Hþ1

1
fn ðvÞ / ð18Þ
v2H21
where fm ðvÞ and fn ðvÞ are spectrum for fBm and fGn, respectively, and v is the
frequency.
For Brownian motions BH ðtÞ traces, where H ¼ 1=2; the standard deviation s
scales with the square root of time t

s / t1=2 ð19Þ

that is, it is a Fickian diffusive process. Therefore, we have the following equation
pffiffiffiffi
s ¼ 2Dt1=2 ð20Þ

which is the standard deviation of the solution to the diffusion equation from a
point source with diffusion coefficient D: Ndumu and Addison (2001) extended
Eq. (20) to the fBm processes, where H – 1=2; and obtained
pffiffiffiffiffiffi
sH ¼ 2DH tH ð21Þ

where DH is the generalized fractal diffusion coefficient. The Fickian diffusion

coefficient D is defined from Eq. (20) using the Einstein relation

1 d s2
D¼ ð22Þ
2 dt
Applying Eq. (22) to Eq. (21) yields a time-dependent Fickian-type diffusion
coefficient DðtÞ in terms of the fractal diffusion coefficient DH

DðtÞ ¼ 2HDH t2H21 ð23Þ

By assuming DðtÞ is the product of a time-dependent dispersivity aðtÞ and a mean

pore water velocity v
; Ndumu and Addison (2001) further reaches an expression
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 235

for a time-dependent dispersivity

2HDH 2H21
aðtÞ ¼ t ð24Þ
v

For H ¼ 1=2; a constant dispersivity is recovered.
Arya et al. (1985) obtained a field-scale dispersivity expression in terms of the
fractal dimension Df : Their model was based on the idea of fractal Brownian
trails. The key assumption they made is that particles move along a path with x
and y coordinates generated by separate fBm functions. A particle’s position,
zðx; yÞ; at time t is given by
zðx; yÞ ¼ ðBH1 ðtÞ; BH2 ðtÞÞ ð25Þ

where BH1 ðtÞ and BH2 ðtÞ are fBms. As pointed out by Tyler and Wheatcraft
(1992), the Hurst coefficient of both the x and y coordinates (H1 and H2 ) must be
equal for a fractal trail developed this way. In addition, the Hurst coefficient must
lie in the range of 1=2 # H , 1 as discussed earlier. Meanwhile, the fractal
dimension Df for trails is equal to 1=H; which is the same as mentioned earlier.
Based on the fractal trails of a cloud of contaminant particles and Taylor’s (1921)
concepts of dispersion, they developed a model of the growth of field scale
dispersivity, am ; of the form (cited in Tyler and Wheatcraft, 1992)

am ¼ 1
2 CD x
ðð2=Df Þ21Þ ð26Þ

where x
is the average displacement of the center of mass of tracer. Considering

the relationship between the Hurst coefficient H and fractal dimension Df ; it is
easy to show that dispersivities described by Eqs. (24) and (26) grow with the
same rate. Later, Arya et al. (1988) suggested a different expression that is also
based on fBm

am ¼ ð2 2 Df ÞCD x
ð322Df Þ ð27Þ

Eq. (27) is derived from a functional expression of the variance of the
displacement. Arya et al. (1988) also pointed out that totally uncorrelated
variations produce classic Brownian motion with Df ¼ 1:5: In this case, Eq. (27)
reduces to a constant. The field data considered by Arya et al. (1988) were best fit
with a value of Df ¼ 1:12: Philip (1986) used a Lagrangian approach to
dispersion. Based on the autocorrelation function of flow velocity and the
famous Taylor’s dispersion theory (Taylor, 1921), he reached a result that the
apparent dispersivity should increase as t2H21 at large time. However, his work
also predicted that the dispersivity grew linearly with time at short times, i.e.,
aðtÞ , t: Hewett (1986) used fGn to model the variability of porosity. The use of
fGn results in the autocovariance of the porosity scaling as l2H22 where l is the
spatial lag. Hewett found the Hurst coefficient lies between 0.7 and 0.8 for actual
data. Interestingly, his work also predicts that the dispersivity varies with time
236 L. ZHOU AND H. M. SELIM

as t2H21 : Assuming an average H of 0.75 based on the range of H from actual

data, he actually predicts that dispersivity increases with time as t1=2 :
Wheatcraft and Tyler (1988) made an effort to explain the scale dependency of
dispersivity based on the fractal dimension of tortuous stream tubes. In their
random walk model, the distance actually traveled by the plume, xf ; is related to
straight-line distance, xs ; through a fractal scaling relation of the form (given in
Tyler and Wheatcraft, 1992)
1=Df
xs ¼ xf ð28Þ
By assuming a Fickian diffusion process along the tortuous fractal path, they
developed a scale-dependent dispersivity of the form
am ¼ C
xDf 21 ¼ C
x12H ð29Þ
where 1 , Df , 2 and 0 , H , 1: Tyler and Wheatcraft (1992) further pointed
out that H is expected to be limited to 1=2 , H , 1; and thus the fractal
dimension satisfies 1 , Df , 1:5: We also refer to Eq. (29) as Wheatcraft –
Tyler’s single tube model. Eq. (29) reduces to a constant, i.e., scale-independent
dispersivity for Df ¼ 1: Wheatcraft and Tyler (1988) also looked at an ensemble
of such stream tubes and developed a scale-dependent dispersivity of the
following form
am ¼ 1=2s 21 x s2Df 21 ¼ 1=2s 21 x 322H
s ð30Þ
Eq. (30) is often referred to as Wheatcraft – Tyler’s multiple tube model. This
model reduces to the Mercado (1967) model as discussed previously for non-
fractal media. Strangely, this reduced model is contradictory to their single tube
model. Besides, the Mercado model describes dispersion in layered systems, and
it is difficult to argue that layered systems would behave non-fractally.
Neuman (1990) developed a universal scaling law for dispersivity in geologic
media based on the quasi-linear theories (Neuman and Zhang, 1990; Zhang and
Neuman, 1990) of non-Fickian dispersion. He used a fBm to describe the log
conductivity field. The proposed semi-variogram, g ðsÞ; of log hydraulic
conductivity is given by
g ðsÞ ¼ C0 s26 ð31Þ
where 6 is equivalent to the Hurst coefficient H in Eq. (16). Using a value of 0.23
for 6; Neuman explained the higher than linear growth rate of apparent
dispersivity with scale, i.e., am , L1:46 s : It should be pointed that the Hurst
coefficient in Eq. (31) 6 is different from that in Eq. (16). 6 is only a parameter of
the log hydraulic conductivity field whereas H in Eq. (16) is the characteristic
exponent of the fractal trails of particles. The value of H can be directly
associated with the type of dispersion processes as discussed above. However, the
value of 6 cannot be associated directly with the type of dispersion processes
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 237

in porous media. For example, an H value of 0.23, which is less than 0.5, should
model a subdiffusion process as we discussed above; a 6 of 0.23 for the log
hydraulic conductivity still describes a superdiffusion process, i.e., dispersivity
increases non-linearly with scale.
Ababou and Gelhar (1990) developed a model based upon a truncated
spectrum of variation in hydraulic conductivity and the macrodispersivity theory
of Gelhar (1987). According to their analysis, the spectra of vertical conductivity
of three boreholes suggested the following power law
f ðkÞ ¼ S0 k2a ð32Þ
where k is wave number. Eq. (32) has the same form as that suggested by Philip
(1986) to describe the velocity spectra. From core data given by Bakr (1976),
Ababou and Gelhar suggested a ø 1: Based on this spectral form, Tyler and
Wheatcraft (1992) inferred that the resulting macrodispersivity grows linearly
with time.
Based on the above discussion, we have several forms of fractal models that
describe time-dependent dispersivity. As pointed out by Tyler and Wheatcraft
(1992), the fractal dimensions are not directly comparable among different
models. There are three different categories of fractal dimensions involved in our
discussion. The first one is the fractal dimension of the particle trails as discussed
in Arya et al. (1985, 1988) and Ndumu and Addison (2001). The second is the
fractal dimension of the tortuous flow paths in Wheatcraft and Tyler (1988). The
third is the fractal dimension of the spectra of some parameter field, for example,
porosity, velocity, and conductivity (Philip, 1986; Ababou and Gelhar, 1990;
Neuman, 1990). Interestingly, the time-dependent dispersivity can be derived in
some way based on any of the three different fractal dimensions.

C. TRANSFER FUNCTION MODELS

The convection – dispersion equation (CDE) is the most commonly used

model to describe solute movement through soils where perfect lateral mixing is
assumed. Thus, the CDE predicts a linear increase of travel time variance with
distance. With an approach similar to that used by Mercado (1967) and Jury and
Roth (1990) developed a stochastic-convective model. The stochastic –
convective process assumes that the solute moves at different velocities in
isolated stream tubes without lateral mixing. Use of lognormal distribution of
travel time results in a convective lognormal transfer function model (CLT). The
CLT model describes solute transport characterized by a quadratic increase in the
travel time variance with depth. However, the travel time variance is often
reported to increase non-linearly with distance because of heterogeneity of the
media (Zhang et al., 1994). To account for the non-linearity in the relationship
between travel time variance and distance or depth, Liu and Dane (1996)
238 L. ZHOU AND H. M. SELIM

proposed an extended transfer function model (ETFM). They introduced an

additional parameter to represent the degree of lateral solute mixing. The ETFM
can be considered as a transition between the CDE and the CLT. Zhang (2000)
also proposed an extended convective lognormal transfer function model
(ECLT). Meanwhile, he attempted to unify all types of transfer function model
with a generalized transfer function model (GTF). More important, Zhang (2000)
showed that the distance-dependent dispersivity model could be associated with
the parameters of the GTF.
The mean and variance of the travel time are two important characteristic
elements in transfer function theory. A third parameter related to scale effects is
the coefficient of variation (CV) of the travel time. The squared CV at depth z is
given by

Varðt; zÞ
CVðt; zÞ2 ¼ ð33Þ
Eðt; zÞ2

where Varðt; zÞ and Eðt; zÞ are the variance and mean of the travel time at depth z:
For the CDE, the means, variances, and CVs of the travel time at depth z and l are
related by

Eðt; zÞ z Varðt; zÞ z CVðt; zÞ2 l

¼ ¼ 2
¼ ð34Þ
Eðt; lÞ l Varðt; lÞ l CVðt; lÞ z

For the CLT, similar relationships can be established as

 2
Eðt; zÞ z Varðt; zÞ z CVðt; zÞ2
¼ ¼ ¼1 ð35Þ
Eðt; lÞ l Varðt; lÞ l CVðt; lÞ2

For the ECLT (Zhang, 2000), the relationships are

 l  2l
Eðt; zÞ z Varðt; zÞ z CVðt; zÞ2
¼ ¼ ¼1 ð36Þ
Eðt; lÞ l Varðt; lÞ l CVðt; lÞ2

The exponent l is introduced by Zhang (2000) to describe transport processes in

which the travel time of solute may increase with depth non-linearly. For the
ETFM by Liu and Dane (1996), the relationships are
 2a  2ð12aÞ
Eðt; zÞ z Varðt; zÞ z CVðt; zÞ2 l
¼ ¼ ¼ ð37Þ
Eðt; lÞ l Varðt; lÞ l CVðt; lÞ2 z

The value of parameter a in the above equation lies in the range between 0.5 and 1
ð0:5 # a # 1Þ: Based upon the observation of the above relationships for
different transfer function models, Zhang (2000) proposed a generalized
relationship of means, variances, and CVs of the travel time at depth z and l
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 239

for a GTF
 l1  l2  2ðl1 2l2 Þ
Eðt; zÞ z Varðt; zÞ z CVðt; zÞ2 l
¼ ¼ 2
¼ ð38Þ
Eðt; lÞ l Varðt; lÞ l CVðt; lÞ z

where l1 and l2 are parameters of the time moments.

The dispersivity can be estimated based on the CV at depth z and is given by
z
a¼ CVðt; zÞ2 ð39Þ
2
Substituting the CV for the GTF (Eq. (38)) into the above equation gives

a / z1þ2ðl2 2l1 Þ ð40Þ

If the two parameters l1 and l2 in Eq. (40) satisfy l2 2 l1 ¼ 20:5; dispersivity

is constant with the distance (CDE). Otherwise, Eq. (40) describes a distance-
dependent dispersivity.

III. ANALYTICAL SOLUTIONS OF CDE WITH

SCALE-DEPENDENT DISPERSIVITY

Based on the above discussion, dispersivity at the field scale often depends on
time or distance. Pickens and Grisak (1981b) suggested four types of time-
dependent dispersivity functions in terms of mean travel distance. These four
types are linear, parabolic, asymptotic, and exponential functions. Similar forms
of functions have been extended to distance-dependent dispersivity. The classical
CDE has been modified to incorporate the different dispersivity models. The
generalized CDE is assumed to better describe the transport process in soils or
aquifers. Similar to normal CDE, these generalized CDE can be solved
analytically or numerically. The analytical solutions are obtained through
Fourier or Laplace transforms. The obtained solutions are often represented by
special functions or integral of some complex functions. Numerical evaluation of
special functions or integrals is often inevitable. However, the analytical
solutions can be used as a benchmark for assessment of numerical solutions. In
this section, we summarize efforts that have been carried out to obtain analytical
solutions to the generalized CDE.

A. SOLUTIONS OF CDE WITH TIME-DEPENDENT DISPERSIVITY

Pickens and Grisak (1981b) summarized several field and laboratory transport
studies and identified four types of functions for time-dependent dispersivity in
240 L. ZHOU AND H. M. SELIM

terms of mean travel distance. When time t is chosen as the independent variable,
rather than mean travel distance, these functions are as follows

(1) Linear
aðtÞ ¼ dt ð41Þ

(2) Parabolic
aðtÞ ¼ dtb ð42Þ

(3) Asymptotic
aðtÞ ¼ A½1 2 B=ðt þ BÞ ð43Þ

(4) Exponential
aðtÞ ¼ E½1 2 e2Ft  ð44Þ

where aðtÞ is dispersivity; t is time; b; d; B; and F are constant; A; and E are

asymptotic or maximum dispersivity values. We assume that the dispersion
coefficient can be expressed by
D ¼ av ð45Þ
where v is longitudinal pore water velocity and a is the dispersivity as described
by Eqs. (41) –(44). For a two-dimensional system, the longitudinal and transverse
dispersion coefficients are given by (Zou et al., 1996)
Dx ðtÞ ¼ vax ðtÞ ð46Þ
Dy ðtÞ ¼ vay ðtÞ ð47Þ
where ax and ay are longitudinal and transverse dispersivities, respectively. The
generalized two-dimensional CDE for non-reactive solutes in systems with time-
dependent dispersion coefficients now reads
›C ›C ›2 C ›2 C
¼ 2v þ Dx ðtÞ 2 þ Dy ðtÞ 2 ð48Þ
›t ›x ›x ›y
For a parabolic dispersivitiy model (Eq. (42)), the longitudinal and transverse
dispersivity coefficients are given by
Dx ðtÞ ¼ vdx tbx ð49Þ
by
Dy ðtÞ ¼ vdy t ð50Þ
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 241

where dx and bx are constants for longitudinal direction; dy and by constants for
transverse direction. For an instantaneous injection of a slug of mass M at t ¼ 0 at
the origin of an infinite two-dimensional domain, the initial condition is given by

M
Cðx; y; 0Þ ¼ dðx; y; 0Þ ð51Þ
n

where n is the porosity; dð Þ is the Dirac delta function. The boundary conditions
are

lim Cðx; y; tÞ ¼ 0; lim Cðx; y; tÞ ¼ 0 ð52Þ

x!^1 y!^1

For non-reactive solutes, the total mass contained in the system at any time
satisfies

ðþ1 ðþ1
BnCðx; y; tÞdx dy ¼ M ð53Þ
21 21

where B is the aquifer depth and n is the porosity as in Eq. (51). The governing
Eq. (48) subject to initial and boundary conditions Eqs. (51) and (52) can be
solved using Fourier transform or a series of variable transformation (Basha and
El-Habel, 1993; Zou et al., 1996; Aral and Liao, 1996). Zou et al. (1996)
developed an analytical solution of Eq. (48) for a system with dispersion
coefficients described by Eqs. (49) and (50) using Fourier transform. Here, we
just follow their steps to derive the analytical solution.
The Fourier transform of function Cðx; y; tÞ; denoted by CðS~ x ; Sy ; tÞ; and its
inverse are defined by the following two equations
ðþ1 ðþ1
~ x ; Sy ; tÞ ¼ F½Cðx; y; tÞ ¼
CðS Cðx; y; tÞexp½2iðSx x þ Sy yÞdx dy ð54aÞ
21 21

and

Cðx; y; tÞ ¼ F 21 ½CðS
~ x ; Sy ; tÞ

1 ðþ1 ðþ1 ~
¼ CðSx ; Sy ; tÞexp½iðSx x þ Sy yÞdSx dSy ð54bÞ
ð2pÞ2 21 21

where Sx and Sy are the wavenumbers in x and y directions, respectively.

Applying the Fourier transform on both sides of Eq. (48) gives

dC~
¼ 2½Dx ðtÞS2x þ Dy ðtÞS2y þ ivSx C~ ð55Þ
dt
242 L. ZHOU AND H. M. SELIM

For a parabolic model as given by Eqs. (49) and (50), Eq. (55) can be written as
dC~
¼ 2vðdx tbx S2x þ dy tby S2y þ iSx ÞC~ ð56Þ
dt
The general solution to Eq. (56) reads
~ x ; Sy ; tÞ ¼ A exp½lðSx ; Sy ; tÞ
CðS ð57Þ
where
!
dx bx þ1 2 dy by þ1 2
lðSx ; Sy ; tÞ ¼ 2v t Sx þ t Sy þ iSx t
bx þ 1 by þ 1

and A is a constant that can be determined from Eq. (53). After complex
computation of the inverse Fourier transform on both sides of Eq. (57), Zou et al.
(1996) gave the following solution to the original equation (Eq. (48))
" #
M ðx 2 vtÞ2 y2
Cðx; y; tÞ ¼ pffiffiffiffiffiffiffiffiffiffiffiffi exp 2 2 ð58Þ
4pBnv a~x ðtÞa~y ðtÞ 4va~x ðtÞ 4va~y ðtÞ

where a~x and a~y are defined by

dx bx þ1
a~x ðtÞ ¼ t ð59aÞ
bx þ 1
dy by þ1
a~y ðtÞ ¼ t ð59bÞ
by þ 1
Zou et al. (1996) also gave the solution to the one-dimensional plume
" #
M=n ðx 2 vtÞ2
Cðx; tÞ ¼ pffiffiffiffiffiffiffiffiffiffiffi exp 2 ð60Þ
4pva~x ðtÞ 4va~x ðtÞ

It is not difficult to verify that solutions (58) and (60) reduce to their scale-
independent counterparts when bx ¼ by ¼ 0: Based on the solution (58), Zou
et al. (1996) proposed two approaches for estimating time-dependent dispersion
parameters. The first approach was based on solute distribution maps at two
specific times, and the second one was based on solute breakthrough curves from
two sampling wells. They applied the first approach to the Borden site field
experiment (Mackay et al., 1986) and obtained the following set of dispersion
parameters: dx ¼ 0:0216; bx ¼ 0:572; dy ¼ 0:0236; by ¼ 0:231:
Basha and El-Habel (1993) developed an analytical solution of the one-
dimensional time-dependent transport equation based on a series of variable
transformations. They also accounted for solute reactivity during transport.
Specifically, linear equilibrium adsorption and first-order decay were included in
the governing equation. They also took into consideration molecular diffusion.
Unlike Zou et al. (1996), Basha and El-Habel (1993) worked directly on
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 243

the dispersion coefficient model. First, they derived a general integral solution
for arbitrary dispersion function. After that, they provided particular solutions for
four types of dispersion models such as constant, linear, asymptotic, and
exponential functions. Their asymptotic dispersion coefficient model is expressed
as
t
DðtÞ ¼ D0 þ Dm ð61Þ
tþK
where D0 is the maximum dispersion coefficient, Dm the molecular diffusion
coefficient, and K is equal to the time corresponding to 0:5D0 þ Dm : In their
further numerical illustration, they analyzed the effects of asymptotic dispersion
coefficient on transport processes. Their results clearly showed the dependency of
the concentration profile on the value of K (equivalent to B in Eq. (43)) at early
times. Specifically, for an instantaneous injection, the maximum concentration
for K ¼ 50 at T ¼ 5 with D ¼ 0:1D0 is almost five times the maximum
concentration corresponding to the constant dispersion K ¼ 0; while at T ¼ 50
with D ¼ 0:5D0 ; the ratio is just double. Basha and El-Habel (1993) attributed the
difference in the maximum concentration to the large variation in the value of the
dispersion coefficient in the pre-asymptotic range. In the asymptotic range,
however, the concentration profiles for various values of K become almost
indistinguishable. Therefore, the scale dependence of the dispersion coefficient is
necessary only for prediction of solute transport at early times. For a continuous
injection, they found that the front of the concentration profile for large K is
relatively slower and less dispersed. At large time, the difference among the
profiles for different K values becomes insignificant. In short, significant
differences exist between the concentration distribution during early times in a
scale-dependent hydrogeologic system and that in a scale-independent system.
Basha and El-Habel (1993) thus concluded that the results of the inversion
method based on information collected from a scale-dependent hydrogeologic
system is likely to be in error if the inversion model assumes a constant dispersion
coefficient rather than a scale-dependent dispersion coefficient. Aral and Liao
(1996) extended the study of Basha and El-Habel (1993) to two-dimensional
media and obtained similar results. An analytical solution of the solute transport
in a semi-infinite spatial domain with time-dependent transport coefficient was
obtained by Barry and Sposito (1989). They also used variable transformation
approaches to solve the governing equation.

B. SOLUTIONS OF CDE WITH DISTANCE-DEPENDENT

DISPERSIVITY

Expressions of time-dependent dispersivity have been extended to distance-

dependent dispersivity. Three forms of distance-dependent dispersivity models
244 L. ZHOU AND H. M. SELIM

have been considered in the literature. These are linear, linear asymptotic, and
exponential models. The expressions for these models are given in the following
equations

(1) Linear
aðxÞ ¼ ax ð62Þ

(2) Linear asymptotic

(
ax x # x0
aðxÞ ¼ ð63Þ
aL x . x 0

(3) Exponential
aðxÞ ¼ aLð1 2 e2bx=L Þ ð64Þ

where aðxÞ is dispersivity; x distance or depth from source; a; x0 ; L; and b are

constants; aL ¼ ax0 is the asymptotic dispersivity; L is a characteristic length.
For all models, as x ! 0 the dispersivity is zero. For the linear model (Eq. (62)),
dispersivity increases linearly with distance without bound. For the linear
asymptotic model (Eq. (63)), dispersivity initially increases linearly with distance
and becomes constant after some critical distance is reached. For the exponential
model, as x ! 1; the dispersivity approaches a constant value equal to aL: If
molecular diffusion is included, the dispersion coefficient can be written as
DðxÞ ¼ aðxÞv þ Dm ð65Þ
where DðxÞ is the dispersion coefficient; aðxÞ the distance-dependent dispersivity;
v is pore water velocity; Dm is the molecular diffusion coefficient. Yates (1990)
developed an analytical solution to one-dimensional transport equation in a
system with linear distance-dependent dispersion coefficient. In his paper, the
molecular diffusion coefficient is expressed in terms of pore water velocity. Thus
the dispersion coefficient now reads
DðxÞ ¼ aðxÞv þ Dm ¼ ðax þ LbÞv ð66Þ
where b is strictly a constant; L is a characteristic distance.
The transport of solutes in a one-dimensional system including linear
equilibrium adsorption and first-order decay is
 
›C › ›C ›C
R ¼ DðxÞ 2v 2 mRC ð67Þ
›t ›x ›x ›x
where C is the solute concentration as a function of distance and time; R a
retardation coefficient and m is a first-order decay coefficient. Eq. (67) with
proper initial and boundary conditions can be solved by finding its similarity
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 245

solution (Su, 1995; Hunt, 1998) or by using Laplace transform methods (Yates,
1990). The other way to solve it is to convert it to a form that has known solution
through a series of variable transformation (Logan, 1996). Here we just follow
Yates (1990) to develop an analytical solution for Eq. (67).
Two sets of initial and boundary conditions are considered by Yates (1990). In
both cases, the initial concentration is assumed to be zero for all x: The boundary
condition at x ¼ 0 is either a constant concentration or constant flux condition
and as x ! 1; the gradient ›C=›x ! 0: For transport with a constant
concentration boundary condition, the initial and boundary conditions can be
expressed as

Cðx; 0Þ ¼ 0 ð68aÞ

Cð0; tÞ ¼ C0 ð68bÞ

›C
¼0 ð68cÞ
›x x!1
where C0 is the constant solute concentration applied to the system. For a
constant flux boundary condition, the initial and boundary conditions are
Cðx; 0Þ ¼ 0 ð69aÞ

›C
2DðxÞ þvClx¼0 ¼ vC0 ð69bÞ
›x x¼0

›C
¼0 ð69cÞ
›x x!1
where DðxÞ is distance-dependent dispersivity as given in Eq. (66); C0 is the same
as in Eq. (68). Incorporating (66) into the governing equation, i.e., Eq. (67),
rearranging and using the following dimensionless variables
t ¼ vt=RL h ¼ x=L j2 ¼ ah þ b b ¼ mLR=v g ¼ 1=a ð70Þ
gives a dimensionless transport equation
›C ›2 C ›C
¼ ðah þ bÞ 2 2 ð1 2 aÞ 2 bC ð71Þ
›t ›h ›h
Depending on the value of the parameter a; Eq. (71) can take one of several
forms. If a ¼ 0; Eq. (71) reduces to common CDE with molecular diffusion only.
This case is not of our interests because no distance-dependent dispersivity is
present in the governing equation. If a ¼ 1; the advective term in Eq. (71) drops
out and Eq. (71) reduces to an equation that is mathematically identical to a one-
dimensional diffusion equation. As pointed out by Yates (1990), this would occur
for a situation that the reverse dispersion at any point equals the forward
246 L. ZHOU AND H. M. SELIM

advection and dispersion. Physically, this may not be possible because the
longitudinal water flow is always present in the transport processes we are
interested in. If 1 , a , 1; Eq. (71) takes the form with a net negative advection
term. This case is also against our laboratory and field observations and is not of
our interests. Therefore, it seems improbable that the growth of the dispersion
process will be so strong that the reverse dispersion will be greater than the
forward advection and dispersion (Yates, 1990). Based on the above discussion,
the parameter, a; will be restricted to the range 0 , a , 1: This range is different
from that by Yates (1990), which is 0 # a # 2: Huang et al. (1996) also
discussed the range of parameter, a; and they suggested that a should be in the
range 0 # a # 1: However, as mentioned earlier, both ends, 0 and 1, should not
be included.
The Laplace transform of function Cðh; tÞ; denoted by Cð
h; sÞ; and its inverse
transform are defined by the following pair of equations
ð1

h; sÞ ¼ L½Cðh; tÞ ¼
Cð e2st Cðh; tÞdt ¼ C
ð72aÞ
0

and
ðGþi1

h; sÞ ¼ 1
Cðh; tÞ ¼ L21 ½Cð
h; sÞest ds
Cð ð72bÞ
2pi G2i1

where G is a line in the complex domain to the right of all poles. Taking the
Laplace transform of Eq. (71) and using the transformations given by Eq. (70)
yields the distance-dependent solute transport equation in the Laplace domain

d2 C
dC

j2 2
þ ð1 2 2=aÞj 2 ð2=aÞ2 ðb þ sÞj2 C
¼ 0 ð73Þ
dj dj

The range of a suggested earlier, i.e., 0 , a , 1; will guarantee that Eq. (73)
takes the form of a radial dispersion equation with a net positive advection. The
general solution to Eq. (73) reads (Yates, 1990)

j; sÞ=C0 ¼ jg {AðsÞKg ½2ðs þ bÞ1=2 j=a þ BðsÞIg ½2ðs þ bÞ1=2 j=a}

Cð ð74Þ

where Kg ðxÞ and Ig ðxÞ are modified Bessel functions of order g ¼ 1=a; AðsÞ and
BðsÞ are Laplace-transformed integration functions and s a complex variable in
the Laplace domain. Since ›Cð
j; sÞ=›j must remain finite as j ! 1; BðsÞ must be
set to zero (Yates, 1990).
For constant concentration boundary condition case, the initial and boundary
conditions in the Laplace domain are

j0 ; sÞ ¼ C0 =s
Cð
j; 0Þ ¼ 0
Cð ð75Þ
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 247
pffiffi
where j0 ¼ b: Incorporating these conditions into Eq. (74) gives
 g
Kg ½2ðs þ bÞ1=2 j=a

j; sÞ=C0 ¼ j
Cð ð76Þ
j0 sKg ½2ðs þ bÞ1=2 j0 =a
The inverse Laplace transform of Eq. (76) is obtained by Yates (1990) and reads
 g ( pffiffi )
j Kg ½2 bðj=aÞ 2
Cc ðj; tÞ=C0 ¼ pffiffi 2 Ic ð77Þ
j0 Kg ½2 bðj0 =aÞ p
pffiffi
where Ic is an integral for the constant concentration case on the interval ½ b; 1
of Bessel functions of order g and is
" #
ð 1 exp½2x2 t J ð1ÞY ð1 Þ 2 J ð1 ÞY ð1Þ
g g 0 g 0 g
Ic ¼ dx ð78Þ
pffi x Jg ð10 Þ2 þ Yg ð10 Þ2
b

where
1 ¼ 2ðx2 2 bÞ1=2 j=a 10 ¼ 2ðx2 2 bÞ1=2 j0 =a ð79Þ
and Jg ðxÞ and Yg ðxÞ are Bessel function of the first and second kind, respectively.
For the constant flux boundary condition, the initial and boundary conditions
in the Laplace domain are

2aj dC

j¼j ¼ C0 =s
j; 0Þ ¼ 0
þCl Cð ð80Þ
2 dj j¼j 0
0

Incorporating these conditions into Eq. (74) gives

 g
Kg ½2ðsþbÞ1=2 j=a

j;sÞ=C0 ¼ j
Cð ð81Þ
j0 s{Kg ½2ðsþbÞ1=2 j0 =aþðsþbÞ1=2 j0 Kg21 ½2ðsþbÞ1=2 j0 =a}
Yates (1990) obtained the inverse Laplace transform of Eq. (81) as
 g " pffiffi #
j Kg ½2 bj=a 2
Cf ðj;tÞ=C0 ¼ pffiffi pffiffi pffiffi 2 If ð82Þ
j0 Kg ½2 bj0 =aþ bj0 Kg21 ½2 bj0 =a p
pffiffi
where If is an integral of Bessel functions of order g on the interval ½ b;1 for
the constant flux case and is given by
ð1 e2x2 t
If ¼
pffi x
b

" #
Jg ð1Þ½Yg ð10 Þ2fðxÞYg21 ð10 Þ2Yg ð1Þ½Jg ð10 Þ2fðxÞJg21 ð10 Þ
 dx ð83Þ
½Jg ð10 Þ2fðxÞJg21 ð10 Þ2 þ½Yg ð10 Þ2fðxÞYg21 ð10 Þ2
where fðxÞ ¼ ðx2 2bÞ1=2 j0 and 1 and 10 are defined in Eq. (79). Yates (1990)
compared the distance-dependent solution where a ¼ 0:1 and the classical
248 L. ZHOU AND H. M. SELIM

convection –dispersion solution with a constant dispersion coefficient for a

constant concentration boundary condition and a pulse input. His results
demonstrate that the classical convection –dispersion equation can produce a
curve with the same shape as a curve that results from a scale-dependent
dispersion model. However, if the classical solution, with the parameter estimates
obtained at x ¼ 100m; is used to predict the breakthrough curve at a larger
distance, for example, x ¼ 500m; a significant deviation from the scale-
dependent solution results. The most significant difference lies in that the
classical solute transport equation produces a higher peak concentration and less
overall dispersion (Fig. 3). Yates (1990) thus concludes that if the dispersion
process is more accurately characterized with a distance-dependent model, any
aquifer parameters obtained from the inverse approach that uses the classical
transport model may produce values for the parameters that are significantly
different from the scale-dependent aquifer values. Using similar procedure, Yates
(1992) developed an analytical solution for one-dimensional transport in porous
media with an exponential dispersion function. The solution is expressed in
hypergeometric functions. He also conducted a comparison between the linear
dispersion solution (Yates, 1990) and the exponential dispersion solution. The
results show that for appropriately chosen parameters the two solutions will
produce essentially the same results at early times, when their respective
dispersion functions are approximately the same, but differences occur at
intermediate and large times. Yates (1992) thus concluded that it would be

Figure 3 Concentration as a function of time for a pulse-type, constant concentration boundary

condition at distance x ¼ 100 and 500 m. The solid curve results from Eq. (77) with a ¼ 0:1: The
curve with triangle is the convection-dispersion solution with constant coefficients that were obtained
by fitting the model to the solid curve at x ¼ 100 m. After Yates (1990). Copyright q 1990 by the
American Geophysical Union. Reproduced with permission of AGU.
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 249

difficult and expensive to determine whether the behavior of the dispersion

process is linear or asymptotic at early or intermediate times. Meanwhile, a clear
distinction between the dispersion processes may be possible but only at very
large times. Logan (1996) extended the studies of Yates (1992) and developed an
analytical solution for solute transport in porous media with an exponential
dispersion function and periodic boundary conditions. The solution was also
expressed in hypergeometric functions. He gave the solution surface for a specific
set of parameter values. According to his illustrations, at each fixed distance x the
concentration is a periodic function of t; phase shifted from the original input
signal at the boundary; the actual phase shift increases with distance x: His
illustrations also show that for each fixed time t; the concentration profile is an
oscillating, decaying function of the distance x downstream. Logan (1996) also
compared the transport processes in the heterogeneous medium with an
exponential dispersion function and those in the homogeneous medium. An
unexpected finding is that the presence of either decay or adsorption can
sometimes force a smaller amplitude of the concentration in the heterogeneous
medium when compared with the homogeneous case (Fig. 4).
Huang et al. (1996) developed analytical solutions for one-dimensional
transport with linear asymptotic scale-dependent dispersion. The distance-
dependent dispersion problem with Eq. (63) for the dispersivity is reformulated in
terms of a two-layer transport problem involving separate concentrations for
the two subregions. In region 1, dispersivity increases linearly; in region 2,

Figure 4 Graphs of the amplitude of c as a function of distance x: The dashed curve refers to
constant dispersion and the solid curve refers to scale-dependent dispersion. After Logan (1996).
Copyright q 1996 by Elsevier Science B.V. Reproduced with permission of the publisher.
250 L. ZHOU AND H. M. SELIM

dispersivity takes constant asymptotic value aL : The two regions are coupled
using two different approaches. In one approach, they assume that region 1
ðx , x0 Þ is an effectively semi-infinite system and hence that concentrations in
region 1 are not affected by what happens downstream in region 2. After
obtaining the solution for region 1 in this manner, the solution for transport in
region 2 can be derived using the concentration of region 1 at x ¼ x0 as the inlet
condition for region 2. Either continuity in concentration or continuity in the
solution flux can be applied at the interface. In a second approach, they consider
region 1 to be finite and invoke simultaneously both concentration continuity and
flux continuity at the interface, x ¼ x0 : Huang et al. (1996) conducted a
comparison among their linear asymptotic dispersivity (LAD) model, linear
dispersivity model of Yates (1990), and the classical CDE model with constant
dispersivity. Their results show that a much stronger scale-dependent effect exists
when the parameters a and x0 increase in value. However, they found that for the
same asymptotic dispersivity value, the effect of a on the calculated
concentration distribution is relatively small at large distances. When compared
with Yates’ linear dispersivity model, their LAD model always gave the same
concentration distributions as those obtained with the Yates’ linear model when
0 # x # x0 while deviating substantially from the Yates model for x . x0 : They
also found that predicted concentration profiles obtained with the CDE model in
most cases differed greatly from those calculated with the LAD model as well as
the Yates’ linear model, except at larger times when an asymptotic dispersivity
value was considered.
Hunt (1998) obtained similarity solutions of the dispersion equation with
dispersivities that increase linearly with distance in one, two, and three spatial
dimensions. Unsteady flow from instantaneous sources and steady flow from
continuous sources were considered. He also ignored the molecular diffusion.
Based on analytical solutions, he compared the breakthrough curves at a fixed
distance for constant and linear dispersivity models. The dispersivities in the
constant dispersivity model are calculated from the maximum distance. Hunt
(1998) found that the breakthrough curves from the variable dispersivity model
started to increase later, reached a higher peak, and then fell off more rapidly
(Fig. 5). He explained that this was because the variable dispersivity model had a
much smaller dispersivity and, thus, less dispersion for smaller distances than the
constant dispersivity model, which used a dispersivity based on the maximum
value of distance for the entire calculation. He also gave an illustrative example
of simulations for a two-dimensional instantaneous source. Contours of constant
concentration from the variable dispersivity model showed that larger and
smaller dispersivities at larger and smaller values of distance, respectively,
created more dispersion for larger values of distance and less dispersion for
smaller values of distance. Meanwhile, he found that concentration contours for
the variable dispersivity model were neither parallel nor symmetric with respect
to x about the point of maximum concentration, which contrasted with the
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 251

Figure 5 Variations of c with t for x held constant for an instantaneous point source in one
dimension. Courtesy of Bruce Hunt. After Hunt. B., (1998). Copyright q 1998 by ASCE. Reproduced
with permission of ASCE.

concentric ellipses obtained for a constant dispersivity solution. Finally, Hunt

concluded that variable dispersivity models were likely to give an accurate
description of experiment over a range of values for distance. Pang and Hunt
(2001) derived analytical solutions for a one-dimension linear dispersion model
using similar approach as that by Hunt (1998). They reached an integral
expression of the solution for a continuous constant source, and their solution has
a much simpler form than the solutions obtained by Yates (1990) and Huang et al.
(1996). They applied their analytical solutions to their 8 m long column transport
experimental data. Interestingly, the dispersivity – distance ratio (a in Eq. (62))
they obtained for one experiment was very small and in the order of 1023.

IV. NUMERICAL APPROACHES

In Section III, we discussed how the CDE was solved analytically in systems
with either a time-dependent or distance-dependent dispersivity. The analytical
solutions obtained are usually expressed as special functions such as Bessel
functions and hypergeometric functions. These special functions are very
complicated, and thus numerical evaluation is inevitable. Although analytical
solutions have advantages over numerical solutions, these solutions are actually
very difficult to implement. As an alternative, numerical methods are often used
to solve the CDE with special dispersivity functions. The most significant
advantage of numerical methods lies in that numerical methods are easier to
252 L. ZHOU AND H. M. SELIM

implement than the analytical solutions. On the other hand, numerical methods
are capable of handling complicated boundary conditions. The most significant
problem encountered in numerical methods is the convergence and stability. For
CDE with constant parameters, the criteria for convergence, i.e., time and space
increments, have been studied thoroughly, and conclusive results have been
reached. However, the discussion on convergence conditions for CDE with time-
or distance-dependent dispersivities is very complicated if not impossible.
Probably this is the reason why few studies on transport in systems with time- or
distance-dependent dispersivities using numerical methods have been reported in
the literature.
The two numerical methods used to solve transport equation are the finite
element and finite difference methods. Sometimes, these two methods are
combined together in studies of solute transport. Generally, finite element method
is used in space domain whereas finite difference method is used in time domain.
Finite element approximation is often based upon Galerkin’s method using linear
and quadratic interpolation functions. Finite difference approximation is often
carried out using the Crank –Nicholson scheme. Pickens and Grisak (1981b) gave
a detailed description of finite element approximation. Here we give only the final
matrix equation and it reads
   
Dt tþDt Dt Dt
½H þ ½K ½C ¼ ½H 2 ½K ½Ct þ {½Ft þ ½FtþDt } ð84Þ
2 2 2

where ½C is the concentration vector; ½H a “mass” matrix; ½K a “stiffness”
matrix, and ½F is a “load” vector. The superscripts refer to the time levels. The
matrix equations can be solved using Gaussian elimination or other methods.
Jayawardena and Lui (1984) summarized studies on stability conditions of the
finite element-finite difference model. The criteria for no oscillations can be
expressed in terms of the Peclet number (sometimes referred to as the Reynolds
cell number), p; defined as

p ¼ vDx=D

and the Courant number, c; defined as

c ¼ vDt=Dx

where Dx and Dt are the space and time increments, respectively. The criterion
for the Peclet number for several finite-difference and finite-element formulations
given by Jenson and Finlayson (1978) is: p , b where b ¼ 2 both for finite
differences centered in space and for finite elements using linear interpolation
functions, and b ¼ 4 for finite elements using quadratic interpolation functions
(cited in Jayawardena and Lui, 1984). Ehlig (1977) has obtained the criterion for
no oscillations as c , 0:1 for formulations involving finite elements in space and
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 253

central finite differences in time. Through numerical tests, Jayawardena and Lui
(1984) found that the criterion given by Ehlig (1977) seemed to be on the safer
side. Their results indicated that the first appearance of numerical oscillations
about the dimensionless unit concentration line for p ¼ 1 and p ¼ 2 occurred at
c ¼ 2:5 for a time t ¼ 40 s.
The above finite element-finite difference model was used by Pickens and
Grisak (1981b) to model dispersion in hydrogeologic systems with time-
dependent dispersivity in terms of mean travel distance. They verified the
numerical model with an asymptotic dispersivity function. From the simulated
concentration distributions for various times, the standard deviation of the solute
distribution was determined as one-half the distance between the positions
corresponding to relative concentrations C=C0 of 0.841 and 0.159. The variance
of the solute distribution was thus calculated as the square of the standard
deviation. Pickens and Grisak (1981b) found that the finite element model could
give a variance pattern excellently comparable with the theoretical variance
pattern. They also examined the effects of time-dependent dispersivity on
breakthrough curves. For a linear time-dependent dispersivity model, they found
that the breakthrough curves would exhibit slightly greater tailing. In a
comparison between constant and LAD models (similar to Eq. (63)), they
found that the importance of time-dependent dispersion at early times or short
mean travel distance was minimal on long-term predictions of solute transport.
Mishra and Parker (1990) applied a Galerkin-type mass-lumped linear finite
element method in a study on solute transport in systems with a hyperbolic
distance-dependent dispersivity. They suggested a hyperbolic distance-depen-
dent dispersivity model of the form

1
1¼ ð85Þ
1 1
þ
11 bx

where 11 is an asymptotic dispersivity attained at large distances, b is a scale

factor describing the linear growth of the dispersion process near the origin, and x
is distance from an injection point. Mishra and Parker (1990) pointed out that
when b ! 1; Eq. (85) reduces to the classical constant dispersivity (i.e.,
1 ¼ 11 ); whereas 11 ! 1 results in a linear increase of dispersivity with
distance, i.e., 1 ¼ bx: They referred to the dispersivity defined in Eq. (85) as the
local dispersivity because it reflects spreading characteristics of the porous
medium at the local scale. They referred to the dispersivity that reflects spreading
characteristics of the entire soil profile up to the location at which the tracer BTC
was monitored as the effective dispersivity. They postulated that the two
definitions could be related by simple averaging since the effective dispersivity is
essentially an integrated measure of the variable local dispersivity. For the
hyperbolic distance-dependent dispersivity model as described by Eq. (85),
254 L. ZHOU AND H. M. SELIM

the effective dispersivity, 1;

can be expressed as

 
lnð1 þ bx=11 Þ

¼ 11 1 2
1ðxÞ ð86Þ
bx=11

They used both dispersivity models to simulate the breakthrough behavior at

different depths. For each depth, the effective dispersivity was used to predict
tracer breakthrough behavior assuming a constant effective dispersivity up to that
depth. The two sets of breakthrough curves obtained based upon local and
effective dispersivities indicated excellent agreement for the following parameter
values: 11 ¼ 0:2 m and b ¼ 0:1: Mishra and Parker (1990) applied the finite
element model to the field experiment conducted by Butters (1987) and obtained
optimal parameters. The dispersion scale factor was estimated to be b ¼ 0:19;
and the asymptotic dispersivity was estimated to be 11 ¼ 3:67 m. These
parameters indicate that dispersivity essentially increases linearly with distance
over the scale of the experiment. The resulting local and effective dispersivity-
distance relationships were compared with effective dispersivity-distance
relationship estimated by Butters (1987) in Fig. 6.

Figure 6 Local and effective dispersivity–distance relationships estimated by Mishra and Parker
(1990), and effective dispersivity– distance relationship estimated by Butters (1987). After Mishra and
Parker (1990). Copyright q 1990 by John Wiley & Sons, Ltd. Reproduced with permission of the
publisher.
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 255

V. FRACTIONAL ORDER ADVECTION– DISPERSION

EQUATION (FADE)

In the above presentation, we discussed how to characterize transport in

heterogeneous media using a modified classical advection – dispersion equation
(ADE) with time- or distance-dependent dispersivities. The underlying process
associated with the classical ADE is the Fickian diffusion or the Brownian
motion. However, time- or distance-dependent dispersivities indicate the
existence of non-Fickian behavior. This non-Fickian behavior can be addressed
in a statistical physics framework. One approach to describe particle transport is
the continuous time random walks (CTRW) (Montroll and Weiss, 1965; Scher
and Lax, 1973). Berkowitz and Scher (1995) and Benson (1998) both present a
review of CTRW in a hydrologic context. Each motion that a particle undertakes
is described by the probability of moving a random distance in a random amount
of time (Pachepsky et al., 2000). On the other hand, the CTRW allow descriptions
of particle motions that have extremely long-range temporal and/or spatial
correlation (Benson et al., 2000a). Benson et al. (2000a) also pointed that the
CTRW eventually converge to Brownian motion unless some infinite moments of
the particle excursion time and/or variance are assumed. In these cases, the
CTRW converge to Lévy motion (Benson, 1998; Benson et al., 2000a). Particles
undergoing Lévy motion can be simply characterized as behaving mostly like
in Brownian motion except for occasional large jumps (Pachepsky et al., 2000).
To describe this type of motion, one needs Fokker – Planck equations with
fractional derivatives. In the hydrogeologic context, a fractional advection –
dispersion equation (FADE) could be developed (Benson et al., 2000a).

A. DERIVATION OF FADE

The FADE can be derived in several different ways. Zaslavsky (1994)

developed a Fokker– Planck – Kolmogorov equation (FPKE) based on the
Markov process equation. He generalized the FPKE to fractional order and
obtained the fractional Fokker – Planck –Kolmogorov equation (FFPKE). It is
worth noting that the FFPKE developed by Zaslavsky has fractional derivatives
in both time and space. Following a procedure similar to that used by Zaslavsky
(1994), Benson (1998) derived a fractional order FADE to describe Lévy motion.
Chaves (1998) proposed a fractional diffusion equation to describe Lévy flights
using a different approach. He first defined a fractional derivative linear operator
based on the eigenvector equation. He then generalized Fick’s law to the form
expressed in terms of left and right fractional derivatives. The equation Chaves
obtained is quite similar to that by Benson (1998). Recently, Schumer et al.
(2001) suggested another method to develop the FADE. They proposed an
256 L. ZHOU AND H. M. SELIM

expression of dispersive flux based on the generalized Taylor series, which is

valid for both integer and fractional-order derivatives (Osler, 1971). A fractional
Fick’s law, in which flux is proportional to a fractional derivative, is thus
obtained. They replaced the classical Fick’s law with the fractional Fick’s law in
a Eulerian evaluation of solute transport in a porous medium and obtained a
fractional ADE. Cushman and Ginn (2000) showed that the FADE is a special
case of the convolution-Fickian non-local ADE proposed by Cushman and Ginn
(1993).

B. SOLUTIONS

The FADE developed by Benson (1998) reads

›C ›C 1 ›a C 1 ›a C
¼ 2v þ ð1 þ bÞD a þ ð1 2 bÞD ð87Þ
›t ›x 2 ›x 2 ›ð2xÞa
where the dimension of D is La T 21 : The fractional order a takes the range:
1 , a # 2: The parameter b describes the skewness of transport processes. The
fractional derivatives in Eq. (87) are defined by
!
›a C 1 ›k ð x
¼ ðx 2 jÞk2a21 Cðj; tÞdj ð88aÞ
›x a Gðk 2 aÞ ›xk 21
!
›a C ð21Þk ›k ð1
¼ ðj 2 xÞk2a21 Cðj; tÞj ð88bÞ
›ð2xÞa Gðk 2 aÞ ›xk x

where a is the order of the fractional derivative, a . 0; G is the gamma function,

k is the smallest integer number larger than a: The properties of the fractional
derivative are discussed by Benson et al. (2000a). When a ¼ 2; the classical
ADE is recovered since ›2 C=›x2 ¼ ›2 C=›ð2xÞ2 : Solutions to common solute
transport boundary value problems (BVPs) are gained through Laplace or Fourier
transforms. Benson et al. (2000a) presented a solution in Fourier domain to Eq.
(87) for instantaneous injection of a Dirac delta function spike of solute. The
solution in Fourier domain reads
h i
^ tÞ ¼ exp 1 ð1 2 bÞð2ikÞa Dt þ 1 ð1 þ bÞðikÞa Dt 2 ikvt
Cðk; ð89Þ
2 2

where i2 ¼ 21 and k is wavenumbers. The close-form inverse Fourier transform

of this solution is very difficult to obtain if not impossible. However, through
some variable transformations, Benson et al. (2000a) showed that Eq. (89)
is similar to the characteristic function for a-stable probability densities.
This means that the solution concentration profile follows a a-stable distribution
or Lévy distribution of ath order (Benson, 1998). Lévy’s random variable arises
from the generalization of the central limit theorem (Gnedenko and Kolmogorov,
1954). Based on a a-stable distribution, Benson et al. (2000a) proposed an
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 257

analytical solution to the step function BVP as

  
C0 x 2 vt
C¼ 1 2 serf a ð90Þ
2 ðBtÞa
where B ¼ lcosðpa=2ÞlD; and serf a is the a-stable error function similar to the
error function. The a-stable error function is defined as twice the integral of a
symmetric a-stable density from 0 to the argument ðzÞ and is given by
ðz
serf a ðzÞ ¼ 2 fa ðxÞdx ð91Þ
0

where fa ðxÞ is the standard, symmetric, a-stable density and reads

x1=ða21Þ a ð1
fa ðxÞ ¼ U ðwÞexp½2xa=ða21Þ Ua ðwÞdw ð92Þ
2l1 2 al 0 a
where 1 , a # 2 and the auxiliary function Ua ðwÞ is defined as
 
sinðpaw=2Þ a=a21
Ua ðwÞ ¼
cosðpa=2Þ
The error function erfðzÞ and serf 2:0 ðzÞ are related by

erfðzÞ ¼ serf 2:0 ð2zÞ ð93Þ

C. APPLICATIONS

Because the FADE is proposed recently, only few examples on its application
are available in the literature. Benson et al. (2000b) gave two case studies in the
same time they proposed the FADE. They applied the FADE to the Cape Cod
natural gradient tracer test and a laboratory sandbox tracer test. They found that
the sandbox tracer test could be described by a dispersion term of order 1.55
and the Cape Cod bromide plume could be modeled by an equation of order
1.65– 1.8. They also proposed a method to estimate parameters fractional order a
and fractional dispersion coefficient D: It should be pointed out that they
estimated both parameters separately. That is, one parameter is estimated firstly
whereas the other parameter is computed based upon the known value of the first
parameter. Specifically, the fractional order a is estimated as
a ¼ 2=ðm þ 1Þ ð94Þ
where m is the slope of the increase of apparent dispersivity versus time on a
log – log graph. The classical CDE is recovered if the apparent dispersivity does
not increase with time. That is, a ¼ 2 if m ¼ 0: The fractional dispersion
coefficient D is then estimated according to the relationship between the Gaussian
distribution and the a-stable distribution together with the known a value from
258 L. ZHOU AND H. M. SELIM

Eq. (94). The fractional dispersion coefficient D can be roughly estimated as

 a=2
VAR 1
D< ð95Þ

2 cos pa t
2

where VAR is the measured plume variance. From Eq. (95), we recognize that the
estimated D is dependent on time. This is contradictory to the purpose of FADE
because the idea of using FADE is to use more general random motion to account
for the heterogeneity of the media so that a time- or distance-dependent
dispersivity could be avoided. Benson et al. (2000b) calculated a; D doublets of
1.65, 0.21 m1:65 =d and 1.8, 0.25 m1:8 =d from the early data.
Pachepsky et al. (2000) made efforts to simulate scale-dependent solute
transport in soils with the FADE. They summarized several laboratory and field
transport experiments and found that the solute dispersivity defined from the
classical ADE increased as the length of a soil column or the soil depth increased.
They suggested that the heterogeneity of soil was a physical reason for this scale
dependence. They justified the application of FADE to transport processes in
soils based upon the heterogeneity of soils. They fitted the analytical solution of
one-dimensional FADE (Eq. (90)) to the BTCs at different lengths in Toride et al.
(1995) and obtained different values of a for different lengths. The optimized
values of a vary from 1.574 to 1.683 for unsaturated experiments and 1.846 to
1.913 for saturated experiments. They also compared the predictability of both
classical ADE and FADE based on some statistical test and they found that the
FADE simulated BTCs better than the classical ADE.

VI. CONCLUDING REMARKS

In this section, we want to emphasize the differences between a scale-

dependent dispersivity and time- or distance-dependent dispersivities. The scale-
dependent dispersivity is identified based upon the estimated apparent
dispersivities at different times or distances. However, in both cases, the
obtained dispersivities are actually constant for time up to the point when the
estimation is obtained or for distance up to the place where the breakthrough
curves are measured. The point here lies in that for different times or distances,
different values for dispersivity must be used. In other words, one cannot infer
from the observed trend of apparent dispersivities that dispersivity is time- or
distance-dependent. Though stochastic theories explain why dispersivity tends to
be time dependent, no development of a time-dependent dispersivity directly
based upon the statistical properties of aquifers has been implemented. On the
other hand, no theory has been advanced to explain or support the concept of a
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 259

distance-dependent dispersivity. The birth of distance-dependent dispersivity can

perhaps be attributed to replacement of mean travel distance with distance in the
time-dependent dispersivity functions.
We summarized several approaches dealing with studies of solute transport in
porous media when time- or distance-dependent dispersivities are considered.
The stochastic approaches give a theoretical basis for use of a time-dependent
dispersivity. However, the use of a distance-dependent dispersivity lacks
theoretical foundation. Nevertheless, the generalized CDE when incorporated
with different forms of time-dependent or distance-dependent dispersivity
functions has been solved analytically and numerically. According to analytical
or numerical solutions of the generalized CDE with time- or distance-dependent
dispersivities, the transport processes in these types of media express distinct
differences from those in media with constant dispersivities. An alternative
approach to account for scale effects in transport processes is the use of a different
governing equation to describe solute transport in heterogeneous media. This new
approach uses a fractional derivative in the governing equation. The underlying
movement of solute particles is Lévy motions rather than Brownian motions. The
order of FADE can be estimated from the relationship between apparent
dispersivity and time.

REFERENCES

Ababou, R., and Gelhar, L. W. (1990). Self-similar randomness and spectral conditioning: analysis of
scale effects in subsurface hydrology. In “Dynamics of fluids in hierarchical porous media”
(J. H. Cushman, Ed.), Academic Press, San Diego.
Aral, M. M., and Liao, B.-Sh. (1996). Analytical solutions for two-dimensional transport equation
with time-dependent dispersion coefficient. J. Hydrol. Engng 1(1), 20–32.
Arya, A., Hewett, T. A., Larson, R., Lake, L. W. (1985). Dispersion and reservoir heterogeneity. Paper
SPE14364 presented at the 60th Annual Technical Conference, Soc. of Pet. Eng., Las Vegas, NV,
September 22– 25.
Arya, A., Hewett, T. A., Larson, R., and Lake, L. W. (1988). Dispersion and reservoir heterogeneity.
SPE Reser. Engng 3(2), 139 –148.
Bakr, A. A. (1976). Effect of spatial variations of hydraulic conductivity on groundwater flow. PhD
Dissertation. N.M. Institute of Minerals and Technology, Socorro.
Barry, D. A., and Sposito, G. (1989). Analytical solution of a convection –dispersion model with time-
dependent transport coefficients. Water Resour. Res. 25(12), 2407–2416.
Basha, H. A., and El-Habel, F. S. (1993). Analytical solution of the one-dimensional time-dependent
transport equation. Water Resour. Res. 29(9), 3209–3214.
Bear, J. (1972). In “Dynamics of Fluids in Porous Media”, pp. 764, Elsevier, New York.
Benson, D. A. (1998). The fractional advection–dispersion equation: development and application.
PhD Thesis, University of Nev., Reno.
Benson, D. A., Wheatcraft, S. W., and Meerschaert, M. M. (2000a). The fractional order governing
equation of Levy motion. Water Resour. Res. 36(6), 1413–1423.
Benson, D. A., Wheatcraft, S. W., and Meerschaert, M. M. (2000b). Application of a fractional
advection–dispersion equation. Water Resour. Res. 36(6), 1403– 1412.
260 L. ZHOU AND H. M. SELIM

Berkowitz, B., and Scher, H. (1995). On characterizations of anomalous dispersion in porous and
fractured media. Water Resour. Res. 31(6), 1461–1466.
Burns, E. (1996). Results of 2-dimensional sandbox experiments: longitudinal dispersivity
determination and seawater intrusion of coastal aquifers. Masters Thesis. Univ. of Nev., Reno.
Butters, G. L. (1987). Field scale transport of bromide in unsaturated soil. PhD Dissertation,
University of California, Riverside, pp. 243.
Butters, G. L., and Jury, W. A. (1989). Field scale transport of bromide in an unsaturated soil, 2.
Dispersion modeling. Water Resour. Res. 25, 1583–1589.
Chaves, A. S. (1998). A fractional diffusion equation to describe Lévy flights. Phys. Lett. A. 239, 13 –16.
Cushman, J. H., and Ginn, T. R. (1993). Nonlocal dispersion in media with continuously evolving
scales of heterogeneity. Transpp. Porous Media 13(1), 123–138.
Cushman, J. H., and Ginn, T. R. (2000). Fractional advection–dispersion equation: a classical mass
balance with convolution-Fickian flux. Water Resour. Res. 36(12), 3763– 3766.
Dagan, G. (1982). Stochastic modeling of groundwater flow by unconditional and conditional
probabilities, 2. The solute transport. Water Resour. Res. 18, 835 –848.
Dagan, G. (1984). Solute transport in heterogeneous porous formations. J. Fluid Mech. 145, 151– 177.
Dagan, G. (1988). Time-dependent macrodispersion for solute transport in anisotropic heterogeneous
aquifers. Water Resour. Res. 24(9), 1491–1500.
Dieulin, A. (1980). Propogation de pollution dans un aquifere alluvial: l’effet de parcours. D. Ing.
Thesis, University of Sciences and Medicine of Grenoble, Grenoble.
Domenico, P. A., and Robbins, G. A. (1984). A dispersion scale effect in model calibrations and field
tracer experiments. J. Hydrol. 70, 123– 132.
Ehlig, C. (1977). Comparison of numerical methods for solution of the diffusion–convection equation
in one and two-dimensions. In “Finite Elements in Water Resources”, pp. 1.91–1.102. Pentech
Press, London.
Feder, J. (1988). In “Fractals.” Pleum Press, New York.
Freyberg, D. L. (1986). A natural gradient experiment on solute transport in a sand aquifer, 2. Spatial
moments and the advection and dispersion of nonreactive tracers. Water Resour. Res. 22,
2031–2046.
Fried, J. J. (1972). Proceedings of the International Symposium on Modeling Techniques in Water
Resources Systems. In “Miscible pollution of ground water: a study of methodology.”
(A. K. Biswas, Ed.), Vol. 2, pp. 362–371. Environment Canada, Ottawa, Ont.
Fried, J. J. (1975). In “Groundwater Pollution”, pp. 330, Elsevier, New York.
Gelhar, L. W. (1987). Stochastic analysis of solute transport in saturated and unsaturated porous
media. In “Advances in Transport Phenomena in Porous Media”. (J. Bear and M.Y. Corapcioglu,
Eds.), pp. 656–700. Martinus Nijhoff, Dordrecht, Netherlands.
Gelhar, L. W., and Axness, C. L. (1983). Three-dimensional stochastic analysis of macrodispersion in
aquifers. Water Resour. Res. 19(2), 161–180.
Gelhar, L. W., Gutjahr, A. L., and Naff, R. L. (1979). Stochastic analysis of macrodispersion in a
stratified aquifer. Water Resour. Res. 15, 1387–1397.
Gelhar, L. W., Welty, C., and Rehfeldt, K. R. (1992). A critical review of data on field-scale dispersion
in aquifer. Water Resour. Res. 28(7), 1955–1974.
Gnedenko, B. V., and Kolmogorov, A. N. (1954). In “Limit Distributions for Sums of Random
Variables”. Addison-Wesley, Reading, MA.
Hewett, T. A. (1986). Proceedings of the 61st Annual Technical Conference, SPE Paper 15386. In
“Fractal distribution of reservoir heterogeneity and their influence on fluid transport.” Society of
Petroleum Engineers, New Orleans.
Himmelsbach, T., Hotzl, H., and Maloszewski, P. (1998). Solute transport processes in a highly
permeable fault zone of Lindau fractured rock test site (Germany). Ground Water 36(5), 792– 800.
Huang, K.-L., van Genuchten, M. T., and Zhang, R.-D. (1996). Exact solutions for one-dimensional
transport with asymptotic scale-dependent dispersion. Appl. Math. Model. 20, 298–308.
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 261

Hunt, B. (1998). Contaminant source solutions with scale-dependent dispersivities. J. Hydrol. Engng
3(4), 268 –275.
Hurst, H. E. (1951). Long term storage capacity of reservoirs. Trans. Am. Soc. Civ. Engng 116,
770 –808.
Hurst, H. E. (1957). A suggested statistical model for some time series that occur in nature. Nature
180, 494–495.
Jayawardena, A. A., and Lui, P. H. (1984). Numerical solution of the dispersion equation using a
variable dispersion coefficient: method and applications. Hydrol. Sci. J. 3(9), 293– 309.
Jenson, O. K., and Finlayson, B. A. (1978). Proceedings of the Second International Congress on
Finite Elements in Water Resources. In “Solution of the convection diffusion equation using a
moving coordinate system”, pp (4).21– (4).32.
Jury, W. A., and Roth, K. (1990). In “Transfer Functions and Solute Movement Through Soil: Theory
and Applications.” Birkhauser Verlag, Basel.
Khan, A. U. H., and Jury, W. A. (1990). A laboratory study of the dispersion scale effect in column
outflow experiments. J. Contam. Hydrol. 5, 119 –131.
Leland, D. F., and Hillel, D. (1982). A field study of solute dispersion in a shallow, unconfined aquifer.
Soil Sci. Soc. Am. J. 46, 905–912.
Liu, H. H., and Dane, J. H. (1996). An extended transfer function model of field-scale solute transport:
model development. Soil Sci. Soc. Am. J. 69, 986 –991.
Logan, L. D. (1996). Solute transport in porous media with scale-dependent dispersion and periodic
boundary conditions. J. Hydrol. 184, 261 –276.
Mackay, D. M., Freberg, D. L., Roberts, P. V., and Cherry, J. A. (1986). A natural gradient experiment
on solute transport in a sand and gravel aquifer, 1. Approach and overview of tracer movement.
Water Resour. Res. 22, 2017–2029.
Mandelbrot, B. B., and van Ness, J. W. (1968). Fractional Brownian motions, fractional noises and
applications. SIAM Rev. 10, 422 –437.
Matheron, G., and De Marsily, G. (1980). Is transport in porous media always diffusive? A
counterexample. Water Resour. Res. 26(5), 901–917.
Mercado, A. (1967). The spreading pattern of injected water in a permeability-stratified aquifer. IAHS
AISH Publ. 72, 23–36.
Mishra, S., and Parker, J. C. (1990). Analysis of solute transport with a hyperbolic scale-dependent
dispersion model. Hydrol. Process. 4, 45 –57.
Molz, F. J., Liu, H. H., and Szulga, J. (1997). Fractional Brownian motion and fractional Gaussian
noise in subsurface hydrology: a review, presentation of fundamental properties, and extensions.
Water Resour. Res. 33(10), 2273–2286.
Montroll, E. W., and Weiss, G. H. (1965). Random walks on lattices, II. J. Math. Phys. 6(2), 167–181.
Ndumu, A. S., and Addison, P. S. (2001). Scale-dependent subsurface dispersion: a fractal-based
stochastic model. J. Hydrol. Engng 6(1), 34–42.
Neuman, S. P. (1990). Universal scaling of hydraulic conductivities and dispersivities in geologic
media. Water Resour. Res. 26(8), 1749– 1758.
Neuman, S. P., and Zhang, Y. -K. (1990). A quasi-linear theory of non-Fickian and Fickian subsurface
dispersion, 1. theoretical analysis with application to isotropic media. Water Resour. Res. 16(5),
887 –902.
Neuman, S. P., Winter, C. L., and Newman, C. M. (1987). Stochastic theory of field-scale Fickian
dispersion in anisotropic porous media. Water Resour. Res. 23(3), 453– 466.
Osler, T. J. (1971). Taylor’s series generalized for fractional derivatives and application. SIAM
J. Math. Anal. 2(1), 37–47.
Pachepsky, Y., Benson, D., and Rawls, W. (2000). Simulating scale-dependent solute transport in soils
with the fractional advective–dispersive equation. Soil Sci. Soc. Am. J. 64, 1234–1243.
Pang, L., and Hunt, B. (2001). Solutions and verification of a scale-dependent dispersion model.
J. Contamin. Hydrol. 53, 21–39.
262 L. ZHOU AND H. M. SELIM

Peaudecerf, P., and Sauty, J. P. (1978). Application of a mathematical model to the characterization of
dispersion effects of groundwater quality. Prog. Water Technol. 10(5/6), 443–454.
Philip, J. R. (1986). Issues in flow and transport in heterogeneous porous media. Transp. Porous
Media 1, 319–338.
Pickens, J. F., and Grisak, G. E. (1981). Scale-dependent dispersion in a stratified granular aquifer.
Water Resour. Res. 17(4), 1191–1211.
Pickens, J. F., and Grisak, G. E. (1981). Modeling of scale-dependent dispersion in hydrogeologic
systems. Water Resour. Res. 17(6), 1701–1711.
Porro, I., Wierenga, P. J., and Hills, R. G. (1993). Solute transport through large uniform and layered
soil columns. Water Resour. Res. 29, 1321–1330.
Russo, D. (1998). Stochastic modeling of scale-dependent macrodispersion in the vadose zone. In
“Scale Dependence and Scale Invariance in Hydrology.” (G. Sposito, Ed.), pp. 266– 290.
Cambridge University Press, Cambridge.
Sahimi, M. (1993). Fractal and superdiffusive transport and hydrodynamic dispersion in
heterogeneous porous media. Transp. Porous Media 13(1), 3–40.
Sauty, J.-P. (1980). An analysis of hydrodispersive transfer in aquifers. Water Resour. Res. 16(1),
145–158.
Scher, H., and Lax, M. (1973). Stochastic transport in a disordered solid. Phys. Rev. B. 7(10),
4491–4502.
Schumer, R., Benson, D. A., Meerschaert, M. M., and Wheatcraft, S. W. (2001). Eulerian derivation of
the fractional advection– dispersion equation. J. Contamin. Hydrol. 48, 69–88.
Serrano, S. E. (1992). The form of the dispersion equation under recharge and variable velocity, and its
analytical solution. Water Resour. Res. 28(7), 1801–1808.
Silliman, S. E., and Simpson, E. S. (1987). Laboratory evidence of the scale effect in dispersion of
solutes in porous media. Water Resour. Res. 23(8), 1667–1673.
Smith, L., and Schwartz, F. W. (1980). Mass transport 1. A stochastic analysis of macroscopic
dispersion. Water Resour. Res. 16(2), 303–313.
Su, N.-H. (1995). Development of the Fokker– Planck equation and its solutions for modeling
transport of conservative and reactive solutes in physically heterogeneous media. Water Resour.
Res. 31(12), 3025– 3032.
Sudicky, E. A., and Chery, J. A. (1979). Field observations of tracer dispersion under natural flow
conditions in an unconfined sandy aquifer. Water Pollut. Res. Can. 14, 1–17.
Sudicky, S. A., Cherry, J. A., and Frind, E. O. (1983). Migration of contaminants in groundwater at a
landfill: a case study, 4. A natural-gradient dispersion test. J. Hydrol. 63, 81– 108.
Wheatcraft, S. W., and Tyler, S. W. (1988). An explanation of scale-dependent dispersivity in
heterogeneous aquifers using concepts of fractal geometry. Water Resour. Res. 24(4), 566– 578.
Taylor, G. I. (1921). Diffusion by continuous movements. Proc. Lond. Math. Soc. 20, 196–212.
Taylor, S. R., and Howard, K. W. F. (1987). A field study of scale-dependent dispersion in a sandy
aquifer. J. Hydrol. 90, 11–17.
Toride, N., Leij, F., and van Genuchten, M. Th. (1995). Version 2.0. Research Report 137. In “The
CXTFIT Code for Estimating Transport Parameters from Laboratory or Field Tracer
Experiments.” US Salinity Laboratory, Riverside, CA.
Tyler, S. W., and Wheatcraft, S. W. (1992). Reply. Water Resour. Res. 28(5), 1487–1490.
Yaglom, A. M. (1987). In “Correlation Theory of Stationary and Related Random Functions: Basic
Results.” Springer, New York.
Yates, S. R. (1990). An analytical solution for one-dimension transport in heterogeneous porous
media. Water Resour. Res. 26, 2331– 2338. (Correction, Water Resour. Res. 27, 2167, 1991).
Yates, S. R. (1992). An analytical solution for one-dimension transport in porous media with an
exponential dispersion function. Water Resour. Res. 28, 2149–2154.
Zaslavsky, G. M. (1994). Renormalization group theory of anomalous transport in systems with
Hamiltonian chaos. Chaos. 4(1), 25–33.
 SCALE-DEPENDENT DISPERSION IN SOILS: AN OVERVIEW 263

Zhang, R. (2000). Generalized transfer function model for solute transport in heterogeneous soils. Soil
Sci. Soc. Am. J. 64, 1595–1602.
Zhang, Y.-K., and Neuman, S. P. (1990). A quasi-linear theory of non-Fickian and Fickian subsurface
dispersion, 2. Application to anisotropic media and the Borden site. Water Resour. Res. 26(5),
903 –913.
Zhang, R., Huang, K., and Xiang, J. (1994). Solute movement through homogeneous and
heterogeneous soil columns. Adv. Water Resour. 17, 317 –324.
Zou, S.-M., Xia, J. H., and Koussis, A. D. (1996). Analytical solutions to non-Fickian subsurface
dispersion in uniform groundwater flow. J. Hydrol. 179, 237– 258.
 Index
A B

acid soils bacterial wilt (Burkhoderia solanacearum) 158,

see also soil pH 159, 160, 188–9
rice cultivation 67–8, 74, 90 boron (B) management, lowland rice cultivation
advection-dispersion equation (ADE), solute 123–7
transport 223–4, 255 Brazil, rice production 66, 68
Aegilops cylindrica see jointed goatgrass breeding methods, groundnut 174
aflatoxins, groundnut crops 159, 160, 161, Bromus tectorum see downy brome
189 –90 Brownian motion, scale-dependent dispersion
Africa, groundnut production 155 –8 231, 255
agricultural chemicals Burkhoderia solanacearum (bacterial wilt) 158,
contaminated soil dust 15–16, 20–1 159, 160, 188–9
fertilizer pollution 69, 90–1
agricultural workers, soil dust exposure 9, 10, 13, C
16 –19, 21– 3, 25–7
air pollution, fertilizer use 69 calcium (Ca) management, lowland rice
allergic reactions, soil dust 20, 21 cultivation 107–15
American Conference of Government Industrial CDE see convection–diffusion equation
Hygienists, threshold limit value cereals
(TLVw), inhalable dust 16–18 see also corn; rice; wheat
Ammonium sulfate fertilizer 74 chromosome number and genome size
anaerobic soils, rice cultivation 65, 70–2, 135–6 variation 180
analytical approaches, scale-dependent dis- genetic linkage mapping 181, 184
persion 239–51 children
animal studies, agricultural dust effects 22 protective effects of farm environments 24–5
annual grasses, weed management 35, 38–44 soil dust exposure 15, 20– 1
aquifers, solute transport 223, 226–7, 230 chromosome number variation, crop plants 180
Arachis hypogaea see groundnut CLT see convection lognormal transfer
Arachis wild species coefficient of variation (CV), dispersion travel
A. hypogaea crosses 185–6 time 238–9
gene introgression to A. hypogaea 195 column experiments, solute transport studies
genetic variability 162– 3, 164–7 225–6
arid regions comparative mapping, crop plants 183–5
see also semi-arid competitiveness, crops and weeds 45–53
soil dust generation 3, 7 computer simulations, solute transport 224
Asia construction activities, soil dust dispersion 10
groundnut production 155–8 contaminants
rice cultivation 64–5 aflatoxins in groundnut 159, 160, 161,
asymptotic function 189–90
see also linear asymptotic model scale-dependent dispersion in soil 223, 224
time-dependent dispersivity 240 soil dust 3, 15– 16, 20–1
atmosphere continuous cropping rotations, semi-arid
air pollution 69 environments 35, 38 –40
soil dust quantities 2, 3 continuous time random walks (CTRW), particle
Australia, herbicide resistance 36 transport 255

265
266 INDEX

convection lognormal transfer (CLT), function disease, groundnut crops 158–61

model of dispersion 237–8 disease resistance in groundnut
convection–diffusion equation (CDE) conventional breeding 169–71, 175, 198,
scale-dependent dispersion 259 200
analytical solutions 239– 51 genetic basis 187–90
numerical approaches 251 –4 genetic transformations 186– 7, 198
stochastic approaches 231, 237–9 molecular breeding 187 –90, 194 –9,
conventional genetic improvement, groundnut 200– 1
168–73, 174–8, 197–8, 200 species variability 163 –6
corn (Zea mays L.), weed management 34–5, dispersion mechanisms
38–40, 46–8, 49 soil dust 2– 11
costs, herbicide inputs 36 solute transport 223–63
critical nutrient levels, rice 79– 80, 93–7, distance-dependent dispersivity, solute transport
103–4, 110–13, 119–20, 125 –6 224– 7, 231, 243 –51, 258 –9
crop competitiveness, weed management downy brome (Bromus tectorum L.), weed
45–53 management 39, 40, 41–2, 43–4
crop plants drought tolerance, groundnut 171, 176, 190 –1
chromosome number and genome size dust see soil dust
variation 180
genetic linkage mapping 179–85 E
crop rotations
cycle-of-four benefits 53–7 early leaf spot (ELS) in groundnut 158, 160, 161
lowland rice fields 67, 68, 100 resistance 164, 175, 187 –8
semiarid Great Plains 34–5, 38–44, 53–7 early maturing cultivars, groundnut 172, 177
weed management 38–44, 53–4 ecologically based weed management, semiarid
crop yields environments 33 –62
ecological weed management 49, 51, 54 efficiency, nutrient use in rice 80– 6, 89–90,
groundnut genetic traits 171 –2, 174, 177– 8, 97– 8, 104–5, 107
192, 197 –8, 201 elite germplasms, groundnut 168–74, 200
rice nutrient management 75 –6, 87, 90, ELS see early leaf spot
134–6 endotoxins 13, 25
CTRW see continuous time random walks environmental impacts, N fertilizer use 69, 91
CV see coefficient of variation erosion, wind 3–7, 10, 19–20
cycle-of-four crop rotations, benefits 53– 7 experimental evidence, scale-dependent
dispersion 224– 7
D exponential function
distance-dependent dispersivity model
deficiency in rice 244
boron 123 –4, 125 time-dependent dispersivity 240
calcium 110 weeds growth 41 –2
iron 127, 129
magnesium 110 F
manganese 132, 133, 134
micronutrients 115–16 FADE see fractional order advection–dispersion
nitrogen 73 equation
phosphorus 90, 93 fallow, weed control 34 –5, 38–9, 58
potassium 102–3 farm environments
sulfur 110 see also agricultural workers
zinc 116, 118–19 allergy protection 24–5
development, rice plants 74–6, 79 –80, 93–7, fBm see fractal Brownian motion
111 Fe see iron
 INDEX 267

fertilization practices genetic linkage mapping

see also nutrient management crop plants 181
lowland rice cultivation groundnut 178–82
boron 126–7 soybeans 182– 3
calcium 113 genome size variation, crop plants 180
iron 130– 2 genotypes
magnesium 113 groundnut traits 187–92
manganese 134 nutrient efficient rice 89 –90, 98, 107, 123
nitrogen 86 –9 tall cultivars 45– 6
phosphorus 98– 9 germplasm lines, groundnut 168– 74
potassium 106 –7 Glycine max see soybean
sulfur 113–15 grain yields see crop yields
zinc 120–2 grasses, weed management 35, 38–44
N banding for weed management 45–6, 48, GRD see groundnut rosette disease
49 –50, 51– 2 Great Plains, ecologically based weed
pollution 69, 90 –1 management 33–62
fGn see fractal Gaussian noise green foxtail (Setaria viridis), weed management
Fickian diffusion, scale-dependent dispersion 35, 38–40, 41–2, 43–4
234, 255 groundnut (Arachis hypogaea L.) 153–221
field studies, scale-dependent dispersion 224, biotechnology applications 178 –87
226 –7, 257 –8 comparative mapping 183–5
flooding, lowland rice cultivation 65– 136 conventional genetic improvement 168–73,
Fokker–Planck equations, scale-dependent 174–8, 197 –8, 200
dispersion 255 crop production 154–8
fractal Brownian motion (fBm), solute transport genetic basis of traits 187–92
231 –5 genetic linkage mapping 178–82
fractal Gaussian noise (fGn), solute transport genetic transformations 186–7, 198
231, 232, 234 –5 genetic variability 162 –74
fractal models, scale-dependent dispersion 227, marker-assisted backcross breeding 195–7,
231 –7 198, 199
fractional order advection– dispersion equation marker-assisted selection 194– 9
(FADE) 255– 8 PCR-based markers 192–4
fugitive dust pests and diseases 158–61
see also wind erosion production constraints 158– 61, 199
exposure 19–20 wide crosses 185–6
health effects 24 groundnut rosette disease (GRD) 158– 9, 188
fungal diseases growth and development, rice 74–6, 79–80,
groundnut 158, 160, 161 93–7, 111
resistance 164, 169 –70, 170, 175, 187–8
H
G
harvesting, soil dust emissions 9 –10
gene banks, ICRISAT 162–3, 168, 178, 186, health effects, soil dust 2, 12 –13, 15, 20 –7
193, 200 Helianthus annuus see sunflower
genetic diversity analysis, groundnut 178 –9 herbicides
genetic improvement of groundnut crop injury 51
biotechnology applications 178– 87 no-till systems 34
conventional breeding 168 –73, 174– 8, resistance 35–6, 47, 55–6
197 –8, 200 weed management 57–8
genetic transformations 186 –7, 198 household dust, contaminated 15, 20–1
molecular breeding 187–201 human activity, soil dust emission 7–11
268 INDEX

hybridization, wild Arachis species 162–3 longspine sandbur [Cenchrus longsopnus

hydrodynamic dispersion, solute transport (Hack.) Fern.], weed management 42
223–4 lowland rice production, nutrient management
63– 152
I
M
ICRISAT gene bank, Arachis species 162 –3,
168, 178, 186, 193, 200
macrodispersivity theory, scale-dependent
inorganic compounds, dust effects on health
dispersion 227, 228–31
12–13, 21–2
magnesium (Mg) management, lowland rice
insect pests of groundnut 158, 159–61
cultivation 107 –15
resistance 163 –6, 171, 175 –6, 186 –7
manganese (Mn) management, lowland rice
iron (Fe)
cultivation 132 –4
availability to groundnut 158
marker systems, soybeans 182–3
lowland rice cultivation 127– 32
marker-assisted backcross breeding, groundnut
irrigated-lowland rice, nutrient management
195– 7, 198, 199
63–152
marker-assisted selection (MAS), groundnut
J 194– 9, 200–1
maturity time, groundnut 172, 177, 197, 198, 200
jointed goatgrass (Aegilops cylindrica Host), model systems, legumes 182–5
weed management 39, 42, 43 –4 molecular breeding, groundnut 153–221
K
N
K see potassium
natural gradient tracer tests, scale-dependent
L dispersion 227, 257
nitrogen harvest index (NHI), rice 78–9
laboratory studies, scale-dependent dispersion nitrogen (N)
224, 225 –6, 257– 8 fertilizer banding for weed management
land improvement, cycle-of-four rotations 54 45– 6, 48, 49– 50, 51–2
land preparation, soil dust emissions 8–9 lowland rice cultivation 68 –90
late leaf spot (LLS) in groundnut 158, 160 no-till systems, weed management 34–5, 43–4,
resistance 164, 169–70, 175, 188 47, 51–3, 57– 8
legumes non-Fickian behaviour, scale-dependent
see also groundnut dispersion 255
chromosome number and genome size numerical approaches, scale-dependent
variation 180 dispersion 251– 4
comparative mapping 183– 5 nutrient management in lowland rice cultivation
genetic linkage mapping 181, 182 63– 152
Lévy motion, scale-dependent dispersion 255 boron 123–7
liming, rice cultivation 67–8, 113 calcium 107 –15
linear asymptotic model, distance-dependent iron 127–32
dispersivity 244 magnesium 107–15
linear function manganese 132–4
distance-dependent dispersivity model 244 micronutrients 115 –34
time-dependent dispersivity 240 nitrogen 68 –90
livestock, protective effects of contact 25 phosphorus 90–9
LLS see late leaf spot potassium 100– 7
longitudinal dispersivities, variation with study sulfur 107–15
224–5 zinc 116–23
 INDEX 269

O Q

oil content, groundnut 163, 167 quality traits in groundnut

oilseeds conventional breeding 173, 176– 7
see also groundnut genetic basis 191– 2
chromosome number and genome size variability 163, 167, 168
variation 180
genetic linkage mapping 181 R
organic compounds, soil dust 6, 13, 21–2
Oryza sativa see rice reduction reactions, flooded rice soils 71,
oxygen status of soil, flood-irrigated rice 65, 135
70 –2, 135 resistance
see also disease resistance
P herbicides 35 –6, 47, 55 –6
insect pests 163 –6, 171, 175 –6, 186 –7
P see phosphorus resource-poor farming, groundnuts 158,
Panicum miliaceum see proso millet 199
parabolic function, time-dependent dispersivity resource-use-efficiency, cycle-of-four rotations
240 54–5
particle motion, scale-dependent dispersion 224, respirable dust
231 –7, 255 agricultural operations 7–10, 16 –17
particle size health effects 21–4
agriculturally dispersed dust 8
particle size effects 13–14
contaminated soil dust 15–16
rice (Oryza sativa)
respirable dust 13–14
boron management 123 –7
wind erosion dust 3 –5, 6, 14 –15
calcium management 107–15
partitioning, nitrogen uptake by rice 74–9
classifications 64–5
paved roads, dust emissions 19– 20
cultivation 64– 5
peanut see groundnut
growth and development 74–6, 79 –80,
pesticides, soil dust exposure 15, 20–1
93–7, 111
pests see disease; insect pests
iron management 127 –32
phosphorus (P)
lowland production nutrient management
availability to groundnut crops 158
lowland rice cultivation 90–9 63–152
photoperiod sensitivity, groundnut 177 magnesium management 107 –15
planting time, crop competitiveness with weeds manganese management 132– 4
48 –9, 51 micronutrient toxicity 125, 127–8, 129, 132,
pollution 133
see also contaminants micronutrients management 115–34
fertilizers 69, 90–1 nitrogen management 68–90
porous media, scale-dependent dispersion nutrient efficient genotypes 89–90, 98, 107,
224 –63 123
potassium (K), lowland rice cultivation 100–7 phosphorus management 90–9
precipitation potassium management 100–7
resource-use-efficiency 54–5 sulfur management 107 –15
soil dust 5, 7 zinc management 116 –23
tillage effects in semiarid environments 47 roads, dust emissions 10 –11, 19– 20
proso millet (Panicum miliaceum L.), row widths, crop competitiveness with weeds 48,
weed management 35, 38 –9, 49–50
50 –3 rust in groundnut 158, 160, 161
protein content, groundnut 163, 167 resistance 164, 169 –70, 170, 175, 187
270 INDEX

S phosphorus 91–2
potassium 100–7
S see sulfur sulfur 108–9
saltation 4, 5, 8 zinc 116–17
sandbox experiments, solute transport studies soil dust 1 –32
226, 257 agricultural operations 7–10, 15, 16– 19
SAT see semi-arid tropics characteristics 11–16
scale-dependent dispersion 223 –63 contaminated 3, 15– 16, 20–1
analytical solutions of CDE 239–51 dispersion mechanisms 2–11
distance-dependent dispersivity 224–7, 231, dust storms 3–4, 19, 24
243–51, 258–9 health effects 2, 12–13, 15, 20–7
field studies 224, 226–7 human activity 7–11
fractal models 227, 231–7 nuisance dust 16
fractional order advection–dispersion occupational exposure 16 –19
equation 255–8 particle size distribution 13–15
interpretation of evidence 227–8 reducing exposure 25–7
laboratory studies 224, 225– 6 roads 10 –11, 19–20
macrodispersivity theory 227, 228–31 wind erosion 3– 7, 10
numerical approaches 251 –4
soil moisture, affecting dust emissions 7, 9
stochastic approaches 227, 228– 39
soil pH
time-dependent dispersivity 227, 237,
liming rice fields 67 –8
239–43, 258–9
nitrogen fertilizer choice 74
transfer function models 237–9
nutrient availability, groundnut crops 158
seasonal variation, N uptake in rice 74–9
potassium availability 100
seed bank dynamics, weed management 41, 43– 4
zinc availability 122
seed emergence, weed management 48 –9, 50
soil types
seed quality in groundnut
rice cultivation 65–7, 90, 100
conventional breeding 173, 176 –7, 197–8,
wind erosion 6– 7
200
genetic basis 191 –2 solute transport, scale-dependent dispersion
variability 163, 167, 168 223– 63
seed yields see crop yields soybeans (Glycine max), genetic linkage
seeding rates/patterns, crop competitiveness with mapping 182 –3
weeds 45–6, 48, 49–50, 51 –2 stochastic approaches, scale-dependent
semi-arid environments, weed management dispersion 227, 228–39
33–62 stress tolerance see disease resistance; drought
semi-arid tropics (SAT), groundnut breeding tolerance; insect pest resistance
153–221 sulfur (S) management, lowland rice cultivation
Setaria viridis see green foxtail 107– 15
soil sunflower (Helianthus annuus L.), weed
anaerobic conditions 65, 70–2, 135–6 management 35, 48–50, 51
contaminant transport 223, 224 surface creep, wind erosion 4
scale-dependent dispersion 223– 63 sustainability, lowland rice nutrient management
wind erosion 3–7, 10 63– 4, 136
soil chemistry
lowland rice fields 135–6 T
boron 124
calcium 108–9 tall cultivars, crop competitiveness with weeds
iron 128 45– 6
magnesium 108–9 threshold friction velocity, wind erosion 4, 6–7
manganese 132– 3 threshold limit value (TLVw), inhalable dust,
 INDEX 271

American Conference of Government water pollution, fertilizers 69, 90–1

Industrial Hygienists 16–18 water regime, rice cultivation 65
tillage, weed management 43–4, 47, 51– 3, weathered soils, nutrient content 90, 100
57 –8 weed management
time-dependent dispersivity, solute transport crop rotation effects 38– 9, 41–4, 53–4
227, 237, 239 –43, 258 –9 ecological processes 36 –58
TLVw see threshold limit value groundnut crops 161
tortuous flow paths, solute transport 236–7 herbicide resistance 35 –6, 47, 55 –6
toxic chemicals increasing crop competitiveness 36, 37,
aflatoxin contaminated groundnut 159, 160, 45–53
161, 189–90 reducing weed density 36–7, 38–44
contaminated soil dust 3, 15 –16, 20– 1 seed bank dynamics 41
rice micronutrients 125, 127–8, 129, 132, semiarid Great Plains 33–62
133 wheat (Triticum aestivum L.), weed management
tracer transport, laboratory studies 224, 225–6 34–5, 37, 38–40, 45–6, 47
transfer function models, scale-dependent whole plant nutrient content, rice 76 –8, 94–5,
dispersion 227, 237–9 111–13
transgenics, groundnut 186 –7 wide crosses, Arachis species 185–6
Triticum aestivum see winter wheat wild Arachis species
tubers, chromosome number and genome size
A. hypogaea crosses 185 –6
variation 180
gene introgression to A. hypogaea 195
two-year intervals in crop rotation, weed
genetic variability 162 –3, 164– 7
management 38 –9, 41– 4, 53–4
wind erosion
U agriculture effects 6, 10
factors 5–7
United States of America (USA) fugitive dust exposure 19–20
Great Plains weed management 33 –62 impacts 3–4
inhalable dust limits 16–17 mechanisms 4–5
road dust 10 –11, 19– 20 winter wheat (Triticum aestivum L.), weed
wind erosion 3, 4 management 34–5, 37, 38–40, 45–6,
unpaved roads, dust emissions 10 –11 47
urea fertilizer 74
Y
V
yields see crop yields
virus diseases, groundnut 158 –61, 164, 170–1,
188
Z
W
Zea mays see corn
water management, fertilizer utilization 68–9, zinc (Zn) management, lowland rice cultivation
70, 135 116–23