Advances in Agronomy, Volume 35

ADVANCES IN

AGRONOMY VOLUME 35

CONTRIBUTORS TO THIS VOLUME C. R. ADAIR PHILLIP BARAK

T. T. CHANG YONACHEN

S. K . DEDATTA

D. L. FR~ESNER C. HAGEDORN R. J . HANKS

G. HUCK MORRIS

T. H . JOHNSTON M. B. KIRKHAM W. E. KNIGHT

F? MIEDEMA V. P. RASMUSSEN N . K. SAVANT

HOWARD M. TAYLOR

V. H. WATSON

ADVANCES IN

AGRONOMY Prepared in cooperation with the AMERICAN SOCIETY OF AGRONOMY

VOLUME 35

Edited by N. C. BRADY Science and Technology Bureau Agency for International Development Department of State Washington, D . C .

ADVISORY BOARD H. J. GORZ,CHAIRMAN

E. J. KAMPRATHT. M. STARLING

J. B. POWELL J. W.BIGGAR M. A. TABATABAI M. STELLY, EX OFFICIO, ASA Headquarters 1982

ACADEMIC PRESS A Subsidiary of Harcourl Brace Jovanovich, Publishers

New York London Paris San Diego San Francisco SBo Paulo Sydney Tokyo Toronto

COPYRIGHT @ 1982, BY ACADEMIC PRESS, INC. ALL RIGHTS RESERVED. NO PART OF THIS PUBLICATION MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM OR BY ANY MEANS, ELECTRONIC OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT PERMISSION IN WRITING FROM THE PUBLISHER.

ACADEMIC PRESS, INC.

111 Fifth Avenue, New York. New York 10003

United Kitigdotii Editiori published by ACADEMIC PRESS, INC. ( L O N D O N ) LTD. 24/28 Oval Road, London NWI 7DX

LIBRARY OF CONGRESS CATALOG CARD NUMBER: 50-5598 ISBN 0-12-000735-5 PRINTED IN THE UNITED STATES OF AMERICA

82 83 84 85

98 76 5 4 3 2 1

CONTENTS CONTRIBUTORS ..................................................... PREFACE ...........................................................

ix xi

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

Morns G . Huck and Howard M . Taylor I . Introduction ................................................ I1. Physical Designs: General Types .............................. 111. Construction Details and Design Features ...................... IV. Some Techniques for Observing and Recording Root System Parameters .......................................... V. Experimental Design: Data Acquisition and Analysis ............ VI . Summary: Advantages and Disadvantages of Rhizotrons for Use in Root Investigations ........................................ References .................................................

1 2 10 20 27

32 33

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

T. T. Chang. C . R . Adair. and T. H . Johnston I. I1 . Ill . IV. V. VI . VII .

38 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Diversity in Rice Genetic Resources ........................... 45 Recent Efforts in Genetic Conservation ........................ Dissemination and Evaluation of Germ Plasm . . . . . . . . . . . . . . . . . . 58 68 Preservation of Germ Plasm .................................. 71 Use of Germ Plasm .......................................... 80 Endeavors for the Future ..................................... 85 References .................................................

THE EFFECTS OF LOW TEMPERATURE ON Zea mays

I? Miedema 1. Introduction ................................................ 11. Freezing Injury ............................................. V

93 94

vi

CONTENTS

Ill . IV. V. VI .

Damage by Low Nonfreezing Temperatures .................... Growth and Development at Suboptimal Temperatures . . . . . . . . . . Breeding for Low-Temperature Adaptation ..................... Summary ................................................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

95 103 119

124 124

AGRICULTURAL USE OF PHOSPHORUS IN SEWAGE SLUDGE

M . B . Kirkham I. I1 . I11 . IV.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Concentration of Phosphorus in Sludges ....................... Agricultural Use of Phosphorus in Sludges ..................... Summary and Conclusions ................................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

129 131 144 154 156

SUBTERRANEAN CLOVER IN THE UNITED STATES

W. E . Knight. C . Hagedorn. V. H . Watson. and D . L . Friesner I . Introduction ................................................ 11. Potential Use of Subclover

................................... Ill . Breeding Subclover . . . . . . . . . . . . . . 1v. Seed Characteristics . . . . . . . . . . . . . ...................................... V. Nitrogen Fixation

166

167

171

v1. Fertilization and N

VII . VIII . IX . X.

Climatic Variations ....................... Establishment and Management . . . Morphological Char Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References ....................... ....

183

188 1x9

PREDICTING CROP PRODUCTION AS RELATED TO PLANT WATER STRESS

R . J . Hanks a n d V. I? Rasmussen I. I1 . Ill . IV. V. VI .

Introduction ................................................ Review of the Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Measuring ET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Estimating ET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Estimating Yield ............................................ Growth Stage Effects ........................................

193 194 199

201 203 205

CONTENTS

VII . Rasmussen and Hanks Spring Wheat Model . . . . . . . . . . . . . . . . . . . . VIII . Rasmussen and Kanemasu Winter Wheat Model . . . . . . . . . . . . . . . . IX . Hill. Johnson. and Ryan Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Morgan. Biere. and Kanemasu Model for Corn . . . . . . . . . . . . . . . . . XI . Other Models with Moisture Stress Included . . . . . . . . . . . . . . . . . . . XI1 . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

x

vii 205 207 209 211 212 213 214

IRON NUTRITION OF PLANTS IN CALCAREOUS SOILS

Yona Chen a n d Phillip Barak I . Introduction

................................................

217

XI . Soil Iron Compounds and Methods for Their Extraction . . . . . . . . . 218 111 . Iron Nutrition of Plants ...................................... Iv. Correction of Iron Deficiency ................................. References .................................................

222 230 238

NITROGEN TRANSFORMATIONS IN WETLAND RICE SOILS

N . K . Savant a n d S . K . De Datta I . Introduction ................................................ I1 . Chemical Nature of Soil Nitrogen ............................. 111. Physical and Physicochemical Processes Relevant to Nitrogen Transformations .................................... I v. Biochemical Nitrogen Transformations ......................... V. Fate of Fertilizer Nitrogen ................................... VI . Regulating Nitrogen Transformation Processes . . . . . . . . . . . . . . . . . . VII . Unresolved Challenges ....................................... References .................................................

INDEX .............................................................

241 244 249 261 286 291 293 294 303

This Page Intentionally Left Blank

CONTRIBUTORS

Numbers

in

parentheses indicate the pages on which the authors’ contnbutions begin

C. R. ADAIR* (37), Agricultural Research Service, U . S . Department of Agriculture, Beltsville, Maryland 20705 PHILLIP BARAK (217), The Seagram Centre f o r Soil and Water Sciences, The Hebrew University of Jerusalem, Rehovot, Israel T. T. CHANG (37), Department of Plant Breeding, International Rice Research Institute, Manila, Philippines YONA CHEN (217), The Seagram Centre f o r Soil and Water Sciences, The Hebrew University of Jerusalem, Rehovot, Israel S. K. D E DATTA (241), Department of Agronomy, International Rice Research Institute, Manila, Philippines D. L. FRIESNER (165), Department of Agronomy, Mississippi State University, and Mississippi Agricultiml and Forest Experiment Station, Mississippi State, Mississippi 39762 C. HAGEDORN (163, Department of Agronomy, Mississippi State University, and Mississippi Agricultural and Forest Experiment Station, Mississippi State, MissisJippi 39762 R. J . HANKS (193), Department of Soil Science and Biometeorology. Utah State University, Logan, Utah 84322 MORRIS G. HUCK (l), Agricultural Research Service, U . S . Department of Agriculture, Auburn University, Auburn, Alabama 36849 T. H . JOHNSTONt (37), Agricultural Research Service, U . S . Department of Agriculture, University of Arkansas Rice Research and Extension Center, Stuttgart, Arkansas 72160 M. B . KIRKHAM (1291, Evapotranspiration Laboratory, Kansas State University Waters Annex, Manhattan, Kansas 66506 W. E. KNIGHT (163, Crop Science Research Laboratory, USDA-ARS, MissiJAippi State, Mississippi 39762 I? MIEDEMA (93), Foundation f o r Agricultural Plant Breeding, 6700 A C Wageningen, The Netherlands V. I? RASMUSSEN (193), Department of Soil Science and Biometeorology, Utah State University, Logan, Utah 84322

*Present address: 3 Bedwell Lane, Concordia, Bella Vista, Arkansas 72712 ?Present address: 13 C & H Circle, Stuttgart. Arkansas 72160.

ix

X

CONTRIBUTORS

N. K . SAVANT* (241), Department of Agronomy, International Rice Research Institute, Manila, Philippines HOWARD M. TAYLOR (I), Department of Agronomy, Iowa State University, Arnes, Iowa 50011 V. H. WATSON (165), Department of Agronomy, Mississippi State University, and Mississippi Agricultural and Forest Experiment Station, Mississippi State, Mississippi 39762

*Present address: International Fertilizer Development Center. PO. Box 2040. Muscle Shoals, Alabama 35660.

PREFACE During the past 25 years, the developing countries of the world have doubled their food production. Increased use of food-producing inputs such as irrigation, fertilizers, and monetary credit is responsible for much of this remarkable achievement. But in most countries increased food production has been based on the development of new and improved technologies and on policies which encourage farmers to use these technologies. Agricultural scientists in both the developing and developed countries have produced these new technologies, and soil and crop scientists certainly have done their share. This volume provides evidence of agronomists’ contributions to the world’s ability to produce food. Rice, the food crop for most of the world’s poor, is the subject of two articles. One deals with the genetic resources of this crop, and the other with transformations of nitrogen in paddy soils. The effects of stress on crop production are addressed in three articles: one concerned with low temperatures, one with plant water stress, and one with calcareous soils. They illustrate continuing attempts to address the problems of large and important food-producing areas. The remaining three articles likewise focus on practical problems facing food producers. One concerns an important forage crop, subterranean clover, and the role it plays in modern agriculture. A second reviews the use of rhizotrons in root research and reminds us of the significance of roots, especially in relation to water utilization. The third summarizes work on sewage sludge as a source of phosphorus for agriculture. We express appreciation to the scientists from different countries who have made these important contributions.

N. C . BRADY

xi

This Page Intentionally Left Blank

ADVANCES IN AGRONOMY, VOL 35

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH Morris G. Huck* and Howard M. Taylor? *Agricultural Research Service, U S . Department of Agriculture, Auburn University, Auburn, Alabama and ?Department of Agronomy, Iowa State University, Ames, Iowa

I. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............. 11. Physical Designs: General Types. . . . . . . . . . . . . A. Simple Pits or Boxes with Glass Walls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B , Multicompartment BelobGround Observation Facilities C. Special Adaptations to Investigate Specific Problems ..................... 111. Construction Details and Design Features. . . . . . . . . . . ............ A. Window Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Control of Soil Physical Properties in Reco IV. Some Techniques for Observing and Recording A. Measurement of Root System Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Population and Spatial Distribution over Time: The Sum of Growth and Death ................. Rates in Each Localized Area . . . . C. Validation of Root Density Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Experimental Design: Data Acquisition and Analysis ..... A. Controlled Environments versus Characterization of Natural Environments. . . . B. Measurement of Root Functions: Water Removal, Mineral Uptake, and Biological Oxygen Demand . . . . . . . . . . . . . . . . . . . . . . . . . C. Selection of Data for Analysis and Storage. . . . . . . . . . . . . D. Questions Which Can Be Addressed in Rhizotron Experim VI. Summary: Advantages and Disadvantages of Rhizotrons for Use in Root

...................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 2 2 3 S 10 I1 1s 20 21 26 21 21 29 29 30 30 32 33

1. INTRODUCTION Roots of higher plants perform several important functions: they provide anchorage, supply water and minerals, and have a regulatory role as well. Because they grow underground and are not easily accessible, roots have been much less completely studied than plant shoots. Many techniques have been used to increase the accessibility of plant roots. Bohm (1979) reviewed the methods for studying root growth and distribution in 1

ISBN 0-12-000735-5

2

MORRIS G . HUCK AND HOWARD M . TAYLOR

field soils. One method he reviewed was the observation of roots growing in soil behind transparent walls. Sachs (1873), one of the pioneers in the study of plant roots, used a simple soil-filled box with one glass wall. The facilities for studying root growth gradually became more complex. Recently, various kinds of sophisticated underground chambers have been constructed which permit plant roots to be studied under replicated conditions while shoots are exposed to field environments. These larger root observation laboratories have been called “rhizotrons,” the Greek “rhizos,” meaning root and “tron,” meaning device for studying. (The word was coined to parallel the words “phytotron,” “edaphotron,” “cyclotron,’’ and other specialized facilities used in modern science.) This article will confine its coverage to facilities where plant roots can be visually observed growing in soil while their tops are growing either under outdoor conditions or in transparent enclosures exposed to sunlight. Although we shall mention other facilities as well, we shall concentrate more heavily on design features and operating characteristics of the Auburn, Alabama, and Ames, Iowa, rhizotrons where the authors have conducted root research.

I I . PHYSICAL DESIGNS: GENERAL TYPES

A. SIMPLE PITSOR BOXESWITH GLASSWALLS

The simplest rhizotron design is a transparent panel covering a vertical face of soil that contains growing roots. The roots are observed from an adjoining pit that usually is covered to exclude light and rainfall and to partially control temperature fluctuations. Kolesnikov (197 l ) , Schuurman and Goedewaagan (197 I ) , and Bohm (1979) have reviewed the extensive observational literature that appeared in the early part of this century dealing with root observations from transparent wall pits. These pits are inexpensive to construct, easy to operate, and yield data suitable for demonstrations or for obtaining qualitative ideas about behavior of root systems. Pearson and Lund (1968), for example, used a soil pit with a transparent face, of which the bottom was recessed more than top, to show that root growth of cotton generally preceded extensive shoot development. In another simple design, plants are grown in soil-filled boxes with a transparent side which is usually covered to exclude sunlight. During the experiment, the box can be either outside under field environmental conditions or inside a greenhouse or growth chamber under controlled conditions. Even with boxes large enough to permit plants to grow to maturity, construction costs of these

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

3

transparent-sided boxes are relatively low; a major disadvantage is that root temperatures are much closer to above-ground air temperatures than to soil temperatures encountered under field environments. B . MULTICOMPARTMENT BELOW-GROUND OBSERVATION FACILITIES

Because of the variability inherent in all biological research, controlled and replicated experiments are required. Under different soil conditions, root growth and function can vary widely. A logical approach to the study of interactions between soil factors and root function was construction of larger, more elaborate facilities where many replicated (or replicable) treatments could be installed in the same location under identical climatic conditions. In the usual arrangement, viewing surfaces of adjacent compartments line the sides of a long tunnel, permitting more economical construction than if the same number of compartments were constructed individually and buried separately in the soil. The central walkway is generally covered with some type of roof to exclude sunlight and to moderate below-ground temperature fluctuations. Instrument shelters and soil preparation areas are easily accessible from all soil compartments.

I . Observation Windows Covering Intuct, Native Soil Profiles The root observation laboratories built at East Malling, Kent, England, in 1961 and 1966 (Fig. 1 ) were prototypes for most of the larger root observation facilities constructed since then. These facilities have been described by Rogers (1969) and Rogers and Head (1963a,b, 1968), and were also mentioned in reviews by Kolesnikov (1971) and Bohm (1979). The basic arrangement consists of a long trench excavated with a mechanical scoop, leaving an undisturbed soil profile on either side. A framework of interlocking precast concrete pillars and lintels was erected in each trench, and a curved roof was cast in place, slightly above ground level. Finally, observation windows were fitted between the concrete pillars, and soil, shaved from the trench wall at an appropriate depth, was screened and replaced to ensure close contact between soil and the glass viewing windows. An above-ground entrance hut and stairway were built at one end of the tunnel. It is also possible to construct a root observation facility containing a large number of different, intact, native-soil profiles in which root growth could be observed. Tackett et al. (1965) and Cannell et al. (1980) have published procedures for obtaining cylinders of relatively undisturbed soil 40-80 cm in diameter. These cylinders could be transported to a central installation for study (Cannell et al., 1980). A cross section of the cylinder plus soil would then be sliced

4

MORRIS G. HUCK AND HOWARD M. TAYLOR

FIG. 1. Observation tunnel in rhizotron at East Malling, Kent, England.

off the cylinder from top to bottom. The remainder of the cylinder would be fitted with a viewing panel, which would be sealed to the remainder of the cylinder wall. The cylinder bottom, with ceramic plates or small cylinders for drainage, would permit maintenance of the desired water potential profile. Although this technique should be possible, we know of no instance where it has been attempted on a large scale. Facilities with a native-soil profile located behind the observation panels expose roots to the heterogeneous soil environment always found in the field. The biological, chemical, and physical environments in this type facility are disturbed much less than in those facilities utilizing reconstructed profiles. 2. Observation Window Covering Artificially Constructed Soils Some experiments require precise information about the volume of soil occupied by root systems and the quantities of water, nutrients, or contaminants located in that soil volume. Some of the newer rhizotrons have sacrificed the ability to observe root growth in undisturbed soil to have better control over experimental conditions. The rhizotron-lysimeter facilities at Guelph, Ontario

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

5

FIG. 2. Entrance hut and stairway leading to observation tunnel in rhizotron at Guelph, Ontario, Canada.

(Hilton et al., 1969, Fig. 2); Auburn, Alabama (Taylor, 1969, Fig. 3); Muscle Shoals, Alabama (Soileau et al., 1974); Ames, Iowa (Taylor and Bohm, 1976); and Columbus, Ohio (Karnok and Kucharski, 1979) use reconstructed soil profiles to attain the physical and chemical soil properties of interest for a particular experiment. C. SPECIAL ADAITATIONSTO INVESTIGATE SPECIFIC PROBLEMS

1, Lysimeters with Root Observation Windows

Some rhizotrons can also be used as lysimeters. The Auburn rhizotron, for example, was fitted with porous plates and suction tubes at the bottom of each compartment to ensure adequate soil drainage. In many experiments conducted there, 20-liter bottles have been installed in the drainage line to collect leachate for analysis or water balance studies (Long and Huck, 1980b) (Fig. 4). Two facilities have been constructed, however, to act specifically as combined

6

MORRIS G . HUCK AND HOWARD M. TAYLOR

FIG. 3. (a) Soybeans and corn growing in the rhizotron at Auburn, Alabama. Observation compartments are surrounded by soil-filled borders at a slightly different height. The access stairway, instrument trailer, and soil preparation building are visible in the background. Access tubes for soil moisture measurement by neutron meter are visible in each root observation compartment. (b) Soil has been removed from observation compartments in the foreground in preparation for another experiment using a different soil profile. Sheet-metal dividers added during the filling process permit separate observations to be made on each of four different species of turf grass in the next compartment, while remaining area is fallow in preparation for planting soybeans. Neutron probe access tubes are capped to prevent entry of rainwater; vertical rods will support apparatus to measure stem diameter (Huck and Klepper, 1977).

lysimeter-rhizotrons: Muscle Shoals, Alabama (Soileau et al., 1974), and Temple, Texas (Arkin et al., 1978). The Muscle Shoals facility consists of 18 compartments, each 1 .0 m side-toside, I .2 m front-to-rear, and I .9 m high. Electrically controlled drainage and water sampling equipment are located in a 0.7-m-high space below each compartment. Suction can be applied to a 30 X 30 m, porous ceramic plate located in the bottom of each compartment. This facility allows researchers to (a) evaluate fertilizer use-root growth-nutrient leaching, and nutrient balance relationships, (b) measure mobility in soils and percolation losses of potentially toxic ions from organic wastes, and (c) study effects of soil chemical and physical properties on root development and crop yields.

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

7

The Temple facility consists of 24 weighable steel chambers, each with one vertical side of glass. These root observation chambers are inserted vertically into concrete retaining liners buried in the soil. An A-frame and load-cell arrangement is used to weigh the chamber periodically and to lift the chamber from the concrete liner for root observations. When compared to the Auburn or Ames

8

MORRIS G. HUCK AND HOWARD M. TAYLOR

FIG.4. (a) Corn (Zea mays L.) in a rhizotron experiment described by Long and Huck (1980b) measuring vertical ion migration. Tensiometers and soil solution sampling tubes (Long, 1978) permit detailed measurements as a function of depth and time. Vertical rods provide mechanical support to help avoid injury to plants during sampling and measurement operations. (b) Observation tunnel in the rhizotron at Auburn, Alabama. Drainage water from each compartment is collected in bottles along the walls (see text).

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

9

rhizotrons, where evapotranspiration rates must be determined from differences in soil water content with time, the Temple facility provides opportunity for more frequent, and relatively more precise, short-term measurements of water loss from the soil profile.

2. Above-Ground Enclosures to Sample or Modify the Aerial Environment In yet another level of measurement sophistication, it is possible to collect water vapor transpired from plants growing in a rhizotron compartment and compare it with water loss as calculated from differences in water content profiles. This procedure requires that a transparent cover such as those described by Leafe (1972), Lange et al. (1975), or Alberda (1977) be placed over the rhizotron compartment. Air inside these chambers is circulated through a nearby air-conditioning unit to maintain temperature and humidity control. Water vapor that condenses on the air-conditioning coils is trapped and measured (King, 1980; Wheeler, 1980) in the unit operating at Auburn. The Auburn unit follows a design prototype described by Musgrave and Moss (1961) and modified by

10

MORRIS G . HUCK A N D HOWARD M. TAYLOR

Phene et al. ( 1 978). In the SPAR (Soil-Plant-Atmosphere-Research) chambers of Phene et al. (1978) both above- and below-ground plant organs are monitored in individually controlled environmental chambers. Similar facilities are located in the Crop Root Research Laboratory, Department of Greenhouse Cultivation, Vegetable and Ornamental Crops Research Station, Taketoyo, Japan. In these units, however, a mist or liquid nutrient solution is used instead of soil in the root portion of the controlled-environment chambers. When the above-ground microenvironment is controlled and/or measured, as in the enclosed-canopy facilities described above, it is possible to measure not only transpiration but also net photosynthetic rates of plants growing in the rhizotron bins. By monitoring the amount of CO, which must be added to maintain a constant level in the circulating gas inside the above-ground chamber, it is possible to estimate instantaneous net photosynthetic rates with much more accuracy than in open systems (Samish and Pallas, 1973) or by inferences drawn from changes in plant size.

Ill. CONSTRUCTION DETAILS AND DESIGN FEATURES Most of the research groups constructing rhizotrons have modified the original design of the East Malling root observation laboratory to fit specific research requirements or to take advantage of more readily available construction materials. Examples are the concrete block retainer walls of the Mlanje, Malawi (Fordham, 1972), Ames, Iowa (Taylor and Bohm, 1976), and Columbus, Ohio (Karnok and Kucharski, 1979), rhizotrons or the brick and waterproof plaster walls in Griffith, Australia (Freeman and Smart, 1976). Some rhizotrons, such as those at Temple, Texas, Ames, Iowa, Columbus, Ohio, and Griffith, New South Wales, are built flush with the soil surface to minimize disturbance of airflow patterns and reduce reflectance of bare concrete and metal surfaces. The rhizotrons at Woodward, Oklahoma (Shoop, 1978), and at Mlanje, Malawi, have their passageways between compartments covered with grass mulches to reduce the extra thermal load which would be created by a concrete roof. The Ames, Columbus, and Auburn rhizotrons have soil-covered roofs which allow plants growing in the experimental compartments to be surrounded by similar plants in guard rows. Figure 5 illustrates the manner in which guard rows are planted on all sides of the test plants in the instrumented observation compartments at Auburn.

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

11

A. WINDOWDETAILS

1. Orientation (Azimuthal, Vertical Angle, Height, and Unsupported Width} In rhizotrons designed to study the roots of row crops, the rows are customarily planted perpendicular to the glass windows so that the glass may be considered to represent a vertical slice across a typical row. Both depth and lateral spread of the root system can be measured. Root density (as defined later) is averaged in two spatial directions, and calculations of root growth, water movement, and other root activity parameters are expressed on a “per unit root length” or “per unit soil volume” basis. On the other hand, if the experiment is one in which rates of extension of a particular root tip are desired (as opposed to mean population density of roots in an average soil volume), then windows with a negative 10” slope (or more) can be used (Pearson and Lund, 1968). In this orientation, geotropism will constrain any root tips intercepting the window to grow along the glass and not turn back into the soil. This configuration is useful for studying root elongation rates or branching and branch initiation in anatomical studies of contiguous flow paths from specific root tips to the base of a given plant’s stem. Autoradiographs are readily prepared from the “bisected” root system resulting from this planting and window configuration (see Fig. 9). When all experimental plants grow adjacent to the window, and sloping glass confines their root system, nearly half the entire root system can be seen at the window. If the plants are grown in a row perpendicular to the window, on the other hand, roots measured at vertical windows represent those intersecting a typical plane through bulk soil. Many root observation laboratories have reflective roofs made from such materials as concrete or gravel in asphalt, which increase transpiration from plants growing in adjacent rhizotron bins to levels far above that of plants growing in comparable field plots. Transpiration rates of cotton plants grown in the Auburn rhizotron before that facility was modified to eliminate a bare concrete roof above the walkway were about twice those of cotton plants growing in an adjacent field soil (Taylor and Klepper, 1975). One solution to this advective energy problem is to lower the passageway or tunnel roof line so that its upper surface is flush with the surrounding soil surface level and then to cover the roof with soil; however, this structural configuration creates a root-viewing problem. If the roof line is dropped so that soil above the roof is flush with the surrounding soil surface, those roots growing in the surface soil (to a depth equal to the roof thickness) cannot be seen from the tunnel windows (see side view, Fig. 6 ) .

12

MORRIS G . HUCK AND HOWARD M . TAYLOR

FIG.5. (a) Plastic tubing supplies irrigation water to guard rows of soybeans and maize which have been planted on either side of observation compartments in the Auburn rhizotron. The central observation tunnel is covered by only 15 cm of soil, so it must be resupplied with water at I - to 2-day intervals to ensure normal growth of plants in the guard rows. (b) By midsummer, maize and soybeans shown in Fig. 5a have grown into a closed canopy, shading exposed concrete boundaries. Only the top of the constant environment chamber is visible above surrounding maize plants; airsupply ducts are completely shaded at this time.

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

13

On the other hand, if the lower edge of the roof is raised to the same height as surface soil in the glass-front compartment, permitting all roots to be seen through the window, then the roof will protrude several centimeters above the soil in the bins (see Figs. 2 and 3) causing turbulence in wind passage and altering the absorption and reflection coefficients for radiation from the surrounding soil. Rhizotrons at Ames (Taylor and Bohm, 1976) and Columbus (Kamok and Kucharski, 1979) have the soil on top of the roof at the same level as the surrounding soil to minimize atmospheric disturbances around the plants, while those at Auburn, Guelph, and East Malling have an elevated roof design, permitting all roots to be seen and photographed. Plants growing in soil on top of an elevated roof help to minimize back-scattering of incident solar radiation and shade the lower levels on adjacent experimental plants after canopy closure (Fig. 5b).

2 . Choice of Window Materials and Sealing Techniques The choice of window materials is of particular importance. Glass from silica sand is chemically similar to the SiO, of soil particles; thus water films adhere to the glass at the soil-glass interface (O’Brien, 1970). Acrylic plastic (polymethyl

14

MORRIS G. HUCK AND HOWARD M. TAYLOR

PLANT OVBR

:p

PLANT I N

PABBAO%WAY

"Ds

PLEXICLAB

WINDOW

CIRAYIC l l L T l R 8

FRONT VIEW

S I D E VIEW

GUARD ROW OVBR PASSAGRWAY

0

GUARD ROW

TOP VIEW

FIG. 6. Schematic diagram of Ames rhizotron, illustrating construction details and manner in which plants growing above the observation tunnel are recessed to completely avoid exposed concrete or metal surfaces. Radiation balance and convective transport of heat and moisture are very similar to field conditions with this arrangement.

methacrylate) is an organic substance with a small dipole moment and, thus, has a different wetting angle than soil particles and glass. Water is not tightly bound to the plastic surface, so a plane of weakness will exist at the soil-plastic interface. Roots tend to grow along this path of least resistance, especially in higher strength soils; thus, an artificially high estimate of root population will be observed along acrylic plastic windows, compared with that in the bulk soil (Taylor and Bohm, 1976). If the soil under investigation has appreciable amounts of montmorillonitic clay, if there is substantial freezing-thawing activity, or if large organs such as potato tubers o r sugar beets grow near the window, glass panes are easily broken. Thus, the flexibility of plastic makes it a more desirable window material from a convenience and longevity standpoint. It is also easier to machine and install. Use of Plexiglas windows is not the answer to all problems, however; the choice of window materials will depend upon the type of information desired from the experiment. There may be no ideal solution.

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

15

Many experiments require watertight compartments. The viewing panel must, therefore, be sealed to the side walls. The type of sealant to be used depends upon the longevity required. If the experiment is to last only a year or two, a satisfactory seal has been attained at the Auburn rhizotron using a silicone rubber caulking compound (Fig. lob). In experiments involving compaction pans at high density, the silicone rubber caulking beads are sufficiently elastic to provide a zone of weakness for root penetration along the caulk bead. In these experiments, epoxy must be substituted for the silicone rubber caulking. Epoxy resin is impermeable to both air and water, and it forms a very rigid seal which usually requires that the glass be broken to remove panels at the termination of an experiment. Viewing windows of the Ames rhizotron are sealed in a slightly different manner. Three sides and the bottom of a sheet-steel compartment liner were welded together. A 2-cm-wide edge was formed from top to bottom on each side. This edge was used to attach the front viewing panel to the compartment. The viewing panel was clamped to the edge, and bolt holes were drilled through both at 5-cm spacings, as shown in Fig. 6 . Bolts were welded into place with the threads outward. The entire steel liner was painted with epoxy, and a bead of silicone caulking was laid in such a way as to join the viewing panel to the rest of the liner. Nuts were placed on the bolts and tightened. When water leaks were discovered, the nuts nearest to the leak were retightened until the leak stopped. Seals of this type have been satisfactory for 5 years. B. CONTROL OF SOILPHYSICAL PROPERTIES IN RECONSTRUCTED PROFILES

I . Bulk Density and Mechanical Impedunce to Root Penetration

Construction of a soil profile in a rhizotron compartment requires a great deal of empirical, trial-and-error skill developed at each location, and the examples discussed below are intended to be representative of the authors’ experience, not universally applicable rules. Generally, reconstructed soil profiles are made from soil materials collected from each horizon of a representative field soil at some particular location. Reconstruction of a soil profile must be done with great care because small differences in mechanical resistance to root penetration resulting from channeling or separation of different-sized particles during the packing operation will cause large differences in root growth. At the Auburn rhizotron, soil materials from each horizon of interest are collected in the field and transported to the rhizotron site for drying, sieving, and mixing with appropriate chemical amend-

16

MORRIS G. HUCK A N D HOWARD M. TAYLOR

ments. After rewetting to some specific moisture content, the sieved soil material from each horizon is carefully packed into a rhizotron compartment with rigid concrete walls and a window that faces the central observation tunnel. Only a few centimeters of soil are added at one time, and then a vibrating tool is used to compact each soil increment until the prescribed bulk density is achieved. From time to time the mechanical resistance can be checked with a penetrometer. Any soil layer found to be too dense, or not uniform, must be removed and repacked until a soil profile layer with the desired soil properties has been attained. Parametric studies (e.g., variable bulk density, pH) are undertaken by simply packing several adjacent bins in a consistent manner, with all soil physical and chemical parameters except the one of interest being held constant. Despite careful packing, some soils settle into a uniform density with depth when saturated and then drained. Such soils are simply used as they are, after putting them through several more wetting and drying cycles before planting the experiment. The compartments in the Ames rhizotron are filled with loose, screened soil that is then wetted very slowly to saturation from top to bottom. Suction is then applied to porous cylinders located in the bottom of each compartment. The soil settles during drainage, with the amount of settling varying with the particular soil material. When the soil matric potential has reached about -0.05 bars at the drainage cylinders, the upper surface of the settled soil is scarified to eliminate any possible interface discontinuity, and more loose soil is added. This process is repeated until the compartment remains filled even after the drainage has ceased. Obviously, different levels of bulk density or penetrometer resistance cannot be attained when this method of packing is employed. 2. Drainage and Irrigation Systems (with Effects upon Soit Aeration) Soil water content influences many fundamental soil physical properties. Control of water relations in a rhizotron compartment offers the possibility for control of a number of soil physical properties, but because of multiple interactions, it is not possible to control each parameter separately. An increase in soil water content, for example, will simultaneously increase unsaturated hydraulic conductivity and reduce the gaseous diffusion rate, while markedly altering the rheologic properties and specific heat capacity of the soil. A more complete discussion of the interacting relationships can be found in standard soil physics textbooks and handbooks for highway engineers. Nearly all rhizotron facilities have some provison for drainage; a number have elaborate devices, such as those discussed earlier, for maintaining a specific soil water potential at the drainage outlet of each compartment. When a rhizotron compartment is covered by a transparent photosynthesis chamber, or when the soil is covered to prevent entry of rainwater, some provi-

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

17

sion must be made for adding water to replace that removed by transpiration of the growing plants. This addition of water is generally accomplished by installing trickle-irrigation emitters on the surface of each compartment or by adding water manually from a sprinkler nozzle. Simply pouring water onto a compartment from a container is generally unsatisfactory because this usually produces an uneven wetting front, with channeling to deep-soil layers, while pockets of dry soil remain near the surface. A large-area sprinkler system has no provision for ensuring that each individual compartment will receive the correct amount of water. Accounting for water added to each compartment by natural rainfall can also be a problem, for large plants often have such dense foliage cover that rainfall can be deflected outside the rhizotron bin area unless special precautions are taken. Provision must also be made for frequent, small irrigations of the border area crops growing in the soil covering the walkway roof. Because of the small storage capacity of this very shallow soil, careful attention is required. Excessive watering can produce runoff which will alter the water and mineral balance of adjacent plots, and must be prevented if possible. Metal or cast epoxy-fiberglass borders around each individual compartment are helpful in this regard, especially in areas where heavy rainfall is common. 3. Radiation Balance, Heat Flows, and Soil Temperature Control

Temperature influences the rate of all biological activity in the soil and, thus, is an extremely important parameter. Because most rhizotrons are outdoor facilities, however, temperature control is difficult. Surface radiation exchange from an individual rhizotron compartment, because of its relatively small area, is not as easily measured as in a field experiment. In addition, the soil in a compartment is surrounded by walls of Plexiglas, glass, metal, or concrete to ensure hydraulic isolation. These walls influence heat exchange with the surroundings, however, and also conduct heat vertically through the metallic structural components. Soil temperature in a rhizotron compartment may vary significantly from that of soil in the surrounding field at the same depth and time. Efforts have been made in some facilities, such as in the Muscle Shoals rhizotron, to control soil temperature by circulating water through metal pipes embedded in the soil, but most present-day facilities simply measure soil temperature by networks of embedded thermistors or thermocouples. Temperature variation can also be introduced by altering soil water content or the amount of shading from plant canopies or artificial soil covers. This results in a range of fluctuating soil temperatures whose minimum and maximum values depend upon ambient climatic conditions.

18

MORRIS G. HUCK AND HOWARD M. TAYLOR

4. Illumination of Roots All examination methods require some illumination of the roots. Questions frequently arise about the effects of this light on root growth and function. Continuous lighting slows root elongation, hastens suberization, and hinders lateral root formation of some species, but short periods of lighting have little effect on root growth. Some plants, such as peanuts, are particularly sensitive to light (Pearson, 1974), while other, such as soybeans, are relatively insensitive (Fig. 7). Lighting effects on root growth should be checked for each species to be used in rhizotron studies. Most researchers using rhizotrons have installed opaque covers over the windows, removing them only when measurements are made. We found that this technique was desirable to reduce algal growth (which obscures roots) in the soil, even where the roots themselves were insensitive to light.

TIME AFlER PLANTING (hr)

lighted Soybeans

dark

-4- -0-

32

FIG. 7. Effects of illumination upon soybean and peanut root elongation. The values shown are the means of eight radicles grown in soil at 29°C in glass-front boxes (data of Pearson er a/., 1970). “Lighted” roots were continuously exposed to fluorescent and incandescent light, while “dark” roots were kept under an opaque cover at all times except for approximately 60 sec of exposure to the same light while measurements were being made. Note that while “lighted” peanut roots grew substantially slower than “darkened” roots, the light had little effect upon the growth of soybean roots.

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

19

FIG. 8. The cortex of this cotton root has been ruptured by growth of a root-knot nematode feeding from vascular tissue in the enlarged stele. Entry of disease organisms is facilitated by cortical rupture.

20

MORRIS G . HUCK AND HOWARD M. TAYLOR

IV. SOME TECHNIQUES FOR OBSERVING AND RECORDING ROOT SYSTEM PARAMETERS Some experiments conducted in rhizotrons require only qualitative observations of root branching patterns, root color, direction of root growth, or duration of root growth during the life of a crop. Figures 8 and 9 illustrate qualitative information obtained from rhizotron experiments at Auburn. For other purposes, however, quantitative data on rooting density in the bulk soil, transpiration rates, or nutrient uptake rates may be required. When these data are needed, a more extensive set of measurements must be made.

FIG.9 . Cotton roots at the rhizotron window photographed under visible light (left) have been heavily labeled with [32P]orthophosphate. After 72 hr of metabolism, a 32P autoradiogram (right) shows that most of the labile phosphate is accumulating in regions of active cell division such as root primordia or the long root at upper right which may be undergoing radial expansion from secondary cambial activity.

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

21

A. MEASUREMENT OF ROOTSYSTEMSIZE

I. Root Length: Direct Measurement The primary function of a rhizotron, by definition, is to measure changes in root system size or activity. A stable coordinate system on the glass window provides a reference for making successive measurements. Changes over time, then, represent root growth or death. In the simplest systems, the locations of roots are marked on the outer surface of the viewing panel (or on a transparent overlay if the viewing panel is too fogged by condensation), and growth is noted by measuring the change in rooting pattern from one measurement to the next. Because the marks are relatively crude, precision of measurement is somewhat limited. At the next level of sophistication, rectangular grids are imposed on the viewing surface. Grid lines may be inscribed on panels of acrylic plastic or standard industrial safety glass. Imbedded wire grids, such as that illustrated in Fig. 10, can also be used as a stable coordinate reference system to which successive measurements are referred. When the grid system is embedded inside the viewing panels, the substantial parallax errors that can occur if measurements are made with reference to a grid on the outside of a viewing panel are reduced. Parallax errors vary with distance between the grid embedded in the glass and the actual root; window thickness is usually substantial because of the need for great strength to support the soil mass. Finally, optical measurements from a grid imposed upon projections of serial photographs can be used for extreme precision in measurement of growth rates (e.g., Huck, 1969; or Huck et a f . , 1970). When measurements of physical dimensions can be made to an accuracy of 0.01 cm or better, it is possible to estimate short-term changes on the order of a few minutes. If less measurement precision is available, root growth must be averaged over longer periods of time. Detailed responses to short-term changes in the microenvironment, such as daily fluctuations in water stress or hourly responses to variation in evaporative demand of the shoot, are obscured, but variation from day to day can still be determined. 2 . Root Length: Estimates by Statistical Methods

As proposed by Taylor et al. (1970), the length of roots visible on the viewing surface can be determined from the number of roots that intersect horizontal grid lines at specified depths. This method is a modification of Newman’s (1966) method. Instead of randomly oriented transect lines, Taylor et ai. (1970) used fixed transect lines and considered that the roots themselves were more or less random in their orientation.

22

MORRIS G . HUCK AND HOWARD M. TAYLOR

FK. 10. (a) Camera and microscope-mount platform on a viewing window in the Auburn rhizotron. Glass panels, supported by steel frame, contain a wire grid for precise location of roots in

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

23

the soil. (b) View of soybean roots, with nodules, growing across the border between two adjacent glass panels. All joints are caulked with silicone rubber or epoxy resin to produce a watertight and gastight seal (see text). (c) Further enlargement of roots shown in Fig. lob. Some roots are dying, while new ones grow to take their places. Shadows from wire grid (just below and to the left of each wire) illustrate the potential for parallax error at high magnification even with the wire grid embedded in the glass.

24

MORRIS 0.HUCK AND HOWARD M. TAYLOR

Lang and Melhuish (1968) described a possibility for error introduced by this method: anisotropy in the angle of roots intersecting transect lines can alter the relationship between the number of transect intersections and root length per unit viewing surface. In the extreme case where all roots are growing vertically, as with early growth of the adventitious roots of a densely planted grass, the Taylor et af. (1970) method tends to overestimate root length. It underestimates rooting density when there is more horizontal than vertical growth, as when many horizontal laterals grow from a few major taproots. Thus, a calibration is required for each crop under each given set of soil and climatic conditions encountered. Further details and an example of the method are given by Browning e r a / . (1975) and Williams et al. ( I 982). After root length per unit viewing surface has been established, it is often desirable to know the root length per unit volume of soil. An estimate of the average distance to which roots can be seen behind the glass is required for converting root length per unit viewing surface to root length per unit soil volume. Usually, the vernier depth-of-field focusing scale on a long workingdistance microscope such as that shown in Fig. 11 is used to obtain this relationship. The operator focuses on the inner surface of the glass and notes the reading on the microscope focus knob. He then focuses into the soil at high magnification (to minimize depth-of-field), and again notes the focus knob reading when he is at the bottom of the deepest void in which he is able to see a root. The difference, as read from the microscope scale, represents the distance (depth) behind the glass to which roots are being sampled by visual observations at the viewing surface. This depth is then multiplied by rooting intensity (cm/cm2) to give rooting density (cmkm3). Results from this procedure are approximate, but generally reasonable, and at least as accurate as those obtained by any other method for repetitive measurement. Factors for both the transect intercepthoot length per unit area, and for the depth-of-view will change with soil porosity, texture, and packing density. The anisotropy factor, expressed as coefficients in a polynomial regression equation, will also vary with the growth habit of the roots under investigation which, in turn, is influenced both by the plant itself and the soil in which the roots are growing.

5 . Root Diameter and Sugace Area Root diameter and surface area require precision measurements, usually accomplished at magnification with a long working distance microscope. True surface area measurements are difficult to obtain because of the large contribution from root hairs. If the contribution of root hairs is neglected, however, an estimate of root surface can be made from combined length and diameter measurements, using the mathematic formula for the surface area of a cylinder.

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

25

FIG. 11. Trolley mount for microscope in the Guelph, Ontario rhizotron. Black curtains are rolled up during observations, but are lowered to exclude light the rest of the time.

Root diameter can be estimated by direct visual observation with an eyepiece reticle in a hand lens. A trolley mount such as that shown in Fig. 11 facilitates rapid movement of a microscope from one area to the next if more measurement precision is required. There is great variation in root diameter from one root to the next and even diameter of the same root varies with time (Huck et al., 1970). Some assumptions regarding statistical averaging must be made before such data are useful.

26

MORRIS G. HUCK AND HOWARD M. TAYLOR

An alternative rapid measurement of root surface involves photographs of roots and rapid electronic scans using a Vidicon camera-integration approach such as that used successfully by Eguchi and Matsui (1977) for shoot measurements in the Biotron at Kyushu, Japan. These measurements, applied to a projected microscope image, permit estimates of the entire root surface area, including root hairs. Contrast and edge discrimination have caused problems in all automatic methods reported to date, because roots are normally grown in a reflective medium, such as shiny soil particles. B. POPULATION AND SPATIAL DISTRIBUTION OVER TIME:THESUM OF GROWTH AND DEATH RATESIN EACHLOCALIZED AREA

Measurements of root length or surface area per unit soil volume are useful in estimating a growing plant’s ability to extract water or minerals from each region of soil. Such a distribution function, called an “extraction term” (Molz and Remson, 1971), can be estimated by simply observing the spatial distribution of the plant’s root system, either by excavation in the field or by direct observation at a rhizotron window. When root distribution measurements are made in a rhizotron, they can be repeated again and again on the same plant, giving an estimate of changes in root population distribution over time. Young root initials enter a region of soil from adjacent regions. If conditions are favorable, they branch, rebranch, and continue exploration so long as growth can be sustained (Brouwer, 1977). After a few days of active absorption, the smallest feeder rootlets, which are most active in water and nutrient uptake, generally die back, while other new rootlets take their place. The result is a relatively stable root population resulting from a dynamic equilibrium between new root growth and death of older roots (Huck, 1977). For many purposes, such as running a total carbon budget, it may be necessary to know the growth rate of new roots, as distinguished from changes in the total root population. An estimate of root death rate is also important for certain problems such as estimating the food available to soil microorganisms or computing mineralization rates. A population distribution function, obtained by simply observing the length, number, or surface area of roots present at any particular time is not sufficient to deduce either the root growth rate or the root death rate. To obtain an accurate estimate of root growth (accounting for death and growth independently), it is necessary to tag individual roots so that those formed since the last observation can be positively identified. In rhizotron experiments, this is generally accomplished by simply marking the outline of each root on the glass window. In some cases, each rootlet is assigned a unique number so that it can be identified in subsequent computer summarizations. The usual procedure

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

27

in experiments at the Auburn rhizotron involves tracing the outline of all roots visible at the glass window at some fixed interval, usually 1-3 days, depending upon the accuracy required and the root growth rate. Then, at the next measurement, all new roots formed since marking the glass will be readily identifiable white lines. Their length and location is noted, and then the new roots are again marked so that additional new growth can be identified the following day. By changing colors at weekly intervals, a map can be constructed directly on the rhizotron window showing the location of new root growth occurring during each week of the growing season. This can be correlated with rainfall patterns, flowering, or other events of interest to the experiment. Although root death rate might also be estimated from daily observations, the demise of a root is generally gradual, and there are no visually distinct symptoms accompanying root death. One approach is to record both new growth and instantaneous population each day. Then, the difference between the cumulative sum of new root growth in each soil region and the instantaneous population represents the day’s root death. Karnok (unpublished personal correspondence) has reported success in distinguishing live roots from dead by viewing under ultraviolet illumination.

c. VALIDATION

OF

ROOT DENSITYMEASUREMENTS

All rhizotron measurements are based upon the hypothesis that root concentrations at the soil-viewing interface are similar to those in bulk soil. This hypothesis must always be checked. Taylor et al. (1970, Table I) and Klepper et al. (1973) found only a slight, if any, increase in the rooting density at the glass wall as compared to the density in bulk soil (centimeter roots per cubic centimeter soil) in a loamy sand soil at the Auburn rhizotron. As anticipated, for the reasons already discussed in Section III,A,2, Taylor and Bohm (1976) found about five times as much root length at an acrylic plastic-soil interface as in bulk Ida silt loam soil (Table 11). Each soil and window combination should be examined quantitatively to determine if roots are more concentrated at the interface than in the bulk soil behind the window.

V. EXPERIMENTAL DESIGN: DATA ACQUISITION AND ANALYSIS In most cases, experiments designed to measure root growth and function (the reason for existence of rhizotrons) are affected by microenvironmental parameters such as temperature, chemical composition of the soil solution, and other

28

MORRIS G. HUCK AND HOWARD M. TAYLOR

Table I Root Dry Weights from 250-cm3 Soil Samples at Various Locations within Compartments at the Auburn, Alabama, RhizotroncJ,h Corn root dry weights (g) at depth of

20 cm

Sample location

85cm

160cm

Tomato root dry weights (g) at depth of

20 cm

160cm

85 cm ~~

At glass-soil interface (front) At stainless steel-soil interface (side walls) At epoxy-soil interface (rear wall) In bulk soil

0.075

0.017

0.014

0.042

0.041

0.014

0.040

0.026

0.020

0.005

0.025

0.025

0.062 0.102

0.015

0.015 0.015

0.024 0.124

0.015

0.026 0.025

0.008

0.020

"Data from Taylor er al. (1970). bThe viewing surface was glass and the soil was a loamy fine sand.

Table I1 Rooting Density at 10 Depths and Five Distances from an Acrylic Plastic Viewing Surface at the Ames, Iowa, Rhizotron"." Rooting density (cmicm') at distance from observation surface (cm) Depth (cm)

c0.2

0.2-2.0

l9.c-21.O

37.8-39.8

39.8-40.0

37.5-52.5 52.5-67.5 67.5-82.5 82S97.5 97.5-112.5 112.5-127.5 127.5-142.5 142.5-157.5 157.5-1 72.5 172.5-187.5

2.59 2.39 3.49 3.32 4.24 6.61 6.67 7.96 2.30 0.83

0.39 0.79

0.91 0.58 0.94 0.63 1.49 0.99 0.92 0.20 0.07 0.01

0.30 1.62 1.63 1.91 1.20 1.16 0.52 0.20 0.58 0.I9

3.17 5.00 2.40 4.45 6.33 7.01 12.36 5.59 2.77 2.69

I .04

0.84 I .03 1.09 0.85 0.61 0.51

0.13

UData from Taylor and Bohm (1976). bThe crop was soybeans; the soil was Ida silt loam, a loess.

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

29

properties of the microenvironment, both above and below ground. Thus, root activity can be reported as a family of curves in some dependent variable, with all other experimental variables held constant or compared at equivalent levels of measured variables. A. CONTROLLED ENVIRONMENTS VERSUS CHARACTERIZATION OF NATURALENVIRONMENTS

In the simpler, lower cost rhizotrons with outdoor bins under natural conditions, such microenvironmental variables as temperature and soil water-content fluctuate in response to variation in the natural environment (weather). The experimenters measure these changes and sort out the results later, usually with the aid of a computer. Root growth is treated as a function. of microclimate, as well as a function of such plant factors as age, species, or genotype. In other facilities, environmental variables are held constant by manipulation of the physical environment (controlled-environment conditions). Equipment and operating costs for these experiments, however, are generally quite high. The SPAR units of Phene et af. (1978), described in Section II,C,2, have been modified to permit direct visual observations of root growth while controlling key parameters affecting the photosynthetic and evapotranspiration processes. Similar controls can be used in photosynthesis chambers above compartments of a more conventional rhizotron, such as the Auburn facility, as illustrated in Fig. 5. B . MEASUREMENT OF ROOTFUNCTIONS: WATERREMOVAL, MINERAL UPTAKE,AND BIOLOGICAL OXYGENDEMAND

Information about root functions, such as removal of water and minerals from the soil, is as important for understanding a plant’s performance as knowing the size and physical location of its root system. Movement of dissolved ions can be inferred from comparisons of successive soil solution analyses (Long, 1978; Long and Huck, 1980b). In some of the more sophisticated rhizotron facilities, automatically recording transducers under computer control record such soil properties as water potential (Long and Huck 1980a) or oxygen content (Melhuish er d . , 1974) as a function of space and time. From changes in successive measurements, it is possible to compute water or oxygen movement and, thus, to assess the activity of a living root system subjected to a weil-defined microclimate, characterized by simultaneous measurements of relevant soil physical and chemical properties from other transducers in the same system.

30

MORRIS G. HUCK A N D HOWARD M . TAYLOR

C. SELECTION OF DATAFOR ANALYSIS AND STORAGE

The soil microenvironment in which roots grow and function is heterogeneous in both space (generally three-dimensional space) and time. Measured variables often change rapidly, as when a wetting front passes through the soil profile, or when oxygen diffusion into an active root zone is blocked by water. Spatially distributed measurements, therefore, must be frequently repeated if the state of the system is to be adequately described. Huck and Davis (1975), Snyder (1976), and others, have described the nature of the storage and retrieval problem for variable-density data used as multidimensionally continuous functions. Beyond mathematical formalism, there are also practical problems of instrumentation. The initial choice of a transducer, in effect, limits both the precision with which a measurement can be made and the time resolution which can be achieved in response to rapid changes in the measurement values. In order to answer particular questions, an experiment will generally require data with a certain level of accuracy, recorded at some specified interval. Measurement and recording of excessive data can rapidly fill all available storage even when large general-purpose computers with high-density magnetic tapes are available, but failure to record necessary information can greatly reduce the value of an experiment. An initial decision on which data to record must be made at the time of experimental design-an after-the-fact decision is often too late. When data must be recorded manually, critical data are often overlooked. With automated systems, however, many scientists tend to err on the side of excessive data recording. This tendency leads to excess data storage and difficulty in obtaining access to specific data items of interest. On-line data reduction by microprocessor, and careful, rational experimental design are the key to easy data access. Graphic display capability is extremely helpful. C. QUESTIONS WHICHCAN BE ADDRESSED I N RHIZOTRON EXPERIMENTS

The unique feature of a rhizotron experiment, as compared with a conventional plot experiment, is the possibility for making repeated measurements on the same root system as opposed to sampling different individuals from a large population at each successive measurement. Nondestructive measurements reduce the variance introduced by plant-to-plant variability and generally increase sensitivity to differences between treatments. In addition, an intuitive feeling for the qualitative nature of below-ground processes can often be gained from serial

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

31

photography or from repeated observations upon the same root system at a rhizotron window. When both environmental parameters and root population data are repeatedly observed (serial measurements) in the same experimental arrangement, changes between successive measurements can be attributed to processes such as water movement, growth, or death of roots in the soil. By comparing successive measurements, active physical processes can be measured and their rates computed. A continuous record of changing soil properties forms the basis for estimating the physiological activity of an imbedded root sytem. The nature of observations which can be made in a fully instrumented rhizotron laboratory are such that a number of questions relevant to root growth, development, and activity can be answered. They might be classed as follows, with references suggesting specific papers in which work applicable to the given subject was reported.

I . Root Growth and Development (a) Morphogenetic changes (Taylor and Lund, 1970; Taylor et d.,1970; Huck, 1979; Taylor and Terrell, 1981) (b) Variation in growth habit influenced by genetic factors (Elkins et a / . , 1979; Haaland et al., 1978; Williams et al., 1982) (c) Root growth in response to variation in soil microenvironment (Camp and Lund, 1968; Pearson et al., 1973; Taylor et al., 1972) (d) Partitioning of carbon between roots and other plant organs (Lund, 1978; Huck, 1977) 2 . Root Functioning (a) Water uptake (Taylor and Klepper, 1971; Klepper et al., 1973; Huck et ul., 1970; Browning et al., 1975; Fiscus and Huck, 1972; Taylor and Klepper, 1974) (b) Mineral uptake and ion movement (Long and Huck, 1980b; Long, 1981) (c) Biological oxygen demand (BOD) and gaseous diffusivity (Melhuish et al., 1974) 3. Characterization of Soil Properties

(a) Lysimetry and estimation of infiltration and evaporation (Moore et al., 1974) (b) Groundwater recharge and ion migration through the soil profile (Molz and Browning, 1977; Long and Huck 1980b)

32

MORRIS G . HUCK AND HOWARD M. TAYLOR

4. Integration of Production Practices under Fieldlike Conditions (a) Water relations and irrigation practices (Stansell et al., 1973, 1974; Molz and Klepper, 1972; Klepper and Browning, 1971; Huck and Klepper, 1977; Klepper et al., 1971) (b) Pest management practices (Kappelman et ul., 1971) (c) Calibration and verification of predictions from simulation models (Niniah and Hanks, 1973; Taylor and Klepper, 1975; Alberda, 1977)

VI. SUMMARY: ADVANTAGES AND DISADVANTAGES OF RHIZOTRONS FOR USE IN ROOT INVESTIGATIONS Rhizotrons have several advantages over most other root study methods when extensive measurements are required. Successive measurements are made on the same plants each time, and estimates of root growth at the root-soil interface can be obtained quite rapidly. At the Ames facility, for example, two people were able to determine soybean rooting intensity to a 2.1-m depth for 48 compartments in an average time of 4 hr, or 10 person-minutes per compartment. When root growth is determined by differences between successive measurements on the sume plants, the normal plant-to-plant variability does not introduce as much error as when plants from similar treatments are destroyed at each sampling period. Instruments and sensors to measure soil or plant properties are easier to install and maintain in a rhizotron than in field plots. Sensors can be installed horizontally from the walkway through the viewing surface. This horizontal installation guards against water flow down the access hole, as often occurs with vertical installation of sensors. The hydrologically isolated compartments of some rhizotrons allow accurate estimates of the volume and properties of soil explored by the plant roots. Rhizotrons have certain disadvantages that should be considered in any decision to construct such a facility. First, there is a significant construction cost. At Auburn, the initial construction cost in 1969 was about $40,000. An additional $50- 100,000has been expended on instrumentation, control systems, and several updatings of associated computer equipment during 13 years of operation. The initial cost of the Ames rhizotron was about $20,000 when built in 1973. Although these construction costs may seem high, there is no other practical means for measuring rooting intensity as it changes over time. When the data volume is high, the cost per measurement is less than with most alternative methods.

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

33

The aerial environment around a rhizotron may be sufficiently different from that in field plots to significantly affect plant growth and function. Rhizotrons can be constructed in such a way that they will not upset the normal transpiration and plant growth patterns, but often certain compromises in experimental design are required. Rhizotron compartments can be constructed so that root concentrations at the soil-viewing surface interface are about equal to those in bulk soil. All rhizotron designs should be validated to determine whether presence of the rhizotron structure itself significantly disturbs field environmental conditions. Rhizotrons are best suited to answer specific questions. Data needed to answer these questions must be accurate, timely, and reproducible. They can, for example, be very useful in a root morphology or root physiology program when specific questions are formulated. They are not particularly useful if the researcher is only interested in general growth patterns. When general observations are sufficient, field techniques (Bohm, 1979) are less expensive and may be equally satisfactory. Finally, the initial cost dictates that the facility should be used for several years and, preferably, located where several scientists are interested in root morphology and function. Usefulness of the rhizotron, then, will not depend on the presence and interest of only a single scientist. Trained technicians, whose permanent assignment is at the rhizotron, are needed to collect these data on a routine basis. These requirements likely will best be fulfilled at large research centers rather than at one- or two-scientist locations.

REFERENCES Alberda, Th., ed. 1977. Versl. Landbouwk. Ondet-z (Agr. Rex Rep.) 865. Arkin, G . V . , A. Blum, and E. Burnett. 1978. Tex. Agric. Exp. Sra. Misc. Pub/. 1386. Bohm, W. 1979. “Methods of Studying Root Systems.” Springer-Verlag, Berlin and New York. Brouwer, R. 1977. In “Environmental Effects on Crop Physiology” (J. J. Landsburg and C . V. Cutting, eds.). Academic Press, New York. pp. 229-245. Browning, V. D . , Taylor, H. M., Huck, M. G . , and Klepper, B. 1975. Auburn Univ. Agric. Exp. Sm. Bull. 467. Camp, C. R., and Lund, Z. G. 1968. Trans. Am. Soc. Agric. Eng. 11, 188-190. Cannell, R. Q., Belford, R . K . , Gales, K., and Dennis, C. W. 1980. J . Sci. Food Agric. 31, 105-116. Eguchi, H., and Matsui, T. 1977. Environ. Control Biol. (Jpn.) 15, 37-45. Elkins, C . B., Haaland, R . L., Rodriguez-Kabana, R., and Hoveland, C. S. 1979. Agron. J . 71, 497-500. Fiscus, E. L., and Huck, M. G. 1972. Plant Soil 37, 197-202. Fordham, R. 1972. J . Horr. Sci. 47, 221-229. Freeman, B. M., and Smart, R. E. 1976. J . Enol. Vitic. 27, 36-39. Haaland, R. L., Elkins, C. B., and Hoveland, C. S . 1978. Crop Sci. 18 339-340. Hilton, R . J., Bhar, D. S . , and Mason, G. F. 1969. Can. J . Plant Sci. 49, 101-104. Huck, M. G. 1969. Agron. J . 22, 815-818.

34

MORRIS G. HUCK AND HOWARD M. TAYLOR

Huck, M. G. 1977. In “The Belowground Ecosystem: A Synthesis of Plant Associated Processes” (J. K . Marshall, ed.). Range Sci. Dep. Sci. Ser. ( C o b . State Univ.) 26, 215-226. Huck, M. G . 1979. I n “The Soil-Root Interface” (J. L. Harley and R. S. Russell, eds.). pp. 273-274. Academic Press, New York. Huck, M. G . , and Davis, J. M. 1975. In “New Directions in the Analysis of Ecological Systems” (G. S. Innis, ed.), pp. 93-100. Society for Computer Simulation, La Jolla, California. Huck, M . G., and Klepper, B. 1977. Agron. J . 69, 593-597. Huck, M. G . , Klepper, B., and Taylor, H. M. 1970. Plant Physiol. 45, 529-530. Kappelman, A. J., Buchanan, G. A,, and Lund, Z. F. 1971. Agron. J. 63, 3-5. Karnok, K. J . , and Kucharski, R. T. 1979. Agron. ADstr. p. 122. King, M. O., Jr. 1980. M. S. thesis, Dept. of Mechanical Engineering, Auburn University, Alabama. Klepper, B., and Browning, V. D. 1971. Highlights Agr. Res. 18, 16. Klepper, B., Browning, V. D., and Taylor, H. M. 1971. Plant Physiol. 48, 683-685. Klepper, B., Taylor, H. M., Huck, M. G., and Fiscus, E. L. 1973. Agron. J. 65, 307-310. Kolesnikov, V . A. 1971. “The Root Systems of Fruit Plants” (Ludmilla Aksenova. transl.), pp. 1-268. MIR Publ., MOSCOW. Lang, A. R. G., and Melhuish, F. M. 1968. Eiometrics 26, 421-431. Lange, 0. L., Schulze, E. D., Kappen, L., Buschbom, U . , and Evenari, M. 1975. In “Perspectives of Biophysical Ecology” (D. M. Gates and R. B. Schmerl, eds.), pp. 121-143. SpringerVerlag, Berlin and New York. Leafe, E. L. 1972. In “Crop Processes in Controlled Environments” (A. R. Rees, K. E. Cockshull. D. W. Hand, and R. G . Hurd, eds.), pp. 157- 175. Academic Press, New York (Appl E o t . Ser. 2.) Long, F. L. 1978. Soil Sci. Soc. Am. J. 42, 834-835. Long, F. L. 1981. Agron. J. 73, 537-546. Long, F. L., and Huck, M. G. 1980a. Soil Sci. 129, 305-310. Long, F. L., and Huck, M. G. 1980b. Soil Sci. Soc. Am. J. 44, 787-792. Lund, 2. F. 1978. J . Environ. Qua/. 7, 473-477. Melhuish, F. M., Huck, M. G., and Klepper, B. 1974. Ausr. J. Soil Res. 12, 37-44. Molz, F. J., and Browning, V. D. 1977. Ground Water 15, 409-415. Molz, F. J., and Klepper, B. 1972. Agron. J . 64, 469-473. Molz, F. J . , and Remson, I. 1971. Agron. J. 63, 72-77. Moore, C. L., Molz, F. J., and Browning, V. D. 1974. Transpiration drying: An aid to the reduction of sanitary landfill leaching. Proc. Annu. Environ. Eng. Sci. Con$, 4th, Loui.rvi//~~ March 4-5. Musgrave, R. B., and Moss, D. N. 1961. Crop Sci. 1, 37-41. Newman, E. I. 1966. J . Appl. Ecol. 3, 139-145. Nimah, M. and Hanks, R. J. 1973. Soil Sci. Soc. Am. Proc. 37, 522-527. O’Brien, W. J . 1970. Surface energy of liquids isolated in narrow capillaries. Surf: Sci. 19, 387-395. Pearson, R. W. 1974. In “The Plant Root and Its Environment” (E. W. Carson. ed.), pp. 247-270. Univ. of Virginia Press, Charlottesville, Virginia. Pearson, R. W., and Lund, Z . F. 1968. Agron. J . 60, 442-443. Pearson, R. W., Ratliff, L. E., and Taylor, H. M. 1970. Agron. J. 62, 243-246. Pearson, R. W., Childs, J., and Lund, Z. F. 1973. Soil Sci. Soc. Am. Proc. 37, 727-732. Phene, C. J., Baker, D. N., Lambert, J. R., Parsons, J. E., and McKinion, J. M. 1978. Trans. ASAE 21, 924-930. Rogers, W. S. 1969 In “Root Growth” (W. J . Whittington, ed.), pp. 361-367. Butterworth, London. Rogers, W. S., and Head, G. C. 1963a. Rep. E. Mulling Res. Sta. 1962 pp. 55-57.

THE RHIZOTRON AS A TOOL FOR ROOT RESEARCH

35

Rogers, W . S. and Head, G. C. 1963b. Rep. Int. Hort. Congr. 16th. 1962. Brussels 111, 31 1-318. Rogers, W. S . , and Head, G . C. 1968. 176-185. I n “Methods of Productivity Studies in Root Systems and Rhizosphere Organisms” (M. S. Ghilarov. V. A. Kovda. L. N. NovichkovaIvanova, L. E. Rodin, and V. M. Sveshnikova, eds.). Nauka, Leningrad. (English text revised by V. M . Ponyatovskaya.) Sachs, J . 1873. Arb. Bot. I n s t . Wurzburg 3 , 395-477, 584-634. Samish, Y . B., and Pallas, J . E. 1973. Phorosyntheticu 7 , 345-350. Schuurman, J . J., and Goedewaagan, M. A . J . 1971. Vrrsl. Lurzdhourrk. Oiideri pp. 1-86. Shoop, M. G. 1978. Ph.D. dissertation, Colorado State University, Ft. Collins. Colorado. Snyder, W. M. 1976. ARS Special Publ. S-76, USDA. Soileau, J. M., Mays, D. A,, Khasawneh, F. E. and Kilmer, V. J . 1974. Agron. J . 66, 828-832. Stansell, J . R., Klepper, B., Browning, V. D., and Taylor, H. M. 1973. Agron. J . 65, 677-678. Stansell, J . R., Klepper, B . , Browning, V. D., and Taylor, H. M. 1974. Agron. J . 66, 591-592. Tackett, J. L., Bumett, E., and Fryrear, D. W. 1965. Soil Sci. Soc. Am. Pro(.. 29, 218-220. Taylor, H. M. 1969. Auburn Univ. Agric. Exp. Sin. Cirr. 191. Taylor, H . M., and Bohm, W. 1976. Agron. J . 68, 693-694. Taylor, H. M. and Klepper, B. 1971. Aust. J . Biol. Sci. 24, 853-859. Taylor, H. M., and Klepper, B. 1973. Agron. J . 65, 965-968. Taylor, H. M., and Klepper, B. 1974. Agron. J . 66, 584-588. Taylor, H. M.. and Klepper, B . 1975. Soil Sci. 120, 57-67. Taylor, H. M., and Lund, 2. F. 1970. The root system of com. Annu. Corn Sorghum Re.\. Con/: P r o f . , 25zh 25, 175-179. Taylor, H . M . and Terrell, E. E. 1981 I n “Rooting Fattern and Plant Productivity Handbook of Agricultural Productivity.” Vol. I , Plant Productivity (M. Rechcigl, J r . , ed.), pp. 185-200. CRC, Boca Raton, Florida. Taylor, H. M . , Huck, M. G., Klepper, B., and Lund, Z. 1970. Agron. J . 62, 807-809. Taylor, H. M . , Huck, M. G., and Klepper, B. 1972. In “Optimizing the Soil Physical Environment toward Greater Crop Yields” (D. Hillel, ed.), pp. 57-77. Academic Press. New York. Wheeler, R. E. 1980. MS thesis, Dept. of Electrical Engineering, Auburn University, Alabama. Williams, C. B., Elkins, C. B . , Hoveland, C. S., and Haaland, R. L. 1982. Agron. J . (in preparation).

This Page Intentionally Left Blank

ADVANCES IN AGRONOMY. VOL . 35

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES T . T . Chang.* C. R. Adair.?.' and T . H. Johnston$.* *Department of Plant Breeding. International Rice Research Institute. Manila. Philippines. ?Agricultural Research Service. U.S. Department of Agriculture. Beltsville. Maryland. and $Agricultural Research Service. U S. Department of Agriculture. University of Arkansas Rice Research and Extension Center. Stuttgart. Arkansas I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . Increasing Importance of Rice . . . . . . . . . . ......................... B . Comparative Late Start in Rice Breeding and Conservation . . . . . . . . . . . . . . . . C . Importance of Genetic Resources to Rice Improvement . . . . . I1 . Diversity in Rice Genetic Resources . . . .......................... ry Pathway . . . . . . . . . . . . . . . . . A . Gondwana Origin of the Genus and B. Diversity in 0. sariva and Its Wild Relatives., . . . . . . . . . . . . . . . . . . . . . . . . . . C . Diversity in 0. glaberrima and Its Wild Relatives . . . . . . . . . . . . . . . . . . . . . . . . 111. Recent Efforts in Genetic Conservation . . . . . . . . . . . . . . . . . . . . . A . Status of Genetic Conservation in the United States ...................... B . Status of Conservation by Other National and Regional Centers . . . . . . . . . . . . . C . Conservation of the Wild Species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D . Collaborative Conservation Efforts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV . Dissemination and Evaluation of Germ Plasm . ......................... A . Distribution and Evaluation by United Stat rkers .................... B . The Genetic Evaluation and Utilization Program of IRRI . . . C . Evaluation by Workers in National Rice Research Centers of Asia . . . . . . . . . . D . International Efforts on Dissemination and Evaluation . . . E . Problems in Dissemination and Evaluation . . . . . . . . . . . . V . Preservation of Germ Plasm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . Preservation by USDA and Recent Efforts for Improvement B . Status at Other National Centers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C . The International Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D . Problems Encountered in Preservation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI . Use of Germ Plasm . . . . . . . . . A . By United States Workers B . By Other National Centers ........................................... C . By International Centers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D . Global Sharing of Improved Germ Plasm., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E . Public Efforts versus Private Enterprise in Rice Breeding . . . . . . . . . . . . . . . . . . F . Problems in Using Diverse Germ Plasm . . . . . . . . . . . . . . . . . . . . .

38 38 39 40 42 42 43 44 45 45 48 55 56 58 58 63 66 66 68 68 68 69 70 70 71 71 72 75 76 77 79

'Present address: 3 Bedwell Lane. Concordia. Bella Vista. Arkansas 72712 *Present address: 13 C & H Circle. Stuttgart. Arkansas 72160 . 37

ISBN 0-12-000735-5

38

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

VII. Endeavors for the Future A. Completion of Field B. Consolidation of Existing Major Collections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Conservation of Wild Species . . , D. The Completion of Evaluating the Collected Materials . . . . . . . . . . . . . . . . . . . . E. Characterization of Environments F. Innovative Breeding Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G. Restoration of Genetic Diversity to the Improved Cultivars . . . . . . . . . . . . . . . . H. Training of Rice Researchers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

80 80 81 81 81

82 82 84

84 85

I. INTRODUCTION A. INCREASING IMPORTANCE OF RLCE

Rice and wheat share equal importance as the world’s staple cereals. Rice is the primary source of energy and protein for 4.5 billion people in the most populous nations of Asia. Rice is a secondary staple for another 450 million people. The demand for rice is rapidly rising in Africa and Latin America. The order of preference is rice, corn, yam, and cassava. However, the average annual income of rice consumers in the developing nations is less than $200. Rice produces more calories and carbohydrates per hectare than any other cereal under normal production practices. Its protein yield per unit of land and amino acid balance among the cereals are only below those of oats (Lu and Chang, 1980). Despite significant increase in the rice yields of Bangladesh, China, Colombia, India, Indonesia, Korea, Pakistan, Philippines, and Sri Lanka since the advent of the high-yielding semidwarfs and the associated production technology beginning in the late I960s, the world’s supply of rice continues to lag behind the demand. First, the production of rice in the developing nations from 1967 to 1975 rose only at a rate of 2.4% a year while the population increased at 2.5%. Recent trends in crop production suggest that the developing countries may find it difficult to maintain the same rate of growth in production (IFPRI, 1977). Second, the developing countries have been suffering from a deficiency of 4-5% per capita in the daily energy supply (Zwartz and Hautvast, 1979). It would be a tremendous task to make up this deficit even if the world’s population were to stay at the present level. The projected increase in food consumption of the food-deficient developing countries between 1975 and 1990 will amount to 241 million metric tons at low rates of increase in per capita income. The increase would represent about twothirds of the 1975 consumption level. About 182 million tons are needed to meet

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

39

the increased need due to population growth. For all developing countries, increased consumption above the 399 million tons of 1975 would range between 262 and 286 million tons, of which 200 million tons would be due to population growth. Asia would bear the largest share of the projected increase in consumption between 1975 and 1990: about 48% (IFPRI, 1977). Because of limited arable land area in Asia, future increase in food production will depend largely on increases in crop yield or cropping intensity as experienced during the last two decades. Considerable areas of uncropped or marginal production may be brought into rice cultivation, but serious climatic, hydrologic, or edaphic constraints generally exist in such areas. For Asian countries, the production increase of cereals would have to surpass the 1960-1970 rate of 2.5% in order to meet the projected increase in consumption. For Africa and Latin America, a brighter prospect looms as a potential expansion in production area is feasible, although the per capita consumption of rice will also continue to rise (IFPRI, 1977). Rice will continue to make up nearly one-half of the major staples in Asia; about 30% for all developing countries (IFPRI, 1977). Therefore, it is imperative that the increase in rice production be sustained through improved yield or stabilized production or both. At places where the growing season permits, increasing the number of rice crops planted in a year offers another means of augmenting rice production. B. COMPARATIVE LATESTART I N RICE BREEDING AND CONSERVATION

Recorded history may not mention man’s earliest efforts to improve the rice plant. However, long before the advent of science, man undoubtedly had made good use of natural variability in the crop and its wild relatives, spontaneous mutations, natural hybrids, and introductions from foreign lands. Susrutha (ca. 1000 B .c.), in his Ayurvedic “Materia Medica,” recognized the differences among rices existing then in India and separated them into groups based upon their growth duration, water requirements, and nutritional values (Ramiah and Rao, 1953). Chinese classics show that Emperor Wen Ti of the Wei Dynasty ( A . D . 186-226) discussed with his cabinet staff about a quality rice having strong and fragrant aroma. Another emperor, K’ang Hsi (1662-1723) of the Ching Dynasty, selected an early maturing and aromatic mutant for a crop of rice grown in the imperial garden which later became the main staple of his household. The new strain was named “Imperial Rice.” The large scale introduction, testing, and extension of the early maturing Champa rices in central and east China during the eleventh century marked the first massive government-sponsored efforts to utilize efficient and productive genotypes (Chang and t i , 1980).

40

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

On the other hand, rice has a relatively short history of scientific improvement when compared to wheat. The first rice variety bred by hybridization, “Ominishiki,” was developed in Japan around 1906. In tropical Asia, breeding began in the 1950s. Genetic conservation of the crop began even later. Conservation efforts gained momentum in the mid-1960s. A combination of the late start in both plant breeding and genetic conservation led to the situation in the early 1960s that tens of thousands of rice cultivars were grown by Asian farmers over vast rice lands. Rice breeding in the United States was started in a minor way early in the twentieth century. The program at the start consisted of introducing a few rice cultivars from foreign countries and growing them in small plots in farmer fields in Louisiana and later in California. Selections were made from these plantings by federal and state agriculturists and by local farmers. Some selections from these introductions were increased and quite widely grown for several years. Cultivars selected by farmers included “Blue Rose,” “Early Prolific,” “Edith,” and “Lady Wright.” Some of those selected by government personnel included “Fortuna,” “Nira,” “Rexoro,” “Colusa,” and “Caloro.” These cultivars were widely grown until they were gradually replaced by improved cultivars that were developed after about 1935. However, Colusa and Caloro were still of major importance in California through the late 1960s. The rice-breeding program initiated in the United States was a cooperative endeavor of the U.S. Department of Agriculture (USDA) and State Agricultural Experiment Stations. These projects were expanded by the addition of personnel in Louisiana and California about 1915. The work was further expanded by the hiring of federal rice breeders in Arkansas, Louisiana, and Texas in 193I . Rice was a minor crop in the United States in the earlier years, thus more funds and personnel were assigned to breeding the major cereals than were allotted for rice breeding. As a result, more comprehensive programs were established for wheat, corn, barley, and oats than for rice breeding, so rice improvement was much delayed. The conservation of rice germ plasm also was much neglected in the earlier years. This was due to the meager number of cultivars available and the lack of personnel and storage facilities. C. IMPORTANCEOF GENETICRESOURCES TO RICEIMPROVEMENT

Crop researchers generally recognize that a wealth of genetic resources provides the building blocks for effective crop improvement. In the case of rice, the germ plasm of the two cultigens (Asian and African rices) is unusually rich in genetic diversity, much of which remains to be tapped. Until a decade or two ago, the full spectrum of germ plasm in the genus Oryza (see Chang, 1979) could

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

41

be readily found in their indigenous habitats and was available to rice researchers. The rich diversity in rice stemmed largely from its wide geographic dispersal and ecogenetic diversification. Among major food crops rice is one of the few staples which are continually expanding into new production areas. Rice can be grown in vast uncropped areas in Africa and Brazil. Rice is also being grown experimentally at numerous locales where irrigation water is available, but cool temperatures and blast disease at high elevations are major constraints. Numerous requests for seed recently received at the International Rice Research Institute (IRRI) for deepwater rices indicate another trend in area expansion. Moreover, in South and Southeast Asia large tracts of land can be brought into profitable rice production if the adverse soil factors such as salinity and acidity are overcome by adequate levels of cultivar tolerance and soil management. The rich diversity in rice presents great promises for rice researchers to alleviate the various production constraints, most of which are known while a few more remain to be identified. As rice expands into new areas, new environments will present challenges to the research community. Among cereals that can be grown in a new area of cultivation, rice has the unique advantage over other staples in that it can thrive in flooded or waterlogged soils. Thus, rice will continue to be the main staple of numerous subsistence farmers in the humid tropics and subtropics. The rapid spread of the high-yielding semidwarfs since 1966-1967 has greatly narrowed the genetic base of the rice crop. The preservation and use of the diverse germ plasm in rice is not only vital to the further improvement of the crop but also serves as a means to safeguard the crop against its vulnerability to attack by diseases and insects which is imminent because of the genetic uniformity among rice cultivars. It has been estimated by plant physiologists that the maximum yield potential of rice in the tropics may be as high as 15.9 tons/ha for a single crop in the dry season and 9.5 tons/ha in the wet season (Yoshida and Oka, 1981). Yield was 25.65 tons/ha total for a four-crop experiment (Yoshida et al., 1972). The estimate is certainly much higher than the 11.7-tons/ha record obtained at IRRI (De Datta and Malabuyoc, 1976). Moreover, the yield potential of the current improved rices far exceeds the yield levels being obtained on the farms of tropical Asia (cf. IRRI, 1977, 1979). The main biological constraints are lack of resistance to insects and diseases, resistance to drought, and competitiveness with weeds. The biological constraints attributable to insects and weeds are being tackled by rice researchers. On the other hand, the socioeconomic constraints related to land tenure, availability of capital, infrastructure, government pricing policy, technical knowhow, and traditions require the concerted efforts of the farmers, private indus-

42

T. T. CHANG, C. R . ADAIR, AND T. H. JOHNSTON

tries, and government sectors to remove a portion of the limiting factors. Coordinated efforts in such directions would undoubtedly raise farm yields.

II. DIVERSITY IN RICE GENETIC RESOURCES The genetic diversity existing in rice germ plasm is amazingly enormous. The 2 cultigens and 18 wild species of the genus Oryza are not only widely distributed (Chang, 1976a) but also represent an extremely ancient grass in the plant kingdom (Stebbins, 197 1). Evidence on the diversification process is presented as follows: A . GONDWANA ORIGIN OF THE GENUS A N D EVOLUTIONARY PATHWAY

The pan-tropical and widely scattered distribution of the 18 wild species in the genus covers Africa, Asia, Australia, and Central and South America (cf. Chang, 1976a, for their geographical distribution). The dispersal pattern has intrigued many botanists and geneticists since the nineteenth century and has given rise to divergent hypotheses about the progenitors of the two cultigens (0. sativa L. and 0. glaberrima Steud.) and the centers of origin for the genus. Thus, various workers postulated that 0. sativa was derived from 0. fatua Koenig, 0.perennis Moench, 0. officinalis Watt, or 0. minuta J . S. Presl. (see summary in Chang, 1964), which are distinct taxa involving different genomes and ploidy levels. Roschevicz (1931) postulated Africa as the center of origin for the section Sativa Rosch., to which both 0. sativa and 0. glaberrima belong. Oka (1977) has lumped the ancestral forms of the two cultigens into one speciescomplex, 0. perennis Moench. Chang (1976a,c,d) has pointed out that the geographical distribution of the wild species strongly suggests the Gondwana (Gondwanaland) supercontinent as the original habitat of the genus. The pantropical distribution of the Oryza species lies within the Paleozoic regions of the fossil fern Glossopteridae (Melville, 1966); the latter is believed by many botanists to be one of the principal progenitors of the angiosperms. Thus, the distant progenitors of the Oryza species had likely differentiated before the Cretaceous period. When the known genomic symbols of 15 Oryza species are placed by their present locations, they match well with their adjoining positions on the Gondwana components (Chang, 1976d). Fracture and drift of the major Gondwana components led to separate entities which are now known as Africa, Antarctica, Australia, Malagasy, South Amer-

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

43

ica, and South Asia along with its associated portion of Southeast Asia. The geological changes can readily explain the finding of the immediate wild relatives of the two cultigens (all of which have the A genome) present in Africa, Asia, Australia, and South America. Because human activities in cultivation and selection began early in Africa and Asia, the cultigens evolved at least 7000 years ago, while in Australia and South America only the wild forms are found (Chang, 1 9 7 6 ~ ) . As in other cereals, the evolutionary pathway of the two rice cultigens may be conceptualized as: wild perennial + wild (prototype) annual + cultivated annual. The evolutionary trend was parallel and independent for the African and Asian rices, but the Asian cultigen was further differentiated into three ecogeographic races. Detailed discussions were provided by Chang (1976~). B. DIVERSITY IN 0.SATIVA

AND

ITS W I L D RELATIVES

The rich diversity found in the Asian rices is probably unparalleled in other crop plants. Dispersal and selection by man have extended rice cultivation from the banks of the Amur River (53”north latitude) on the Sino-Russian border to central Argentina (40” south latitude). As the tropically based and semiaquatic plant was introduced into the cooler regions to the North or the high elevation areas of the tropics and subtropics, ecogenetic differentiation and human selection have greatly intensified the diversification process. Cultivation in different soil-and-water regimes led to extreme types which are adapted to pluvial (rainfed-dryland), phreatic (wetland) or fluxial (deepwater) conditions. Varietal sensitivity to photoperiod and temperature regimes at different growth stages became markedly divergent along latitudinal and altitudinal inclines. In the Ganges River delta, crops for three seasons (winter, summer, and autumn) evolved to fit into varying water and temperature regimes. Such variety groups are known as “boro,” “aus,” and “aman,” respectively. Rice cultivators’ personal preferences and socioreligious traditions have further added morphologic diversity to the cultivars (Chang, 1976c; Chang and Oka, 1976; O’Toole and Chang, 1979). Undoubtedly natural crossing between land races followed by differentiation and both natural and human selection have accelerated the diversification process. The rich diversity of the land races found in their native habitats of Bangladesh, Indonesia, and Sri Lanka has been described by Vaughan and Chang ( 1 980). The wild perennial (0.rufpogon Griff.) and the wild annual (0. nivara Sharma et Shastry) relatives of 0. sativa are generally found in the humid tropics of Asia. The wild forms have retained a high rate of natural crossing (5-40%) which enabled the populations to maintain their diverse and heterozygous com-

44

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

position (Oka and Morishima, 1967). However, the wild rices do not have the extreme range of adaptation to dryland or deepwater conditions found in the Asian cultivars (Chang et al., 1977). Intercrossing among the wild perennial, wild annual, and the cultigen has produced numerous intergrades which could be conveniently grouped as the spontanea forms of 0. sativa or the weedy annuals (Chang, 1976a,b). Hybridization among the four types in areas where they coexist continues to add genetic diversity to the hybrid swarms. Their dispersal is assisted by flowing water and seed exchange of the contaminated cultivars. Scientific plant breeding in many rice-growing countries has further enriched the diversity of the cultivars. In recent decades it has become increasingly difficult to classify rice varieties into their ecogeographic races (lndica, Sinica or Japonica, and Javanica) by their morphologic traits or grain characteristics. Many United States and Korean cultivars belong to this intermediate category. C. DIVERSITY IN 0.GLABERRIMA

AND

ITS WILDRELATIVES

The African cultigen also contains great diversity. 0. glaberrima is found in a full range of water regimes varying from dryland to deepwater conditions (Porteres, 1956; Oka, 1974). On hydromorphic soils the African rice is a selfperpetuating crop in admixture with other grasses and bushes. But its total range of diversity is less than that of 0. sativa. The smaller magnitude of varietal diversity in 0. glaberrima as compared to 0. sativa may be traced back to a low population density, a narrow north-to-south distribution, a rather flat topography, and a dearth of iron farm implements, irrigation facilities, and draft animals in Africa since the dawn of civilization (Chang, 1976~).With the rapid expansion of the Asian rices in the favorable areas of Africa where progressive agriculture is practiced, 0. glaberrima has been frequently reduced to the status of a weed race in fields planted to 0. sativa (Moormann and Van Breemen, 1978). As a result of natural crossing between the two cultigens, an increasingly large proportion of 0. glaberrima samples has lost their typical features (glabrous leaves and hulls) and meanwhile acquired a few other 0. sativa characteristics (Chang et al., 1977). The wild perennial 0. longistaminata A. Chev. et Roehc. is widely distributed in Africa, strongly rhizomatous, and incompatible upon self-fertilization. The wild annual (0. barthii A. Chev.) is quite diverse in morphologic and electrophoretic properties, probably exceeding that of the African cultigen but less than that of 0. sativa (Bezancon et al., 1977). Intercrossing among the three African taxa has given rise to a weedy race known as 0. sfapfii (Chang et al., 1977).

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

45

Ill. RECENT EFFORTS IN GENETIC CONSERVATION

A survey made in 1970-1971 indicated that agricultural research centers of major rice-producing countries had made varying attempts to assemble and conserve the indigenous rice cultivars shortly after World War 11. Commercial varieties in the readily accessible areas were the main target in field collection. Minor varieties in the remote areas and wild rices in their natural habitats were largely overlooked, however (Chang, 1972a). Systematic collection efforts in tropical Asia were initiated when the highyielding varieties spread quickly from 1967 to 1970 and threatened to replace numerous traditional (unimproved) varieties in the irrigated areas. Several national centers joined IRRI, USDA, USAID, and the Rockefeller Foundation country programs in developing plans for field collection. Systematic plans were formulated at the Rice Breeding Symposium held in 1971 and the IRRI-IBPGR (International Board of Plant Genetic Resources) Workshop on the Genetic Conservation of Rice convened in late 1977, both of which were hosted by IRRI (IRRI, 1972, 1978b). One hundred rice breeders present at the 1971 Symposium requested IRRI to initiate and coordinate field collection activities (IRRI, 1972). Priority for implementation and for IRRI’s participation in a country program is based on: 1. The rate at which improved cultivars replace local or traditional varieties 2. The richness of genetic diversity and the range of environments within countries or areas 3. The time and extent of past collection and preservation efforts 4. The accessibility of potentially rich germ plasm areas to field collectors 5. The extent of local (in-country) support for collection 6. Funding from an outside source Thus, the high-priority areas have been Bangladesh, Burma, Southwest China, India, Indonesia, Kampuchea, Laos, Philippines, Sri Lanka, and Vietnam. Fortunately, flexibility and opportunities in timing have enabled IRRI to work cooperatively with most of the countries mentioned (Chang, 1980). A.

STATUS OF GENETIC CONSERVATION IN THE UNITED STATES

No species of Oryza including the Asian cultigen (0.sativa L.) are indigenous to continental United States. As a result, all the materials in the USDA collection are foreign introductions or lines derived from them. In the earlier years the only

46

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

material received was seeds sent to the United States by embassy personnel or collected by plant explorers who were interested primarily in collecting other crops. It was not until about 1926 that J. W. Jones, then the USDA rice breeder in California, made a trip to Asia and collected seeds of leading commercial and experimental rice varieties in Japan, China, and Indonesia. Through contacts made by Mr. Jones, he continued to receive new varieties from rice breeders in the countries he visited. For a few years immediately following World War I1 many agricultural scientists were on temporary assignments in Asia and collected rice seeds and sent them to the USDA. From this source United States rice workers received seeds of many varieties from China, Japan, Korea, Taiwan, and some Middle East countries. In the early years of this century United States agricultural personnel stationed in the Philippines sent seeds of many of the commercial rice varieties to the United States. In recent years the USDA has sponsored rice collection programs in Assam State in India and in Pakistan. Seeds of the varieties collected were sent to the United States and to IRRI, in addition to being added to the rice collections in the country of origin. Until about 1957 no rice seed storage facility was available in the United States so the seed had to be regenerated frequently to retain viability. Because of this condition many entries were lost and the genetic structure probably was altered for some entries. In 1957 a medium-term storage facility was established at the Beltsville (Maryland) Research Center, and soon after that the National Seed Storage Laboratory (NSSL) was established at Fort Collins, Colorado. Space was made available in both of these facilities for rice storage. Since these facilities were established, it has been possible to retain viable seed on a somewhat more reliable basis than in earlier years. A recent check of the material in the USDA rice collection illustrated this fact. Of the varieties received before 1919, less than 7% are still in the collection; of those received from 1919 to 1933 about 9% are still available; of those received from 1934 to 1948 about 37% are still available; but of those received after 1948 practically all were in the collection. Most of the accessions introduced recently but not now in the collection are for the most part long-season, daylength-sensitive, tropical cultivars. The number of entries in the United States rice cultivar and genetic stocks collections is now in excess of 13,000. Included in this number are all cultivars that have been grown commercially in the United States. The accessions in the USDA collection were received from about 75 countries throughout the world. The USDA collection is probably deficient in land races and wild species of Oryzu and some of the long-season tropical types. However, this collection does contain a fairly good representation of the rice germ plasm of most countries. In 1962-1963, seed of all varieties in the USDA rice collection were sent to IRRI for testing and inclusion in the genetic stocks collection at that institution. Since then all new entries in the USDA collection have been sent to IRRI. In

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

47

turn, USDA also has received all released varieties and many experimental lines from IRRI. Since about 1963 small quantities of seed of all entries in the USDA rice collection have been stored in the NSSL. Small samples of seed of most varieties in the IRRI collection also are stored in the facility at NSSL. The introduction of rice seed for growing in rice-producing areas of the United States is restricted by a federal law. Before about 1968, all rice introductions were grown in a greenhouse at Arlington Farm Virginia or at the Beltsville Research Center (Cooper, 1976). Starting in 1968 a cooperative arrangement was developed with the California Agricultural Experiment Station to grow new introductions at the Imperial Valley Field Station at El Centro (Lehman er al., 1970). This location is far removed from commercial rice-production areas. Before planting at this location the introduced seeds were treated in hot water and with pesticides to reduce the risk of introducing diseases, nematodes, or insects. Recently another method for introducing seeds by cooperating research stations has been developed. Under this arrangement a qualified plant pathologist can obtain an official permit from USDA Plant Quarantine officials to import seed to his location. The introduced seeds are treated as above, then dehulled and the hulls burned, the naked seeds sterilized and then seeded in pots isolated in the greenhouse. The seed produced from these isolated plants then can be seeded in the field, provided the greenhouse plants remain healthy. For the past two to three decades rice researchers have been requesting additional funding, facilities, and qualified personnel to properly maintain and classify the entries still available in the USDA Rice World Collection and to computerize the available data. Progress has been slow, but Oakes (1980) reported that some additional space in a rejuvenated USDA laboratory recently had been provided for improved storage of rice germ plasm at Beltsville. In addition, available data are being computerized for more ready access. More recently (A. J . Oakes, personal communication, May 1982) word was received that the rice germ plasm (USDA Rice World Collection) has been consolidated with and made a part of the USDA Small Grains World Collection at Beltsville. The National Plant Genetic Resources Board (NPGRB) (1979) in a report to the Secretary, USDA, described a seven-phase program for conserving and using plant genetic resources and stated as follows: “The program includes plant introduction, classification, screening, basic genetics, developmental research, applied research, and finally production of seeds of improved cultivars for sale to farmers.” In discussing this program they further stated: This seven-phase program represents an outstanding example of State, Federal, and private industry cooperation and planning. The program has been and continues to be extremely successful, but it tends to receive low priority ratings in the budget process in spite of the fact that it is fundamental to all agricultural missions. Plant germplasm resources and their use are central to a multitude of national goals, including increasing exports; boosting farm income and enhancing the

48

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

national economy; protecting the environment; conserving energy; helping with soil conservation; minimizing cost of food; providing safe, nutritious food; developing pest-resistant crops and crops better adapted to less favorable environments; and minimizing cost of building materials and other forest products.

This NPGRB report cites a “white paper” by several outstanding crop scientists who caution that well-proven plant-breeding methodologies that make use of the variability existing in domestic and exotic germ plasm must not be discarded for the more glamorous “genetic engineering. ” They state that genetic engineering is a potentially useful tool that must concurrently employ plant-breeding techniques if it is to be effectively applied. They contended that plant germ plasm conservation and use and conventional plant breeding do not now receive the high priority in research funding that they deserve. Reports by P. R. Mooney (1980) and by the Comptroller General of the United States General Accounting Office (1981) review the United States germ plasm conservation programs and offer suggestions for improvement. Many of the suggestions are good but, for the most part, only reiterate what researchers have been proclaiming for many years. B. STATUS OF CONSERVATION BY OTHER NATIONAL AND REGIONAL CENTERS

The following is a survey on the status of genetic conservation in major riceproducing countries or by regional centers based mainly on the proceedings of the 1977 IRRI-IBPGR Workshop on the Genetic Conservation of Rice (IRRI, 1978a).

I . Bangladesh Varietal diversity was rich in Bangladesh because the boro, aus, and aman types were grown in three seasons over different hydrological regimes. Dryland rices are also found in the hilly tracts. Assemblage of farmers’ varieties began in 191 1 when an economic botanist was appointed by the East Bengal Department of Agriculture at Dacca. Workers of the former Agricultural Research Institute collected thousands of cultivars from 1918 to 1960. Purification, selection, and removal of duplicate samples reduced the total number to 1442 entries. A substantial portion of the collection was of the aus type which is the summer crop. Since 1972 about 2450 samples were collected in collaboration with IRRI. Local efforts brought together 2780 more samples. The collection phase is being continued by the staff of the Bangladesh Rice Research Institute and extension workers. Replacement of traditional varieties was much greater in the aus and trans-

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

49

planted-aman varieties than in the broadcast-aman (deepwater) and boro varieties because breeding efforts for the latter categories have been small.

2. Burma Rice lands extend from 10 to 28" north latitude and include many distinct ecotypes. About 736 accessions were maintained by the rice experiment stations. During 1973-1974, 1100 samples were gathered from the less accessible areas with the cooperation of IRRI and local extension workers. This was followed by local efforts, and 960 more samples were collected from 1975 to 1980. The rate of replacement by improved cultivars is rather slow, but many hilly areas along the border remain inaccessible to the field collectors.

3. China Rich diversity existed in China as rice cultivation dated back to 7000 years ago or even earlier, and the production areas extend from tropical Hainan Island to the Amur River at 53" north latitude. The temperate race Keng (Sinica or Japonica) was differentiated inside China. Three wild species are found in China (Lu and Chang, 1980). Thousands of farmers' varieties were assembled before World War I1 and many samples were purified. About 710 samples were sent by the National Agricultural Research Bureau staff to the USDA in the late 1940s. In the early 1950s, China began a massive campaign to collect, conserve, and use its vast crop genetic resources. Between 1955 and 1958, about 40,000 rice samples were gathered from various provinces and maintained by each province concerned. During the Cultural Revolution (1967- 1976) a number of seedstocks were lost due to lack of rejuvenation. About 33,000 samples remained mostly in small seed quantities. The national Crop Germ Plasm Resources Institute was established in 1978. It holds the regional rice collection of 2500 accessions for North China and aspires to be the national depository for crop germ plasm (Rockefeller Foundation, 1980). Field collection efforts were renewed in 1979. A team canvassed extensively in the hills of Yunnan province in the southwest where Indica and Sinica races are distributed along altitudinal clines (Ting, 1961). The Taiwan Agricultural Research Institute separately maintains a rice collection of about 2500 accessions, of which 1662 are Chinese varieties originating from the mainland or from Taiwan. The replacement of traditional varieties on the mainland had been rapid since the semidwarfs were developed during 1959-1963 and the F, hybrids were extensively grown in 1978. On the island of Taiwan, traditional varieties of the Indica type practically vanished after Taichung Native 1 was released in 1961. The wild rices became extinct in the 1970s.

50

T. T. CHANG, C. R. ADAIR, A N D T . H. JOHNSTON

4. India Rice lands cover a vast array of ecological niches, and the diversity of germ plasm, both cultivated and wild, is enormous. An assemblage of farmers’ varieties began in 1911 when an economic botanist for rice and other crops was appointed at the Paddy Experiment Station of Madras State at Coimbatore. The Coimbatore station also served as the central depository for rice until the Central Rice Research Institute (CRRI) was established in 1946. Since 1929 the Indian Council of Agricultural Research (ICAR) has assisted the states of Bihar, Orissa, Madhya Pradesh, and Uttar Pradesh to add rice botanists to their agricultural departments. The rice botanists made the early collection of local varieties for breeding purposes (Parthasarathy, 1972). By 1946-1947, about 2000 varieties were assembled. During the early 1950s CRRI was designated by FA0 as the regional center for Indica varieties. Hundreds of foreign varieties were assembled by CRRI for use under the F A 0 Indica/Japonica hybridization project. National awareness of genetic conservation was spurred by a conference held in 1951 (Ramiah and Ghose, 1951) which led to a systematic survey of the Jeypore Tract by the CRRI staff during 1955-1956 and 1958-1960. More than 1500 samples were collected. This was followed by collection in Manipur State during 1965-1970. An intensive collection project in northeast India was assisted by PL 480 funds. About 6630 samples were assembled under the ARC (Assam Rice Collection) code, from which many promising sources of useful traits were later identified. Collection in rainfed and dryland areas was implemented during 1976-1977. The Indian Council of Agricultural Research sponsored a massive collection project during 1978-1979 and about 5400 samples were collected from 14 states. The collection efforts are being continued by several state and regional centers. The national collection held at CRRI is approximately 15,000 accessions. Holdings at various state levels amount to 35,000 or more. Considerable duplication exists in the state collections. Field collectors need to canvass many small ecological pockets where traditional types are grown, especially in the northeast region. The rate of genetic erosion has been rapid in the irrigated areas. 5 . Indonesia

Although climatic conditions of the Indonesian islands were generally similar, the hydroedaphic variation is marked. Indonesia is also the home of the Javanica ecogeographic race (locally called the bulu and gundil types) which coexists with the predominant Indica race (tjereh or cere). Field collections were made in the 1930s, but no rigorous effort was made to

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

51

preserve the collection. Since 1972, systematic efforts were executed with IRRI’s participation, and a total of 10,000 samples has been assembled from the major islands. Recent collection efforts were directed toward the smaller islands in the east. The rate of replacement of traditional varieties by semidwarfs has been rapid. About 60% of the rice area is planted to the high-yielding cultivars.

6. Japan All of the Japanese varieties belong to the temperate-zone race, Japonica or Sinica. Twenty leading varieties are grown on about 60% of the rice land. In 1962, about 1300 local varieties were collected from farmers’ fields. Japanese scientists have actively collected cultivars and wild taxa in South and Southeast Asia, Australia, Africa, and Latin America. The National Institute of Agricultural Sciences at Tsukuba maintains a collection of about 10,500 native rices, while six major experiment stations have 9000 stocks. Other collections maintained by Kyushu University, the National Institute of Genetics, and other universities amount to 16,000; collections of prefectural agricultural experiment stations totals about 8000 varieties.

7. Kampuchea Local collections were maintained at different stations. During 1973, IRRI assisted the local workers in assembling 948 samples around Phnom Penh and Lake Battambang. Areas formerly controlled by the Khmer Rouge have not been covered by systematic canvassing; however, the adoption of high-yielding varieties is negligible.

8. Korea (South) Prior to the release of “Tong-il” in 197 1, all Korean varieties were of the Sinica race. Field collection of local varieties began in 1906 with the establishment of the national agricultural experiment station. By 1977 the collection had grown to 4227. High-yielding cultivars of IndicdSinica parentage now occupy 60-75% of the total rice area. 9. Laos

In the early 1970s the government had a collection of about 250 varieties. The USAID Mission to Laos assisted extension workers during 1972- 1973 in assembling 898 samples, a substantial portion of which was harvested from preceding

52

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

crops and was therefore nonviable. The canvassed areas were confined to those controlled by the former royalist government. The spread of semidwarfs in Laos was negligible. 10. Malaysia

Conservation of local rices grown in West Malaysia was largely completed in the late 1960s. About 3130 varieties were maintained by the Malaysian Agricultural Research and Development Institute (MARDI). Missionary workers in Sarawak State of East Malaysia collected 166 varieties from the hilly areas and sent the seed to MARDI and IRRI. Further collection in Sarawak and Sabah States is needed. 11. Nepal

Despite Nepal’s small size, the hilly topography encompasses climatic conditions ranging from tropical to temperate. Thousands of varieties were grown. In 1971 the USAID financed collection in 57 of the 75 districts, and 780 rices were gathered. Collection efforts are being renewed for the less accessible mountainous areas. Genetic erosion is serious in the Kathmandu and Chitawan valleys where improved varieties of foreign origin occupy more than 80% of the rice area. In the hills the traditional varieties are predominant. 12. Pakistan

Rice is largely grown in the hot and semiarid regions, but varieties grown in the northernmost and northwest corners represent different ecotypes. Collections made during the early 1930s, early 1940s, and in 1955 totalled 555 accessions. From 1972 to 1977 a United States-sponsored PL 480 project helped the staff of the Agricultural Research Council and rice experiment stations to assemble and evaluate about 900 rices. These latter accessions included current and obsolete cultivars and primitive varieties or land races. Insofar as possible the accessions were classified for botanical, morphological, and agronomic characteristics as well as for resistance or tolerance to diseases, insects, and adverse environmental conditions (Husain, 1978). Modern varieties are rapidly replacing traditional types in the major production areas. Renewed collection efforts are needed to cover central Punjab, Swat Valley, and both banks of the Indus River.

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

53

13. Philippines

Ecological variation is considerable among the hundreds of islands stretching from 5 to 19" north latitude. Foreign introductions were extensively obtained before and after World War 11. Local varieties of the Indica type probably amounted to 1500 while the Javanica race predominates in the Mountain Province. Before the war 828 varieties were collected and maintained. The national collection of 607 accessions was turned over to IRRI in 1962. Another total of 805 samples was collected from 1962 to 1980. An anthropology team from Yale University collected intensively in the mountainous areas and donated 466 Javanica varieties to IRRI. Canvassing in a number of isolated areas is being continued by Philippine extension workers in cooperation with IRRI staff. In tropical Asia, the Philippines has the largest proportion of rice area (about 90%) planted to the high-yielding cultivars. But a substantial proportion of the replaced varieties was foreign introductions. 14. Sri Lanka

Although Sri Lanka is a small island, it has marked variations in climatic, edaphic, and hydrological conditions. The island is a microcenter of varietal diversity. Its rices are also a principal source of resistance to several major insects and to adverse soil factors. Collaborative collection efforts with IRRI from 1972 to 1980 led to the acquisition of 2200 samples. The canvassing is essentially complete for the island. The national collection holds 25 16 land races. Genetic erosion has rapidly taken place since the early 1970s. Locally improved cultivars are grown on over 90% of the rice area. 15. Thailand

Similar to neighboring Burma and Vietnam, Thailand is rich in the genetic diversity of shallow- and medium-depth lowland varieties, deepwater rices, and hill (dryland) rices. An extensive collection campaign from 1950 to 1967 yielded 6739 samples from wetland areas. Evaluation, purification and selection reduced the collection to 2434 stocks. Field collection in the 1960s and 1970s were largely for the dryland types (hill rices). The collection efforts are being sustained with IBPGR-IRRI assistance. Genetic erosion has proceeded at a rather slow pace as the improved cultivars are grown largely in the dry season in irrigated areas. The wet-season crop is dominated by the traditional rices.

54

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

16. Soviet Union Rices found in the Soviet Union are either foreign introductions or locally improved cultivars. The national collection initiated by N. I. Vavilov now totals 3200 accessions, most of which are introductions.

17. Vietnam Owing to the long north-to-south configuration, Vietnam is also rich in varietal diversity. The early-maturing Champa rices of central Vietnam in the eleventh century contributed much to China’s increased rice production during the ensuing centuries (Ho, 1956). No information is available on the national collection. About 387 accessions were maintained in Saigon during the 1960s. A collaborative venture with IRRl in early 1975 led to the assemblage of 760 samples from the sourthern provinces. Collection efforts have been implemented in recent years. The University of Cantho in the south has assembled a collection of 1350 samples. Genetic erosion had already taken its toll in the late 1960s and early 1970s when the semidwarfs spread rapidly in the favored areas. 18. Central and South America

During the period from the fifteenth to the eighteenth centuries, numerous rices were introduced from Europe, Asia, and North America. Extensive exchanges among Latin American countries were known (Lu and Chang, 1980). Local selection and breeding efforts have enriched the limited diversity of the 0. sativa introductions. Among the Latin American countries, Brazil has maintained a collection of 4000 accessions since 1974. A survey and collection project was implemented in 1976-1977, and it led to the assemblage of 533 unimproved cultivars. Argentina has a collection of 3200 samples. Genetic erosion has been nearly complete in the irrigated and mechanized upland areas. Unimproved varieties can be found only in the unmechanized upland areas.

19. East and North Africa Malagasy probably has the longest history of growing 0. sativa in East Africa because of its proximity to South and Southeast Asia. The Indica race dominates the plains while the Javanica race occupied the central plateau. Malagasy has a collection of about 2000 accessions (Toll et al., 1980).

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

55

Mozambique also has a long history of rice cultivation (Lu and Chang, 1980). but the germ plasm is composed of foreign introductions. Rice cultivation in Egypt dates back to the seventh century. Egypt has about 2500 entries in a working collection for breeder’s use (personal communication from Dr. M. I. Maximos). 20. West Africa The introduction of Asian rices has been rather recent, no earlier than the fifteenth century (Lu and Chang, 1980), but the Asian cultivars underwent considerable genetic changes and diversification after local selection took place in the widely scattered production areas of diverse ecosystems. On the other hand, no breeding work has been carried out on the African cultigen. Systematic collection in the Francophone countries was first implemented by staff from the Institut de Recherches Agronomiques et des Cultures Vivrieres (IRAT) and the Office de la Recherche Scientifique et Technique d’Outre-Mer (ORSTOM), both of France, since 1950. Parts of Liberia, Tanzania, and Malagasy were also included. The collections assembled up to 1979 totaled 996 0. sativa samples, 636 0. glaberrima populations, and 500 wild taxa. The UNDP/FAO project staff in Liberia mounted an extensive collection operation from 1971 to 1974 and gathered 1734 0. sativa and 135 0. glaberrima samples. The International Institute of Tropical Agriculture (IITA) in Nigeria began extensive collection efforts in 1976 with the support of IBPGR. More than 10 countries in West, Central, and North Africa were covered during a 4-year period. The collections amounted to 2363 samples of 0. sativa land races, 0. glaberrima samples, and wild taxa. The West Africa Rice Development Association (WARDA) located in Liberia has also assembled 7650 samples from six African countries. West African countries including Ghana, Liberia, Nigeria, and Sierra Leone also maintain rice collections at their national agricultural research centers.

C. CONSERVATION OF THE WILDSPECIES The wild species are widely distributed in the humid tropics of Africa, Asia, Australia, and Central and South America. Initial collection was made by botanists primarily for taxonomic studies. With funds from The Rockefeller Foundation, scientists of the National Institute of Genetics in Japan extensively collected in South and Southeast Asia, West Africa, Latin America, and East Africa during the late 1950s and early 1960s. Several university professors of Japan also collected wild rices while they systematically surveyed the genetic diversity of

56

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

rice cultivars in South and Southeast Asia (cf. Chang, 1975). Field collectors of IRRI have also gathered wild rices while collecting cultivars in remote areas of tropical Asia. Volunteer efforts of workers in Australia, Brazil, and Liberia have expanded the collection maintained at IRRI. The most extensive collection of the African taxa was made by the staff of IRAT and ORSTOM. IRRI has practically all of the collected materials. The total size is around 1100 samples, many of which are heterozygous populations rather than pure strains. The size also indicates that only a small segment of the wild taxa has been conserved. The wild rices of tropical Asia are rapidly dwindling or decreasing in population size in their traditional habitats: canals, ponds, and roadside ditches. Moreover, the wild rices are losing many of their characteristic features because of introgressive hybridization with the rice cultivars grown in their vicinity. A recent study by the IBPGR-IRRI Rice Advisory Committee urges that immediate steps be taken to conserve the rich gene pools in the wild relatives (IBPGR-IRRI, 1982). D. COLLABORATIVE CONSERVATION EFFORTS

From the beginning of its research operations in 1962, IRRI has served and continually expanded its role as a central depository for rice germ plasm. The growth of the IRRI germ plasm bank into a genetic resources center has been described (Chang, 1972b, 1980; Chang et al., 1975b). Since 1972 IRRI staff has directly participated in the collection programs of Bangladesh, Burma, Indonesia, Kampuchea, Philippines, Sri Lanka, and Vietnam, leading to the acquisition of 10,352 samples from many remote areas (Table I). During their field travel, the teams inquired about the existence of ecoedaphic constraints in the canvassed areas. Such efforts have enabled the collectors to acquire cultivars having tolerance to one or more of the adverse factors (IRRI, 1981). Through IRRI’s promotion and technical assistance, workers in 14 Asian countries have implemented field collection activities and assembled more than 22,000 samples (Table I), although a higher proportion of duplicate and nonviable samples exists in the collections than those canvassed by the team of national and IRRI staff members. The collaborating workers include not only research scientists, college professors and their students, and extension workers in the national program concerned, but also foreign missionary workers, United States Peace Corps members, other service volunteers, and anthropologists. The national center concerned and IRRI share the collected seed samples. IRRI’s catalytic role in pooling financial and manpower resources for systematic field collection efforts has been discussed (Chang and Perez, 1975; Chang,

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

57

Table I Indigenous Rice Varieties Collected with IRRI's Direct or Indirect Participation in 14 Collaborating Asian Countries, 1971 to 1981 Indigenous varieties collected (number) ~~

Country

Years

Bangladesh Bhutan Burma India Indonesia Kampuchea Laos Malaysia Nepal Pakistan Philippines Sri Lanka Thailand Vietnam

1973-1981 1975-1976 1973-1974, 1976, 1980 1976, 1978-1981 1972-1976, 1978-1 98 1 1973 1972-1 973 1973-1 979 1971-1972, 1979-1981 1972-1973, 1976, 1979 1973-1976, 1977-1980 1972, 1975-1976, 1978-198 1 1973, 1975-1976, 1978-198 1 1972-1 975, 1978-1980 Total

With direct participation of IRRI

245 1 225 5103 280 -

5 10 I675 -

108 10,352

With indirect participation of IRRI

2857 121 967 4828" 4681 898 972 1688 772 1971 550 2650 1650 24,605

UPartialestimate based on seed samples received by IRRI.

1980). The development of a manual for field collectors (Chang et al., 1972) has proved to be most helpful to field collectors in national programs. Since its establishment in 1973, the International Board for Plant Genetic Resources (IBPGR) has given high priority to the collection of rice in tropical Asia and in West Africa (IBPGR, 1976). Recently the Board has channeled funds through IRRI to assist a number of Asian centers in continuing the field collections, upgrading seed storage facilities, and training field collectors (IRRI, 1981). The Rice Advisory Committee cosponsored by IBPGR and IRRI has assisted in developing plans for field collection and in producing a set of uniform descriptors for rice cultivars (IBPGR-IRRI, 1980). Field collection in 12 West African countries has been implemented by the Food and Agriculture Organization (FA0)-United Nations Development Programme (UNDP) Project in Liberia, and by IITA, IRAT, WARDA, and ORSTOM. The IBPGR has financed the collection activities of IITA. The Centro Nacional de Recursos Geneticos (CENARGEN) of Brazil has surveyed and collected indigenous varieties. Each of the above institutions has provided IRRI with duplicate sets of collected samples for preservation. IRRI has received about 3000 duplicate seed samples from the above centers. The overall and cooperative efforts on the genetic conservation of rice un-

58

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

doubtedly excel those of other major food crops. However, collaborative efforts have been seriously hampered in those areas where politicomilitary strife exists. The rate of collection has been rather low in other areas which are frequently inaccessible to collectors during the wet season, for instance, the deepwater areas and high elevations.

IV. DISSEMINATION AND EVALUATION OF GERM PLASM

A. DISTRIBUTION AND EVALUATION BY UNITEDSTATES WORKERS

In the earlier years rice germ plasm evaluation and distribution were conducted by USDA in cooperation with State Agricultural Experiment Stations in Arkansas, California, Louisiana, and Texas. Also, some introduced cultivars were grown in farmers’ fields and in some cases these growers selected, increased, and distributed selections from these materials (Jones, 1936). Until about 1930, the evaluation program for rice consisted of general observation on the adaptability, yield, and milling quality. During that period many varieties were distributed which were superior to the ones currently being grown. The varieties developed and distributed during that period were described by Jones (1936). In 1931, additional federal and state employees were added so there was a gradual increase in the germ plasm evaluation program. During the past halfcentury many of the entries in the USDA rice germ plasm collection have been evaluated for reaction to disease and nematodes and to field- and stored-grain insects (Rush et al., 1977, 1978a,b; Marchetti, 1975; Atkins, 1974; Cogburn, 1977a; Anon., 1980). Blast (Pyricularia oryzae Cav.) is one of the most serious diseases of rice. This disease has been the subject of much research in recent years in the United States. The reaction of cultivars and advanced lines to the causal fungus was observed at many locations in North America. There was a wide range in varietal reaction and a differential reaction among varieties at different locations. These observations led to studies of different strains of the fungus and to the setting up of a group of differential varieties to categorize these strains or physiologic races (Atkins et al., 1967). The reaction of rice varieties to P. oryzae when artificially innoculated has been studied in Texas (Atkins, 1965). For several years in the late 1960s and early 1970s, G. E. Templeton and T. H. Johnston (unpublished) had reasonably good success in annually screening about 2000 varieties and breeding lines for resistance to rice seedling blast in

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

59

Arkansas under conditions of late seeding, sprinkler irrigation, and natural infection. Two lines found resistant in these tests were later released as the cultivars “Nova 76” and “Mars” (Johnston, et al. 1979a,b). More recent research by Marchetti (1978) dealt with screening for general resistance to rice blast. For the past several years Fleet N. Lee in Arkansas and M. A. Marchetti in Texas (both unpublished) have screened hundreds of varieties and breeding lines under artificial inoculation to find the best possible sources of blast resistance in adapted types. These are being used as parents in the cooperative breeding programs. A large number of varieties were evaluated for reaction to the stem rot fungus Magnaporthe salvanii (Catt.) Krause and Webster by Cralley (1 936) in Arkansas. Additional studies to evaluate the reaction of genetic stocks to M. salvanii were conducted in Louisiana by Amad et al. (1974). Hoff et at. (1976) reported on evaluation of varietal resistance to the stem rot fungus in Louisiana studies. Stem rot studies also were conducted in California by Ferreira and Webster (1975). Figoni et al. (1981) reported that Oryza rufipogon has greater resistance to stem rot than found to date in 0. sativa. Inheritance of resistance in the interspecific cross (0. satival0. rufpogon) appears to be polygenic (Figoni, 1981). Brown leaf spot (Helrninthosporium oryzae B. de H.) is sometimes a serious disease in the sourthern United States. Varietal differences were noted and the heritability of reaction toH. oryzae was studied (Adair, 1941). More recently the reaction of many varieties to H . oryzae has been studied by Harahap et a f . (1974) in Louisiana and by Atkins (1974) in Texas and other southern states. According to Atkins (1974), Tullis in 1935 noted that United States varieties showed a wide range in reaction to the brown spot fungus. Adair and Cralley (1 952) tabulated reactions of numerous varieties to H . oryzae in Arkansas. Sheath blight of rice incited by Rhizoctonia solani Kuehn has become of increasing importance in the United States in recent years. Templeton and Johnston (1969) reported varietal differences in resistance among United States varieties. Eleven hundred and fifty rice cultivars and advanced lines were evaluated for response to this disease in Louisiana (Masajo et a/., 1974). Some mediumgrain varieties were more resistant than long-grain varieties, but no United States variety was highly resistant. A moderately high level of resistance was found in less than 1% of the 1150 varieties tested. In recent years hundreds of varieties and breeding lines have been evaluated under inoculated conditions for resistance to the sheath blight organism in separate extensive reaction tests by M. C. Rush and W. 0. McIlrath and co-workers in Louisiana, by M. A. Marchetti in Texas, and F. N. Lee in Arkansas. Although the reaction data have not been formally published, they have been made available to rice breeders and other interested persons for use in current improvement programs.

60

T. T. CHANG, C. R. ADAIR, AND T. H . JOHNSTON

Kernel smut of rice [Neovossia barclayana Bref. = Tilletia barclayana (Bref.) Sacc. and Syd.] is usually a minor disease of rice in the United States. This disease has been studied intensely (Tullis and Johnson, 1952), and an inoculation technique has been developed (Whitney, 1974). Reaction of varieties to this fungus has been studied (Templeton and Johnston, 1970b). Varietal differences in reaction were found. Two Arkansas selections which carry the designations C.I. 9695 and C.I. 9808 have shown a high degree of resistance to kernel smut in several cooperative tests in Arkansas and elsewhere (Templeton and Johnston, 1970a). Hoja blanca is a virus disease of rice that is transmitted by the planthopper Sogatodes orizicola Muir. This disease was a serious menance to rice production in many areas of Central and South America. It was first noted in the United States in 1957 (Atkins and Adair, 1957) and subsequently for a few years, but it did not become a serious disease in the United States because the vector did not become established. The United States collection of rice varieties was grown in Cuba and Venezuela. Several varieties were found to be resistant to hoja blanca (Atkins and Adair, 1957). Resistant short- and medium-grain varieties were distributed and a breeding program was conducted in the United States to develop resistant long-grain varieties (Adair et al., 1973). Lamey (1969) summarized information then available on varietal resistance to hoja blanca. Straighthead, usually described as a “physiological disturbance” has occurred in the United States for the past 70-80 years. Damage to highly susceptible varieties may range from slight to almost total loss of grain yield, depending on several environmental conditions. Reports by Atkins et al. (1956a,b, 1957) list reaction to straighthead of numerous varieties and describe testing methods used. Johnston et al. (1959) discussed straighthead and rice varieties in Arkansas and pointed out the differential reactions. Johnston and Templeton (1 970) reported on the comparative performance of rice varieties grown under severe straighthead and normal conditions. For the past several years Johnston (unpublished) has tested from 600 to 2000 varieties and breeding lines annually using a method partially based on research by Wells and Gilmour (1977). Application of arsenic as MSMA to the soil surface and working it into the upper few inches of soil is used along with continuous flooding to artifically induce “straighthead. Reactions of several of the older varieties tested under these conditions correlate very well with results obtained previously under “natural” field conditions conducive to straighthead. White tip disease of rice is caused by the seed-borne nematode Aphelenchoides besseyi Christie (Cralley, 1949). This disease was observed and reaction of many varieties was recorded (Atkins, 1974) before the causal agent was determined. In recent years the disease has been of relatively little commercial importance in the United States because the old susceptible varieties are no longer grown and most new varieties are resistant. However, “Brazos” and several of the currently or ”

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

61

recently grown California cultivars frequently show rather severe white tip symptoms when conditions are favorable and when the white tip nematodes are present. Atkins and Todd (1959) reported on the effect of white tip on yield of about 20 resistant and susceptible cultivars. Reaction of rice cultivars and advanced lines to the principal field insects has been studied. Many of these studies have been conducted in Louisiana and Texas since most insects are more prevalent in that region although studies also have been conducted in other areas. In a recent general report, Bowling (1980a) discussed the field insect pests of rice and the progress made in finding resistance among the thousands of accessions in the USDA World Collection of rice. Reaction of varieties to the rice water weevil (Lissorhoptrus oryzophilus Kuschel) was reported by Bowling (1963, 1973) in Texas; by Oliver et al. (1970, 1971), Gifford et al. (1974), Latson et al. (1976), Gifford and Trahan (1976), Robinson et al. (1978), and Smith et al. (1980) in Louisiana; and by Grigarick (1974) and Grigarick et al. (1976) in California. The stink bug (Oebaluspugnan Fabriciis) is a major pest of rice in the southern rice-producing states. Varietal reaction to infestation by this insect was studied by Latson et al. (1976) in Louisiana. Bowling (1980b) reported on the differential reactions of a number of cultivars and advanced lines from the USDA World Collection and the Regional Uniform Rice Nursery. The stem borer (Chilo plejadellus Zencken) causes serious damage to rice in some fields in the Gulf coast areas of Louisiana and Texas. Resistance in rice to this pest was studied by Oliver et al. (1970, 1971) in Louisiana. Differential susceptibility of varieties of stored rice to losses caused by storage insects was reported by Cogburn (1974, 1977a, 1980). In other studies Cogburn ( 1977b) found resistance to the angoumois grain moth (Sitotroga cerealella O h . ) in some varieties of rice from the USDA World Collecton. Cogburn et al. ( 1 980) also studied the effect of field environment on the grain of some of these varieties and its ultimate resistance to this insect. Another study conducted in Texas dealt with varietal differences and the nature of the resistance to this insect (Cogburn and Bollich, 1980). Russell and Cogburn (1977) reported on differences in resistance to seed penetration by this insect (S. cerealella) among USDA World Collection varieties. The reaction of rice varieties to low temperatures in the seedling and flowering stage is an important characteristic, especially in California. This characteristic has been studied (Adair, 1968; Li, 1975; Lin and Peterson, 1975; Peterson et al., 1972, 1974) and varietal differences have been found. Seedling vigor over a range of temperatures also was found to differ among varieties (Jones et al., 1974). Peterson et al. (1979) reported on cool-temperature screening of rice varieties and lines for seedling vigor. Rutger and Peterson (1979) discussed techniques for determining cold tolerance in rice varieties in California. Peterson et al. (1972) discussed differential panicle blanking under California conditions.

62

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

Li and Rutger (1980) reported on the inheritance of cool-temperature seedling vigor. Research studies have been made to determine susceptibility of rice varieties to commonly used herbicides. In a 2-year field test by Johnston and Smith (1976) an experimental high-protein variety from the cross “Zenith”/ ‘‘ADT3” showed 70% injury from molinate whereas 15 other advanced lines or cultivars showed little or no injury. Other breeding lines and many of the newer cultivars developed at IRRI and elsewhere also have shown high susceptibility to damage from molinate. Differential varietal responses to molinate and other commercial herbicides also have been noted by other researchers at other locations (unpublished observations). The physicochemical properties and cooking characteristics of the milled rice have been determined. The cooking characteristics of rice are correlated with the physicochemical characters of the grain. The physicochemical characteristics commonly used to screen rice varieties for cooking quality are: amylose content, reactions in dilute iodine and dilute (2.0, 1.7, or 1.5%) potassium hydroxide solutions, gelatinization temperature, viscosity of the rice flour paste, and parboil-canning stability (Adair et al., 1971, 1973). Many varieties in the United States collection have been evaluated for these characteristics (Webb et al., 1968; Adair et al., 1971) and the information is used in breeding programs. An example of the use of these tests in rice breeding is the development of “Newrex, ” a cultivar with higher amylose content than other United States long-grain types and thus more suitable for canning (Webb and Bollich, 1980). The data obtained in the evaluation tests are entered in the USDA World Collection Rice Data Bank (Johnston et al., 1978, 1980). This data bank is patterned after the Rice Data Bank System of Louisiana State University (Schilling and Parenton, 1976). Seed samples of public rice cultivars in the United States are available to seedsmen and farmers. Each state in the United States has a program of seed certification and distribution. The methods used in seed production and certification are described by Adair et al. (1973). New cultivars developed by public institutions and private breeders are registered and a description of each is published in Crop Science. The procedure was instituted for rice in 1958. All major rice cultivars that had been distributed before that date were registered and a description published that year (Johnston, 1958). Varieties developed since that date have been registered and a description published at the time of release. Not only are commercial varieties (cultivars) registered, but breeding and genetic lines (elite germ plasm) are registered and a description published, and seed is made available to all rice breeders (Rutger et al., 1979a). The entries in the USDA rice collection are available for distribution to public and private breeders after they have cleared quarantine.

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

63

B. THEGENETIC EVALUATION AND UTILIZATION PROGRAM OF IRRI

Since IRRI began research operations in 1962, its staff has actively screened large numbers of accessions for reactions to the blast fungus, bacterial blight pathogen, the stem borers, and for grain quality. The evaluation program was soon expanded to include bacterial streak, the virus diseases, sheath blight, green leafhopper, brown planthopper, other planthoppers, rice whorl maggot (Chang et al., 1975a,b), adverse soil factors (Ponnamperuma and Castro, 1972), nutritive quality (Juliano, 1972), and drought resistance (Chang et al., 1974). The multidisciplinary evaluation program was expanded and systematized during 1974 as the Genetic Evaluation and Utilization (GEU) Program (IRRI, 1974; Brady , 1975). One of the major focuses under the GEU Program is to develop rices that can cope with the biological and physical constraints found in the unfavored rice production areas. Farmers in such areas were bypassed by the modern technology associated with the semidwarf rices and access to irrigation water. Their numbers represent nearly three-fourths of the rice growers of tropical Asia (IRRI, 1976a). Interdisciplinary and problem-oriented rice breeding is the backbone of GEU , while the IRRI germ plasm bank provides the genetic inputs. Under the program, plant breeders are teamed up with “problem area” scientists, such as plant pathologists, entomologists, soil chemists, and cereal chemists, in planning and developing evaluation and breeding operations. Each team member contributes his specialized knowledge to the joint effort in identifying, screening, and hybridizing diverse rices to meet specified objectives. By so doing, a spectrum of the desired characteristics can be incorporated into a high-yielding or an adapted background which could meet a wider array of rice-growing environments, especially those of the rainfed areas. The major GEU problem areas include: resistance to diseases, resistance to insects, high levels of grain and nutritive quality, resistance to drought, tolerance to adverse soils, tolerance to submergence, ability to survive deep water, and tolerance to extreme temperatures. The team of each research area includes at least one plant breeder and one or more scientists in the problem area. As many as five disciplines may be involved in one problem area, such as nutritive quality. The team approach covers the development of mass screening techniques, identifying the outstanding sources of resistance or tolerance, research on the mechanism of resistance or tolerance, investigations on the interpopulational variation in a pathogen or insect, and study on the inheritance of resistance and the allelism among donor genes. In-depth research on complex traits is being jointly undertaken by scientists in academic institutions of advanced nations and by IRRI staff on a collaborative basis.

64

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

Systematic evaluation under the GEU Program includes 5 tests on diseases, 1 1 tests on insects, 4 tests on grain quality, 10 scores on drought resistance, 4 tests on adverse soil factors, 2 tests on flood tolerance, and 1 test on low temperature tolerance. Tests of other traits on smaller scales are also made by GEU scientists, but the data have not been fully computerized. In the breeding process, the rice breeders provide the expertise and leadership in choosing breeding approaches, selecting the cross combinations, planting the breeding nurseries, selecting the desirable progenies, processing the selected seeds for multidisciplinary testing for further selection, and collating the experimental findings for the next cycle of crossing and selection. Integration of collaborative testing efforts begins as soon as sufficient seed is obtained to test for a specific trait. The massive research data are entered into computerized files, summarized, analyzed, and updated by the statisticians. Research findings are discussed among members of each team before planting the next generation. The GEU scientists meet once a month to discuss and review current findings and problems. Visits to experimental plots form a part of the monthly gathering. Annual and 5-year program reviews help to set or modify breeding strategies and related investigations. The annual International Rice Research Conferences sustain the dialogue and collaboration with workers in national centers. Figure 1 exemplifies the operational flow of research in the drought-resistance component. Resistant or tolerant reactions are verified by repeated testing or expanded testing at additional sites and under diverse hydrologic-edaphic regimes (Chang et al., 1982). The evaluation techniques and findings of promising sources have been summarized by several IRRI scientists (Ou, 1972; Ling, 1974; Chang et al., 1974, 1975a, 1977; Chang, 1976e; Vergara et al., 1976; Juliano, 1977; Ponnamperuma, 1977; Pathak, 1977; Khush, 1977). Computerized data systems have been developed for the massive GEU program and for linking the GEU data with the morphoagronomic data of the germ plasm bank (Gomez et al., 1978; IRRI, 1980~). Certain GEU tests are accelerated when the germ plasm staff feeds newly harvested seeds of accessions reputed to have special characteristics to the concerned GEU scientist in order to speed up the evaluation process. Such special characteristics include tolerance to cool temperatures, floating ability, and tolerance to adverse soil factors. Among cultivars collected from high elevations, Japanese breeders on Hokkaido island have identified Silewah (Acc. 257 18) of Sumatra, Indonesia, as the most cold tolerant (Satake and Toriyama, 1979). Several saline-resistant accessions were collected from salty areas. The rapid growth of the GEU program may be indicated by the number of seed samples requested by the GEU scientists from IRRI’s germ plasm bank for

65

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES Doto Flies ond lnlofmotlon Exshonge

Germ

@om

hdvonced Field Testing

Sources

Wcilond breedlnp n m w c s

Ihe Philippines -------8n

In!ernotmnol Rice

Terlmq Program (IRTP)

FIG. 1. Schematic representation of the activities collectively representing the drought resistance component of IRRI’s Genetic Evaluation and Utilization Program. (*)Progeny evaluation includes mass screening under different hydrologic-edaphic-cultural regimes.

testing purposes: 2,300-10,600 packages per year during 1971-1973 to 20,500-50,350 packages during 1974-1977. The number of crosses made in a year likewise leaped from 287-485 crosses during 1971- 1972 to 4,525-5,780 during 1976-1977 along with the development of a vacuum emasculation technique (Herrera and Coffman, 1974). More than 100,000breeding lines are tested annually for various GEU traits. Since 1975 a GEU training program has provided hundreds of young scientists in national programs with specialized knowledge on different facets of the GEU program. Moreover, the trainees are frequently recruited on a team basis so that they can implement multidisciplinary national programs after returning to their home country. While the young scientists are residing at IRRI, each trainee completes a certain number of crosses which are planned to meet the breeding needs of his home country. The trainee takes the F, seeds with him when returning home. The GEU Program together with the International Rice Testing Program (IRTP) now takes up 40% of IRRI’s annual research funds.

66

T. T. CHANG, C. R . ADAIR, ANDT. H. JOHNSTON

C. EVALUATION BY WORKERS I N NATIONAL RICERESEARCH CENTERSOF ASIA

Historical records of China documented varietal differences and their descriptions as early as the first century A . D . (Ting, 1961). Empirical evaluation of economic traits undoubtedIy was made when rice breeders of Asia began to select among farmers’ varieties in the early part of this century. Scientific breeding of rice began rather late. In Japan, the first variety developed by hybridization, “Ominishiki,” came out in 1906 (Matsuo, 1957). One of the early examples of systematic varietal evaluation is Sasaki’s (1922) recognition of the variation in the varietal resistance to the pathogenicity of the blast fungus. Screening for varietal differences in stem borer infestation was initiated at the Central Rice Research Institute in India during 195 1 - 1955 (Israel, 1967). Studies with academically oriented objectives have also contributed to the understanding of other economic traits such as growth duration, seed dormancy, and grain type (Matsuo, 1952; Chandraratna, 1955). Intensified national efforts on systematic evaluation were markedly stimulated by IRRJ’s early success in finding and using valuable traits found in IRRI’s collection of cultivars and wild taxa (Pal, 1972; Zaman et al., 1972). Staff of the All-India Coordinated Rice Improvement Project (AICRIP) instituted regional testing of breeding lines and systematic screening for disease and insect resistance under several nurseries (Shastry et al., 1971; Freeman and Shastry, 1972; AICRIP (undated)). Other Indian workers have carried out mass screening for a specific pest in its endemic areas or “hot-spot” nurseries (Shastry et at., 1972; CRRI, 1980). More recently, national GEU Programs have been instituted in Bangladesh, Indonesia, and Thailand. National efforts on multidisciplinary evaluation of breeding lines and donor parents have been further intensified since the establishment of the International Rice Testing Program (IRTP).

D. INTERNATIONAL EFFORTSON DISSEMINATION AND EVALUATION

International exchange and collaboration on the evaluation of rice germ plasm began on a modest scale when the International Rice Commission of the FA0 implemented cooperative variety trials on adaptability during the period from 1956 to 1958 (Parthasarathy, 1972). This was followed by the International Blast Nurseries which were initiated under FAO’s sponsorship and turned over to IRRI in 1963 for coordination. Shortly after the IRRI germ plasm bank began its operations, it supplied rice researchers all over the world with seed samples of Asian rices, African rices,

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

67

wild species, and genetic testers. The total number of samples distributed during the past 20 years in response to more than 2500 requests amounted to about 80,000 seed packets. Between 1972 and 1981, the number of samples sent abroad in a year ranged between 2770 and 10,200. These statistics may indicate the magnitude of experiments for specific objectives being conducted by rice researchers in different countries. Since the mid 1960s IRRI has supplied numerous IR breeding lines to rice researchers all over the world. The number of lines sent abroad each year ranged from 5000 to 15,000. An international yield nursery of 32 entries was composed in 1972 and distributed to 17 countries in 1973. IRRI has continued to supply rice researchers with breeding lines. The number of seed samples of IR lines sent abroad each year remains at a level between 10,000 and 17,000. In 1966 a set of 303 varieties and breeding lines were sent by IRRI to 10 countries for testing and selection. The testing not only established the superiority of the semidwarf Indicas over other types of breeding lines but also led to the finding of many promising parents for use in various countries (Beachell et af., 1972). The International Rice Testing Program (IRTP) was established in 1975 with the enthusiastic support of rice researchers in many countries and funded by the UNDP. The types of nurseries have increased from 12 in 1975 to 37 in 1981. During 1981, 2250 sets of seeds were distributed to hundreds of workers in 50 countries. The nurseries consist of yield trials; observational trials; disease and insect nurseries; and tests for adverse soils, cool temperatures, internode elongation ability, and drought resistance. The nurseries cover different types of rice culture: irrigated, rainfed-wetland, rainfed-dryland, and deepwater. In the beginning, IRRI breeding lines and accessions of the IRRI germ plasm bank formed the bulk of the entries in IRTP nurseries. The contributions from national programs have rapidly increased from 30% in 1975 to slightly over 50% of the total entries in 1979 (IRRI, 1980a). Along with distribution of the seed for testing, a uniform scoring system was developed and provided to place recorded data on a standard basis (IRRI, 1976b, 1980~).The data returned by collaborators are processed, analyzed, and reported by IRRI staff. For Latin America and West Africa, the nurseries are handled with the collaboration of the Centro Internacional de Agricultural Tropical (CIAT) in Cali, Colombia; IITA in Ibadan, Nigeria; and WARDA in Monrovia, Liberia. Some regional nurseries are individually developed to cope with the needs of local ecoedaphic environments. On the basis of the IRTP tests, workers in entomology, plant pathology, and crop physiology have developed collaborative research projects to probe deeper into the specialized research problems. Such projects involving several countries include investigations on the biotypes of the brown planthopper and the gall

68

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

midge, pathotypes of the tungro virus and the bacterial blight pathogen, varietal resistance to the bacterial blight pathogen, cooperative testing and breeding for tolerance to extreme temperatures, site characterization and testing rices on adverse soils, cooperative breeding nurseries for drought resistance, exchange of early-generation breeding materials, and studies on tissue culture (IRRI, 198 1). E. PROBLEMS I N DISSEMINATION AND EVALUATION

The process of massive and collaborative evaluation efforts was also marked by handicaps and problems in the initial stages. Gradually some of the problems were resolved by periodic workshops, exchange of visits, and cooperative efforts. Plant quarantine restrictions of some countries and regions have slowed the flow of genetic materials among research centers. Special arrangements often were made to allow an accelerated release of the introduced seeds after appropriate inspection and observation. For disease and insect nurseries, a light epidemic or a lack of the outbreak renders notes taken on those nurseries of little practical value. Some of the nurseries have been relocated in endemic or “hot-spot’’ areas. In the case of light infections, the data are adjusted on the basis of the control varieties. For nurseries grown under the rainfed culture, vagaries of weather conditions at different locales across seasons have limited the comparative study and summarization of research data. Workers are now taking supplementary data on weather, hydrological conditions of the soil, and soil characteristics of the site, all of which will enhance the interpretative aspect of the tests. In recent years Thailand has added and expanded field facilities to test the internode elongation ability and submergence tolerance of materials intended for deepwater rice. Korea (South) has built special fields to test for tolerance to cold water. Such specialized facilities also take in a certain amount of breeding materials generated by other centers.

V. PRESERVATION OF GERM PLASM A. PRESERVATION BY USDA FOR

AND RECENTEFFORTS IMPROVEMENT

The rice germ plasm preservation program is centered at the Agricultural Research Center, USDA at Beltsville, Maryland. The program is conducted in cooperation with State Agricultural Experiment Station and federal personnel in

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

69

Arkansas, California, Louisiana, and Texas and with the National Seed Storage Laboratory (NSSL) at Ft. Collins, Colorado. The varieties for the most part are grown by USDA personnel. However, state employees and personnel of the California Cooperative Rice Research Foundation (CCRRF) cooperate in many of the evaluation tests that are conducted in field and greenhouse experiments. Seed of new entries in the collection are sent to IRRI and to NSSL each year. There is an informal cooperative arrangement between IRRI and NSSL for the conservation of all rice genetic materials in the IRRI collection. The plan is to send a small sample of each entry in the IRRI collection to the long-term storage facility at NSSL. The long-term storage at NSSL was installed in 1978-1979 (personal correspondence from L. N. Bass, Director, NSSL, November 9, 1978). The seeds are accepted for storage in the NSSL in accordance with the standard policies of NSSL which were adopted September 13, 1977. B. STATUSAT OTHERNATIONAL CENTERS

The size of rice collections maintained by different national and regional centers has been enumerated in Section III,B. The statistics were based on recent surveys made by IRRI (1978b) and IBPGR (Toll et af., 1980). Germ plasm is preserved at all centers in the form of seed. Only a few centers maintain small collections of wild species as live plants. The viability of rice seeds seldom exceeds one year in the humid tropics when the seeds are stored in paper bags under ambient temperatures and humidity inside unrefrigerated storerooms. Rice workers found it necessary to rejuvenate the seeds by growing the whole collection every year. Such a laborious process has led not only to mixtures and errors in the rejuvenated stocks but also to serious constraints on the germ plasm workers. Under such conditions several national collections have dwindled in size, and this has reduced the capability for international exhange. Among germ plasm-rich countries in the Asian tropics, nearly all countries have recently installed refrigerated seed storage rooms to insure that the rice stocks can last for 3-5 years (Toll et af., 1980). With the help of the Japanese government a modem facility for medium- and long-term storage has been constructed in Thailand. Japan, the Republic of Korea, and the Soviet Union have medium-term and long-term storage facilities, while China is beginning to develop such facilities at provincial and national levels. An encouraging development is that during the last decade nearly every riceproducing country in tropical Asia has donated their rice collections to IRRI for preservation. The IRRI collection is incomplete for a few state collections of India and for the traditional varieties of China, North Korea, and North Vietnam.

70

T. T. CHANG, C. R . ADAIR, AND T. H. JOHNSTON C. THEINTERNATIONALNETWORK

At a 1977 IRRI-IBPGR workshop on genetic conservation held in IRRI, scientists of Japan, the United States, IITA, IRAT, ORSTOM, and 12 country programs agreed to join IRRI in an international network of conservation in which IRRI would preserve the entire (base) rice collection of all institutions concerned so that at least one duplicate set of rice seed stocks is safely stored outside the country of origin. Moreover, IRRI would continue to deposit a duplicate set of its seedstocks at the United States National Seed Storage Laboratory in Fort Collins, Colorado, providing further security to the world's rice germ plasm (IRRI, 1978b). Staff members of IRRI, USDA, and NIAS of Japan are comparing their holdings so that redundancy among different centers can be reduced while no individual accession will be overlooked in the process of providing duplicate or triplicate sites of storage. The national centers are urged to construct mediumterm storage facilities and to help the international seed banks on rejuvenation (IRRI, 1978b). Both the National Institute of Agricultural Sciences (NIAS) of Japan and IRRI have modern facilities for medium- and long-term seed storage. At NIAS the rice seeds are dried to 6-7% moisture content at 50°C packed in vacuum inside tin cans, and stored at - 10°C (IRRI, 1978b). At IRRI the seeds are dried (at 38°C) down to 6% moisture content under a low-risk procedure, packed under partial vacuum inside aluminum cans, and kept at 2"C, and -10°C for medium- and long-term storage, respectively (IRRI, 1980d). IRRI presently holds approximately 60,000 accessions of 0. sativa, 2600 samples of 0. glaberrima, 1096 populations of wild species, and 680 genetic testers. During the last decade, several nations have lost a substantial portion of their rice collections because of civil strife or difficulties in maintaining their collections. IRRI has returned hundreds of accessions to the national centers of China, India, Indonesia, Kampuchea, Kenya, Nepal, Sri Lanka, and Thailand. D. PROBLEMS ENCOUNTERED IN PRESERVATION

Complexity in the genetic makeup of rice germ plasm and the magnitude of the operations have posed certain difficukies for the large centers. The following problems in characterization, seed multiplication, and preservation have been encountered (Chang, 1980). 1. Inviable incoming seed samples 2. Duplicate samples, morphologic variants, or ecostrains with identical varietal names

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

71

3. Inherent genetic variation within accessions 4. Incomplete information on the unique features of an incoming accession 5. Loss of the true-to-name features 6. Low seed yield of unimproved cultivars 7. Inefficient seed increase of ecologically unadapted or pest-susceptible accessions 8. Need for separate plantings for seed increase and rejuvenation of photoperiod-sensitive accessions 9. Unpredictable demand for seeds of certain accessions 10. Differential loss of seed viability among varieties during storage 1I . Misidentification or errors in processing accessions 12. Lack of biochemical criteria to differentiate between morphologically similar accessions or samples 13. Insufficient field space and manpower to concurrently handle three types of seed increase: (a) seed increase of new accessions, (b) rejuvenation of exhausted stocks, and (c) rejuvenation of old stocks for storage in the new mediumand long-term facilities. Therefore, seed increase activities are expanded in number and must be conducted over a longer period of time. IRRI staff also found it necessary to rejuvenate seed every few years for the often-requested or difficult-to-multiply accessions. On the other hand, the pitfalls of frequent rejuvenation pointed out by Chang et al. (1979), include: 1. Errors and mechanical mixtures 2. Loss of unadapted or susceptible accessions 3. Changes in genetic composition, especially for small populations 4. Increased workload and requirements for field and storage space One remedial measure for the above problems is to multiply sufficient seed in one planting, confirm the varietal identity, and preserve the seed under cold storage. Such an approach has been adopted at IRRI but poor seed production of unadapted or highly susceptible accessions remains the main constraint.

VI. USE OF GERM PLASM A. BY UNITEDSTATESWORKERS

Rice is not indigenous in the United States so all United States cultivars were derived by some means or another from introduced materials in the germ plasm collection of USDA. The history of the development of the cultivars grown in the United States has been documented by Chambliss and Jenkins (1923), Jones (1936), Jones et al. (1941, 1953), Adair et al. (1973), Johnston (1958), and by

72

T. T. CHANG, C . R. ADAIR, AND T. H. JOHNSTON

other registration statements in the Agronomy Journal and Crop Science. The genetic base of the United States cultivars is somewhat narrow (National Academy of Sciences, 1972). However, concerted efforts are being made to broaden the genetic base of the United States rice cultivars (Johnston et al., 1978). Several varieties in the collection are being used in the cooperative rice-breeding programs. For example, a short-grain improved variety introduced from Taiwan about 20 years ago (Tainan-iku 487, P.I. 215936) as a “Ponlai” variety has been an outstanding parent in the development of the high-yielding cultivar “Nortai” (Johnston et al., 1973). Several other varieties from a number of countries have been utilized in recent years to contribute desirable genes and help to broaden the genetic base of United States cultivars being released to growers. New varieties imported from these programs will help to broaden the genetic base of available parent material. The recently released cultivar “Newrex” (Bollich, 1979) is an example of a cultivar that has a somewhat different genetic background than older United States cultivars. Rice genetic stocks that have unique characteristics but are not suitable for commercial production can now be registered by the Crop Science Society of America. This program allows all rice breeders to get seed of these genetic stocks if they are needed (Rutger et al., 1979a). B. BY OTHERNATIONAL CENTERS

Nearly all of the national centers have made profitable use of the semidwarfing gene (sd-I) contributed by Dee-geo-woo-gen and varying numbers of the pest resistance genes derived from IRRI lines or IR varieties. Moreover, through local screening and selection, several national centers have incorporated additional resistance or tolerance genes from other sources into their improved cultivars. Among the major rice-producing countries in Asia, researchers in India have conducted the most extensive testing of indigenous germ plasm to pests and diseases (cf. AICRIP, [undated]; CRRI, 1980). Therefore, Indian rice breeders have made effective use of the indigenous genepools which provide resistances to pests or tolerance to ecoedaphic stresses. The drought-resistant ‘“22” was used in breeding “Bala.” “TKM6,” which has multiple resistance to insects and diseases, became a parent of “Ratna,” “Saket 4,” “Parijat,” “CR44-1,” and other improved cultivars (CRRI, 1980). TKM6 was also extensively used at IRRI as a donor of disease- and insect-resistance (Khush, 1977). The gall midgeresistant “Eswarkorra” was used to breed W1251, W1256, and W1263; the latter lines were widely used inside India as well as in Sri Lanka and Thailand (Rice Division, 1976; Dalrymple, 1978; CRRI, 1980; Directorate of Extension, [undated]). The tungro virus resistance of “PTBIO” has been bred into improved cultivars such as “Aswini,” “Bharathi,” “Jyothi,” “Rohini,” “Sabari,” and “Triveni.” Similarly, “PTB18,” possessing multiple re-

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

73

sistance, has been widely used in India (CRRI, 1980) and at IRRI (Khush, 1977, 1980). For tolerance to submergence by flood waters, FR13A is an outstanding source. Indians breeders were also developing saline-tolerant varieties from Indigenous sources such as “Pokkali,” “Getu,” and “Dasal” (CRRI, 1980). R575, a local variety of H.P. State, was used to breed “Himdhan” which is adapted to altitudes above 1000 m (Sharma and Kaushik, 1978). Indian workers have also made numerous exchanges with workers in other countries and benefited early from foreign introductions such as “Taichung Native 1 , ” “IR8,” “Mahsuri,” “Leb Mue Nahng,” and “China 1039.” Taichung Native 1 and IR8 were the principal sources of semidwarfism during the mid-1960s (Hargrove, 1976). Mahsuri of Malaysia and Leb Mue Nahng of Thailand were used to develop photoperiod-sensitive varieties (CRRI, 1980). “Rajendra Dhan 20” and “Pusa 4-1-1 1 derived their disease resistances from “Tadukan” of the Philippines (Chaudhary et al., 1979). “BG79” of Sri Lanka was extensively used by Indian breeders in Maharashtra State to develop “Satya,” “Surya,” and “Suhasini” (Directorate of Extension, [undated]). Rice breeders of Sri Lanka have used the outstanding levels of resistance to insects and tolerance to adverse soil factors found in their diverse germ plasm. BG400-1 derived from OB678//IR20/H4 has resistance to gall midge, blast, and the bacterial diseases. BG276-5 originating from OB678/2*BG34-8 is resistant to gall midge and bacterial blight. BW 100 selected from H501//Podiwee-A8/HS has resistance to blast as well as tolerance to iron toxicity (Rice Improvement Program of Sri Lanka, 1980). Malaysian breeders extensively used the germ plasm of other Asian countries to develop improved cultivars. Mahsuri is one of the few Indica/Japonica hybrids that has won wide acceptance in several countries. “Sri Malaysia Dua 1 1 ” was derived from IR8/Pankhari 203 (cf. Dalrymple, 1978). Breeders in the Philippines continued to draw heavily on germ plasm supplied by the USDA and locally improved materials (Hargrove et al., 1979). Recently C1064-5 was selected from C22/IR26//C22/OS4. “OS4” came from West Africa and is known as a drought-resistant upland variety. Rice breeders in Bangladesh and Thailand have made substantial progress in improving their deepwater i c e s from indigenous germ plasm (Zaman, 1977; Prechachat et al., 1980). Thai breeders have retained much of the excellent grain quality of their export cultivars in the improved rices (Rice Division, 1976; Kongseree, 1979). Two sister lines (BR5 1-91-6 and BR5 1-46-C l), both derived from IR20/ IR5-114-3 and selected by the Bangladesh Rice Research Institute, have been recommended for release in Burma and India because of their promise in yield performance (IRRI, 1980a). An Indonesian breeding line (Kn-lb-361-1-8-6) having a high level of tolerance to cool temperatures was recommended for planting in the Mt. Province of ”

74

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

the Philippines as “RP KN-2” (Ronduen and Villareal, 1976). The Indonesian variety “Remadja” was used to breed “BG90-2” of Sri Lanka; the latter has been renamed as new releases in Burma and Nepal (IRRI, 1980b). Improved cultivars bred by Indonesian workers in the mid 1970s involved parents coming from foreign countries such as “Latisail” of Bangladesh, “Puang Nahk 16” of Thailand, “Basmati-370’’ of India, and IR varieties or lines (cf. Dalrymple, 1978). Chinese workers have made extensive use of indigenous semidwarfs, pestresistant sources, and foreign introductions at many breeding centers. The semidwarfs were derived from “Ai-zai-shan” or “Ai-Jiao-Nan-Te,” both of which were indigenous to China. A Japanese introduction led to many improved “keng” (Sinica race) varieties bred from “Nong-Keng 58.” IRRI varieties and lines have been extensively used in conventional breeding programs and in the development of recent hybrid rices (Shen, 1980). Chinese breeders have successfully incorporated the cytoplasmic-genic mate sterility found in a sterile plant (Wild Aborted) belonging to the annual weedy race, 0. sativa f. spontanea, into many productive backgrounds. This sterile source proved to be highly compatible in crosses with Indica varieties of China and with IRRI varieties. Since 1977 the planted area of hybrid rices has been expanded to nearly 8 million hectares (Lin and Yuan, 1980). Aside from the cytoplasmic sterile source found in 0. sativa f. spontanea, Chinese breeders had earlier used another spontanea plant in developing “Yatsen 1.” This variety was later used as a parent in breeding the “Bao-tanai” semidwarfs which are widely adapted and pest resistant (Kwangtung Agriculture and Forestry College, 1975). “Bao-tan-ai” and “Bao-xuan 2” were recently found by IRRI virologists to have a higher level of tolerance to the grassy stunt virus than other traditional varieties which do not have the Gs gene for resistance derived from 0. nivara (IRRI, 1982). Wide crosses of rice involving sorghum, maize, wheat, bamboo, and barnyard grasses also have been investigated in China (Shen, 1980), but the continually segregating progenies have yet to prove their agronomic worth. Korean workers have made dramatic advances in yielding ability by crossing IR lines with local varieties of the Sinica (Japonica) type in a collaborative venture with IRRI. A series of high-yielding cultivars led by “Tong-il” (selected from IR667-98) dramatically increased rice yields from 3.35 to 5.01 tons/ha in the demonstration fields during 1971. Tong-il came from the cross of “Yukara”/ TNl//IR8, and it was used to develop “Yushin.” The next series of Milyang varieties (Milyang 21, 22, and 23) involved “Jin Heung” of Korea and IR262-43-8 and “IR24” of IRRI. The series of 15 interracial hybrids developed between 1971 and 1977 had increased the national rice yield from 3.30 to 4.94 tons/ha and achieved self-sufficiency for the country in 1975 (ORD, [undated]).

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

75

However, serious disease epidemics made importation of rice necessary a few years later. Grain yield of the Ponlai (Keng or Sinica) type in Taiwan was raised to a new level by the development of “Tainung 67” which involved Taichung Native 1 in its parentage (Huang, 1979). Other breeders drew on local Ponlai varieties, Taiwan’s semidwarf Indicas, IR varieties and lines, and early-maturing varieties introduced from Japan to further improve both the Keng and Indica types. However, to cope with the major virus diseases and the brown planthoppers, rice breeders in tropical Asia rely heavily on IRRI’s resistant varieties and lines having a high-yielding background (Hargrove, 1978; IRRI, 1980b). C. BY INTERNATIONAL CENTERS

IRRI scientists have made extremely profitable use of the genepools present in the IRRI germ plasm bank. The initial and best known success is the incorporation of the semidwarf gene (sd- 1) into the background of tropical varieties, leading to the development of IR8 (see Chandler, 1968). The dramatic advances in yield performance and range of ecogeographic adaptedness associated with the improved plant type have been described and analyzed (Chang, 1967; Yoshida et al., 1972; Chang and Oka, 1976). The second stage, improvements in grain quality, was contributed largely by TKM6 of India and Tadukan progeny of the Philippines (Beachell et al., 1972). Successful improvements in disease and insect resistance were largely contributed by TKM6, a strain of 0. nivara, “CR94-13” (derived from “PTB 18” and “PTB 21”), “Mudgo,” and ‘“22,” all of which came from India. “Gam Pai 15” of Thailand furnished resistance to the tungro virus (Khush, 1977). Earliness was derived mainly from Chinese and Indian varieties (IRRI, 1980a). Resistance to drought was transferred from “Rikuto Norin 21” of Japan; “E425,” “Moroberekan,” and “OS4” of West Africa; “Nam Sagui 19” and “Khao Dawk Mali 105” of Thailand; “Khao Lao” of Laos; the Gorai varieties of India; and traditional upland varieties of the Philippines (Chang et al., 1982). The Kn lines of Indonesia and Chinese introductions selected in India have provided promising sources of tolerance to cool temperatures (IRRI, 1980b). Tolerance to salinity was derived primarily from varieties of South India and Sri Lanka, such as “Pokkali,” “Nona Bokra,” and “Getu.” Outstanding sources of tolerance to alkalinity such as “Damodar” and “Dasal” also came from India. Varieties tolerant to iron toxicity in wetland soils came from Sri Lanka and Vietnam. Sources able to withstand aluminum and manganese toxicities in aerobic soils were identified from the dryland varieties of the Philippines, Latin America, and West Africa (IRRI, 1981). On the other hand, many IR lines tolerant to adverse

76

T. T. CHANG, C. R . ADAIR, AND T. H. JOHNSTON

soil factors were obtained from crosses involving nontolerant parents (Ikehashi and Ponnamperuma, 1978). Breeders at CIAT in Colombia have used IR varieties and lines, selections bred at the Instituto Colombiano Agropecuario (ICA) of Colombia, blast-resistant varieties from Vietnam, and BG varieties from Sri Lanka in the breeding program. At IITA in Nigeria, rice scientists have utilized IR varieties and lines, African upland varieties, Taichung Native 1, and “LAC 23” in the breeding program. Varieties developed by the IRAT staff in Ivory Coast include in their parentage upland varieties from Africa, Indica varieties from Taiwan, and upland varieties from Brazil. D. GLOBALSHARING OF IMPROVED GERMPLASM

The rapid dispersal and adoption of IR varieties by different countries in Africa, Asia, and Latin America from 1965 to 1977 have been summarized by Dalrymple (1978). During 1976-1977 the total land area planted to IR varieties and other high-yielding varieties amounted to about 25.27 million hectares. Diffusion of the IR varities and lines into national breeding programs has been documented by IRRI staff (Hargrove, 1978; Hargrove et al., 1979; IRRI, 1981). Since late 1975 IRRI ceased to name and release new rice cultivars under the IR series-I 1 varieties were so named, numbering from “IR5” to “IR34.” The revised policy was adopted at a time when national rice-breeding programs and international cooperation through the IRTP had reached an expanded stage which made the practice of IRRI’s naming of varieties no longer necessary. This change further encourages national centers to fully utilize IRRI’s breeding materials and other selections included in the IRTP nurseries. Several national centers have named selections chosen from the above sources either under their own codes or as IR varieties. Since 1975, 2 1 varieties have been released in 23 countries under the cooperative testing program (IRRI, 1978a, 1980a). The establishment of the IRTP in 1975 marked a giant step in the systematic and coordinated exchange and sharing of improved germ plasm by rice researchers all over the world. The testing of a uniform set of genotypes across many environments has provided credibility to the initial experimental findings as well as information on genotype X environment interactions and growth stage-specific responses. A 5-year report of the IRTP (IRRI, 1980b) furnished an extensive list of stressresistant or tolerant selections confirmed by many rice researchers. Some of the outstanding performers or resistant donors in the international nurseries were subsequently released by national programs as recommended cultivars. Since 1979 more than 30 IRTP entries have been released in that manner in more than

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

77

20 countries. The use of other promising entries as parents in national breeding programs provides an even greater potential in using the useful genetic pools of diverse origin (IRRI, 1980b). E. PUBLICEFFORTSVERSUS PRIVATE ENTERPRISE IN RICEBREEDING

Most of the rice breeding in the United States has been done at public institutions although in the early years the development of cultivars was largely by private breeders. These private breeder efforts consisted primarily in selecting true breeding lines from mixed populations, increasing the seed, and distributing it to rice growers. Using this method, S. L. Wright, an independent rice breeder in Louisiana (Jones, 1936), developed several cultivars that were widely grown from about 1912 to 1935 in the Southern States. Public rice-breeding programs were started in most rice-producing states early in this century. These programs were cooperative projects conducted by the USDA and State Agricultural Experiment Stations. Cultivars developed in these cooperative projects gradually replaced the older ones developed in private rice-breeding programs. The acreage of rice in the United States is small compared to that of other cereals so the demand for seed is limited. Because of this, and also because the public ricebreeding programs were fairly successful, there was little effort by private breeders to conduct rice-breeding programs. In recent years rice acreage has expanded, thus there is an increased demand for rice seed. The Plant Variety Protection Act (PVPA) enacted in 1970 enabled plant breeders to retain control of the varieties they developed (Rollin, 1972). These two items gave added incentive for private breeders to initiate rice-breeding programs. There has not been a big rush of private breeders into the breeding of rices, but a few companies have entered this field, and their activity increased noticeably in 1980 and 1981. The most active nongovernmental rice-breeding project is conducted by the California Cooperative Rice Research Foundation (CCRRF). This agency is supported by marketing-order funds raised by self-imposed rice-grower assessments. No other public funds (since the marketing-order funds are collected by the California Department of Food and Agriculture, they are considered “public funds”) are directly involved in this rice-improvement project but several University of California and USDA/ARS personnel are involved in cooperative projects which provide assistance to the breeding and testing efforts of CCRRF and others. Recent varietal releases have continued to be joint, cooperative releases by CCRRF, the University of California, Davis, and USDA/ARS, with CCRRF controlling the production and sale of foundation seed. Some privately developed rice cultivars have been marketed in recent years in

78

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

California, Louisiana, and to a lesser extent in Mississippi and Arkansas. For the most part, however, such cultivars sold in the Southern United States are still lacking in resistance to diseases, especially blast, and to date occupy a relatively small percentage of the total rice acreage. The present system of conducting rice-breeding research with public agencies in charge of the introduction, maintenance, and evaluation of genetic stocks is working quite well. This system entails publication from time to time of reports on the material available and the characteristics of these materials. The personnel involved in these projects also continue their own genetic and breeding projects. This arrangement makes all the basic breeding material available to all breeders, but private breeders are able to control distribution of the cultivars and breeding material they develop under provisions of the PVPA. The recent appearance of a booklet (Mooney, 1980) has led to some rather detailed discussions on the future domain of plant breeding. The principal emphasis of the booklet is that plant breeding of cash crops in the advanced countries has been largely brought under the plant breeder’s rights and more recently controlled by multinational companies dealing with agricultural chemicals. The rapid acquisition of major seed companies by such industrial giants has alarmed agricultural scientists as well as many people in the public sectors. The situation in rice is not so alarming as it is in hybrid maize or in the highpriced vegetable crops. In all rice-growing countries the breeding phase has been traditionally the responsibility of the government sector. This is even true in the major rice-growing states of the United States. Commercial rice-breeding ventures currently are found only in the United States and Latin American countries, and their percentage of the seed market to date, is insignificant. Moreover, the international and national centers are continuing the free exchange of elite germ plasm and unrestricted naming of varieties selected from such materials (IRRI, 1980b). Since rice is largely a self-pollinated crop, the prospects of profitable private enterprise in breeding and seed production do not appear overwhelmingly attractive. Even with the recent development of hybrid rice in China, the margin of profit rests with the price differential between ordinary seed and F, hybrids as well as the increase in yield of the hybrids over the pureline varieties. For rice varieties of nonaromatic grain quality, the price differential likely will remain small. Moreover, the diverse rice-growing environments would require a wide array of improved genotypes to meet the needs of specific environments. China was able to expand hybrid rice production because it is a government-supported campaign. Therefore, the threat of private companies to monopolize the breeding programs does not appear eminent in rice. However, two large seed companies operating in the United States and elsewhere reportedly have obtained “exclusive rights” to the component lines of the cytosterile sources used to constitute the major rice hybrids currently being grown in China. The cooking and process-

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

79

ing characteristics of these current hybrids are not satisfactory for the markets where United States rice is now marketed. Meanwhile, research on hybrid rice also is being conducted at public institutions of several countries and at the international centers. In addition, rice breeders of public institutions will likely continue their practice of sharing both the unimproved and improved germ plasm without the restrictions imposed by the plant breeder’s rights act. F.

PROBLEMSIN

USINGDIVERSE GERMPLASM

It is apparent from the preceding sections that rice researchers and particularly the IRRI staff, have been using diverse donor parents in the hybridization programs. On the other hand, wide crosses among cultivars within one species or between two species have hampered progress in breeding because of incompatibility problems. The best known example of intervarietal sterility within 0. sativa has been documented in the crosses between the tropically based Indica race and the temperate zone Japonica (or more appropriately Sinica) race. Early findings on the sterility and incompatibility aspects have been summarized by Chang (1964) and Oka (1964). Extensive cytogenetic investigations on the partially sterile F, hybrids were summarized Chang (1964). The partial sterility, weakness of F, seedlings, breakdowns of F, seedlings, disturbed segregation ratios in the F,, and aberrant recombinations have hampered the recovery of useful progenies in the Indica X Japonica crosses. The problems associated with the interracial crosses may partly explain why only a few promising varieties were obtained from the FAO-sponsored Indica X Japonica Hybridization Project which involved several Asian countries during the 1950s (Parthasarathy, 1972). Similarly crosses between tropical Indicas and the Javanica varieties of Indonesia have shown partial sterility and chromosomal aberrations (Engle et al., 1969). The incompatibility phenomenon in Asian rice is not confined to interracial crosses. It has been demonstrated that similar difficulties showed up when crosses were made between tall tropical Indica varieties and the Chinese semidwarfs and between traditional dryland varieties and the semidwarfs (Engle et al., 1969; Hung and Chang, 1976; Lin and Chang, 1981). Cytological and chromosomal abberrations of the inter- and intraracial hybrids are essentially similar (Demeterio et al., 1965; Engle et al., 1969, 1970). The well-publicized findings on inter- and intraracial incompatibility have discouraged some rice breeders from attempting wide crosses, but the difficulties were not so unsurmountable as to exclude wide crosses from being used for specific objectives. Breeders in Korea, Taiwan, the United States, and at IRRI have obtained useful progenies from wide crosses, but only after long cycles of crossing and selection (Johnston ef al., 1972; ORD, [undated]; Chang, 1976f; Chang et al., 1982).

80

T. T. CHANG, C. R. ADAIR, A N D T . H. JOHNSTON

Research in China has shown that anther culture has produced higher proportions of plantlets in Indica X Japonica hybrids than in Indica X Indica hybrids (Zhang, 1981). Tissue culture may offer an alternative approach to make more efficient use of the progenies of wide crosses. Crosses between the Asian cultigen 0. sutivu and the African cultigen (0. gluberrima) have not been rewarding because of sterility and related problems. The African rices possess certain biotic resistances and ecoedaphic tolerances, but the levels were generally not superior to the Asian cultivars (Chang et al., 1977). A limited amount of crossing, selection, and evaluation for insect resistance has been conducted at IRRI since 1976. Crosses between the cultivated rices and their wild relatives generally present serious difficulties in recovering desirable progenies, especially when the genomes of the parents differ (cf. Chang, 1964; Chang et al., 1977). But one strain of 0. nivuru (IRRI Acc. 101508),an annual Asian weed race, has been successfully used in incorporating resistance to grassy stunt virus into the semidwarfs (Khush, 1977). This strain, which was the only one of 40 strains in the same species that was resistant to the grassy stunt virus, appears to be an exceptional sample in the weed race. Its hybrids with the Asian cultivars showed less chromosomal aberrations than several crosses among Asian cultivars (Dolores et af., 1979).

VII. ENDEAVORS FOR THE FUTURE A. COMPLETION OF FIELDCOLLECTION FOR THE Two CULTIGENS

The collaborative efforts of many research centers in canvassing and assembling the uncollected rice germ plasm during the last decade is a unique venture among the major crops. The total number of seed samples collected was more than 32,000, but the coverage is still incomplete insofar as the widely distributed rice cultigens are concerned. Moreover, a number of the collected samples were either obvious duplicates or nonviable. The 5-year collection program developed at IRRI in late 1977 has served as a useful guide in implementing systematic collection in Asian countries. Although more than one-half of the accessible areas of the target countries have been covered since the 1977 workshop, vast areas in Laos, Cambodia, and Nepal remain unexplored. Field collection in Southwest China has been initiated but the area has not been completely canvassed. The northeastern states of India also remained largely unexplored. There are many pockets in West Africa which could yield rice germ plasm, if canvassed. The assemblage of the wild relatives of the two cultigens remain as the most obvious gap in the conservation efforts.

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

81

It will require periodic workshops and vigilant communication to maintain the tempo of the collection activities. Fortunately, the IBPGR is channeling larger allocations of its resources into this phase of conservation. B . CONSOLIDATION OF EXISTING MAJORCOLLECTIONS

Conferences held at IRRI and elsewhere have revealed substantial duplication of conserved seedstocks at different centers. However, some of the superficial duplications may be in varietal names only because the seedstocks may represent different eco-strains of the original variety. Morphologic variants and mutants of the same varieties are also found in the duplicated samples. In addition, in at least a few cases, mislabeling apparently occurred at harvest. In the interest of efficient management, it is essential to minimize the duplications without overlooking the ecostrains, morphologic variants, and mutants. Steps have been taken by the staff of the USDA, the NIAS of Japan, IITA, and IRRI in comparing accession lists as preparations for consolidation. This phase is a time-consuming and laborious operation because seed and plant data need to be included in the process of comparison and consolidation. On the other hand, the process would lead to a more efficient and secure preservation of all conserved stocks at two or more long-term storage sites. Working collections could be maintained in the region of collection as desired. C. CONSERVATION OF WILDSPECIES

The total number of wild forms being conserved by different centers is probably not more than 3000. The wild strains were collected either by botanists or explorers on specific missions or by breeding-oriented collectors while canvassing for the cultivars. More systematic efforts are needed to survey and plan for the conservation of the wild forms which are also threatened by developments in progressive agriculture and public infrastructures. The IBPGR-IRRI Rice Advisory Committee studied the complex problem of collecting and conserving the wild species in late 1981. The conservation of the wild forms would require a synthesis of contributions from botanists, geneticists, conservationists, and government administrators (IBPGR-IRRI, 1982). D. THECOMPLETION OF EVALUATING THE COLLECTED MATERIALS

The massive scale of varietal evaluation by rice workers in different centers on a worldwide basis is unprecedented in crop history. However, the sequence and

82

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

degree of completeness in evaluating the existing collections have not been as systematic and comprehensive as desired. Periodic reviews and discussions are needed to upgrade the coverage and usefulness of the evaluation activities. The material in the collections should be evaluated as rapidly as funds and availability of personnel and equipment permit. The evaluation could be done at several locations. That is, reaction to disease and insects could be studied at locations where the disease or pest occurs naturally. Cooking and nutritional studies must be conducted at institutions that have the necessary equipment and personnel. Because of the enormous diversity in the genetic makeup of the cultigens and of the environments under which they are grown, the interpretation of evaluation results obtained at one location for use at other locations would require in-depth study and analysis by teams of scientists of the several appropriate disciplines. E. CHARACTERIZATION OF ENVIRONMENTS

The diverse environments under which rice is grown requires the coordinated efforts of rice researchers to concurrently describe and characterize the important ecological components so that a better understanding of the ecosystem involved would lead to an efficient use of the experimental data. The international agricultural research centers have been conducting dialogues which will eventually lead to unified and comparable systems of characterizing the rice-growing environments. F. INNOVATIVEBREEDING TECHNIQUES

The great majority of elite rice germ plasm bred at different centers has been developed through the conventional process of hybridization and selection. Induction of mutations as a breeding tool has met with varying degrees of success at different rice research centers. Rutger et al. (1976) reviewed the advances in rice breeding through the use of spontaneous and induced mutations. In the cooperative USDNARS-California Agricultural Experiment Station-Rice Industry program, induced mutations of useful genes have been obtained for shortstature, early maturity, and glutinous character in adapted, cold-tolerant California germ plasm (Rutger et al., 1976, 1977, 1979a,b). Rutger and his co-workers have had considerable success with induced mutation breeding in their unique situation in California where they have rather minor insect and disease problems compared to other areas of the United States and most rice-growing countries. However, they emphasize that this is only one of several breeding methods that should be used. They believe that induced mutation breeding, both on whole plants and in cell culture studies, offers the potential to eventually “generate new genes upon demand.”

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

83

It has been suggested (J. N. Rutger, personal communication) that for better utilization of existing rice germ plasm, serious consideration should be given to establishing an “instant germ plasm reservoir package” which may be quickly put into use in areas with serious new disease or insect problems. Such a package could be used in situations such as the ragged stunt virus outbreak in Indonesia, the Philippines, and elsewhere. If prepackaged germ plasm reservoirs or pools were available for instant distribution, resistance to new pests might be more quickly identified. Perhaps the simplest such package might contain a few seeds from each of the accessions in the IRRI World Collection, perhaps broken down into at least three major categories. More sophisticated pools might be formed by using genetic male steriles to create composite cross populations. Efforts have been made to develop somewhat simpler gene pools but involving hundreds of crosses with a wide range of types of varieties from throughout the world (personal communications from W. R. Coffman and W. 0. McIlrath). The establishment of such gene-pool composites represents a very major undertaking and the involvement of many personnel, extensive funding and coordination, and prolonged dedicated efforts. Tissue culture has been used to a limited extent in rice breeding, often under restricted objectives. Croughan er al. (1980) applied cell culture techniques to selection of salt-tolerant rice in California experiments. Starting with naturally occurring haploid plants, they obtained four cell lines which exhibited salt tolerance. Their research is continuing. Chalef er al. (1978) working at Cornell University and Schaeffer et al. (1978) working at the USDA/ARS, Beltsville Agricultural Research Center have been using tissue culture techniques in an effort to develop rice plants with increased lysine content in the grain. Some progress is reported and the research is continuing. In a recent report, the National Plant Genetic Resources Board (1979) cites a “white paper” entitled “Research Priorities in Plant Breeding” by Sprague, Alexander, and Dudley which cautions against the high budgetary rating currently given research in cell culture and other forms of “genetic engineering. ’’ These well-respected researchers of long standing agree that the field holds some promise and unquestionably deserves support. However, they seriously question the high priority rating of genetic engineering if the research is to be achieved through neglect of those disciplines that continue to improve plant performance and offer promise for the future. They further state, “Even if new products are developed by genetic engineering techniques, the use of them will be accomplished through conventional crop improvement programs. Continued progress in improving the performance of crops cannot reasonably be expected unless this relation is understood and implemented.” The possibility of employing useful genes present in the wild species of Otyza and those of related genera have not yet been attempted extensively other than the efforts of IRRI staff. Figoni et al. (1981) reported on the transfer of stem rot

84

T. T. CHANG, C. R . ADAIR, AND T. H. JOHNSTON

resistance from a wild species (Oryza rufpogon) to 0. sativa. Advanced generation lines are undergoing further testing for resistance to stem rot. Some of the lines also were tested in 1981 in Arkansas for resistance to sheath blight (F. N. Lee and K. S. McKenzie, personal communication). Similarly, screening of wild species for resistance to sheath blight is underway in California (J. N. Rutger, 1981, personal communication). incompatibility and a lack of controlled experimentation have impaired the reproducibility of the wide crosses. A combination of different breeding techniques such as mutation and tissue culture would lead to more efficient use of the specific gene pools to be transferred through the wide crosses. G. RESTORATION OF GENETIC DIVERSITY TO THE IMPROVED CULTIVARS

It is inevitable that as breeding programs advance, there is a concurrent increase in the genetic uniformity of the elite germ plasm. The narrowing of the genetic diversity in the major commercial varieties would increase the genetic vulnerability of the relatively few cultivars to serious epidemics of pests and diseases. Recent events such as the shifts in the brown planthopper biotypes, the sudden emergence of hitherto obscure virus diseases, and the flareup of rice blast in South Korea during 1980 are associated with the rapid spread and large-scale adoption of a few genetically similar cultivars over large areas, especially under continuous cropping in the humid tropics. It is imperative that the breeders as well as public administrators recognize the need to reinstate the rich genetic diversity which was present in rice until a few decades ago. The cooperative breeding programs in the United States in the past 10-15 years have included the use of a much wider range of parental types than in earlier years in order to broaden the genetic base for resistance to blast and other diseases and for other needed characteristics. Concerted efforts are being made in the United States, IRRI, and other rice research centers for different sources of short stature and high- and stable-yielding ability. Seed multiplication and distribution programs may be complicated by the restoration of genetic diversity to the major production areas, but the costs involved may be a small price to compensate for the much larger potential losses due to serious pest damages. Moreover, genetic diversity should be considered both on a genic and cytoplasmic basis. H. TRAINING OF RICERESEARCHERS

The number of rice researchers in different national programs has been greatly expanded during recent years. But the total number and their distribution across disciplines are insufficient for the needs of the near furture. Researchers in the

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

85

areas of pathology, entomology, and physiology are more deficient in number than those in breeding and agronomy. Among many rice breeders trained in advanced countries, the graduate training was largely based on dryland crops. The semiaquatic rice requires an in-depth understanding of those peculiarities associated with the crop. On the other hand, some breeders working on dryland rice were trained under programs designed for irrigated rice. A frequent criticism of existing advanced country training programs for developing country trainees is that the training is highly basic and has little application in the student’s home country. Nevertheless, it is crucial that students become cognizant of the latest techniques in order better to understand how these might be utilized in local situations. Workers trained in genetic conservation are few in number and low on experience. Training in each of the above areas, emphasizing practical experience in handling the rice plant and its associated biotic and physical factors, is essential to developing real expertise in different aspects of rice research. Training will continue to be an equal partner to research on rice. One important aspect in the training of rice researchers is to emphasize the need to look at the rice plant or crop as a whole. Workers in all disciplines need to be aware of the many interrelated traits such as the effect of N-fertilization on diseases and quality of the grain. Also, N-fertilization and cultural management can have profound effects on lodging of the crop, cost of harvesting, and ultimate food value or cash crop value obtained. REFERENCES Adair, C. R. 1941. USDA Tech. Bull. 772. Adair, C. R. 1968. Crop Sci. 8, 264-265. Adair, C. R.. and Cralley, E. M . 1952. Arkansas Agric. Exp. Stn. Bull. 525. Adair, C. R., Bollich, C. N., Jodon, N. E., Johnston, T. H., Scott, J . E., Webb, 8. D., and Wiser, W. J. 1971. USDA Agric. Res. Serv. PSR-14-71, pp. 1-138. Adair, C. R., Bollich, C. N., Bowman, D. H., Jodon, N. E., Johnston, T. H., Webb, B. D., and Atkins, J . G. 1973. USDA Agric. Res. Serv. Agric. Handb. 289, 41-45. Ahamad, M. G., Hoff, B. J., McIlrath, W. O., and Rush, M. C. 1974. Proc. Rice Tech. Work. Group. 15th, pp. 49-50. AICRIP (All-India Coordinated Rice Improvement Project). [undated]. “India’s Rice Revolution. Hyderabad. Anon. 1980. Annual Report to California Rice Growers, 1 1th. Rice Research Board, Yuba City, California Atkins, I. G . 1965. Proc. Rice Tech. Work. Group, IOth, 1964 p. 66. Atkins, J . G. 1974. USDA Agric. Res. Serv. Agric. Handb. 448, 1-106. Atkins, J . G . , and Adair, C. R. 1957. Plant Dis Rep. 41(11), 911-915. Atkins, J . G . , and Todd, E. H. 1959. Phyropathology 49, 189-191. Atkins, J . G., Beachell, H. M. andcrane, L. E. 1956a. Tex. Agric. Exp. Stn. Prog. Rep. 1865, 1-2. Atkins, J . G . , Beachell, H. M . , and Crane, L. E. 1956b. Rice J . 59(6), 36, 38. Atkins, J . G . , Beachell, H. M . , and Crane, L. E. 1957. Int. Rice Comm. Newslett. 6(2), 1-4.

86

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

Atkins, J. G., Robert, A. L., Adair, C. R., Goto, K., Kozaka, T., Yanagida, R., Yamada, M., and Matsumoto, S. 1967. Phytopathology 57, 297-301. Beachell, H. M., Khush, G. S . , and Aquino, R. C. 1972. In “International Rice Research Institute, Rice Breeding,” pp. 89-105. Los Banos, Philippines. Bezancon, G . , Bozza, J., Koffi, G., and Second, G. 1977. Meet. Afr. Rice Species, IRAT-ORSTOM, Paris. pp. 15-64. Bollich, C. N. 1979. Tex. Agric. Exp. Stn. Leaflet L-1719. Bowling, C. C. 1963. J . Econ. Entomol. 56(6), 893-94. Bowling, C. C. 1973. J . Econ. Entomol. 66, 572-573. Bowling, C. C. 1980a. In “Rice: Production and Utilization” (B.S. Luh, ed.), pp. 260-288. Avi Publ., Westport, Connecticut. Bowling, C. C. 1980b. Proc. Rice Tech. Work. Group, 18th p. 48. Brady, N. C. 1975. In “Crop Productivity-Research Imperatives” (A.W.A. Brown, T. C. Byerly. M. Gibbs, and A. San Pietro, eds.), pp. 62-96. Kettering Foundation, Cleveland, Ohio. Chaleff, R., Thompson, J . , and Parsons, M. 1978. Proc. Rice Tech. Work. Group, 17th. p. 17. Chambliss, C. E., and Jenkins, J. M. 1923. USDA Dep. Bull. 1127. Chandler, R. F., Jr. 1968. USDA Yearb. Agric. pp. 252-255. Chandraratna, M. F. 1955, J . Genet. 53, 215-223. Chang, T. T. 1964. Tech. Bull. No. 1. International Rice Research Institute, Los Banos, Philippines. Chang, T. T. 1967. Int. Rice Comm. Newslett. 16(4), 4-12. Chang, T. T. 1972a. Plant Genet. Resources Newslett. 27, 9-14. Chang, T. T. 1972b. In “International Rice Research Institute Rice Breeding,” pp. 177-185. Los Banos, Philippines. Chang, T. T. 1975. In “Crop Genetic Resources for Today and Tomorrow” (0.H. Frankel and J. G. Hawkes, eds.), pp. 159-173. Cambridge Univ. Press, London and New York. Chang, T. T. 1976a. In “Evolution of Crop Plants” (N. W. Simmonds, ed.), pp. 98-104. Longman, London. Chang, T. T. 1976b. Philos. Trans. R. SOC. London Ser. B275, 143-157. Chang, T. T. 1976c. Euphytica 25, 425-441. Chang, T. T. 1976d. Int. Rice Res. Newslett. 2/76, 4. Chang, T . T. 1976e. “Manual on the Genetic Conservation of Rice Germplasm for Evaluation and Utilization.” International Rice Research Institute, Los Banos, Philippines. Chang, T. T. 1976f. SABRA0 J . 8(1), 11-16. Chang, T. T. 1979. In “Replies from Biological Research.” R. de Investigaciones Cientificas. Madrid. Chang, T. T. 1980. In “Rice Improvement in China and Other Asian Countries,” pp. 85-105. International Rice Research Institute, Los Banos, Philippines. Chang, T. T., and Li, C. C. 1980. I n “Rice: Production and Utilization” (B. S. Luh, ed.), pp. 87-146. Avi, Westport, Connecticut. Chang, T. T., and Oka, H. I. 1976. In “Climate and Rice,” pp. 87-114. IRRI, Los Banos. Philippines. Chang, T. T., and Perez, A. T. 1975. In “South East Asian Plant Genetic Resources” (J. T. Williams, C. H. Lamoureau, and N. Wulijarni-Soetjipto eds.), pp. 120- 128. IBPGRIBIOTROP/LIPPI, Bajor, Indonesia. Chang, T. T . , Sharma, S. D., Adair, C. R. and Perez, A. T. 1972. “Manual for Field Collectors of Rice.” IRRI, Los Banos, Philippines. Chang, T. T., Loresto, G. C., and Tagumpay, 0. 1974. SABRA0 J . 6(1), 9-16. Chang, T. T., Ou, S. H., Pathak, M. D., Ling, K. C., and Kauffman, H. E. 1975a. I n “Crop Genetic Resources for Today and Tomorrow” (0. H. Frankel and J. G. Hawkes, eds.), pp. 183-200. Cambridge Univ. Press, London and New York. Chang, T. T., Villareal, R. L., Loresto, G . C., and Perez, A. T. 1975b. I n “CropGenetic Resources

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

87

for Today and Tomorrow” (0.H. Frankel and J. G. Hawkes, eds.), pp. 457-466. Cambridge Univ. Press, London and New York. Chang, T. T . , Marciano, A. P., and Loresto, G . C. 1977. Meet. Afr. Rice Species IRAT-ORSTOM, Paris pp. 67-76. Chang, T. T., Brown, W. L., Boonman, J . G . , Sneep, J . , and Lamberts, H. 1979. In “Plant Breeding Perspectives” (J. Sneep, A. J . T. Hendriksen, and 0. Holbek, eds.), pp. 83-103. PUDOC, Wageningen. Chang, T. T., Loresto, G . C., O’Toole, J. C., and Armenta-Soto, J . L. 1982. “Principles and Methods of Crop Improvement for Drought Resistance: With Emphasis on Rice.” IRRI, Los Banos, Philippines (in press). Chaudhary, R. C., Saran, S . , and Prasad, S. C. 1979. Int. Rice Res. Newslett. 4(5), 4. Cogburn, R. R. 1974. Environ. Entorno/. 3(4), 581-685. Cogburn, R. R. 1977a. J . Stored Prod. Res. 13(1), 29-34. Cogburn. R. R. 1977b. J . Econ. Entomol. 70(6), 753-754. Cogburn, R. R. 1980. In “Rice: Production and Utilization” (B. S. Luh, ed.), pp. 289-310. Avi, Westport, Connecticut. Cogburn, R. R., and Bollich, C. N . 1980. Proc. Rice Tech. Work. Group, 18th p. 29. Cogburn, R. R., Bollich, C. N., Johnston, T. H . , Stansel, J. W., and Mcllrath, W. 0. 1980. Environ. Entomol. 9(5), 689-693. Comptroller General. U.S. General Accounting Office. 1981. CED-81-75. Cooper, F. E. 1976. Proc. Rice Tech. Work. Group, 16th p. 26. Cralley E. M. 1936. Arkansas Exp. Stn. Bull. 329. Cralley, E. M. 1949. Phytopathology 39, 5. Croughan, T. P., Stavarek, S . J., and Rains, D. W. 1980. Proc. Rice Tech. Work. Group. 18th p. 22. CRRI (Central Rice Research Institute). 1980. “Rice Research in India-An Overview.” Central Rice Research Institute, Cuttack, India. Dalrymple, D. G. 1978. USDA For. Agric. Econ. Rep. 95. De Datta, S. K., and Malabuyoc, J. 1976. In “Climate and Rice,” pp. 509-539. IRRI, Los Banos, Philippines. Demeterio, E. G. V., Ando, S . , Ramirez, D. A.. and Chang, T. T. 1965. Philipp. Agric. 49(4), 248-259. Directorate of Extension [undated]. Improved Rice Varieties for Different Regions of India. Ministry of Agriculture & Irrigation, New Delhi. Dolores, R. C., Chang, T. T . , and Ramirez, D. A. 1979. Cytologia 44,527-540. Engle, L. M., Chang, T. T., and Ramirez, D. A. 1969. Philipp. Agric. 53 (5&6), 289-307. Engle, L. M., Ramirez, D. A., and Chang, T. T. 1970. Cytologia 34(4), 572-585. Ferreira, S. A , , and Webster, R. K., 1975. Phytopatho/ogy 65, 968-971, Figoni, R . A. 1981. Ph.D. thesis abstract, Univ. of California, Davis. Figoni, R. A., Rutger, J. N., and Webster, R. K . 1981. Agron. Abstt. p. 61. Freeman, W. H., and Shastry, S. V. S . 1972. In “International Rice Research Institute, Rice Breeding,” pp. 115-132. Los Banos, Philippines. Gifford, J. R., and Trahan, 0. B. 1976. Proc. Rice Tech. Work. Group, 16th p. 63. Gifford, J. R., Oliver, B. F., and Trahan, G. B. 1974. Proc. Rice Tech. Work. Group, 15th p. 40. Gomez, K. A., Chang, T. T., Francisco, C. J., and Chan, M. A. 1978. Proc. Workshop Genet. Cons. Rice p. 26. Grigarick, A. A. 1974. Proc. Rice Tech. Work. Group, 15th p. 43. Grigarick, A. A.. Way, M. O., and Clement, S. L. 1976. Proc. Rice Tech. Work. Group, 16th pp. 63-64. Harahap, Z . , Hoff, B. J., McIlrath, W. O., and Rush, M. C. 1974. Proc. Rice Tech. Work. Group, 15th p. 48.

88

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

Hargrove, T. R. 1976. Int. Rice Res. Newslett. 1/76, 5-6. Hargrove, T. R. 1978. IRRI Res. Paper Ser. 18. Hargrove, T. R., Coffman, W. R., and Cabanilla, V. L. 1979. IRRI Res. Paper Ser. 23. Herrera, R. M . , and Coffman, W. R. 1974. Proc. Crop Sci. Soc. Philipp. 5 , 12-14. Ho, P. T. 1956. Econ. Hist. Rev. 9, 200-218. Hoff, B. J . , Ahmad, M. G., Rush, M. C., Schilling, P. E., and Mcllrath, W. 0. 1976. Proc. So. Assn. Agric. Sci. 3, 15. Huang, C. S. 1979. Int. Rice Res. Newslett. 4(6), 4. Hung, H. H., and Chang, T. T. 1976. SABRA0 J . 8(2), 127-134. Husain, I. 1978. Final Tech. Rep. Agric. Res. Council, Islamabad 1922-77. IBPGR (International Board for Plant Genetic Resources). 1976. “Priorities among Crops and Regions.” IBPGR, Rome. IBPGR-IRRI Rice Advisory Committee, 1980. “Descriptors for Rice” ( O r y a sutiva L , ) , IRRI, Los Banos, Philippines. IBPCR-IRRI Rice Advisory Committee, 1982. Plant Genet. Res. Newsletr. (in press). IFPRI (International Food Policy Research Institute). 1977. IFPRl Res. Rep. No. 3. Ikehashi, H., and Ponnamperuma, F. N. 1978. In “Soils and Rice,” pp. 801-823. IRRI, Loa Banos, Philippines. IRRI (International Rice Research Institute). 1972. “Rice Breeding.” IRRI, Los Banos. Philippines, IRRI. 1974. IRRI Rep. 2174. IRRI. 1976a. Res. Highlights 1975. IRRI. I976b. “Standard Evaluation System for Rice. IRRI, Los Banos, Philippines. IRRI. 1977. “Constraints toHigh Yields on Asian Rice Farms: An Interim Report.” IRRI, Los Banos, Philippines. IRRI. 1978a. “Annual Report for 1977.” IRRI, Los Banos, Philippines. IRRI. 1978b. Proc. Workshop Genet. Cons. Rice. IRRI. 1979. “Farm-Level Constraints to High Rice Yields in Asia: 1974-77.” IRRI, Los Banos, Philippines. IRRI. 1980a. “Annu;?] Report for 1979.” IRRI, Los Banos, Philippines. IRRI. 1980b. “Five Years of the IRTP: A Global Rice Exchange and Testing Network.” IRRI, Los Banos, Philippines. IRRI. 1 9 8 0 ~ .“Standard Evaluation System for Rice,” 2nd ed. IRRI, Los Banos, Philippines. IRRI. 1980d. lRRl Rep. 3/80. IRRI. 1981. “Annual Report for 1980.” IRRI, Los Banos, Philippines (in press). IRRI. 1982. “Annual Report for 1981 .” IRRI, Los Banos, Philippines (in press). Israel, P. 1967. / n “The Major Insect Pests of the Rice Plant,” pp. 391-403. IRRI, Los Banos. Philippines. Johnston, T. H. 1958, Agron. J . 50, 694-700. Johnston, T. H., and Smith, R. J., 1976. Proc. Rice Tech. Work. Group, 16th pp. 91-92. Johnston, T. H., and Templeton, G . E. 1970. Rice J . 73(7), 35-36. Johnston, T. H., Cralley, E. M. and Henry, S . E. 1959. Arkansas Farm Res. 8(2), 2. Johnston, T. H., Jodon, N. E., Bollich, C. N. and Rutger, J. N. 1972. In “International Rice Research Institute, Rice Breeding,” pp. 61-76. Johnston, T. H . , Wells, B. R., Templeton, G. E., Faw, W. F., and Henry, S. E. 1973. Crop Sci. 13, 774. Johnston, T. H., Oakes, A. J., Rutger, J. N., Lehman, W. F., Mcllrath, W. 0 . .and Bollich, C. N. 1978. Proc. Workshop Genet. Cons. Rice, 1977 pp. 18-19. Johnston, T. H., Wells, B. R., Marchetti, M. A., and Henry, S. E., 1979a. Crop Sci. 19, 743. Johnston, T. H., Wells, B. R., Marchetti, M. A,, Lee, F. N., and Henry, S. E. 1979b. Crop Sci. 19, 743-744. Johnston, T. H., Bollich, C. N., Mcllrath, W. O., Webb, B. D., Miller, T. C., Marchetti, M. A,, ”

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

89

Schilling, P. E., Lee, F. N., Rush, M. C., Hoff, B. J., Scott, J. E., Robinson, J . F., and Smith, C. M. 1980. Proc. Rice Tech. Work. Group, 18th pp. 17-18. Jones, D. B., Peterson, M. L., and Rutger, J. N. 1974. Rice J. 77(7), 41. Jones, J. W. 1936. USDA Yearb. Agric. pp. 415-454. Jones, J. W., Jenkins, J . M., Nelson, M., Carter, L. C., Adair, C. R., Wyche, R. H., Beachell, H. M . , Davis, L. L . , and King, B. M. 1941. USDA Circ. 612. Jones, J. W., Adair, C. R., Beachell, H. M., Jodon, N. E., and Williams, A. H. 1953. USDA Circ. 915. Juliano, B. 0. 1972. In “International Rice Research Institute, Rice Breeding,” pp. 389-405. Los Banos, Philippines. Juliano, B. 0. 1977. Cereal Foods World 22, 284-287. Khush, 0 . S. 1977. Adv. Agron. 29, 265-341. Khush, G . S. 1980. In “Rice Improvement in China and Other Asian Countries,” pp. 219-238. IRRI, Los Banos, Philippines. Kongserre, N. 1979. Proc. Workshop Chem. Aspects Rice Grain Quai. pp. 183-190. Kwangtung Agricultural and Forestry College. 1975. Acta Genet. Sin. 2(1), 31-35. Lamey, H. A. 1969. In “The Virus Diseases of the Rice Plant,” pp. 293-31 1. Johns Hopkins Press, Baltimore, Maryland. Latson, L. N., Gifford, J. R., and Trahan, G. B. 1976. Proc. Rice Tech. Work. Group, 16th p. 62. Lehman, W. F., Peterson, M. L., Adair, C. R., Davis, L. L., and Huabrich, R. W. 1970. Calg. Agric. 24(6), 4-5. Li, C. C. 1975. Ph.D. thesis, Univ. of California, Davis. Li, C. C., and Rutger, J . N. 1980. Crop Sci. 20, 295-298. Lin, M. H., and Chang, T. T. 1981. SABRA0 J. 13, 11-23. Lin, S. S., and Peterson, M. L., 1975. Crop Sci. 15, 657-660. Lin, S. C., and Yuan, L. P. 1980. In “Innovative Approaches to Rice Breeding-Selected Papers from the 1979 International Rice Research Conference,” pp. 35-52. IRRI, Los Banos, Philippines. Ling, K. C. 1974. Philipp. Phyfopathol. 10, 19-30. Lu, J . J . , and Chang, T. T. 1980, In “Rice: Production and Utilization” (B. S. Luh, ed.), pp. 1-74. Avi, Westport, Connecticut. Marchetti, M. A. 1975. Rice J. 77(7), 72-73. Marchetti, M. A. 1978. Phytopathol. News 12, 140-141. Masajo, T. M., Hoff, B. J., McIlrath, W. O., and Rush, M. C. 1974. Proc. Rice Tech. Work. Group, 15th p. 49. Matsuo, T. 1952. B . Natl. Inst. Agric. Sci. (Jpn.) Ser. D 3, 1-111 (in Japanese with English summary). Matsuo, T. 1957. “Rice Culture in Japan.” Ministry of Agriculture and Forestry, Japan. Melville, R. 1966. Nature (London) 211, 116-120. Mooney, P. R. 1980. Canadian Council for International Cooperation and the International Coalition for Development Action (London). Moormann, F. R., and Van Breemen, N. 1978. “Rice: Soil, Water, Land.” IRRI, Los Banos, Philippines. National Academy of Sciences. 1972. “Genetic Vulnerability of Crops,” pp. 172-189. National Plant Genetic Resources Board. USDA 1979. M16-008. Oakes, A. J . 1980. Proc. Rice Tech. Work. Group, 18th pp. 18-19. Oka, H. I. 1964. In “Rice Genetics and Cytogenetics,” pp. 158-174. Elsevier, Amsterdam. Oka, H. I. 1974. Generics 78, 475-486. Oka, H. I. 1977. Meet. Afr. Rice Species IRAT-ORSTOM, Paris pp. 57-65. Oka, H. I., and Morishima, H. 1967. In “The Ancestors of Cultivated Rice and Their Evolution,” pp. 145-154. Dep. Applied Genetics, National Institute of Genetics, Japan.

90

T. T. CHANG, C. R. ADAIR, AND T. H. JOHNSTON

Oliver, B. F., Gifford, J. R., and Trahan, G . B. 1970. Louisiana State Univ. Rice Exp. Stn. Annrr. Prog. Rep., 62nd pp. 156-163. Oliver, B. F., Gifford, J. R., and Trahan, G. B. 1971. Louisiana State Vniv. Rice Exp. Stn. Annu. Prog. Rep., 63rd pp. 218-224. ORD (Office of Rural Development) [undated]. “Success in the Green Revolution, 1964- 1977.” ORD, Suweon, Korea. O’Toole, J. C., and Chang, T. T. 1979. In “Stress Physiology in Crop Plants” (H. Mussell and R. C. Staples, eds.), pp. 373-406. Wiley, New York. Ou, S. H. 1972. “Rice Diseases.” Commonwealth Mycological Institute, Kew, Surrey, England. Pal, B. P. 1972. In “Rice, Science and Man,” pp. 89-104. IRRI, Los Banos, Philippines. Parthasarathy, N. 1972. In “International Rice Research Institute, Rice Breeding,” pp. 5-29. Los Banos, Philippines. Pathak, M. D. 1977. Ann. N . Y . Acad. Sci. 287, 287-295. Peterson, M. L., Rutger, J. N., Henderson, D. W., and Lin, S. S. 1972. Cali$ Agric. 26(4), 3-5. Peterson, M. L., Lin, S. S., Jones, D. B., and Rutger, J. N. 1974. Calij Agric. 28(7), 12-14. Peterson, M. L., Jones, D. B., and Rutger, J. N. 1979. I l Riso 27, 269-274. Ponnamperuma, F. N. 1977. IRRI Res. Paper Ser. No. 6. Ponnamperuma, F. N., and Castro, R. U. 1972. In “International Rice Research Institute Rice Breeding,” pp. 677-684. Los Banos, Philippines. Porteres, R. 1956. J . Agric. Trop. Bor. Appl. 3, 341-384, 541, 580, 627, 700, 821, 856. Prechachat, C., Setabutara, C., Kupkanchanakul, K., Amonsilpa, S., Supapoj, N., Kupkanchanakul, T., Boonwite, C., Sirikant, W., Wiengweera, A . , Kongseree, N., and Sarigabutr, A , , 1980. Int. Rice Res. Newslett. 5(1), 7-8. Ramiah, K., and Chose, R. L. M. 1951. Zndian J . Genet. 11, 7-13. Ramiah, K., and Rao, M. B. V. N. 1953. “Rice Breeding and Genetics.” The Indian Council of Agricultural Research. Rice Division. 1976. Int. Rice Res. Newslett. 1/76, 4-5. Rice Improvement Program of Sri Lanka. 1980. Int. Rice Res. Newslett. 5(6), 3-4. Robinson, J. F., Smith, C. M., and Trahan, G . B. 1978. Annu. Prog. Rep., Louisiana State Univ. Rice Exp. Sm., 70th pp. 155-167. Rockefeller Foundation. 1980. “Crop Germplasm Conservation and Use in China: With Special Reference to a National Plant Genetic Resources Center. ” Rockefeller Foundation, New York. Rollin, S. F. 1972. Proc. Rice Tech. Work. Group, 14th pp. 16-18. Ronduen, A., and Villareal, R. L. 1976. Int. Rice Res. Newslett. 2/76, 11. Roschevicz, R. J . 1931. Bull. Appl. Bot. Genet. Plant Breed. 27(4), 1-133. [English summary, pp. 119-133.1 Rush, M. C., Hoff, B. J., McIlrath, W. O., and Morgan, A. 1977. Annu. Prog. Rep. Louisiana State Univ. Rice Exp. Sin., 69th pp. 140-195. Rush, M. C., Hoff, B. J . , McIlrath, W. O., Morgan, A. 1978a. Annu. Prog. Rep. Louisiana State Univ. Rice Exp. Stn., 70th pp. 199-240. Rush, M. C., McIlrath, W. O., and Hoff, B. J. 1978b. Louisiana Agric. 21(3), 8-9. Russell, M. P., and Cogburn, R. R. 1977. J . Stored Prod. Res. 13(3), 103- 106. Rutger, 3. N., and Peterson, M. L. 1979. In “Report of a Rice Cold Tolerance Workshop,” pp. 101-104. IRRI, Los Banos, Philippines. Rutger, J. N., Peterson, M. L., Hu, C. H. and Lehman, W. F. 1976. Crop Sci. 16, 631-635. Rutger, J. N., Peterson, M. L., and Hu, C. H. 1977. Crop Sci. 17, 978. Rutger, J. N., Foster. K. W . , McKenzie, K. S., Mackill, D. J., Peterson, M. L., and Hu, C. H. 1979a. Crop Sci. 19, 299-300. Rutger, J . N., Peterson, M. L., Carnahan, H. L., and Brandon, D. M. 1979b. Crop Sci. 19, 929. Sasaki, R. 1922. J . Plant Prot. (Tokyo) 9, 631-644. Satake, T., and Toriyama, K. 1979. Int. Rice Res. Newslett. 4(2), 9-10.

THE CONSERVATION AND USE OF RICE GENETIC RESOURCES

91

Schaeffer, G. W., Sharp, F., and Oakes, A. J. 1978. Proc. Rire Tech. Work. Group, 17th pp. 17-1R. Schilling, P. E., and Parenton, V . L. 1976. Proc. Rice Tech. Work. Group, 16th p. 32. Sharma, K. D., and Kaushik, R. P. 1978. Int. Rice Res. Newslett. 3(1), 2. Shastry, S. V . S., Sharma, S . D., John, V . T., and Krishniah, K. 1971. Int. Rice Comm. Newstett. 20(3), 1-16. Shastry, S. V. S., John, V. T., and Seshu, D. V. 1972. f i r “International Rice Research Institute Rice Breeding,” pp. 239-251. Los Banos, Philippines. Shen, J. H. 1980. In “Rice Improvement in China and Other Asian Countries,” pp. 9-30. IRRI, Los Banos, Philippines. Smith, C M., Robinson, J . F., and Trahan, G. 9. 1980. Proc. Rice Tech. Work. Group, 18th p. 34. Stebbins, G. L. 1971. “Chromosomal Evolution in Higher Plants.” Arnold, London. Templeton, G . E., and Johnston, T. H. 1969. Arkunsus Farm Res. 18(1), 5 . Templeton, G. E., and Johnston, T. H. 1970a. Proc. Rice Tech. Work. Group, 13th pp. 28-29. Templeton, G. E., and Johnston, T . H. 1970b. Rice J . 73(7), 74. Ting, Y.. ed. 1961. “Chinese Culture of Lowland Rice.” Agricultural Publ. Society, Peking (Chinese). Toll, J . , Anishetty, N. M., and Ayad, G. 1980. “Directory of Germplasm Collections. 3. Cereals: Rice.” IBPGR Secretariat, Rome. Tullis, E. C., and Johnson, A . G. 1952. Mycologiu 44, 773-788. Vaughan, D. A , , and Chang, T. T. 1980. Plant Genet. Res. Newsletr. 42, 2-6. Vergara, B. S., Visperas, R. M . , Coffman, W. R., Villareal, R. L., and Bacalangco, E. 1976. SABRA0 J . 8(2), 97-104. Webb, B. D., and Bollich, C. N. 1980. Proc. Rice Tech. Work. Group, 18th pp. 61-62. Webb, 9. D., Bollich, C. N., Adair, C. R., and Johnston, T. H. 1968. Crop Sci. 8, 361-365. Wells, B R., and Gilmour, J . T. 1977. Agron, J. 69, 451-454. Whitney, G. 1974. Proc. Rice Tech. Work. Group, 15th p. 45. Yoshida, S., and Oka, 1. N. 1981. “Rice Research Strategies for the Future.” IRRI, Los Banos, Philippines (in process). Yoshida, S., Cock, J. H., and Parao, F. T. 1972. In “International Rice Research Institute Rice Breeding,” pp. 455-468. Los Banos, Philippines. Zaman, S. M. H. 1977. Proc. Workshop D e e p w a t e r Rice pp. 127-134. Zaman, S. M. H., Choudhury, M. A,, and Ahmad, M. S . 1972. In “International Rice Research Institute, Rice Breeding,” pp. 151-156. Los Banos, Philippines. Zhang, Z.-H. 1981. “Special Planning Conference on Rice Tissue Culture.” IRRI, Los Banos, Philippines (in press). Zwartz, J. A , , and Hautvast, J. G. A. J. 1979. In Plant Breeding Perspectives” (J. Sneep, A. J. T. Hendriksen, and 0. Holbek, eds.), pp. 1-46. PUDOC, Wageningen.

This Page Intentionally Left Blank

ADVANCES IN AGRONOMY, VOL 35

THE EFFECTS OF LOW TEMPERATURE ON Zea mays P. Miedema Foundation for Agricultural Plant Breeding Wageningen, The Netherlands

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............................................

11. Freezing Injury . . . . . .

..................................

.................................. B. Chlorosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...................... ... ...................... ... IV. Growth and Development at Suboptimal Temperatures. . . , . . . . . . . . . . . . . . . . . . . . ................. A. Introduction.. . . . . . . . . . . . . . B. Growth Phenomena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Water Relations. . . . . . . . . . . , . . , . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Mineral Nutrition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Dry-Matter Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F. C4 Photosynthesis G . Morphogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H. Biochemical Aspe I. Concluding Remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Breeding for Low-Temperature Adaptation . . . . . . . A . Maize Cultivation in a Cool Climate.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Breeding Objectives and Selection Criteria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Screening Techniques. . . . .............................. D. General Remarks and Con . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . VI. Summa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

...............................................

93 94 95 96 99

100 102 103

103 104 109 110 112 113 114 117 119 119 120 120 122 123 124 124

I. INTRODUCTION Maize (Zea mays) originates from subtropical regions, probably from the highlands of Mexico (Wilkes, 1979). The crop is characterized by the high optimum temperature needed for germination and growth and belongs to the socalled thermophilic plant species. Another characteristic of maize is the shortday requirement for flowering. 93

Copyrighl 0 1982 hy Academic Press. Inc. All rights of reproducrion in any form reserved. ISBN 0-12-000735-5

94

P. MIEDEMA

In pre-Columbian times the cultivation of maize was extended to Chile in the south and southern Canada in the north and to altitudes from sea level to 3600 m in the Andes (Weatherwax and Randolph, 1955; Galinat, 1979). A further extension of maize cultivation occurred from 1493 onward, and today maize is a leading crop in many temperate regions. Its cultivation in the cooler climates of higher latitudes requires adaptation to long days and low temperatures. Maize has a great genetic variability with respect to day-length response (Stevenson and Goodman, 1972). Day-neutral types and even a long-day line have been reported (Francis et al., 1969). Early flowering varieties with a short vegetation cycle can be grown in cold regions by avoiding the low temperatures of spring and autumn. Real adaptation to low temperature seems to be rather difficult in maize. This adaptation requires a number of features: resistance to frost, resistance to chilling, resistance to soil fungi during germination at low temperatures, and the ability to germinate, grow, and mature at low temperatures. With the increase of forage maize in northern regions, breeding for low-temperature adaptation has become more important. A rational approach to this breeding work requires adequate selection criteria. Therefore, we need to know first, which plant features limit maize yield in a cool climate, and second, what the genetic variation of those features is. This article describes various physiological effects of low temperature on maize, and evaluates their significance for maize growing and breeding. A division has been made into three parts: (1) effects of freezing temperatures, (2) damage by low nonfreezing temperatures, and (3) effects of suboptimal temperatures on the growth rate and the processes underlying it.

II. FREEZING INJURY Freezing injury limits maize distribution. In agricultural practice, maize can be damaged by frosts in spring and autumn. McRostie ( 1939) demonstrated that frost damage to maize kernels depends on the moisture content. Kernels with a moisture content below 15% withstood -22°C for at least 4 months but all kernels with more than 25% moisture content lost their germination ability. Similar results were obtained by Kiesselbach and Ratcliff (1920) and Rossmann (1949). Dry seeds in wet soil withstood -5 and - 10°C for 2 days, but they were killed if the freezing treatment was preceded by exposure to 20°C for 1 or 2 days (Harper, 1956). After emergence, the aerial parts are easily killed by frosts. Young seedlings, however, may recover if the shoot apical meristem, which is below soil level, is not killed. Usually in reports on freezing injury, the degree of killing is mentioned in relation to air temperature and the duration of temperature exposure. In artificial freezing trials plant temperatures are often higher than air temperatures.

EFFECTS OF LOW TEMPERATURE ON Zea mays

95

With radiation cooling in the field, however, leaf temperatures were found to be about 1°C below air temperature (Rahn and Brown, 1971). The effects of artificial freezing have been investigated in maize seedlings to discover killing temperatures and genetic variation in frost resistance. Data of Buican (1969) showed that 50% of the plants were killed by -6°C for 2 hr, -5°C for 3 hr, -4°C for 3-6 hr, -3°C for 24-36 hr, and -2°C for 48 hr. Koch and Muller (1963) found severe damage when seedlings were exposed to -4.5"C for 3 hr or -3.5"C for 9 hr. Frost sensitivity slightly decreased when seedlings had been "hardened" at 5-10°C (Buican, 1969). Genetic variation in the frost resistance of maize seedlings seems to be very small (Koch and Muller, 1963). Frost is also detrimental in later stages of plant development. With a short period of radiation frost, damage is restricted to the upper leaves; the ears are not frozen since they are protected by the husks. Rahn and Brown (1971) concluded from temperature measurements in and above a canopy of maturing maize that the leaves are killed when their temperature is between -1.5 and -3.O"C At CIMMYT research stations some genetic variation was found in the sensitivity of ripening maize to mild freezing, but with air temperatures below -2°C all accessions were equally damaged (P. G. Goertz, personal communication). Maize belongs to the frost-sensitive plants (Burke et al., 1976). In this type of plant, ice crystals cause structural distortions of cells and tissues. Cary and Mayland (1970) demonstrated in artificial freezing experiments with seedlings of sweet corn that any ice formation in the tissue is detrimental. Ice formation could be delayed until the plants were supercooled to as low as -9°C. Ice formation progressed very rapidly in plants with a high water potential and was impeded in water-stressed plants. External nucleation by snow increased freezing injury in plants with a high water potential but had no effect on plants with low water potential. Under field conditions, external nucleation of plants occurs with hoarfrost. Lindow et al. (1978) showed that certain epiphytic bacteria promote ice nucleation. The small differences between genotypes and between hardened and nonhardened plants may be associated with differential water status or external nucleation. It can be concluded that maize is sensitive to frost in all phases of its growth cycle except as dry seed. Freezing injury depends on the temperature, the duration of freezing, the water status of the plant, and the stability of supercooled water. Genetic variation seems to be very small.

Ill. DAMAGE BY LOW NONFREEZING TEMPE RATUR ES Temperatures below and around the minimum temperature for germination and growth may cause various types of physiological damage in maize. These

96

P. MIEDEMA

low-temperature effects are often referred to as chilling injury. Not all adverse effects of low temperature can be attributed to chilling. Low-temperature chlorosis, for instance, occurs at temperatures above the chilling range. A. CHILLING INJURY

Chilling injury is physiological damage caused by temperatures between 0 and about 12°C (Lyons, 1973). Many thermophilic plants are sensitive to chilling. The degree of injury depends on the temperature and on the duration of exposure. Injury is not usually visible during chilling but appears after an increase in temperature. A wilting and discoloration of the leaves are symptoms of chilling injury; with severe chilling, plants or plant parts are killed. The dysfunction of membranes at low temperatures could be the primary cause of chilling injury (Lyons, 1973; see also Section IV,H). Maize is moderately sensitive to chilling. The effects of chilling before and after emergence will be discussed separately.

I . Chilling before Emergence Imbibed seeds are killed by long-term exposure to chilling temperatures. Segeta (1964) investigated the effect of a 28-day cold treatment in six varieties. Varietal differences in survival were observed. The average mortality was 36 and 21% at 4 and 6"C, respectively; practically no damage was found at 8 and 10°C. The incubation of maize seeds at low temperature resulted in an exudation of sugars (Vedralovh and Segeta, 1970) and amino acids (Sege'fa and Vedralovh, 1970). This exudation was much greater at 6 than at lWC, and may thus be associated with the dysfunction of membranes at the lower temperature. A particular type of chilling injury during imbibition was observed by Cal and Obendorf (1972) in very dry seeds incubated at 5°C. The injury consisted of structural lesions in the radicle during initial hydration (Cohn and Obendorf, 1978). Young seedlings were found to be killed by a 6-day exposure to 1°C and a 8day exposure to 2.5"C (Wheaton, 1963). Klasovh (1980) found that chilling caused ultrastructural changes in the meristematic cells of primary roots. After 3 days of chilling, destruction of the Golgi apparatus and the inner mitochondria1 membranes occurred, the endoplasmic reticulum was reduced, and lipid bodies had accumulated. After 4 days of chilling, double cells consisting of a small cell inside a larger cell were found. Miedema et al. (1982) found a large genetic variation in chilling injury in dark-grown seedlings exposed to 2°C for 6 days. The range of injuries included a slight discoloration of the roots, the killing Qf, the primary root, and even the killing of the base of the mesocotyl. The injury increased with the developmental

EFFECTS OF LOW TEMPERATURE ON Zea mays

97

stage of the seedling, imbibed seeds being the least sensitive. Little injury was found after a 6-day exposure to 4°C. The genetic variation in sensitivity to chilling at 2°C did not correlate with emergence time or survival under cold field conditions. There was, however, a positive correlation between this chilling resistance and seedling vigor after emergence in a group of 10 single-cross hybrids but not in a group of various land races. The significance of chilling in breeding for low-temperature adaptation deserves further investigation. It can be concluded that young maize seedlings are injured or killed by temperatures below about 6°C. The severity of the damage depends on the temperature, the duration of the cold treatment, the stage of development, and the genotype. There are no data about effects of chilling during germination under field conditions. In agriculture maize is sown late in spring when soil temperatures are seldom below 6°C for long periods of time, and it seems, therefore, unlikely that survival and emergence are affected by chilling.

2 . Chilling after Emergence Creencia and Bramlage (1971) investigated the effects of chilling at 0.3"C and low light intensity in 7-day-old maize seedlings. After chilling, the seedlings were transferred to 21°C to study physiological and biochemical aftereffects. Leaf injury began to develop after 36 hr of chilling; after 72 hr of exposure the injury was irreversible. After 24 and 36 hr of chilling, leaf extension at 21°C was markedly reduced. Leaf segments of chilled plants showed increased ion leakage and increased oxygen uptake presumably by uncoupling of oxidative phosphorylation. Three to 4.5 days after transfer to 21"C, the leaf injury in the plants chilled for 36 hr disappeared and leaf extension, oxygen uptake, and ion leakage returned to control levels. This did not occur in plants chilled for 72 hr. Sellshop and Salmon (1928) found that seedlings chilled in a conditioned greenhouse at 2-4°C for 60 hr developed transverse chlorotic bands on the leaf blades 5-10 days after the chilling treatment. The bands occurred on those parts of the blades which formed the curl of the plant at the time of chilling. Miedema et al. (1 982) reported similar necrotic cross bands and other leaf damage in maize seedlings subjected to 4°C in the dark for 3 days. Most of the injury disappeared after transfer to normal temperatures. Irreversible leaf damage occurred ater exposure to 4°C for 6 days. Chlorotic cross bands were also formed after 14 days of exposure to a dayhight temperature of 10/4"C but not with 16/4"C. Tissue in the cell extension zone of the leaves was more sensitive to chilling than fullgrown or meristematic tissue. In some of the cross bands, tissue between the veins was chlorotic, whereas tissue along the veins (bundle sheath) was green. Chlorotic cross bands are found in various thermophilic gramineae. Slack et al. (1974) reported that transverse irreversibly chlorotic bands were formed after a

98

P. MIEDEMA

single cold night in Sorghum bicolor, Paspalum dilatatum, and Digitaria smutsii. Chlorophyll-deficient chloroplasts with disorganized lamellae were found in most of the mesophyll cells in the chlorotic bands; chloroplasts of the bundle sheath cells were green and had a normal structure. No structural abnormalities were found in the nuclei and mitochondria of the chlorotic mesophyll cells. It seems that leaves are more sensitive to chilling than other organs because of the chilling sensitivity of the chloroplasts. Sellshop and Salmon (1928) and Creencia and Bramlage (1971) reported that besides the leaves, the roots of maize seedlings are visibly injured by chilling treatments. Genetic variation in chilling-induced leaf injury was relatively small in maize (Sellshop and Salmon, 1928; Miedema et al., 1982). Chlorotic cross bands are often observed under field conditions after a cold spell. They may be the result of chilling during cold nights or of a combination of low temperature and high light intensities (see Section 111,B).

3. Chilling Injury at High Light Levels Taylor and Rowley (1971) reported that leaves of maize develop necrotic lesions when the plants are exposed to a temperature of 10°C and a light intensity of 170 W m-*. The photosynthetic rate at 10°C steadily decreased with the increase of exposure time to very low levels on the third day. This chilling treatment caused a permanent reduction of the photosynthetic capacity; 1.5 and 2.5 days of chilling reduced photosynthesis at 25°C by 40 and 70%, respectively. Chlorophyll levels were not affected by the chilling treatment. Taylor and Craig (1971) showed that in Sorghum, this type of injury was associated with swelling and changes of the ultrastructure of chloroplasts. Starch grains disappeared, and the membranes of the thylakoids initially closed together, but with more severe stress the thylakoids moved apart and granal stacking disappeared. Similar effects were observed in Paspalurn and soybean. A decline in the photosynthetic rate similar to that described by Taylor and Rowley (1971) was reported by Scott (1970) with maize plants exposed to 13°C and a light intensity of 350 W mP2. Blondon et al. (1980), however, showed that photosynthesis of maize seedlings at 10°C and a light intensity of 105 W m-2 decreased only 30% during a 10-day exposure time, whereas on return to 22°C the photosynthetic rate soon reached the original level. It can be concluded that the type of injury just described is provoked by temperatures of around 10°C in combination with high light levels. The injury may affect seedling growth in the field since conditions at the start of the growing season resemble the temperature and light levels used by Taylor and Rowley (1971). Sometimes silvery leaf necrosis is observed in the field (0.Dolstra, personal communication). This type of injury may be similar to the so-called sun scald,

EFFECTS OF LOW TEMPERATURE ON Zea mays

99

i.e., light gray bands that appear on the leaves of maize when cool, dewy nights are followed by clear mornings (Hoppe, 1961). B. CHLOROSIS

Seedlings become green after emergence by exposure to light. In cool and bright weather conditions, however, seedlings are often partly chlorotic. Young emerging leaves are pale green, yellow by the content of carotenoids, or even white. This chlorosis is also the result of a combination of low temperature and high light intensity but differs from the chilling injury described previously. Sachs (1864) had already studied the effect of temperature on the greening of etiolated maize seedlings after exposure to light. At 13-14°C the seedlings remained yellow, at 16- 17°C they gradually became green, whereas at 25°C they became green within 7 hr. Millerd and McWilliam (1968) estimated the chlorophyll content of seedlings grown until the appearance of the third leaf in naturally lit growth-rooms. Chlorophyll content was very low at 12°C and increased with a rise in temperature. Alberda (1 969) found that after the transfer of green seedlings from 20 to IO"C, only leaves and basal leaf parts formed at 10°C were chlorotic; the chlorophyll content in fully emerged green leaves was not lowered by exposure to 10°C. Miedema et u1. (1982) obtained yellow seedlings at a dayinight temperature of 10116°C; seedlings grown at 16110°C were green. At 10/16"C the leaves did elongate during the night, but greening was inhibited by the low day temperature. Chlorotic seedlings became completely green within 4 days at 20/10"C. McWilliam and Naylor (1967) studied the effect of light intensity on lowtemperature chlorosis in etiolated seedlings. At 16"C, chlorophyll accumulation strongly decreased when the light intensity was increased from about 12 to 175 W m-*. A similar decrease was found in carotenoid content. At 26"C, chlorophyll accumulation was hardly affected by the light intensity. Chlorotic seedlings rapidly greened when the temperature was raised to 26"C, indicating that the chlorophyll synthesizing system was not damaged. The inhibition of chlorophyll accumulation at low temperature and high light intensity was attributed to the photooxidation of the chlorophyll at a rate as fast as it was being synthesized. Once chlorophyll is complexed with chloroplast lamellae, it seems to be protected from photooxidation. Carotenoids may play a role in protecting chlorophyll from photooxidation since a carotenoidless mutant of maize was found to be extremely sensitive to photooxidation of chlorophyll (Anderson and Robertson, 1960). Chlorosis is often observed in the field just after emergence, but cool weather can also cause chlorotic growth in later stages of development. The photosynthetic capacity of chlorotic leaves is strongly reduced (Alberda, 1969; Bird et al.,

100

P. MIEDEMA

1977). Seedlings usually recover after a rise in temperature, but during the period of recovery chlorotic seedlings lagged behind nonchlorotic seedlings in leaf extension rate and dry-matter accumulation (Miedema et ul., 1982). Severe chlorosis results in irreversible leaf injury. It is rather difficult to quantify the adverse effects of chlorosis. At low temperature, plant growth is limited by leaf extension rather than by photosynthesis (see Section IV,D) so that the adverse effects of chlorosis will be small under cool conditions. The above reports showed that the growth of chlorotic seedlings is inhibited at higher temperatures. The degree and duration of this inhibition will depend on the rate and extent of recovery of the chlorotic leaves. Genetic variation in sensitivity to chlorosis is well known (e.g., Stamp, 1981). Sensitivity to chlorosis in 10 single-cross hybrids did not correlate with the sensitivity to chilling of young seedlings (Miedema et al., 1982), suggesting differential physiological backgrounds of the two phenomena. The above chlorosis also differs from chilling-induced chlorotic banding since it occurs at higher temperatures and it depends on light intensity. Under cool field conditions, however, chilling during the night and chlorotic growth during the day may result in a combination of the two types of injury. It can be concluded that chlorosis occurs at a high light intensity in combination with temperatures of 10-15°C. Those temperatures are high enough for some leaf extension. In the growing leaf parts chlorophyll is broken down by photooxidation before it is incorporated in the chloroplasts. Green leaves are resistant to chlorosis. Chlorotic plants may recover at higher temperatures but growth is somewhat inhibited in comparison with nonchlorotic plants. C. OTHERADVERSE EFFECTS

1. Soil Fungi

Cold and wet conditions after sowing may lead to poor stands due to seed rot and seedling blight. At temperatures in the range of 5-15"C, and particularly around 1O"C, imbibed and germinating seeds can be killed by soil fungi (Tatum and Zuber, 1943; Wemham, 1951; Harper, 1956). At those temperatures germination is very slow whereas several species of soil fungi are active. The sensitivity to seed rot depends primarily on seed quality; pericarp injury, frost damage, and insufficient maturity of the seeds facilitate fungal attacks (Tatum and Zuber, 1943; Rush and Neal, 1951). Seed rot is also influenced by the genotype. Seeds of the flint type are in general less sensitive than dent. Inheritance is largely maternal since properties of the pericarp play a predominant role (Pinnel, 1949; Rinke, 1953). Laboratory tests have been designed to measure the survival of seed in pathogenic soil at low temperature. In the Netherlands, the survival of seeds after exposure to 8°C for 17 days followed by 26°C for 3 days

EFFECTS OF LOW TEMPERATURE ON Zea mays

101

showed a high correlation with field emergence (Ebskamp, 1981). Such tests are called “cold tests,’’ a confusing term since seed quality and disease resistance rather than cold tolerance are measured (see also Bunting and Gunn, 1974). Seed rot has become of lesser importance since seed dressing with such fungicides as thiram and captan is common practice, and the quality of commercial seed has been improved. The protective action of the fungicide is restricted to the seed and its environment. Cold conditions after germination may result in seedling blight. Brown lesions on the roots and the basal part of the mesocotyl are symptoms of such a seedling infection (Ullstrup, 1977). Those seedlings progressively wilt and eventually die since the connection between the shoot and the seed and root system is disrupted. The symptoms of seedling blight resemble those of chilling injury. It seems worthwhile, therefore, to investigate whether physiological damage predisposes the seedling to fungal infection. 2 . Seedling Malformations before Emergence From germination to emergence the shoot grows in an upward direction by extension of the mesocotyl and the coleoptile. Normally, the coleoptile encloses and protects the leaves until emergence. At emergence, the growth of the mesocotyl and coleoptile ceases and the first leaf breaks through the coleoptile. Buckle and Grant (1974) reported that with deep sowing (11 cm) and wide diurnal temperature fluctuations the leaves break through the coleoptile below soil level. This reduced seedling emergence as shoot growth was disoriented, and abnormal seedlings developed. In a field experiment, in which seedling emergence was reduced by long-term exposure to low temperature, about 15% of the nonemerged seedlings showed this abnormal shoot morphology (Miedema et al., 1982). The seeds were planted at a depth of 5 cm in this experiment. In some cases of subsoil coleoptile breakage, seedlings emerged with a folded first leaf. Some genotypes were more sensitive to subsoil seedling malformations than others. Buckle and Grant (1974) attributed the subsoil coleoptile breakage to the inhibition of shoot growth particularly of the mesocotyl at low night temperatures. There is no evidence that pathogens are involved (Ullstrup, 1977).

3. Reduced Seedling Vigor after Cold Germination The sowing of maize before the mean soil temperature exceeds 10°C is advised against (Bunting, 1978): Miedema et al. (1982) observed that early sowing followed by cold weather not only reduced survival but also seedling vigor after emergence. In a field experiment 10 single-cross hybrids were sown on March 19, April 2, and April 16. From March 19 to April 16 soil temperature was continuously low. The average dates of emergence were May 11, 9, and 8, . 1

102

P. MIEDEMA

respectively. On June 11, the average shoot dry weight of the first sowing was 28% below the shoot dry weight of the third sowing, and that of the second sowing, 14% below. The extra 2 or 4 weeks of exposure to low temperature had an adverse aftereffect on seedling growth. Similar results were obtained when a moderate temperature stress during germination was compared with germination at 20°C. All genotypes tested showed reduced seedling vigor, but there was some variation in the degree of the effect. The physiological basis of this adverse aftereffect is not known. There is no evidence that the growth reduction is due to depletion of seed reserves. Chilling injury may be involved, but in the above experiments no correlation was found between chilling sensitivity and reduced seedling growth after cold germination. 4 . Low-Temperature-Induced Male Sterility

Heslop-Harrison (1 96 I ) reported that plants exposed to short photoperiods and cool nights (10°C) in a greenhouse showed male sterility during flowering. The degree of sterility depended on the stage of tassel development during the cold treatment. Male sterility caused by low night temperature has also been reported for rice (Board et a l . , 1980) and Sorghum (Brooking, 1976) but to my knowledge not for maize under field conditions. D. CONCLUDING REMARKS

Maize seedlings exposed to temperatures below 16°C are sensitive to various types of physiological damage. In the range from 0°C (or more precisely from the freezing temperature of the tissue) to about 6"C, plants are sensitive to chilling. The severity of the injury depends on the temperature and the duration of the exposure. Short periods of low temperature (cool nights) are in general not harmful. The sensitivity to chilling increases with the stage of seedling development. Chilling injury has been attributed to dysfunction of membranes. A special case of chilling injury, chlorotic cross bands in the leaves, is associated with structural abnormalities of the chloroplasts. High light levels increase the sensitivity to low temperature. Plants exposed to the 10 or 13°C and high light levels showed a permanent reduction of photosynthetic capacity whereas chlorophyll content was not affected. Another type of interaction between low temperature and light causes chlorosis. At temperatures of 1&15"C, which are just high enough for leaf extension, chlorophyll is broken down by photooxidation before it is incorporated into the chloroplasts. Newly formed leaves are chlorotic but they usually turn green when the temperature increases. Imbibed seeds and young seedlings exposed to low temperature are susceptible

EFFECTS OF LOW TEMPERATURE ON Zea mays

103

to some other, physiologically less defined, types of injury. First, seeds and seedlings may be killed by soil fungi; second, seedling malformations can occur as a result of the opening of coleoptiles below soil level; and third, long-term exposure to low temperatures reduces seedling vigor after emergence. Genetic variation has been reported for most types of cold injury but there were few indications of interrelationships. This suggests that different mechanisms are involved in the various types of low-temperature damage.

IV. GROWTH AND DEVELOPMENT AT SUBOPTIMAL TEMPERATURES

A. INTRODUCTION

Sections I1 and 111 dealt with damage caused by temperatures around and below the minimum temperature required for growth. This section concerns the influence of suboptimal temperatures, i.e., temperatures between minimum and optimum, on growth rate and development. A major objective is to discover the limiting processes for the growth of maize at low temperature above the injury threshold. To that purpose temperature-response curves for growth and the underlying processes will be compared. Growth is the result of cell division and cell extension which in turn depend on the supply of water, mineral nutrients, and carbohydrates. If these substances are present in sufficient quantities, growth may still be restricted by biochemical activities such as the synthesis of cell constituents. Growth is expressed as the relative growth rate or RGR, i.e., the increase of biomass in grams of dry matter per gram of dry matter per day. Temperature curves of single biochemical reactions are usually exponential between minimum and optimum since the relationship between reaction rate and temperature is determined by the Arrhenius equation (see Section IV,H and Fig. 3). The temperature coefficient or Q , , of such processes is around 2 to 3. It seems unlikely that biophysical processes restrict growth at low temperature since their Q , , is in the range 1.1-1.3. The relationship between temperature and the rate of plant physiological processes is very complex (Wassink, 1972, 1974). We usually find an optimum curve consisting of a sigmoid part between minimum and optimum and a rapid decline from optimum to maximum (Fig. 1). The rate that is measured at a given temperature is the end result of several biochemical and physical processes which are simultaneously affected by temperature but not necessarily in the same way. The rate of physiological processes not only depend on temperature but also on

104

P. MIEDEMA

other external factors like light, and on internal factors like the stage of plant development, and preceding growth conditions. The duration of exposure is very important. Particularly at low temperatures, long-term exposure may slow down the growth rate or even result in damage, whereas short-term exposure does not. It is clear, therefore, that temperature-response curves have no absolute value since numerous other factors are involved. Many data on temperature effects are obtained under controlled conditions. Those data will be reviewed in this section and as far as possible related to field conditions. B. GROWTH PHENOMENA

I . Cell Division and Cell Extension

One of the first questions to be answered is whether growth-rate reduction at low temperatures is caused by the decrease of cell division, cell extension, or both. In the primary roots of maize seedlings Erickson (1959) found that the rates of cell division and cell elongation were equally reduced when the temperature was decreased from 30 to 15°C; 30°C was the optimum for both processes. At lO"C, however, the rate of cell elongation was much more reduced than the rate of cell division and swellings appeared in the growing region of the roots. These swellings may be caused by an accumulation of nonelongated cells or by lateral expansion. The shoot growth of young maize plants is primarily leaf growth; from about the eighth-leaf stage (eighth leaf just visible), stem extension starts. Whether cell division or cell extension limits leaf growth is not known. It may well be cell extension since the extension zone is more sensitive to low-temperature injury than the dividing cells (Section 111,A). Kleinendorst (1975) obtained evidence that cell division is less affected than is cell extension by a shortage of water, effected by cooling the roots, or by a shortage of carbohydrates, induced by local cooling of the leaves just above the meristem. Direct cooling of the meristem temporarily blocked both cell division and cell extension. In the experiments of Kleinendorst (1975) plant parts were cooled to 5"C, which is below the minimum required for growth. Further research on cell division and cell extension in the shoot meristem is needed at temperatures around 10°C. 2 . Germination There are many reports in the literature on the effect of low temperature on the germination and early seedling growth of maize. Before discussing those data it

EFFECTS OF LOW TEMPERATURE ON Zea mays

105

is necessary to define our terms. “Germination” is considered to be the process that starts with the first metabolic activity during inbibition and ends with the emergence of the radicle from the seed (Heydecker, 1973). The continued development until the emergence of the shoot above the soil will be called “early seedling growth.” The first process that occurs when seeds are set to germinate is the imbibition of water. Metabolic activity starts as soon as the cells are sufficiently hydrated. Under favorable conditions, maize seeds show a rise in the respiration rate 2 hr after the beginning of imbibition (Woodstock and Grabe, 1967). Toole er al. (1956) reported that the elongation of the cells of the coleorhiza starts about 20 hr after the beginning of imbibition. The coleorhiza breaks through the pericarp and extends to about 2 mm beyond the surface. Then the radicle breaks through the coleorhiza. At that time cell division is first observed in the root tip. This may indicate that cell division is not a prerequisite for germination. The effect of temperature on the imbibition of maize seeds was studied by Blacklow (1972a). He showed that the rate of water uptake was very high during the first hours; even at low temperatures the water content of the seed increased considerably in a short time. It seems unlikely, therefore, that temperature restricts germination by its effect on imbibition. The effect of temperature on the rate of germination, i.e., the reciprocal of the time to radicle emergence, is presented in Fig. 1. The minimum temperature was just above 6°C. At 6“C, the proportion of emerged radicles did not exceed 35%, even after 76 days of exposure. This long-term exposure, however, did not kill the seeds. Between 8 and 32°C more of a linear relationship was found between temperature and germination rate. The time to 50% radicle emergence ranged from 10.6 days at 8°C to 17.5 hr at 36°C. Several reports show genetic variation in minimum germination temperature or germination rate at low temperature. Most genotypes, however, do not germinate, or germinate very slowly, at temperatures below 6-8°C (Segefa, 1964; Pollmer, 1969). It appears that the lowest germination temperature is just above the temperature range for chilling injury.

3 . Early Seedling Growth The growth from germination until emergence consists of the extension of the primary root followed by development of seminal roots, and the growth of the shoot. The latter consists of the mesocotyl and the coleoptile enclosing the first leaves. The young shoot grows by cell extension and cell division in the meristematic regions located in the upper part of the mesocotyl and the base of the coleoptile and the first leaves (Avery, 1930; Sass, 1977). The growth of the mesocotyl is inhibited when the coleoptile is exposed to light at seedling emergence (Blacklow, 1972b).

106

P. MIEDEMA

1.5 h

i x

a

D v

Q, c

2 1.0 C

?! a

c

.-C

E

0

0.5

Temperature ("C)

and the shoot elongation rate FIG. 1. The effect of temperature on the germination rate (0) before emergence (0)in the maize variety Fronica. From Miedema et al. (1982).

Early seedling growth can be expressed as the elongation rate of either the primary root or the shoot. The former is rather constant (Erickson, 1959) until it decreases as the root reaches its final length. The elongation rate of the shoot, i.e., of mesocotyl and coleoptile, increases with seedling development (Miedema et al., 1982). The effect of temperature on the elongation rate of the primary root has been described by Erickson (1959) and Blacklow (1972b). In both studies a nearly linear relationship was found between 10 and 25°C. The minimum temperature was around 9"C, the optimum around 30°C. Elongation rates ranged from 0.1 m d h r at 10°C to 3 m d h r at 30°C. Crawford and Huxter (1977) exposed seedlings reared at 20°C to temperatures from 2 to 14°C. At 2 and 6"C, the elongation of the primary root was very slow and ceased after 2 days. Varietal differences in the elongation rate of the primary root at low tempera-

EFFECTS OF LOW TEMPERATURE ON Zea mays

107

ture were reported by McAdam and Hayes (1978) and Clarkson and Gerloff (1979). The effect of temperature on shoot elongation is presented in Fig. 1. The minimum temperature was just above 8°C; at 8°C shoots emerged from the pericarp but elongation ceased after a time. The optimum temperature was 32°C. Between 8 and 32°C the temperature curve consists of two linear parts, from 8 to 18°C and from 18 to 32"C, with a slightly steeper slope at the higher temperature range. The study by Lehenbauer (1914) produced temperature curves largely similar to that of Fig. 1. The increase of shoot growth from 18°C may be explained by a proportionally higher extension rate of the mesocotyl (Miedema et al., 1982). In shoots of about 5 cm, the mesocotyl length at 12°C was 54% of the total shoot length; at 18°C is was 64%, and at 24"C, 70%. At diurnal temperature fluctuations of 18/6"C, 40% of the shoot length was mesocotyl. A similar reduction of mesocotyl growth was found by Buckle and Grant (1974) at diurnal fluctuations of 35/20, 30115, and 25/10°C. Temperature affects the rate of emergence by its effect on germination and shoot growth. An investigation of the effect of constant temperatures on imbibed seeds sown at a depth of 4 cm showed that the time from sowing to emergence was 23 days at IO"C, 8 days at 15"C, 4 days at 21°C and 2 days at 32°C (Miedema et al., 1982). Care should be taken in extrapolating such data to mean temperatures under field conditions because diurnal fluctuations retard emergence by inhibiting mesocotyl extension.

4. Leaf Extension After emergence, the successive leaves appear and grow until their final size has been reached. The extension rate of individual leaves increases until around the time that the next leaf appears. Then the extension rate of the former leaf decreases. The effect of temperature on the rate of leaf appearance was described by Tollenaar et al. (1979). The optimum temperature was 30°C and the extrapolated minimum 7°C. Figure 2 shows the effect of temperature on the elongation rate of the third leaf in seedlings reared under nonstress conditions. It appears that even at 7°C leaves are able to elongate. Prolonged exposure to temperatures of 10°C and lower steadily decreased the leaf extension rate, damaged the leaves, and killed the shoot apical meristem (Miedema et al., 1982). Leaf extension is affected more by the temperature of the shoot meristematic region than by the root or air temperature (Kleinendorst and Brouwer, 1970, 1972; Watts, 1972). The shoot meristematic region is below soil level until about the eighth-leaf stage. Under field conditions the temperature of the upper layer of

108

P. MIEDEMA

I

P D 00 c

c

-aQ h(

0

V

GO v

al c

rd (L

40

U

'2 5 h

b

0 c 0

L

20 a

1

0

5

10 Temperature

15 ( O C

20

)

FIG.2. The effect of suboptimal temperatures on gross photosynthesis (O), net photosynthesis (O), and the extension rate of the third leaf ( 0 )in the single-cross hybrid SH2. From Miedema and Sinnaeve (1980) and Miedema ef al. (1982).

the soil may be higher than that of the air (Duncan et al.: 1973), and in this way leaf extension may be promoted. Miedema et al. (1 982) found genetic variation in leaf extension rate at day/ night temperatures of 15/10, 20/15, and 25/20"C. There was, however, little interaction between genotype and temperature. Two genotypes with differential cold tolerance had nearly the same temperature curve for leaf extension between 7 and 16°C. 5 . Conclusions

The temperature curves of the various growth phenomena in maize are largely similar. Germination, initial shoot extension, root extension, and leaf extension

EFFECTS OF LOW TEMPERATURE ON Zea mays

109

all show an optimum of 30-35°C. The minimum temperature is between 6 and 8°C. At prolonged exposure to low temperature, plants are damaged. There is some evidence that cell extension is more restricted by low temperature than is cell division. C. WATERRELATIONS

In studying the effect of low temperature on plant-water relationship we have to consider (1) uptake and translocation, and (2) evapotranspiration of water. For reviews, see Brouwer (1965), Slatyer (1967), and Kramer (1969). The driving force for water movement in the plant is the gradient of water potential. The major component of this force is delivered by the transpiration of the leaves resulting in the passive uptake of water by the roots. Active uptake, i.e., osmotic absorption following the active uptake of ions, contributes to a small extent to the driving force (root pressure). The main barrier to water flow is the cytoplasm of root cells between the outside of the root and the xylem vessels. The transpiration rate is controlled by stornatal movement. The stomatal aperture of maize plants grown at 15/10"C was about one-third of that at 21/16"C day/ night temperature (Hofstra and Hesketh, 1969). Tschape (1972) exposed the aerial parts of maize plants to temperatures of 2, 5, and 20°C during the dark period. At exposure to light and a temperature of 25"C, the stomatal aperture was only slightly reduced by the cold nights. Raschke (1975) argued that the stomatal opening follows the temperature curve of CO, assimilation and that stomatal response to temperature is mediated by the internal concentration of CO,. It can be concluded that the low temperature of the leaves has no direct effect on transpiration. Many studies have been devoted to the effects of different root temperatures with the temperature of the shoot at a constant level between 20 and 30°C (for a review see Cooper, 1973). In general, a lowering of the root temperature reduces water uptake and transpiration. Low temperatures in the root zone have been assumed to reduce the uptake of water in two ways (Brouwer, 1965). First, increase in the viscosity of water decreases the water flow. Second, low temperature decreases water permeability of the roots. In maize it was found that the water flow through the epidermis and cortex of excised primary roots at 10°C was about one-fourth of that at 20°C (Ginsburg and Ginzburg, 1971). Root permeability, however, is a variable factor; it increases with increase of the water potential gradient (Brouwer, 1965) and may be affected by pretreatment of the plants. Maize plants reared at root temperature of 13°C had a higher water permeability than plants grown at 20°C root temperature (Stephens et al., 1980). Kleinendorst and Brouwer (1970, 1972) showed that the transfer of maize plants from a root temperature of 20 to 5°C resulted in a rapid decline in the water

110

P. MIEDEMA

content of the leaves. A parallel decrease in the leaf extension rate was observed, which was attributed to water deficiency. After about 4 hr, the water content and leaf extension showed a gradual recovery, which was attributed to increased osmotic absorption due to a higher sugar content of the plants. Barlow et al. (1977) cooled roots from 28 to 10°C for periods of up to 10 hr; the shoots were kept at 28°C. Leaf extension was found to be more sensitive to low-root-temperature-induced water stress than photosynthesis and transpiration. Leaf extension ceased at 13°C root temperature. In long-term experiments the effects of root cooling are less dramatic. Grobbelaar (1963) exposed maize plants to root temperatures of 5 to 40°C for 5-10 days; the shoots were kept at 20°C. A linear relationship was found between transpiration rate and root temperature in the range 5-20°C; the transpiration at 5°C was about 30% of that at 20°C root temperature. Leaf extension was strongly reduced but had not yet ceased at 5°C root temperature. In the experiments of Grobbelaar (1963), water uptake was less affected than leaf extension by a lowering of the root temperature. The above experiments (Kleinendorst and Brouwer, 1970, 1972; Barlow et al., 1977) showed that a rapid cooling of the roots to temperatures 15°C below air temperature resulted in water stress and subsequent cessation of leaf extension. In a field situation soil temperatures are seldom so far below air temperatures. Temperature fluctuations are rather slow and the lowest temperatures occur during the night when the stomata are closed. A second difference between the field and most laboratory experiments is in the root medium. In the experiments of Grobbelaar (1963) and Kleinendorst and Brouwer (1970, 1972), maize plants were grown in a nutrient solution. It seems likely that the uptake of water from a nutrient solution is less affected by low temperature than is water absorption from the soil. Long-term exposure to low root-temperature may lead to adaptations by the plants. On the other hand, root growth is minimal at temperatures below 10°C (Grobbelaar, 1963) and that may impede root functioning, particularly in soil. Taylor and Rowley (1971), however, showed that changing the overall temperature of soil-grown maize plants from 25 to 10°C did not lead to a water shortage in the leaves within 3 days. The practical significance of adverse effects of low root temperature should be investigated in field studies. With the present data, however, it seems unlikely that water stress is a major component in growth depression under low-temperature conditions in the field. D. MINERAL NUTRITION

This section aims to answer the question of whether the growth of maize at low temperature is limited by the uptake of mineral nutrients. Only data obtained with intact plants are considered. A distinction is made between the short-term and long-term effects of temperature.

EFFECTS OF LOW TEMPERATURE ON Zen mays

111

In the following short-term experiments maize plants grown in a nutrient solution were exposed to various root zone temperatures for periods up to 24 hr. Van den Honert and Hooymans (1955) determined the nitrate uptake at temperatures from 5 to 40°C. The uptake at 10°C was about 30% of that at 20°C. Shtrausberg (1958) reported that the uptake of 32P at 7°C was 32% of that at 21°C. The behavior of potassium can be studied by investigating the related element 86Rb. Grobbelaar (1963) showed that lowering the root temperature from 20 to 5°C caused a sharp and nearly linear decrease in 86Rb uptake; the 86Rb content of roots at 10°C was about 40% of that at 20°C. The above data show that ion uptake is retarded by lowering the temperature, but appreciable amounts of nitrate, phosphate, and probably potassium are absorbed at around the minimum temperature for growth. A second approach to the problem is the analysis of plants for mineral nutrients after long-term temperature treatments. In long-term experiments the uptake of mineral nutrients is directly affected by temperature and indirectly by temperature-affected changes in root mass, root morphology, and structural adaptations at cell level, if any. To my knowledge, no data have been published about the effect of whole-plant temperature on mineral composition. Several reports describe the effect of a cooling of the root zone with air temperatures at 20-30°C. Walker (1969) showed, with plants grown in a clay loam soil that the potassium content at 12°C was about half that at root-zone temperatures above 15°C. The concentration of N, P, Ca, Mg, and minor elements was little affected by the root temperature. Severe anthocyanin pigmentation of the leaves, a symptom of P deficiency, was observed with root temperatures of 12-14°C although P concentration of the shoots was relatively high. Clarkson and Gerloff (1 979) attributed such pigmentation to the high concentrations of soluble carbohydrates in seedlings with cooled roots. Nielsen et al. (1961) found the highest concentration of N, P, K, and Ca in plants grown in a well-fertilized loam soil at 5°C root temperature. Grobbelaar (1963) reported that in plants grown in nutrient solution the concentration of N, P, and K was much lower at 10 than at 20°C root temperature. Root cooling increased dry matter content and the concentrations showed little differences if expressed on a fresh weight basis. The beneficial effect of phosphate fertilization on the growth of maize seedlings in the field deserves some special attention. It is generally agreed that the drilling of soluble phosphate in a fertilizer band near the seed promotes seedling growth, particularly at low temperatures (Caldwell and Ohlrogge, 1966). Banding of the fertilizer is more effective than a broadcast application because phosphate moves very little in the soil, and it can soon revert into less soluble forms. From those data it can be concluded that the soil around the absorbing part of the root is soon depleted of phosphate. A temperature gradient in the soil suppressing root growth but allowing some growth of the shoot may result in a shortage of phosphate. If, however, the roots have reached a band of soluble phosphate,

112

P. MIEDEMA

large amounts can be taken up without root growth. I assume, therefore, that P deficiency at low soil temperature is due to poor root growth. Considering the results of both short-term and long-term experiments, we may conclude that in general low root-zone temperatures do not result in mineral deficiencies. Phosphate applied in fertilizer bands promotes seedling growth at low soil temperatures because of its improved availability to the slow-growing roots. It should be emphasized, however, that anthocyanin pigmentation is not always due to P deficiency. E. DRY-MATTER PRODUCTION

The relative rate of dry-matter accumulation or relative growth rate (RGR) is affected by several environmental factors. If the supply of water and mineral nutrients is sufficient, temperature and light intensity are the main variables that limit the RGR of maize seedlings in the field (Warren Wilson, 1967; Voldeng and Blackman, 1973). Studies in growth rooms showed that the optimum temperature for the RGR of maize seedlings is around 30"C, and that at 10-12°C the RGR is nil or very low (Warren Wilson, 1966; Rajan et al., 1973; Singh et al., 1976). Bunting (1962) investigated dry-matter accumulation at 12.5"C in seedlings of the cold-tolerant cv Prior. The total dry weight did not increase when the plants were exposed to this temperature from emergence; shoots and roots increased in dry weight, but this gain was at the expense of seed reserves. Older seedlings, first grown at 17°C for 3 weeks, showed a substantial increase in dry weight at 12.5"C. Alberda (1969) obtained similar results with seedlings of different ages exposed to 10°C. Those results, however, are not due to the differential response of younger and older seedlings to low temperature. A poor or even negative RGR is normal in young maize seedlings since dry matter gain by photosynthesis is negligible until the third-leaf stage (Cooper and McDonald, 1970). The RGR is the product of the net assimilation rate (NAR) and the leaf area ratio (LAR) (Radford, 1967). Lowering the temperature from 20 to 10°C (Rajan e t a l . , 1973) or from 24/19 to 18/13"C (dayhight) (Warren Wilson, 1966) caused a sharp decrease in RGR, a similar decrease in NAR, and a slight decrease in LAR. As NAR was much more affected than LAR, attention will first be paid to the effects of low temperature on photosynthesis. A comparison of temperature curves of the net photosynthetic rate (Chmora and Oya, 1967; Bird et al., 1977; More Herrero et al., 1980; Miedema and Sinnaeve, 1980) with those of NAR (Warren Wilson, 1966; Rajan et al., 1973) shows that a lowering of the temperature in the range of 20 to 10°C results in a much stronger decrease of NAR than of photosynthesis. Low temperatures have been shown to result in an increase of dry-matter content and especially in

EFFECTS OF LOW TEMPERATURE ON Zea mays

113

soluble carbohydrates in maize (Brouwer, 1968) and other plant species (Wardlaw, 1968). As noted previously (Section III,A,3), long-term exposure to low temperatures and high light intensity may damage the photosynthetic apparatus of maize. The above data suggest that under cool conditions above the injury threshold, dry-matter production is not limited by the rate of photosynthesis. In the above quoted reports, temperature curves for RGR were similar (Warren Wilson, 1966) or largely similar (Rajan et al., 1973) to those of the relative growth rate of leaf area. Miedema et al. (1982) found similar temperature curves for RGR and leaf extension rate in seedlings of two single-cross hybrids at 7- 16°C. Gross photosynthesis, net photosynthesis, and dark respiration of those hybrids (Miedema and Sinnaeve, 1980) were less affected by lowering of the temperature than was RGR or leaf extension (Fig. 2). Those data show that the effects of temperature on dry-matter production are associated with the growth of the leaves. Low temperature affects the translocation of assimilates as has been shown by feeding I4CO2to leaves of intact plants. Hofstra and Nelson (1969) reported that lowering the temperature from 26 to 7°C considerably decreased the rate of translocation of [ ''C]assimilate from the leaves. Kleinendorst and Brouwer (1972) showed that cooling the base of I4CO,-fed leaves to 5°C inhibited translocation of assimilate from that leaf to the rest of the plant. It is uncertain, however, whether growth limitation at temperatures in the range of 10-15°C is due to translocation (cf. Wardlaw, 1979). Several reports show genetic variation in dry-matter production of seedlings exposed to low temperature (e.g., Bretschneider-Hermann and Schuster, 1970; Stamp, 1980). Growth-room experiments with 10 single-cross hybrids showed that the variation in dry-matter accumulation at 15/10"C (dayhight) was partly associated with the leaf extension rate (Miedema et al., 1982). Duncan and Hesketh (1968) found that maize races from high altitudes grew faster in dry matter and leaf area when exposed to low temperature and high light intensities than accessions from low altitudes. Photosynthesis showed little genetic variation. From the temperature curves for RGR and the underlying processes as well as from data on genetic variation, we may conclude that dry-matter production of seedlings at suboptimal temperatures is restricted by leaf growth rather than by net photosynthesis. F. C, PHOTOSYNTHESIS

Maize belongs to the C, species, a taxonomically diverse group of higher plants that are distinguished from C, species in the initial steps of CO, fixation (Hatch, 1976; Downton, 1971). In C, species, CO, is first assimilated into C,

114

P. MIEDEMA

compounds before entering the Calvin cycle. C, species lack this pathway. The C, pathway has some physiological advantages, one of them being higher rates of net photosynthesis at high temperature and high light intensity. The effects of temperature on the photosynthesis of higher plants have been reviewed recently by Berry and Bjorkman (1980). They conclude that the main function of the C, pathway is to increase the concentration of CO, for the Calvin cycle, which is advantageous in hot, dry climates. Nearly all C, species originate from tropical and subtropical regions. Those species have high minimum and optimum temperatures for growth and are sensitive to chilling. This suggests a relationship between the C, pathway and the thermophilic character of those plants. However, many C, species (e.g., soybean, sweet potato, rice) also require high temperatures for growth and are sensitive to chilling. Some C, species are adapted to cool temperature regions. Those species showed comparatively high rates of net photosynthesis at temperatures as low as 5-1092 (Long and Woolhouse, 1978; Berry and Bjorkman, 1980; Jones et al., 1981). For these reasons C, photosynthesis and poor growth at low temperatures in maize and other thermophilic C, species are probably unrelated phenomena. G. MORPHOGENESIS

In this section the effect of temperature on various aspects of morphogenesis will be discussed. Morphogenesis is the result of an interaction between genotype and environmental factors. Photoperiod, temperature, irradiation, water supply, and mineral nutrition govern or modify the size and shape of the plants. Temperature affects the rate of development. Aitken (1980) showed that an increase in temperature from 15 to 24°C nearly doubled the rate of leaf initiation, tassel initiation, leaf appearance, flowering, and ripening. Such effects are not formative, in that the size and shape of the plants are the same, but rather reflect the influence of temperature on growth rate (Section IV,B,4). Formative or morphogenetic effects of temperature, particularly low temperature, are discussed below.

1 . Seedling Morphology As mentioned previously (Section IV,B,3), low temperature and particularly large diurnal fluctuations inhibit mesocotyl elongation more than coleoptile growth. The coleoptile/mesocotyl ratio is increased. The shoot apex remains deeper in the soil which improves the root standability of seedlings after emergence. Growth-room experiments showed that seedlings in the sixth-leaf stage had

EFFECTS OF LOW TEMPERATURE ON Zea mays

115

longer leaf blades and sheaths when grown at 20/15"C than at 15/10"C (day/ night) (Miedema et al., 1982). This effect, however, was partly attributed to the comparitively low light intensity (60 W m - 2 ) .

2. Shoot/Root Ratio and Root Morphology Van Dobben (1962) reported that the shoothoot ratio of various crops decreases with lowering the growth temperature. Shoot/root ratios of maize plants grown in greenhouses at 25 and 16°C were 5.0 and 3.0, respectively. Similar results were obtained by Rajan et al. (1971). This response may be attributed to an excess of assimilates at low temperatures which promotes root more than shoot growth (Brouwer, 1962a,b). Temperature not only affects the relative amount of root dry matter but also root morphology. The roots of maize plants grown in water culture showed a finer branching and a larger surface area per gram root dry weight at 25 than at 15°C (Brouwer et al., 1973). The root system of plants grown in sand culture was shorter and more branched at 15 than at 25°C (Beauchamp and Lathwell, 1967).

3 . Tillering The lowest axillary buds of a maize plant may develop into tillers. There is a wide genetic variation in the number and size of tillers (Duncan, 1975). Tiller formation is promoted when growing conditions for individual plants are improved. Stevenson and Goodman (1972) reported that tillering was increased by lowering the growing temperature from 26/22 to 18/14"C (dayhight) but then decreased with a further lowering of the temperature. Duncan and Hesketh (19681, however, found that temperatures between 15/10 and 36131°C had no such effect. 4 . Floral Induction, Leaf Number, and Plant Size

Plant height at maturity is largely determined by the number of internodes (or leaves); it ranges from less than 60 cm with eight leaves to 7 m with 48 leaves (Duncan, 1975). Five leaves or leaf initials have already been formed in the embryo of the mature seed (Sass, 1977). The initiation of leaves is resumed during germination and terminates when the stem apex forms the tassel (Bonnet, 1953). The large genetic diversity in final leaf number is determined by the seedling stage at which tassels are initiated. The photoperiod is the principal environmental factor that influences tassel initiation. Short photoperiods generally promote tassel initiation and reduce the final leaf number (Hesketh et al., 1969; Stevenson and Goodman, 1974). Low

116

P. MIEDEMA

temperature may also reduce the final leaf number. Duncan and Hesketh (1968) observed that a lowering of the growth temperature from 36/31 to 15/10"C (day/ night) resulted in a decrease of the average leaf number from 26 to 19 in a group of exotic races. A similar but less pronounced effect was found in hybrids adapted to higher latitudes (Hesketh et al., 1969) whereas certain short-season varieties were hardly affected (Coligado and Brown, 1975). Short-day exposure was more effective than low temperature in reducing leaf number (Stevenson and Goodman, 1974). It can be concluded that at low temperature tassels may be initiated at an earler stage of seedling development. It is not known whether this is a direct effect of temperature or merely the consequence of prolonged exposure to flower-inducing conditions at low temperatures. It is often observed that plant length at maturity decreases with earlier sowing (Becker, 1976). This effect may be attributed to short photoperiods if early sowing coincides with early emergence. It may well be, however, that low temperatures during early seedling development reduce leaf number (Beauchamp and Lathwell, 1966; Hesketh et al., 1969). Low temperature may also cause abnormal flower development. Heslop-Harrison (1961) found that maize plants grown under short photoperiods in a greenhouse formed female flowers in the tassels; this effect was stronger with a night temperature of 10 rather than of 22°C.

5 . Hormones Atkin et al. (1973) investigated endogenous hormones in maize plants grown at various root temperatures for 17 days. Hormone concentrations in the root pressure exudate of decapitated plants were assessed by bioassay methods. The amount of exudate decreased with temperature. Production on a plant basis of substances with cytokinin and gibberellin activity showed a very sharp decline with a decrease of root temperature from 28 to 8°C. The production of unidentified growth inhibitors doubled with a lowering of the root temperature from 28 to 13°C. The production of inhibitors at 8°C was lower than at 13"C, but the concentration was extremely high, which may be related to the very small amount of root exudate at that temperature. Endogenous hormones may modify morphogenesis but it is unknown whether they are involved in temperatureaffected plant responses. 6. Conclusions

Temperature has been shown to modify plant size and shape in various stages of development. The most striking effects of low temperature are the increased coleoptile/mesocotyl ratio, the decreased shoot/root ratio, and the thicker roots. In addition, low temperature may reduce seedling size, leaf number, and final

EFFECTS OF LOW TEMPERATURE ON Zea mays

117

plant height. It seems likely that the higher carbohydrate level at low temperature is responsible for some of the morphogenetic effects.

H. BIOCHEMICAL ASPECTS Growth is the end result of numerous chains of biochemical reactions. To elucidate the mechanism of growth limitation at low temperature we have to answer the following questions: (1) How does temperature affect the rate of biochemical reactions? (2) Which reactions ultimately limit growth rate at suboptimal temperatures? (3) What is the biochemical basis of genetic variation? The rate of a single biochemical reaction is a function of enzyme concentration, substrate affinity, enzyme activity, the concentration of substrate and end products, and temperature. The effect of temperature on the rate of a (bio)chemical reaction is expressed by the Arrhenius equation. The logarithm of the reaction rate is proportional to the activation energy of the reaction and the reciprocal of the absolute temperature (Fig. 3). In biochemical reactions, the Arrhenius equation is valid in a restricted temperature range between minimum and optimum. The enzyme systems of thermophilic plant species often show a higher activation energy at low temperature, usually presented by Arrhenius plots consisting of two straight lines intersecting at a “critical temperature” in the range 10-12°C (Fig. 3). Lyons (1973) and Raison (1974) hypothesized that at this temperature membrane lipids undergo a phase change from fluid to solid, resulting in a conformational change and a higher activation energy of membranebound enzymes, and subsequently, in an imbalance in the metabolism and dysfunction of the membranes. The phase transition of membrane lipids was considered the primary event of chilling injury (see also Lyons et al., 1979a,b). The physical basis of this hypothesis has been criticized by Bishop et u1. (1979) who argued that membrane lipids would not solidify above 0°C in either chillingsensitive or chilling-resistant species. Wolfe and Bagnall (1979, 1980) argued that the experimental data can also be represented by smooth, continuously declining curves; the critical temperatures at straight-line intersections were considered artifacts of the mathematical treatment and the choice of temperatures. Wolfe ( 1978) supposes that several mechanisms might explain low-temperature effects on membranes and membrane-bound enzymes. Although the mechanism is still unclear, membranes and protein-membrane interactions play a role in low-temperature response. An increase of the activation energy at low temperature has been reported for membrane-bound enzyme systems in several thermophilic plant species. In maize, for instance, this has been found in the succinate oxidation of root mitochondria (Raison et al., 1979), and in reactions catalyzed by the C,-pathway enzymes pyruvate Pi dikinase (Shirahashi et al., 1978) and phosphoenolpyruvate carboxylase (Uedan and

118

P. MIEDEMA

30

Temperature ("C] 10

20

0

3.c

-c 2E .-

E

\

E

E 2fi C 0 .-

:2.4

c

.'c

N

0

02.2

>

0

2.0

1.a 1.6

33

3. L

3.5

3.6

I/T (xi03 K)

FIG. 3. Arrhenius plot of the reaction rate (V) of pure phosphoenolpyruvate carboxylase. The activation energy above and below the critical temperature (10.8"C) has been shown. Redrawn from Uedan and Sugiyama (1976).

Sugiyama, 1976; see Fig. 3). In the latter case we are dealing with a soluble enzyme. Arrhenius plots of those reactions showed a discontinuity at 1 1- 12°C. Taylor et al. (1972, 1974) and Brooking and Taylor (1973) obtained evidence for an imbalance of the CO, assimilation by differential inactivation of C,-pathway enzymes in maize and Sorghum exposed to 10°C and high light intensity (see also Section III,A,3). Sugiyama and Boku (1976) suggested a correlation between the in vitro cold lability of pyruvate Pi dikinase and cold sensitivity in the field in a group of Japanese maize varieties. Inactivation of C,-pathway enzymes may reduce the photosynthesis and growth of certain genotypes under certain conditions. In other instances, however, leaf extension rather than photosynthesis is reduced by low temperature (Fig. 2). Considerable attention is being paid to the biochemical aspects of low-temperature stress in thermophilic plant species. The relatively high minimum growth temperatures might be explained by the increase of the activation energy of biochemical processes. However, its connection with the slow growth rate

EFFECTS OF LOW TEMPERATURE ON Zea mays

119

above the injury threshold is not yet known. Unfortunately, very little attention has been paid to the biochemical backgrounds of cell division and cell extension at low temperatures. In conclusion, many questions remain open and a biochemical explanation of the slow growth rate of maize at low temperatures cannot yet be given. I . CONCLUDING REMARKS

Temperature-response curves of the germination, root elongation, leaf extension, and dry-matter production (RGR) of maize seedlings show a great similarity; the minimum temperatures are between 6 and 10°C and the optima around 30°C. The effects of low temperature are strongly affected by the duration of exposure. Long-term exposure of seedlings to temperatures below 10°C results in a steady decrease of growth rate and finally in plant damage. The underlying processes of growth, i.e., the uptake of water and mineral nutrients, and the production of carbohydrates (photosynthesis), are usually less restricted by low temperature than growth itself. Low temperature, therefore, seems to restrict plant dry-matter production by its effect on leaf growth and possibly root growth rather than by affecting leaf and root functioning. There is some evidence that cell extension is more restricted than cell division by temperatures below 15°C. Further research is required on temperature effects at the cell and tissue level. Temperature affects plant morphogenesis in several ways. Low night temperatures reduce mesocotyl length and improve seedling standability . Low temperatures also decrease shootkoot ratios and may reduce plant length. There is considerable evidence that membranes and membrane-bound enzymes are involved in chilling injury. Little information, however, is available on the biochemical backgrounds of low-temperature effects above the injury threshold. Most physiological data reviewed in this article are obtained in growth rooms with constant temperature and light regimes. In the field, temperatures and light levels fluctuate and soil conditions are often suboptimal. Therefore, precise descriptions of plant response under field conditions are required to test the above conclusions and to investigate the physiological basis of genetic variation.

V. BREEDING FOR LOW-TEMPERATURE ADAPTATION A brief account will be given of the temperature requirements of maize cultivation. Reviews on this subject have been given by Shaw (1977) and Major

120

P. MIEDEMA

and Hamilton (1978) for the United States and Canada, and by Bunting and Gunn (1974) and Carr and Hough (1978) for Northwest Europe. With respect to breeding, emphasis will be given to selection criteria and screening techniques. For selection methods see Crosbie et al. (1980) and Dolstra and Jongmans (1981). A. MAIZECULTIVATION IN A COOLCLIMATE

The main areas of maize production have a temperate or subtropical climate with an average temperature during the warmest month ranging from 21 to 27°C and a frost-free season of 120- 180 days; practically no maize is grown where the mean midsummer temperature is less than 19°C (Shaw, 1977). Those data concern grain maize. The requirements for silage maize are less critical. In the Netherlands, for instance, silage maize is successfully grown at an average July temperature of 17°C and a growing season of 160 days. Accumulated temperature units are used to classify maize growing areas (Brown, 1975; Carr and Hough, 1978). Several methods of temperature summation have been reported (Cross and Zuber, 1972; Shaw, 1977). The simplest one accumulates average day temperatures above 10°C during the growing season. Accumulated temperature units are also used to classify varieties for earliness. It should be emphasized that this classification is not determined by a differential temperature response. In the continental climates of higher latitudes, temperature increases quickly in early summer. Maize can be grown in those areas if the summer temperature is relatively high. Adapted varieties are early flowering and maturing but not necessarily cold tolerant. In cool marine climates, however, the rise of the average temperature in spring and early summer is slow, and cold spells often impede growth even in later stages of development. Adapted varieties are early flowering and cold tolerant. In those areas, small improvements of soil temperature by the use of transparent plastic mulch can accelerate vegetative growth and increase final yield (Carr and Hough, 1978). Management practices that have improved the cultivation of maize in cool areas are seed dressing with fungicides and the application of phosphate fertilizer bands. Artificial means of increasing soil temperature have been too expensive up to now. Any further improvements of maize yield in cool conditions, therefore, will probably have to be made by breeding. B. BREEDING OBJECTIVES AND SELECTION CRITERIA

The objectives of breeding for low-temperature adaptation are (1) to increase yield and (2) to improve yield stability. Yield can be increased by lengthening

121

EFFECTS OF LOW TEMPERATURE ON Zea mays

the growing season, and particularly by advancing the formation of a closed canopy (Sibma, 1977). This might be obtained by earlier sowing or by improving the growth rate after emergence. Table I summarizes the main effects of low temperature on maize from sowing until about the eighth-leaf stage. To evaluate the selection criteria we have to consider the genetic variation and the significance for maize growing. The esti+) are based on the data of the preceding sections and on field mates (- to experiments in the Netherlands. Sowing earlier than normal, i.e., at a time when the average soil temperature is below IO'C, increases the risk of various adverse effects. Long-term exposure to low temperatures results in poor stands and reduces seedling vigor after emergence. Occasional warm weather promotes emergence, but then the seedlings may be killed by frost. For those reasons breeding efforts should be primarily directed towards a crop sown at a normal date, and most attention should be paid to improvements after emergence.

++

Table I Responses of Maize to Low Temperatures" Temperature range ("C) Damage before emergence Chilling injury to seeds Chilling injury to seedlings Seedling malformations Reduced vigor Seed rot Seedling blight Rate limitations before emergence Germination Shoot growth (rate of emergence) Root growth Damage after emergence Frost injury Chilling-induced cross bands Chilling injury at high light levels Chlorosis Rate limitations after emergence Water uptake Net photosynthesis Translocation of carbohydrates Root growth (P deficiency) Shoot growth, leaf extension 0

- , -t , +,

0-5 0-5 ca. 10 5-12 8-12 8-12

6-10 8-15 8-15

<-2
+ + , and + + + represent a scale from zero to large.

Genetic variation

+ ++ + + +

Significance for growth

? +?

*

+ +

?

+

+ +

?

f

rf-

+

+

+?

?

++

++

? ? ?

-

++ ++

? ?

+ +++

122

P. MIEDEMA

The most important selection criteria after emergence are chlorosis and growth rate. Breeding for reduced susceptibility to chlorosis appears to be rather straightforward (G. E. van Dijk, personal communication). The sources for this trait are varieties from the high altitudes of South America and European flints. The growth rate after emergence has been improved using the same sources (Dolstra and Jongmans, 1981). CIMMYT gene pools developed for high altitudes may also contain valuable material (Eagles and Hardacre, 1979; Miedema, 1979).

c. SCREENtNG

TECHN~QUES

The evaluation of the selection criteria is based on the climatic conditions of Western Europe. In other regions the situation may be different. Therefore, a complete list of screening techniques is given. 1. Chilling Injury. Chilling injury to seeds or imbibitional chilling injury; incubate dry seeds at 2°C for 6 days and then at 20°C for 5 days. Chilling injury to seedlings: incubate seeds at 20°C for 3 days, then at 2°C for 6 days, and finally at 20°C for 3 days. 2. Seedling Malformations. Seeds are sown at a depth of approximately 8 cm and exposed to cool field conditions. Diurnal fluctuations of 18/6 or 25/10"C can possibly be used. 3 . Reduced Seedling Vigor. From sowing to emergence: cold treatment, e.g., 3-4 weeks under cool field conditions or diurnal fluctuations of 12/8"C; a control seed group is exposed to nonstress conditions. From emergence, the two groups of seedlings are exposed to equal conditions. shoot size or dry weight of the groups is estimated in the sixth- to eighth-leaf stage. 4. Soil Fungi. Various "cold test" methods have been described. Seed rot is examined by incubating dry seeds in cold test soil at 8-10°C for 10-20 days; seedling blight in a similar way after pregermination at 20°C for 3-4 days. Seed should be of good quality. 5. Germination. Seeds are incubated at 8- 10°C and periodically screened for radicle emergence. Incubation at diurnal temperature fluctuations of for instance 10/5"C is probably better and more in agreement with field conditions. 6. The Rate of Emergence (initial shoot growth) is easily screened under cool field conditions. Artificial temperature treatments of 12°C or diurnal fluctuations around 14/8"C can also be used. Initial root growth can be examined under similar conditions in vermiculite supplied with nutrient solution. (Root growth is impeded in media without mineral nutrition.) 7. Frost Sensitivity is rather difficult to screen. Seedlings raised out of doors can be exposed to -4°C for 3 hr in a growth chamber. Frost injury depends on tissue temperature which lags behind air temperature. Geersing (unpublished

EFFECTS OF LOW TEMPERATURE ON Zea mays

123

data) showed that seedlings started freezing at about -2°C at air temperatures of -4 and -5°C. 8. Chlorosis is observed occasionally under field conditions. A good correlation was found between field observations and chlorosis in a greenhouse with an air temperature of 10°C. Screening can also be done in a growth chamber at 14/10"C (dayhight). 9. Early vigor, i.e., the rate of shoot growth after emergence, is usually screened under field conditions after early sowing. To avoid adverse effects before emergence, seedlings raised in soil blocks or peat pots in a greenhouse are planted in the field early in the season. Screening can be done in growth chambers at marginal temperatures (15/10 or 16/8"C) which just avoid chlorosis. Root growth after emergence can be examined by growing plants in slabs of soil between transparent plates placed at an angle. A major difficulty in screening for growth-rate characteristics is the discrimination between specific effects of low temperature and effects of plant vigor independent of temperature. High correlations have been found, for instance, between leaf extension at low and moderate temperature (Miedema et a/., 1982; see also Grosser, 1979). The ratio of growth rate at low and normal temperature may present a better estimate of low-temperature response. Such a ratio, however, may still contain a vigor component since hybrid vigor is found to be more evident at low than at high temperatures (McWilliam and Griffing, 1965). Screening of inbred lines, therefore, is very difficult and should be done in test crosses. The large environmental variation in maize is a second difficulty in screening for growth-rate characteristics. This variation may be smaller in the field than in growth rooms because of unequal light levels and plant temperatures in the latter (see also Rajan et a/., 1973). D. GENERAL REMARKSA N D CONCLUSIONS

Progress in breeding for cold tolerance is slow because (1) we are dealing with a very complex phenomenon, (2) the genetic variation of most plant responses is rather small, and (3) most cold-tolerance traits are genetically unrelated (Pollmer, 1969; McConnell and Gardner, 1979; Miedema et al., 1982). Simultaneous selection for all traits is practically impossible end ineffective. The best approach, therefore, is to select source material for the various traits separately and to recombine the selected sources. Advanced populations might be selected for the whole complex of features in order to recombine and to accumulate the desired genes. Many questions remain open (Table I) and further research is required to improve our knowledge of cold tolerance. The plant responses to low tempera-

124

P. MIEDEMA

ture are manifold, but they differ markedly in their relevance to breeding. Selection for the most important features will facilitate breeding of maize varieties for cool conditions. It seems likely that selection criteria relevant to maize also apply to the breeding of other thermophilic crop species.

VI. SUMMARY In this article the effects of low temperature on the growth and development of maize have been described with the focus on seedling growth before and after emergence. A distinction has been made between low-temperature damage and the effects of suboptimal temperatures above the injury threshold on growth rate and the processes underlying it. Mention has been made of genetic variation, and the significance of plant responses for maize growing and breeding has been evaluated. Seeds and seedlings are sensitive to various types of injury. Soil fungi cause seed rot and seedling blight. Several types of physiological damage may occur, such as freezing injury, chilling injury, and low-temperature chlorosis. Low temperatures retard germination, emergence, and vegetative growth, but also affect morphogenesis. Evidence has been presented showing that seedling growth at suboptimal temperatures is limited by leaf and root extension rather than by photosynthesis or the uptake of water and mineral nutrients, Genetic improvement of the low-temperature adaptation, however, is difficult because the variation is rather small, and most of the desired traits are genetically unrelated. ACKNOWLEDGMENTS Thanks are due to 0. Dolstra, G. E. van Dijk, and R. Brouwer for their useful comments on the manuscript.

REFERENCES Aitken, Y. 1980. Z. Ack. Pjlanz. 149, 89-106. Alberda, Th. 1969. Acru Bor. Need. 18, 39-49. Anderson, I. C., and Robertson, D. S. 1960. Plant Physiol. 35, 531-534. Atkin, R. K., Barton, G. E., and Robinson, D. K. 1973. J . Exp. Eot. 24, 475-487. Avery, G. S., Jr., 1930. Bor. Gaz. 89, 1-39. Barlow, E. W. R., Boersma, L., and Young, J. L. 1977. Agron. J . 69, 95-100. Beauchamp, E. G., and Lathwell, D. J. 1966. Can. J. Plant Sci. 46, 593-601. Beauchamp, E. G., and Lathwell, D. J. 1967. Plant Soil 26, 224-234. Becker, W. R. 1976. “Mais.” Proefstation voor de Akkerbouw, Lelystad, Netherlands

EFFECTS OF LOW TEMPERATURE ON Zea mays

125

Berry, J., Bjorkman, 0. 1980. Annu. Rev. Plant Physiol. 31, 491-543. Bird, J. F., Cornelius, M. J., and Keys, A. J. 1977. J . Exp. Bor. 28, 519-524. Bishop, D. G . , Kenrick, J. R., Bayston, J . H., Macpherson, A. S., fohns, S. R., and Willing, R. I. 1979. In “Low Temperature Stress in Crop Plants. The Role of the Membrane” (J. M. Lyons, D. Graham, and J. K. Raison, eds.), pp. 375-389. Academic Press, New York. Blacklow, W. M. 1972a. Crop Sci. 12, 643-646. Blacklow, W. M. 1972b. Crop Sci. 12, 647-650. Blondon, F., Clabault, G., and Rainguez, M. 1980. Ann. A m d . Plantes 30, 399-410. Board, J. E., Peterson, M. L., and Ng, E. 1980. Agron. J . 72, 483-487. Bonnet, 0. T. 1953. Univ. 111. Agric. Exp. Sta. Bull. 568, 5-47. Bretschneider-Herrman, B., and Schuster, W. 1970. Buyer. Landwirt. Jahrb. 47, 747-763. Brooking, I. R. 1976. Aust. J . Plant Physiol. 3, 589-596. Brooking, I. R., and Taylor, A. 0. 1973. Plant Physiol. 52, 180-182. Brouwer, R. 1962a. Neth. J . Agric. Sci. 10, 361-376. Brouwer, R. 1962b. Neth. J . Agric. Sci. 10, 399-408. Brouwer, R. 1965. Symp. Soc. Exp. Biol. 19, 131-149. Brouwer, R. 1968. Acta Bot. Neerl. 17, 78-80. Brouwer, R., Kleinendorst, A., and Locher, J. Th. 1973. In “Plant Response to Climatic Factors” (R. 0. Slatyer, ed.), pp. 169-174. Unesco, Paris. Brown, D. M. 1975. Factsheet, Agdex 111/31. Ontario Ministry of Agriculture and Food, Toronto. Buckle, J. A,, and Grant, P. M. 1974. Rhod. J. Agric. Res. 12, 149-161. Buican, D. 1969. Ann. Physiol. Vdg. 11, 431-440. Bunting, E. S . 1962. Rep. 2nd Meet. Maize Sorghum Sect. Eucarpia, 2nd, FAO, Rome pp. 83-87. Bunting, E. S . 1978. In “Forage Maize” (E. S. Bunting, B. F. Pain, R. H. Phipps, J. M. Wilkinson, and R. E. Gunn, eds.), pp. 57-85. Agricultural Research Council, London. Bunting, E. S., and Gunn, R. E. 1974. Annu. Rep. 1973, Plant Breed. Insr. Cambridge, pp. 32-74. Burke, M. J., Gusta, L. V . , Quamme, H. A., Weiser, C. J., and Li, P. H. 1976. Annu. Rev. Plant Physiol. 27, 507-528. Cal, J . P., and Obendorf, R. L. 1972. Crop Sci. 12, 369--373. Caldwell, A. C . , and Ohlrogge, A. J. 1966. In “Advances in Corn Production” (W. H . Pierre, S . A. Aldrich, and W. P. Martin, eds.), pp. 237-256. Iowa State Univ. Press, Ames, Iowa. Carr, M. K. V . , and Hough, M. N. 1978. In “Forage Maize” (E. S . Bunting, B. F. Pain, R. H. Phipps, J. M. Wilkinson, and R. E. Gunn, eds.), pp. 15-55. Agricultural Research Council, London. Cary, J. W., and Mayland, H. F. 1970. Agron. J . 62, 715-719. Chmora, S. N . , and Oya, V. M. 1967. Sov. Plant Physiol. (Engl. transl.) 14, 513-519. Clarkson, D. T., and Gerloff, G. C. 1979. Annu. Rep. 1978, Letcombe Lab., Wantage, pp. 52-54. Cohn, M. A., andobendorf, R. L. 1978. Am. J . Bot. 65, 50-56. Coligado, M. C., and Brown, D. M. 1975. Agric. Meteorol. 14, 357-367. Cooper, A. J. 1973. “Root Temperature and Plant Growth-A Review.” Commonwealth Agricultural Bureaux, Slough, England. Cooper, C. S . , and MacDonald, P. W. 1970. Crop Sci. 10, 136-139. Crawford, R. M. M., and Huxter, T. J. 1977. J . Exp. Bor. 28, 917-925. Creencia, R. P., and Bramlage, W. J. 1971. Plant Physiol. 47, 389-392. Crosbie, T. M., Mock, J. J., and Smith, 0. S . 1980. Crop Sci. 20, 649-655. Cross, H. Z., and Zuber, M. S . 1972. Agron. J . 64, 351-355. Dolstra, O., and Jongrnans, M. A. 1981. Proc. Congr. Eucarpia Maize Sorghum Sect., Ilrh, Montreux (in press). Downton, W. J. S. 1971. In “Photosynthesis and Photorespiration” (M. D. Hatch, C. B. Osmond, and R. 0. Slatyer, eds.), pp. 3-17. Wiley (Interscience), New York.

126

P. MIEDEMA

Duncan, W. G. 1975. In “Crop Physiology. Some Case Histories” (L. T. Evans, ed.), pp, 23-50. Cambridge Univ. Press, London and New York. Duncan, W. G., and Hesketh, J. D. 1968. Crop. Sci. 8, 670-674. Duncan, W. G., Davis, D. R., and Chapman, W. H. 1973. Proc. Soil Crop Sci. SOC.,Fiorida 32, 59-63. Eagles, H. A., and Hardacre, A. K. 1979. N . Z. J . Agric. Res. 22, 553-559. Ebskamp, A. G. 1981. Bedrijfsontwikkeling 12, 175-181. Erickson, R. 0. 1959. Am. Nut. 93, 225-235. Francis, C. A,, Grogan, C. O., and Sperling, D. W. 1969. Crop Sci. 9, 675-677. Galinat, W. C. 1979. In “Maize,” pp. 6-12. CIBA-GEIGY, Basel. Ginsburg, H., and Ginzburg, B. 2. 1971. J . Exp. Bot. 22, 337-341. Grobbelaar, W. P. 1963. Meded. Landbouwhogesch. Wageningen 63 ( 5 ) . 1-71. Grosser, 3. 1979. Tagungsber. Akad. Landwirtschafmiss. DDR Berlin 175, 141- 149. Harper, J. L. 1956. New Phytol. 55, 35-44. Hatch, M. D. 1976. In “Plant Biochemistry” (J. Bonner, and J. E. Vamer, eds.), 3rd ed., pp. 797-844. Academic Press, New York. Hesketh, J. D., Chase, S. S . , and Nanda, D. K . 1969. Crop Sci. 9, 460-463. Heslop-Harrison, J. 1961. Proc. Lion. SOC. London 172, 108-123. Heydecker, W. 1973. In “Seed Ecology” (W. Heydecker, ed.), pp. 553-557. Butterworth, London. Hofstra, G . , and Hesketh, J. D. 1969. Can. J . Bot. 47, 1307-1310. Hofstra, G., and Nelson, C. D. 1969. Can. J . Bot. 47, 1435-1442. Hoppe, P. E. 1961. Plant Dis. Rep. 45, 787. Jones, M. B., Hannon, G. E., and Coffrey, M. D. 1981. Plant Cell Environ. 4, 161-168. Kiesselbach, T. A , , and Ratcliff, J. A. 1920. Res. Bull. Nebr. Agric. Exp. Sra. 16, 1-96. Klasovi, A. 1980. BioMgia (Bratislava) 35, 687-689. Kleinendorst, A. 1975. Neth. J . Agric. Sci. 23, 139-144. Kleinendorst, A,, and Brouwer, R. 1970. Neth. J . Agric. Sci. 18, 140-148. Kleinendorst, A,, and Brouwer, R. 1972. Neth. J . Agric. Sci. 20, 203-217. Koch, H. D., and Miiller, H. W. 1963. Zuechter 33, 155-163. Kramer, P. J. 1969. “Plant and Soil Water Relationships: A Modem Synthesis.” McGraw-Hill, New York. Lehenbauer, P. A. 1914. Physiol. Res. 1, 247-288. Lindow, S. E., Amy, D. C., Upper, C. D., and Barchet, W. R. 1978. In “Plant Cold Hardiness and Freezing Stress. Mechanisms and Crop Implications” (P. H. Li, and A. Sakai, eds.), pp, 249-263. Academic Press, New York. Long, S. P., and Woolhouse, H. W. 1978. J . Exp. Bot. 29, 803-814. Lyons, J. M. 1973. Annu. Rev. Plant Physiol. 24, 445-466. Lyons, J. M., Graham, D., and Raison, J. K. (eds.) 1979a. “Low Temperature Stress in Crop Plants. The role of the Membrane.” Academic Press, New York. Lyons, J. M., Raison, J. K., and Steponkus, P. L. 1979b. In “Low Temperature Stress in Crop Plants. The role of the Membrane” (J. M. Lyons, D. Graham, and J. K. Raison, eds.), pp. 1-24. Academic Press, New York. McAdam, J. H., and Hayes, P. 1978. Rec. Agric. Res. 26, 55-61. McConnell, R. L., and Gardner, C. 0. 1979. Crop Sci. 19, 765-768. McRostie, G. P. 1939. Sci. Agric. 19, 687-699. McWilliam, J. R., and Griffing, B. 1965. Aust. J . Biol. Sci. 18, 569-583. McWilliam, J. R., and Naylor, A. W. 1967. Plant Physiol. 42, 171 1-1715. Major, D. J., and Hamilton, R. I. 1978. Can. J . Plant Sci. 58, 643-650.

EFFECTS OF LOW TEMPERATURE ON Zea mays

127

Miedema, P. 1979. Euphytica 28, 661-664. Miedema, P., and Sinnaeve, J. 1980. J . Exp. Bot. 31, 813-819. Miedema, P., Post, J . , and Groot, P. J. 1982. “The Effects of Low Temperature on Seedling Growth of Maize Genotypes.” PUDOC, Wageningen. Millerd, A , , and McWilliam, J. R. 1968. Plant Physiol. 43, 1967-1972. More Herrero, A,, Bethenod, O., and Morot-Gaudry, J.-F. 1980. Physiol. VPg. 18, 301-312. Nielsen, K. F., Halstead, R. L., MacLean, A. J., Bourget, S. J., andHolmes, R. M. 1961. SoilSci. Soc. Am. Proc. 25, 369-372. Pinnel, E. L. 1949. Agron. J . 41, 562-568. Pollmer, W. G . 1969. Bayer. Landwirt. Jahrb. 46, 279-288. Radford, P. J . 1967. Crop Sci. 7, 171-175. Rahn, J . R., and Brown, D. M. 1971. Can J . Plant Sci. 51, 173-175. Raison, J . K. 1974. In “Mechanisms of Regulation of Plant Growth” (R. Bieleski, A . Ferguson, and M. Cresswell, eds.). pp. 487-497. Royal Society of New Zealand, Wellington. Raison, J. K., Chapman, E. A , , Wright, L. C., and Jacobs, S . W. L. 1979. I n “Low Temperature Stress in Crop Plants. The Role of the Membrane” (J. M. Lyons, D. Graham, and J . K. Raison, eds.). Academic Press, New York. Rajan, A . K., Betteridge, B., and Blackman, G. E. 1971. Ann. Bor. 35, 323-343. Rajan, A . K . , Betteridge, B., and Blackman, G . E. 1973. Ann. Bor. 37, 287-316. Raschke, K. 1975. Annu. Rev. Plant Physiol. 26, 309-340. Rinke, E. H. 1953. Rep. Annu. Hybrid Corn fnd. Res. Conf., 8th, pp. 54-58. Rossman, E. C. 1949. Agron. J . 41, 574-583. Rush, G. E., and Neal, N. P. 1951. Agron. J . 43, 112-1 16. Sachs, J . 1864. Flora 22, 497-506. Sass, J. E. 1977. I n “Corn and Corn Improvement” ( G . F. Sprague, ed.), 2nd Ed., pp. 89-1 10. American Society of Agronomy, Madison, Wisconsin. Scott, D. 1970. N. Z . J . Bot. 8, 369-379. Segefa, V. 1964. B i d . Plant. 6, 189-197. Segefa, V., and Vedralovh. 1970. Biol. Plant. 12, 315-326. Sellshop, J. P. F., and Salmon, S. C. 1928. J . Agric. Res. 37, 315-338. Shaw, R. 1977. In “Corn and Corn Improvement” (G. F. Sprague, ed.), 2nd Ed., pp. 591-623. American Society of Agronomy, Madison, Wisconsin. Shirahashi, K., Hayakawa, S., and Sugiyama, T. 1978. Plant Physiol. 62, 826-830. Shtrausberg, D. V. 1958. Sov. Plant Physiol. (Engl. transl.) 5, 226-232. Sibma, L. 1977. Neth. J . Agric. Sci. 25, 278-287. Singh, P. M., Gilley, J. R..and Splinter, W. E. 1976. Truns. Am. Soc. Agric. Eng. 19, 1152- 1155. Slack, C. R., Rougham, P. G . , and Basset, H. C. M. 1974. Planta 118, 57-73. Slatyer, R. 0. 1967. “Plant-Water Relationships.” Academic Press, New York. Stamp, P. 1980. 2. Pflanzenzuecht. 84, 226-239. Stamp, P. 1981. 2. Pflanzenzuecht. 86, 20-32. Stephens, J. S., Sanderson, S., and Clarkson, D. T. 1980. Annu. Rep. Letcombr Lab., Wantage 1979 pp. 48-49. Stevenson, J. C., and Goodman, M. M. 1972. Crop Sci. 12, 864-868. Sugiyama, T., and Boku, K . 1976. Plant Cell Physiol. 17, 851-854. Tatum, L. A , , and Zuber, M. S. 1943. J . Am. SOC. Agron. 35, 48-59. Taylor, A . O., and Craig, A. S. 1971. Plant Physiol. 47, 719-725. Taylor, A. O., and Rowley, J. A. 1971. Plant Physiol. 47, 713-718. Taylor, A. O., Jepson, N. M., and Christeller, J. T. 1972. Plant Physiol. 49, 798-802. Taylor, A. O., Slack, C. R . , and McPherson, H. G. 1974. Plant Physiol. 54, 696-701.

128

P. MIEDEMA

Tollenaar, M., Daynard, T. B., and Hunter, R. B. 1979. Crop Sci. 19, 363-366. Toole, E. H., Hendriks, S. B., Borthwick, H. A., and Toole, V. K. 1956. Annu. Rev. Plant Physiul. 7, 299-324. Tschape, M. 1972. Biochem. Physiol. Pflanzen 163, 81-92. Uedan, K., and Sugiyama, T. 1976. Plant Physiol. 57, 906-910. ULlstrup, A. J . 1977. In “Corn and Corn Improvement” ( G . F. Sprague, ed.), 2nd Ed., pp, 39 1-500. American Society for Agronomy, Madison, Wisconsin. Van den Honert, T. H., and Hooymans, J. J. M. 1955. Acta Bot. Neerl. 4, 376-384. Van Dobben, W. H. 1962. Nerh. J . Agric. Sci. 10, 377-389. Vedralovi, E., and Segefa, V. 1970. Biol. Plant. 12, 265-274. Voldeng, H. D., and Biackman, G. E. 1973. Ann. Bur. 37, 553-563. Walker, J. M. 1969. Soil Sci. SOC. Am. Proc. 33, 729-736. Wardlaw, I. F. 1968. Bot. Rev. 34, 79-105. Wardlaw, I. F. 1979. Proc. Agron. Soc. N . Z . 9, 39-48. Warren Wilson, J. 1966. Ann. Bot. 30, 753-761. Warren Wilson, J. 1967. Ann. Bot. 31, 41-57. Wassink, E. C. 1972. Meded. Landbouwhogesch. Wageningen 72 (25) 1-15. Wassink, E. C. 1974. Meded. Landbouwhogesch. Wageningen 74 (17) 1-7. Watts, W. R. 1972. J. Exp. Bot. 23, 713-721. Weathenvax, P., and Randolph, L. F. 1955. In “Corn and Corn Improvement” ( G . F. Sprague ed.), pp. 1-61. Academic Press, New York. Wernham, C. C. 1951. P a . Agric. Exp. Sta. Prog. Rep. 47, 1-12. Wheaton, T. A. 1963. “Physiological Comparison of Plants Sensitive and Insensitive to Chilling Temperatures.” Ph.D. thesis, University of California, Davis, California. Wilkes, H. G. 1979. Crop lmprov. 6, 1-18. Wolfe, J , 1978. Plant Cell Environ. 1, 241-247. Wolfe, J., and Bagnall, D. 1979. In “Low Temperature Stress in Crop Plants. The Role of the Membrane” (J. M. Lyons, D. Graham, and J. K. Raison, eds.), pp. 527-533. Academic Press, New York. Wolfe, J., and Bagnall, D. J. 1980. Ann. Bot. 45, 485-488. Woodstock, L. W., and Grabe, D. F. 1967. Plant Physiol. 42, 1071-1076.

ADVANCES IN AGRONOMY, VOL. 35

AGRICULTURAL USE OF PHOSPHORUS IN SEWAGE SLUDGE* M. B. Kirkham Evapotranspiration Laboratory, Kansas State University Waters Annex, Manhattan, Kansas

...... .

... . ...

IV. Summary and Conclusions...............................................

141 144

154

1. INTRODUCTION

Phosphorus is one of about 20 elements necessary for the growth, development, and maintenance of plants and animals. Therefore, it is essential that the supply of phosphorus in agricultural soils be maintained. If crops are removed from the land, the supply of phosphorus in the soil may become the limiting factor in crop production. Large quantities of phosphorus compounds are used as fertilizers and an extensive phosphate deposit in a country represents a great natural resource. The discovery of phosphorus is generally attributed to Hennig Brand of Hamburg, who in 1669 obtained it by distilling urine (Corbridge, 1980). Urine remained the only source of the element for nearly 100 years after its discovery. By the end of the eighteenth century, it had been replaced by bones. Supplies of bones became inadequate. But large phosphate mineral deposits soon were found in Morocco, the United States, the Soviet Union, China, and Tunisia. Today in *Contribution of the Department of Agronomy, Evapotranspiration Laboratory, Kansas Agricultural Experiment Station, Manhattan, Kansas 66506. 129

Copyright 0 1982 by Academic Press. Inc. All rights of reproduction in any form reserved. ISBN 0-12-000735-5

130

M. B. KIRKHAM

the United States, phosphorus fertilizers are manufactured from nonrenewable deposits of phosphate rock (U.S. Department of Agriculture, 1978). The Bone Valley formation in Florida is the richest deposit of phosphate ever discovered (Carter, 1980). Because of an annual production rate of over 45 million metric tons of phosphate rock, supplies are declining. The high-grade ores, the most important ones to the phosphate fertilizer industry, are dwindling and miners must dig into lower grade ores, which create problems such as high impurity levels, large volumes of wastes, high energy use, and high costs. Although there is no shortage of phosphate, the industry cannot process low-grade ores into fertilizers with existing technology (National Fertilizer Development Center, 1979; White, 1979). The diminishing amount of easy-to-mine phosphate rock is a strong argument for recycling phosphorus in the waste products of society. Sewage sludge is the by-product of wastewater treatment processes. Handling and disposal of sludge are the most costly phases of sewage treatment. The traditional means of sludge handling and disposal are (1) ocean disposal through barging or pipeline transport, ( 2 ) dewatering and drying with disposal in a landfill, (3) incineration, and (4) lagooning. These means of disposal only displace the wastes and do not get rid of them. Dried sludge sometimes can be sold or given away as a fertilizer, but drying is expensive. Land application of wet sludge is another, final method of disposal which solves the disposal problem in a beneficial way (Vesilind, 1979). Sludge has been spread on agricultural land for disposal, fertilizing, and soil conditioning purposes for decades (Zwerman and de Haan, 1973; Page, 1974; Carroll et al., 1975). With the use of trucking, railroad, barging, and pipeline transport, however, land spreading of sludge is becoming more prevalent among the large municipalities. In a survey of nearly 2000 treatment plants of a size of 1 million gallons daily or larger in the United States, 25% stated that they now are spreading sludge, and 68% have been spreading for less than 10 years (Carroll ef al., 1975). In Canada, about 25% of the sludge also is deposited on land (Black and Schmidtke, 1974). Phosphorus, in the form of phosphate, is one of the major uncontrolled pollutants in wastewater. As a result of the United States Federal Water Pollution Control Act Amendments of 1972 and the 1972 International Agreement between Canada and the United States, signed to restore and protect the water quality of the Great Lakes, phosphorus discharges from all sources must be reduced to the lowest practical level. The effluent objective is 1 mg/liter total phosphorus (Black, 1974; Van Fleet er al., 1974). High phosphorus removal can be achieved from wastewater either chemically or biologically (Convery, 1970). Addition of phosphate-precipitating chemicals such as lime, aluminum sulfate (alum), sodium aluminate, or ferric chloride during sewage treatment minimizes effluent phosphate concentrations. Also, biological technology has been developed to reduce effluent phosphate to an average concentration of 0.55 mglliter P

PHOSPHORUS IN SEWAGE SLUDGE

131

(Levin and Shapiro, 1965; Levin and Shaheen, 1967; Shapiro et al., 1967; Levin et al., 1972, 1975). Under appropriate conditions, which include high dissolved oxygen levels and stirring rates, activated sludge’ can remove more phosphate than the sludge requires for growth. This biological uptake of excess phosphate has been termed “luxury uptake” (Carberry and Tenney, 1973). Removal of large amounts of phosphorus from effluents results in sludges with high concentrations of phosphorus. Consequently, the effect of the high phosphorus-containing sludges spread on land is of interest agriculturally. The purpose of this article is to assess the impact of these sludges on agricultural land. Three types of sludge will be considered: organic sludge, chemical sludge, and combined sludge. Organic sludge is the waste-activated sludge which usually is digested2 and then disposed. Chemical sludge is the sludge produced when treated sewage effluent is further treated for phosphorus removal by the addition of lime, alum, sodium aluminate, or iron salts. Combined sludge is the sludge that results when the chemical sludge is handled along with the waste-activated sludge (organic sludge) to form a single sludge. The article will not consider the agricultural value of phosphorus in effluent or animal manures spread on land because these topics have been reviewed (Bouwer and Chaney, 1974; Porcella and Bishop, 1975; Pratt, 1977; Olsen and Barber, 1977; Loehr et al., 1979b; Sommers and Sutton, 1980). Also, the article will not consider the value of phosphorus in sludge spread on strip-mined land, although this is a useful application (Peterson and Gschwind, 1972; Pietz et al., 1974). Finally, the article will not discuss the value of phosphorus in effluent used as a fertilizer in marine aquaculture systems to produce food (shellfish) (Ryther et al., 1972). The relative merits of land disposal of wastewater versus ocean disposal have been evaluated (Cowlishaw and Roland, 1973). Utilization on land, where there is less hazard of heavy-metal and virus uptake, appears to be the better alternative.

It. CONCENTRATION OF PHOSPHORUS IN SLUDGES A. ORGANIC SLUDGES

Table I gives the concentrations of phosphorus in sludge from treatment plants using conventional techniques of settling and biological treatment to produce “‘Activated sludge” is sludge floc produced in raw or settled sewage by the growth of bacteria and other organisms in the presence of dissolved oxygen. 2“Digested sludge” is produced during “digestion” which is the process in which organic or volatile matter in sludge is gasified, liquefied, mineralized, or converted into more stable organic matter, through the activities of living organisms.

Table I Concentration of Phosphorus in Organic Sewage Sludges Total (%) Location

w N

Treatmento

Total P (%)

Milwaukee, WI United States, 1931-1935

A,HD A

United States, 193 1-1935

D

Milwaukee, WI Urbana, IL Houston, TX Maywood, IL Pasadena, CA United States, all cities Milwaukee, WI

A A A A A I,D A

0.91.7 Av.: 1.4 0.41.7 Av.: 0.9 1.36 1.28 1.22 I .85 I .73 0.71 1.26

Milwaukee, WI Toledo, OH England, 1942-43 England, 1945 New Jersey New Jersey New Jersey New Jersey New Jersey New Jersey

A D AD AD R A D R D HT

1.36 0.89 1.14 0.97 0.99 1.10 0.66 0.66 0.83 1.54

Extractable p (%)

Al

Ca

Fe

Reference Kadish (1925) Anderson (1956)

1.046

Anderson (1956)

1.07 0.87

0.99(citrate) O.OOa(water)

1.58

1.10

4.64

I .70 3.60

1.20 4.74

5.00 3.80

1.11

I .92 1.21 3.34

2.24 5.04 4.20

1.70 2.28 0.56(citrate) 0.69(c itrate) 1.28(citrate)

DeTurk (1935) DeTurk ( 1935) DeTurk (1935) DeTurk ( 1935) DeTurk (1935) DeTurk (1935) Rehling and Truog (1939) Rudolfs (1940) Rudolfs ( 1940) Anonymous (1973) Anonymous (1973) Rudolfs and Gehm (1942) Rudolfs and Gehm (1942) Rudolfs and Gehrn (1942) Rudolfs and Gehrn (1942) Rudolfs and Gehm (1942) Rudolfs and Gehm (1942)

I

New Jersey New Jersey New Jersey New Jersey New Jersey New Jersey New Brunswick, NJ South River, NJ Highland Park, NJ Tenafly, NJ Marion, IN United States, 27 plants United States, 29 plants United States, 17 plants United States, 22 plants England England New Zealand New Zealand Ohio, 10 plants San Diego, CA Washington, DC Beltsville, MD Greenbelt, MD Chicago, IL Houston, TX McKeesport, PA

A R+A D(R + A)

.oo

D(R+ HT)

1

FC,C D(FC,C) R,CL,U R,CAI,CF,CL,U D A,CF,HD

0.70 0.65 0.87 0.65 0.80 1.62 1.6 1.6 1.7 Trace1.8 0.93 .O 0.51.6 0.6 1.o 0.9 1.5 1.06 2.10 0.63 0.72 0.40 3.02 1.36 3.25

D,S,Li I,D,AD D(R +TF),AD A,HD A,D(not HD)

R DAD AD A ,HD D P,D P,D P,D

P J A,U A,U A,U

1.29(citrate) 0.76(citrate) 0.69(citrate) 0.73(citrate) 0.62(citrate) 0.52(citrate)

1.43 0.87 0.84

Rudolfs and Gehm (1942) Rudolfs and Gehm (1942) Rudolfs and Gehm (1942) Rudolfs and Gehm (1942) Rudolfs and Gehm (1942) Rudolfs and Gehm (1942) Bear and Prince ( I 947) Bear and Prince (1947) Bear and Prince (1 947) Bear and Prince ( 1 947) Backmeyer (1948) Adams et a/. (1951) Adams et a/. (1951) Adams et a / . (195 1) Adam et ai. (I 95 1)

-

Adams et a/.(195 1) Adams er al. (195 I ) Adam et al. (195 I) Adams ef al. (1951) Anderson (1955) Men (1956) Merz (1956) Merz (1956) Men ( 1956) Merz (1956) Merz (1956) Merz ( I 956) (continued)

Table I-Conrinued Total (%) Location

-g

Milwaukee, WI Des Moines, IA Hagerstown, MD Hagerstown, MD Los Angeles, CA Beltsville, MD Rochester, NY Indianapolis, IN United States, Av. United States, Av. Washington, DC Washington, DC Baltimore, MD Baltimore, MD Baltimore, MD Baltimore, MD Jasper, IN Jasper, IN Jasper IN Richmond, IN Richmond, IN Richmond, IN Chicago, IL Southwest Southwest Southwest

Treatmentu

Total p (8)

Extractable p (%)

Al

Ca

Fe

Reference

A D R A D

1.74 1.46 1.24 2.18 1.79 0.25 0.51 I .90 2.5 0.49 0.49 0.63 0.57 4.84 1.74 1.31 0.71 1.24 I .54 2.28 1.60 1.91

Merz (1956) Merz (1956) Merz (1956) Merz (1956) Merz (1956) Merz (1956) Merz (1956) Merz (1956) Merz (1956) Merz (1956) Anderson (1956) Anderson (1956) Anderson (1956) Anderson (1956) Anderson (1956) Anderson (1956) Anderson (1956) Anderson ( 1956) Anderson (1956) Anderson ( 1956) Anderson (1956) Anderson (1956)

R,U A,U HD,U

1.19 2.46 2.89

Anderson (1956) Anderson ( I 956) Anderson ( 1956)

A,U A,D A,D A,D A,D 1

I Lagoon A D R D R A HT D,HD

R

t; VI

Milwaukee, WI United States, 1951-1955

HD.U A

United States, 1951-1 955

D

Chicago, IL Baltimore, MD New Haven, CT Stamford, CT Torrington, CT Wallingford, CT Waterbury, CT West Haven, CT United States, 23 cities England England United States, average England England N . Toronto, Canada Fergus, Canada Guelph, Canada Richland. WA Richland, WA Richland. WA Richland, WA Richland, WA

A,U D(P+ A +TF) D ,CF,CL ,VF D,CF,CL,VF,FD D,AD I,D,AD R,P,CF,CL,FD E,D,CF,VF M R D R AD D,Li D,CL,VF,FC D D,An p,u P+TF,U TF,U P,CL,U P+TF,CL,U

1.74 1 .&3.3 Av.: 2.5 0.42.2 Av.: 1.2 3.07 I .25 0.67 0.27 0.58 0.31 0.50 0.60 2.3 0.6 1.2 0.44I .32 I .96 2.4 5.9 I .62 2.67 0.94 1.31 I .04 0.95 1.04

Anderson (1956) Anderson ( 1956)

Anderson (1956)

Anderson (1959) Anderson ( 1959) Irving (1960) Irving (1960) Irving (1960) Irving (1960) Irving (1960) Irving (1960) Irving (1960) Bunting (1963) Bunting (1963) Burd (1968)

0.31 0.24

5.00 5.83

1.25 2.50

Anonymous (1973) Anonymous (1973) Baldock (1 974) Bates et a / . (1974) Bates et al. (1974) Counts et a / . (1975) Counts et al. (1975) Counts et al. (1975) Counts et al. (1975) Counts et al. (1975) (continued)

Table &Continued Total (%) Location

0, W

Richland, WA Sweden Miami, FL Pensacola, FL Chicago, IL Chicago, IL Chicago, IL Calumet Stickney Drury Lane, Canada Elizabeth Gardens, Canada Elizabeth Gardens, Canada Elizabeth Gardens, Canada North Toronto, Canada

Sweden, 162 plants

Treatmenta

Total p (%)

Extractable p (%)

A1

Ca

Fe

Reference

3.73 2.53 I .25

0.67 1.22 5.25 4.37

Counts er al. (1975) Emrnelin (1973) Hammond (1973) Hammond (1973) Hammond (1973) Hammond (1973)

4.70 2.96

3.54 4.82

Hinesly er al. (1974) Hinesly er al. (1974)

D,Ae A(dried) D,Li

0.89 1.1 0.85 1.82 2.29 2.16

D,An,Li D,An,Li

3.69 2.62

P,WA0,500°F

4.6

Hudgins and Silveston (1973)

P,WA0,450"F

7.3

Hudgins and Silveston (1973)

P,WA0,500°F

2.9

Hudgins and Silveston (1973)

A, WA0,45O0F

2.6

Hudgins and Silveston (1973)

P,WA0,SOO"F WAO, 15% COD red. WAO, 45% COD red. WAO, 80% COD red.

5.2 2.16 3.08 4.59 0.23.6 Av.: 1.4

Hudgins and Silveston (1973) Teletzke ( 1966) Teletzke (1966) Teletzke (1966) Jansson (1972)

TF,CL,U D,An or Ae

P

0.29 1.49 4.25 1.85

0.65.4 Av.: 2.5

Wisconsin , 35 plants Athens, GA

D,Li D

2.76.1 0.88

Lebanon, OH

p,u

0.91

Dayton, OH

D

0.361.2

4.218.0 I .47

0.87.8

Keeney et al. (1973)

King and Morris ( I 973) Kirkham and Dotson (1974)

2.40 1.78 4.72 4.80 1.525.41

1971 1973

Ringkoping, Sweden Aahus Viby, Sweden Sweden, 20 plants Chicago, IL West-SW

D,An

Illinois, 24 cities United States, 7 states Milwaukee, WI Denmark, 21 plants

A D,An

2.44 5.54 1.628.72

1.77 2.63 1.214.28

1.18.1

0.3-

I .54-

4.8

4.90

Median:

Median:

Median:

2.8 0.74.9 Av.: 2.4 0.046. I Av.: 1.8 2.29 1.555.42

2.1

1.02 2.06 0.953.99

0.125

Av.: 5.1

3.63 0.049.00 Av.: 1.29

Kirkham (1975) Kirkham (1975) Pauly (1973) Pauly (1973) Pauly (1973)

Lynam et a!. ( 1 975)

Lynam et al. (1975)

0.6-

Lynam et al. (1975)

6.2 Av.: 2.10 1.86 1.86

Milwaukee (197 1) Pauly and Simonsen (1973)

Median: 3.73

Av.: 3.71 (continued),

Table I-Continued ~~

Total (8) Location Hasting, MN St. Paul, MN Chicago, 1L Hanover Calumet West-Southwest Denver, CO Athens, GA Denver, CO Denver, CO Washington, DC Washington, DC Connecticut, 9 plants Iowa, Iowa, Iowa, Iowa, Iowa, Iowa, Iowa,

12 towns 12 towns 12 towns 12 towns 12 towns 12 towns 12 towns

Treatment0

Total p (%)

P,WA,D,An P,WA( 1:2),U

2.61 2.20

P,WA,D,An P,WA,D,An WA,CF,VF,HD P,WA,CF,CL,VF,U P,D,An P,D,An WA,D,Ae P,R D

2.59 3.90 2.49 1.75 0.75 1.3 3.1 0.70 1.69 0.85.0 Av.: 2.5 1.95 1.61 1.48 I .52 2.18 1.22 1.82

Extractable p (%)

A1

Ca

Fe

0.65 0.74

2.97 2.52

0.45 0.76

Peterson et a/. (1973) Peterson et al. (1973)

5.05 4.20 1.4 7.35 1.21

2.22 3.68 5.32 1.48

Peterson et al. (1973) Peterson e t a / . (1973) Peterson er a / . (1973) Peterson et a / . (1973) Peterson e t a / . (1973) Sabey and Hart (1973, 1975) Sabey and Hart (1973, 1975) U.S.D.A. (1972) U.S.D.A. (1972) Sawhney and Norvell (1980)

1.21 1.04

1.7 1 .o

1.o

Reference

Chae Chae Chae Chae Chae Chae Chae

and Tabatabai and Tabatabai and Tabatabai and Tabatabai and Tabatabai and Tabatabai and Tabatabai

(1981) (1981) (1981) (198 1) (1981) (1981) (1981)

OAbbreviations: A, activated; AD, air-dried; Ae, aerobic; Al, alum added; An, anaerobic; C, conditioned; CAI, conditioned with alum; CF, conditioned with iron; CL, conditioned or stabilized with lime; D, digested; De, dewatered; E, elutriated; F, iron added; FC, filter cake; FD, flash-dried; HD, heat-dried; HT, humus tank; HOT, holding tank; I , Imhoff tank; L, lime added; Li, liquid; M, marketed sludge; P, primary: R , raw; S, secondary; TF, trickling filter; U , undigested; VF, vacuum filtered; WA, waste-activated; WAO, wet air oxidation.

PHOSPHORUS IN SEWAGE SLUDGE

139

organic sludges. Primary and raw sludges are listed, as well as activated and digested sludges. The table does not include all references to phosphorus in sludge. Bates (1972) gives an extensive bibliography. Values of aluminum, calcium, and iron have been included, when listed in the original publication, to compare concentrations of these elements in organic sludges with those in chemical and combined sludges. Sometimes alum, lime, and ferric chloride are added to sludge to condition it. This has been shown in Table I, if noted in the original publication. The table shows that the phosphorus concentration in conventional sludges has increased since 1925. The change has resulted from industrial wastes, widespread use of phosphate-containing detergents, and the increased use of home garbage grinders (Burd, 1968). The data also show that, at any one time, digested sludge has less phosphorus than activated sludge. For example, in 1956 Merz reported that the average value of phosphorus in activated sludge was 2.5% and in digested sludge was 0.49%. Bear and Prince (1947) and Anderson (1956), however, found that activated sludge contained about twice as much phosphorus as digested sludge. Digested sludge contains about two times the amount of phosphorus in raw sludge because digestion destroys the unstable organic component. The phosphorus content of sludge resulting from wet-air oxidation has been included in Table I, even though this sludge may not be biologically treated. The wet air oxidation process reduces soluble forms of phosphorus and produces a cake in which nearly all the phosphorus in sludge is precipitated into particulate forms (Teletzke, 1966; Sommers and Curtis, 1977). Therefore, the sludge can contain a high concentration of phosphorus (greater than 7%) (Hudgins and Silveston, 1973). The overall variation in phosphorus content in organic sludges throughout the years ranges from trace amounts to 8.1% (Chicago, West-Southwest Plant). Disregarding the one very high value at Chicago, the range of phosphorus in conventional, digested sludges is about 0.8-6.1%. High values may reflect chemical treatment of the wastewater to precipitate phosphorus, although not noted in the original publication. The average value of phosphorus in digested sludge is now about 2.5% (Hecht et al., 1975; Council for Agricultural Science and Technology, 1976; Sommers et al., 1976; Kelling et al., 1977a; Regional Wastewater Solids Management Program, 1977; U.S. Environmental Protection Agency, 1977, 1978, 1979; Dick et al., 1978; Knezek and Miller, 1978; Leeper, 1978; Loehr et al., 1979a; Shipp et al., 1979; Sawhney and Norvell, 1980). The average value was 0.86% in 1931-1935 and 1.11% in 1951-1955 (Anderson, 1956). Composted, raw sludge has a concentration of phosphorus (about 1.5%) similar to raw, uncomposted sludge (Taylor et al., 1978; Loehr et al., 1979a; Tester et al., 1979). Sommers (1977) reported that aerobically digested sludge has less phosphorus than anaerobically digested sludge. He analyzed sludges

140

M . B . KIRKHAM

from 150 treatment plants in eight states (Illinois, Indiana, Michigan, Minnesota, New Hampshire, New Jersey, Ohio, Wisconsin) and found the medians for total phosphorus concentrations for anaerobically and aerobically treated sludges were 3.0 and 2.7%, respectively. But Mitchell et al. (1978) measured the same or more total phosphorus in aerobically digested sludge (1.71 %) than in anaerobically digested sludge (0.75- I .7 1%) from four sewage treatment plants in the metropolitan area of Syracuse, New York. Table I also shows that much of the total phosphorus in sludge is extractable and in the inorganic form. Rehling and Truog (1939) reported that 79% of the total phosphorus in Milwaukee’s activated sludge was soluble in ammonium citrate. DeTurk (1935) also found that 79% of the Milwaukee sludge and 71% of the Houston sludge were in an “available” form. Rudolfs and Gehm (1942) showed that the various sludges they analyzed contained from 73 to 90% extractable phosphorus. The type of treatment given to the sludge did not affect the amount of extractable phosphorus. Vlamis and Williams (1972) determined that half of the total phosphorus in sludge was soluble in 2% acetic acid. Sommers e f al. (1973) carried out a detailed study of the characteristics of phosphorus in digested sludge. They separated the phosphorus in the sludge into three fractions: soluble inorganic, total inorganic, and total organic. The total phosphorus content ranged from 1.2 to 4.5% and was composed primarily of insoluble, inorganic phosphorus. They said that chemical rather than biological reactions controlled the accumulation of phosphate in sludge. Organic phosphorus constituted a relatively low proportion of the total phosphorus, indicating that extensive mineralization of organic matter occurred during digestion. Organic CIP ratios ranged from 21 to 92, which are similar to ratios found in readily decomposable plant residues. Therefore, they concluded that phosphorus should not be a limiting factor in the decomposition of sewage sludge after incorporation into soil. Little information is available on the forms of organic phosphorus in sewage sludges. Chae and Tabatabai (1981) determined the distribution of phospholipid phosphorus in 12 sewage sludges in Iowa. The sludges contained from 1.11 to 2.64% P (average, 1.72%), from 0.49 to 1.35% organic P (average, 0.91%), and from 0.43 to 1.37% inorganic P (average, 0.80%).The lipid phosphorus in the sludges ranged from 26 to 433 ppm. Expressed as percentage of organic phosphorus in sewage sludges, the lipid phosphorus ranged from 0.24 to 8.84% (average, 2.18%). Cosgrove (1973) extracted inositol polyphosphates from activated sludge in Australia. The extract was a mixture of penta- and hexaphosphates of myo-, chiro-, scyllo-, and neoinositol and resembled the inositol polyphosphate component of soil organic matter. Because the same polyphosphate mixture did not occur in raw sewage, he concluded that it was formed as a result of biological action during the aeration process.

141

PHOSPHORUS IN SEWAGE SLUDGE

B. CHEMICAL AND COMBINED SLUDGES

There are few data on the concentration of phosphorus in chemical sludges that have not undergone biological treatment (Table 11). Rudolfs and Gehm ( I 942) found 1.35% phosphorus in a raw sludge containing alum, iron, and lime. About 90% of the phosphorus was extractable in ammonium citrate. Recent values for the amount of phosphorus in chemical sludges are two to three times higher than the 1942 values published by Rudolfs and Gehm. Hudgins and Silveston (1973) found that the phosphorus content of sludges containing lime or ferric chloride was less, after wet air oxidation of the sludge, than the phosphorus content of sludges without chemicals. Oxidized sludges with no chemical treatment had 2.6-7.3% phosphorus (Table I). Sludges with lime or ferric chloride treatment had 1.5-2.0% phosphorus (Table 11). At the same time that phosphorus concentration of the oxidized solids from chemical sludges was reduced, about a 10-fold increase in orthophosphorus concentration in the filtrate Table I1 Concentration of Phosphorus in Chemical Sewage Sludges Total (%) Location New Jersey

Total Extractable TreatmentU P (%) P (%) R,AI,F,L

1.38

1.23(citrate)

1.6

North Toronto, Canada P,F,WAO, 550°F Samia, Canada P,F,WAO, 450°F Samia, Canada P,F,WAO, 500°F Newmarket, Canada P,L,WAO, 450°F P,Fb Grand Rapids, MI

4.3

Grand Rapids, MI

P,FC

3.3

Lebanon, OH

P,AI

2.67

Lebanon, OH

P,F

2.77

2.0 1.5

1.5

aFor abbreviations, see footnote to Table I. bIron added prior to primary clarification. cIron added prior to secondary clarification.

A1

Ca

Fe

Reference

Rudolfs and Gehm ( 1 942) Hudgins and Silveston (1973) Hudgins and Silveston (1973) Hudgins and Silveston (1973) Hudgins and Silveston (1973) 8.6 Green er al. (1974) 2.2 Green et a / . ( 1974) 1.81 3.79 0.42 Kirkham and Dotson (1974) 0.68 3.00 4.01 Kirkham and Dotson ( 1974)

142

M. B. KIRKHAM

occurred. They speculated that the ash with its high concentration of phosphorus might be usable. The metropolitan district of greater Chicago, however, has abandoned its wet air oxidation process, not only because of expense and safety factors, but also because it is committed to a program of land disposal of sludge (Lue-Hing et al., 1975). Wet air oxidation ash is not suitable for this purpose. As a result of the 1972 Canada-United States International Agreement on Great Lakes Water Quality, the governments of Ontario and Canada have funded research related to the use of chemically treated sludges on agricultural land (Black, 1974; Black and Schmidtke, 1974). Therefore, much of the data in the literature on the phosphorus content of combined sludges comes from Canada (Bates, 1972; Bates et al., 1974, 1977, 1978, 1979, University of Guelph, 1975, 1976; Chawla et al., 1974a,b, 1977; Bryant and Cohen, 1978; Cohen and Bryant, 1978; Soon et al., 1978, 1980; Lee et al., 1981). Scott and Horlings (1974), using sludge from North Toronto, Canada, found that the anaerobically digested sludge, which was produced primarily from domestic sewage, and in which iron had been used for phosphate removal, contained about 7% iron, about 6% calcium, and 11-12% phosphorus (as orthophosphate), based on dry-sludge weight. After conditioning for filtration using lime and ferric chloride, the sludge, on a dry-weight basis, contained 8-9% or more of iron, about I 1- 12% calcium, and about 12% orthophosphate. Between January 1973 and March 1974, Chawla et al. (1974a) sampled digested sludge from four Canadian cities that have incorporated phosphorusremoval systems into existing facilities by the addition of alum, ferric chloride, or lime as precipitating agents. Mean concentrations of total phosphorus and standard deviations are shown in Table 111. Chawla et al. (1974a) noted that metal composition did not vary with sludge type except, of course, for aluminum and iron. The metal concentration was correlated with the industries in a town. Kirkham and Dotson (1974) also found no change in metal (cadmium, copper, manganese, nickel, and zinc) and boron concentrations, due to alum or iron treatment. But, as expected, aluminum and iron concentrations were higher in alum and iron sludges, respectively. Van Fleet et al. (1974) predicted that the use of lime and alum for phosphorus removal would enhance significantly the removal of lead, copper, trivalent chromium, and arsenic, This may be offset, however, by the interference with the precipitation of metals by nitrilotriacetate, one of the detergent phosphorus substitutes. Actual concentrations of metals in chemical sludges will have to be assessed as phosphorus removal facilities come into operation. Green et al. (1974) compared phosphorus and iron concentrations in both primary and secondary (activated) sludges (Tables I1 and 111). When ferric chloride was added prior to primary clarification, the phosphorus content of both sludges increased about one-third, and the iron content increased 3-fold. When ferric chloride was added before secondary clarification, the phosphorus con-

143

PHOSPHORUS IN SEWAGE SLUDGE

Table I11 Concentration of Phosphorus in Combined Sewage Sludges Total ( S ) Treatmento

Total P (%)

New Jersey

D,AI,F,L

2.02

New Jersey

1.10

Milwaukee, WI Texas City, TX

D,ALF,L, FC A,F A,AI

Texas City, TX

A,F

Texas City, TX

AS-

Location

Aurora, Canada Pt. Edward, Canada Newmarket, Canada North Toronto, Canada Samia, Canada North Toronto, Canada

D,AI D,An,Al D,An,L D,An,F D,F D,An,F,FC, CF North Toronto, Canada D,An,F Pt. Edward, Canada

D,An,AI

Newmarket, Canada

D,An,L

Samia, Canada

D,An,F

North Toronto, Canada P+S,D,An, F North Toronto, Canada P+S,D,An, F,CL,CF Grand Rapids, MI A,F" Grand Rapids, MI

A,Fc

Washington, DC

S,AI

See footnotes to Table 11.

(1.h~

Extractable P (%)

AI

Ca

Fe

0.99(citrate) 5.08

2.61 1.6-

0.38-

6.0 3.85.6 1.31.7 4.71 2.50 1.13 4.83 5.65

8.1

8.75 5.56 0.34 1.35 0.58

12.9 2.274 1.1 1.90+ 0.7 0.552 0.3 0.77? 0.5 3.74

0.78

3.85

1.18

1.55.6 6.47 9.05 18.68 3.67 3.70

Rudolfs and Gehm (1942) Rudolfs and Gehm (1942) Milwaukee (1971) Univ. Texas (1971)

Univ. Texas (1971) 9.314.4 Univ. Texas (1971) 0.72 1.24 0.85 9.50 8.10

Bates et al. (1974) Bates e f a / . (1974) Bates et a / . (1974) Bates et al. (1974) Bates et a/. (1974) Baldock (1974) Chawla et a / . ( 1974a) Chawla et a/. (1 974a) Chawla Ct al. ( 1974a) Chawla et al. ( 1974a)

6.42 12.0

7.50 8.75

6.5

7.6

14.4

8.7

4.03

Reference

Scott and Horlings 1974) Scott and Horlings 1974) Green et a/ ( 1974) Green et a / . ( 1974) USDA (1972)

144

M. B. KIRKHAM

centration in the activated sludge was about double compared to the phosphorus concentration in the activated sludge when ferric chloride was added before primary clarification. As expected, the phosphorus concentration of the primary sludge fell when the chemical was added before secondary clarification. The concentration of phosphorus in chemical, nonbiological sludges ranged from 1.4 to 4.3% (Table 11) in the publications reviewed in this article. The concentration of phosphorus in combined sludges ranged from 0.55 to 14.4% (Table 111). The lowest value (0.55%) occurred with an anaerobically digested, limed sludge. The two highest values (12.9 and 14.4%) were associated with an anaerobically digested, alum sludge conditioned with ferric chloride, and an activated, iron sludge, respectively. Disregarding the two high values and wet air oxidation sludges, the average phosphorus concentrations, in percent, of chernical and combined sludges were as follows: Digested, lime Activated, lime Primary, alum Digested, alum Primary, iron Digested, iron Activated, alum Activated. iron

0.84 1S O 2.67 3.04 3.53 3.69 3.80 4.60

The concentration of phosphorus in digested sludges with lime was 3.6 times less than the concentration of phosphorus in digested, alum sludges. The concentration of phosphorus in digested, iron sludges was 1.2 times more than the concentration of phosphorus in digested, alum sludges. The average value of phosphorus in the digested, combined sludges was 2.9%. The average, present (1 97 1- 1981) value of phosphorus in digested, organic sludges was found to be 2.4% (Table I).

111. AGRICULTURAL USE OF PHOSPHORUS IN SLUDGES A. ORGANIC SLUDGES

I . Plant-Growth Studies Bear and Prince (1947) published one of the first reports concerning the fertilizer value of phosphorus in sewage sludge. They used four New Jersey

PHOSPHORUS IN SEWAGE SLUDGE

145

sludges from New Brunswick, South River, Highland Park, and Tenafly. The phosphorus content and characteristics of these sludges are given in Table I. In a greenhouse experiment using Sudan grass (Sorghum vulgare sudanense), they applied sludge at the rate of 1 1.2 tons/ha to Sassafras loam soil. Benefit from added P,O, was small because the sludges were all relatively high in phosphorus. Dry-weight yield with the activated Tenafly sludge was highest compared to yields from the undigested New Brunswick and South River sludges and the digested Highland Park sludge. The activated sludge contained about twice as much phosphorus as the other sludges. Bear and Prince (1947) also spread the sludges for 3 years on a limed Sassafras loam soil in a field experiment. About 243 kg/ha P were added to the soil as a result of sludge applications. Yield of silage corn (Zea mays L.) was similar among the treatments and averaged 29 tons/ha. The quick-test values of phosphorus in the sludged soils showed that phosphorus levels were excellent except in the soil with the Tenafly activated sludge, which had a good level. They concluded that ample quantities of phosphorus had been applied and that sludge was a good source of fertilizer phosphorus. The most extensive and long-term study of the effect of sludge on agricultural land in the United States has been done using sludge from the Metropolitan Sanitary District of greater Chicago, Illinois. In 1970, the District bought land for sludge disposal in Fulton County, Illinois, 300 km southwest of Chicago (Halderson et al., 1974). The sludge is barged to the disposal site on the Illinois River, lagooned, and then spread on the fields using sprinklers. Since 1967, the University of Illinois in Urbana-Champaign has been collaborating with the Metropolitan Sanitary District in studying the effect of sludge on cropland (Hinesly and Jones, 1974; Hinesly et al., 1974, 1977, 1979a,b; Hohla et al., 1978). Phosphorus in corn, reed canary grass (Phalaris arundinaceae L.), alfalfa (Medicago sativa L.), and soybean (Glycine max Merr.) grown with sludge has been measured. Corn and reed canary grass grew on plots with silt loam soil that had received sludge over a 4-year period (1967-1970) (Hinesly et al., 1974). About 102 cm of sludge were applied during the 4-year period to the maximum sludge-treated plots, and 67 cm of sludge were applied to the half-maximum sludge-treated plots. In 1970, the concentrations of phosphorus in the corn leaves and grain and reed canary grass grown on maximum-treated plots were higher than phosphorus concentrations in plants grown on control plots, which received 224 kglha N, 112 kg/ha P,O,, and 112 kg/ha K,O annually in 1968, 1969, and 1970. The concentrations of phosphorus in the corn leaves and grain and reed canary grass, however, were within the normal ranges observed in corn and grass plants (Jones, 1967; Chapman, 1973). Extractable (0.1 N HC1) phosphorus concentrations in the 0- to 10-cm depth of soil at both the maximum sludge rate and half-maximum sludge rate were significantly higher than levels of extractable phosphorus in control plots. In 1970, after 4 years of sludge treatment, the

146

M. B . KIRKHAM

extractable phosphorus concentrations in the maximum, half-maximum, and control plots amounted to 525, 502, and 380 kg/ha P, respectively. In another corn experiment, the Illinois researchers grew corn on a Blount silt loam which received sludge for 4 years (1968-1971). Total phosphorus in the surface layer of the zero, one-quarter, one-half, and maximum-treated plots, after a total of 65.4 cm or 168 dry tons/ha digested sludge had been applied to the maximum-treated plots, were 429, 522, 750, and 1152 ppm, respectively. Extractable phosphorus concentrations for these four treatments were 16, 51, 148, and 376 ppm, respectively. The range of total phosphorus in mineral soils from humid areas is about 88-1760 ppm, with a typical value of 440 ppm. The range of extractable phosphorus in soils is 10-200 ppm, with typical values between 20 and 50 ppm (Buckman and Brady, 1969; Bauer, 1972). Therefore, the maximum sludge treatment had increased the value of extractable phosphorus in the surface of the soil above the amount normally found in soil. Increases in phosphorus concentrations in deeper soil horizons were extremely variable. Sludge applications were highly correlated with increased contents of phosphorus in the corn plants. Corn leaves and grain on the control plot had 0.27 and 0.50% phosphorus, respectively, while corn leaves and grain on the maximumtreated plot had 0.33 and 0.56% phosphorus, respectively. These two values, however, were still within normal concentration ranges: 0.2-0.4% P, leaves; 0.4-0.8% P, grain (Chapman, 1973). Results from an experiment with alfalfa were similar to results from the corn and reed canary grass experiments. Phosphorus concentrations built up in the soil and in alfalfa. But concentrations of phosphorus in the alfalfa were within normal concentration ranges. Kelling ef al. (1977a) in Wisconsin also reported that increasing rates of sewage sludge resulted in marked increases in concentration of phosphorus in tissue of corn, rye (Secale cereale L.), and sorghum-Sudan (Sorghum bicolor Moench. X S . sudanense P. Stapf.). Hinesly et al. (1974) thought that reduced yields of soybean, treated with combinations of sludge, superphosphate, and potassium chloride, were due to toxic concentrations of phosphorus in the plants (Bauer, 1973). Greatest reduction in yield occurred in plots receiving additional inorganic fertilizer. Whether or not phosphorus toxicity is observed depends upon many factors, including the plant species, the phosphorus status of the plant, concentration of micronutrients, and soil salinity. Lower organisms like algae, aquatic weeds, and activated sludge microorganisms apparently can take up high concentrations of phosphorus without impairing metabolic activity (Gerloff and Skoog, 1954; Levin and Shapiro, 1965; Gerloff and Krombholz, 1966). With higher plants, poor growth associated with high phosphorus levels has been related to specific effects caused by excessive phosphorus. For example, greater than optimum amounts of phosphorus can reduce nodulation of legumes (Bingham, 1973), inhibit growth of root hairs (Brewster et al., 1976), and decrease shoothoot ratios (Brewster et al.,

PHOSPHORUS IN SEWAGE SLUDGE

147

1975). Not only do species vary in phosphorus content, but cultivars within a species differ in the amount of phosphorus absorbed. Wallace et al. (1973) reported that two soybean cultivars had different tolerances to excess phosphorus fertilization. O’Sullivan et al. (1974) observed that snapbean (Phaseolus vulgaris L.) strains differed in efficiency in phosphorus utilization. They identified “responders” and “nonresponders” to phosphorus fertilization. Kirkham (1980) found that leaves of a drought-resistant winter-wheat (Triticurn aestivum L. em. Thell.) cultivar had less phosphorus than those of a drought-sensitive winter-wheat cultivar. Yet both cultivars grew equally well. The phosphorus level of plants before they are subjected to high levels of phosphorus also determines whether or not toxicity symptoms are observed. Apparently, under some conditions, as phosphorus status of the plant increases, a feedback mechanism slows down net phosphate flux (Green and Warder, 1973; Loneragan, 1979). The mechanism for phosphate appears to operate at a rather coarse level of control, because several species have been shown to accumulate toxic concentrations of phosphorus in their leaves from low phosphate concentrations in solution (Loneragan, 1979). The literature has many examples of phosphate-induced micronutrient deficiencies. Excess phosphorus results in copper, iron, and zinc deficiency (Bingham, 1973; Ellis, 1973). Wallace et al. (1969) found that not only did high phosphorus levels induce zinc deficiency, but low phosphorus levels induced zinc toxicity. At toxic levels of copper and zinc, phosphorus can reduce uptake of toxic elements like nickel and aluminum (Brown et al., 1950). The availability of nickel is reduced by additions of phosphate to soil (Pratt et al., 1964). Andrew and Vanden Berg ( 1 973) and Andrew et al. (1 973) found that aluminum treatments reduced phosphorus uptake in plant tops and roots of pasture legumes. Phosphorus fertilizers themselves can be contaminated with toxic elements, such as cadmium, that inhibit plant growth (Clark and Hill, 1958; Schroeder and Balassa, 1963; Bingham, 1973; Stenstrom and Vahter, 1974; Gilkes et al., 1975; Lee and Kenney, 1975; Severson and Gough, 1976; David et a/., 1978; Goodroad and Caldwell, 1979; Jung et al., 1979; Webber, 1979; Mortvedt et al., 1981). In general, micronutrient deficiencies associated with high phosphorus are more readily produced in acid soils. Liming promotes phosphorus uptake (Truog, 1953; Amarasiri and Olsen, 1973; Smilde, 1973). Organic matter content in a soil also affects the amount of phosphorus taken up. Nishita et al. (1973) found that additions of organic matter to soil decreased uptake of phosphorus by barley (Hordeurn vulgare L.) seedlings. In addition, number of mycorrhiza (Malajczuk et al., 1975), amount of substances secreted by roots which hydrolyze fertilizer phosphorus (Pammenter and Woolhouse, 1979, and soil temperature (Sheaffer et al., 1979) affect the rate of phosphorus uptake. High soluble-salt concentrations caused by high phosphorus levels also can create phosphorus toxicity. The Illinois workers (Hinesly et al., 1974) observed

148

M. B. KIRKHAM

high levels of salt in the soil where the soybeans grew. Beans are extremely salt sensitive, compared to corn, reed canary grass, and alfalfa, which also were grown on the Illinois plots (U.S. Salinity Laboratory Staff, 1954). Bhatti and Loneragan ( 1970) thought that high concentrations of phosphorus in wheat leaves (200 mM P in the cell sap, 10 atm osmotic pressure) caused the necrotic injury they observed, because of salt damage. High phosphate concentrations disturbed the water relations of the leaf cells.

2. Soil Studies Sabey and Hart (1973) incorporated sludge from Denver, Colorado, into a Truckton loamy sand at rates of 0, 25, 50, 100, and 125 tonsiha and measured the phosphorus distribution in the soil. Millet (Panicurn miliaciurn L.), sorghumSudan, or wheat grew on the sludge-treated soil. The phosphorus distribution in the soil followed a pattern similar to the distribution of organic matter, except that at high sludge-application rates there appeared to be some phosphorus in the 30- to 46-cm depth of soil. This indicated that phosphorus moved down the soil profile. The surface soil had phosphorus concentrations that greatly exceeded the normal need for crop growth at all sludge application rates above 50 tons/ha, but these quantities were not considered detrimental. In the fall, after spring application of the sludge, extractable phosphorus concentrations in the surface of the soil were as high as 230 ppm. The control plot had extractable phosphorus concentrations of about 10 ppm. Sludge application rates did not increase the amount of phosphorus in the wheat plants. Hinesly and Jones (1974) in Illinois and Webber and Hilliard ( 1 974) in Canada also found that extractable phosphorus moved down to lower soil depths in sludge-treated soil. In California, Lund et al. (1976) measured increased extractable phosphorus concentrations to depths as great as 3 m in sewage sludge ponds. Kelling et al. (1977b) found that liquid digested sludge, applied at rates from 3.75 to 60 tons/ha, increased the concentration of extractable phosphorus in a sandy loam and a silt loam in southcentral Wisconsin. The extractable phosphorus, however, decreased with time after sludge application, probably as a result of phosphorus fixation. Significant amounts of phosphorus remained available for at least 2 years at the high treatment rates. The rate of movement of phosphorus through soils treated with sludge depends upon whether or not the soil is anaerobic. Magdoff et al. (1974) found that once soil columns treated with sewage effluent became anaerobic, the phosphorus concentration of the leachate fell. They postulated that the lower phosphorus concentrations of the leachate in anaerobic columns compared to aerobic columns may have been caused partially by formation of iron phosphates when the anaerobic solution came in contact with air in the gravel at the bottom of the columns. Greenberg and McGauhey (1955) also found that phosphate penetrated

PHOSPHORUS IN SEWAGE SLUDGE

149

to greater depths in a Hanford fine sandy loam kept under submergence with effluent than in soil that was not continually ponded. Patrick and Khalid (1974) showed that the rate of phosphate release and sorption by aerobic and anaerobic soils was dependent upon the phosphate concentration of the soil. Anaerobic soils released more phosphate to soil solutions low in soluble phosphate and sorbed more phosphate from soil solutions high in soluble phosphate than did aerobic soils. King and colleagues (King and Morris, 1973; Touchton et al., 1976) applied over a 2-year period a total of 6.93, 13.75,20.0, and 40.0 cm of liquid domestic sludge (0.88% P) to coastal bermudagrass (Cynodon ductylon L.) and rye. A 2.5c d h a application of this sludge supplied about 146 kg/ha P to the soil. They found a large amount of extractable phosphorus in the sludge crust (496 ppm, or about 2.8 times as much phosphorus in the 0- to 15-cm depth of soil), which apparently supplied ample phosphorus for plant uptake. Sludge did not significantly increase extractable phosphorus levels in the soil and most of the applied phosphorus was retained in the soil in an unavailable form. Losses through leaching, they said, would not be a problem, but runoff of phosphorus in the crust could be.

3. Drainage and Runoff Studies Hinesly and Jones (1974) and Hinesly et al. (1974) found that concentrations of phosphorus increased in drainage waters with increased sludge applications. Thirty percent of the drainage water samples from maximum sludge-treated plots (174 tons/ha) contained greater than 1 ppm P. A 10-fold increase in applied phosphorus resulted in a 10%increase in the number of samples with phosphorus concentrations greater than 1 ppm. Uncontaminated, underground waters contain 0.01-0.03 ppm P (Martin et ul., 1970). The major proportion of drainage water samples having the higher concentrations of phosphorus occurred in samples collected from the first drainage water obtained in the fall. Lue-Hing et al. (1974) did not observe increased phosphorus concentrations in leachate collected at monitoring wells at a site near Chicago, which received sludge for 5 years. Concentrations of phosphorus in surface waters, however, increased from 0.22 ppm in 1969 to 3.12 ppm in 1973. Increases in phosphorus concentration of surface waters at the Fulton County sludge disposal site have not been observed (Zenz et al., 1974). Personnel at the USDA (1972) did not find increased phosphorus levels in surface or drainage waters as a result of sludge applications. Brink (1969, 1971) studied the phosphorus content in runoff water from a clay soil with vegetation treated for three winters with digested sludge from Uppsala, Sweden. Concentrations of phosphorus in the runoff water were highly variable and depended upon the sludge application rate and the intensity of rainfalls. High

150

M. B. KIRKHAM

sludge application rates (15-30 tons/ha) and high intensity spring rains resulted in large concentrations in the runoff (1.88 ppm P) and large amounts of phosphorus being lost (1.4 kg/ha P). Brink said that 4-5 tons/ha/year of sludge could be safely applied without seriously increasing concentrations of phosphorus in surface waters. He did a phosphorus balance sheet. Compared to the amount added with sludge, the runoff was small (less than 0.01% of total phosphorus applied, even with maximum runoff on maximum sludge-treated plot), and the stored amounts were large. A negative balance in the case of the control plot caused a deficit in yield of about 20%, while the highest amounts of phosphorus added as a result of sludge applications did not cause an increase in yield. Coker (1965) also noted that large amounts of phosphorus in sludge applied to already fertile cropland had no effect on crop yield. B. CHEMICAL AND COMBINED SLUDGES

1. Plant-Growth, Soil, and Drainage Studies

Large quantities of sludges containing phosphate precipitants will be produced in the future. In 1972, 82,600 metric tons of chemical sludges were produced (Farrell, 1974; Smith and Rosenkranz, 1975). In 1985, 413,000 metric tons of chemical sludges will be produced. At the same time, more sludge will be utilized on land than it has been in the past. Therefore, one wonders what effect these sludges containing alum, iron, or lime will have on cropland. If plants are grown on land spread with sludge containing phosphate precipitants, the phosphate precipitated into sludge during wastewater treatment may be in a form unavailable for plant growth. In fact, the excess metal hydroxide may react with the natural soil phosphate and make it less available. Jansson (1970, 1972) at the University of Uppsala in Sweden was one of the first people to study phosphorus availability in soils spread with chemically treated sludges. His results indicated that iron and alum sludges with excess hydroxides did not bind exchangeable phosphate in the soil, and phosphorus availability was not decreased. Under field conditions and with adequately limed soils, applications of alum and iron sludges resulted in a slow but positive phosphate reaction. On acid, lime-poor soils, the effect was unfavorable. But such soils, Jansson pointed out, were not consistent with rational land use and they should be limed. In greenhouse studies, he found that corn grew just as well on a loam soil fertilized with superphosphate as on the same soil treated with alum sludges during a 3year period. Personnel of the USDA (1972) did a series of greenhouse experiments using raw, digested, alum and iron secondary sludges from Washington, D.C. The phosphorus content of the first three sludges is shown in Tables I and 111. In the first experiment, they mixed digested sludge, raw sludge, alum secondary

PHOSPHORUS IN SEWAGE SLUDGE

151

sludge, and iron secondary sludge into loamy sand at the following four rates: 0, 2, 4, and 6%, on a dry-weight basis. They grew field corn, rye, and soybean in the pots 34 days after the initial mixing of the soil and sludge. The seedlings died if they were planted in the soil immediately after the sludge had been added. All the sludge-treated soils produced better plants than the control. Alum-treated sludge, however, produced about half as much dry weight as the digested, raw, or iron sludges. In another greenhouse experiment, the USDA investigators constructed boxes (1 22 X 61 X 15 cm) to simulate sludge behavior in trenches. They placed digested sludge from Washington, D.C., in one box and a raw alum-sludge stabilized with lime in the other box. They planted corn and observed root growth through the Plexiglas-covered front of the boxes. They collected drainage water from the boxes. Growth of the corn in the digested sludge was excellent, and the roots penetrated into the sludge layer within 2-3 weeks. Drainage water from this box was clear and almost odorless. Corn did not grow as well on the alum-lime sludge as it did on the digested sludge. The plants were chlorotic and showed marginal burning of the lower leaves. The corn roots did not penetrate the alum-lime sludge layer until 3 months after planting. The drainage water from the alum-lime sludge box was reddish-brown, turbid, and malodorous. The raw alum-sludge seemed to allow water to pass through it and leach it, in contrast to the digested sludge which appeared relatively impervious in the wet state. Dolar et al. (1972) also found that an alum-treated sludge applied to oat (Avena sativa L.) plants grown under greenhouse conditions markedly depressed plant growth in soil with high rates (224 tonsiha) of sludge. Phosphorus and nitrogen contents of plants grown on alum sludges were low at all rates of sludge application. Kirkham and Dotson (1974) grew barley for 16 weeks in pots of loam soil which they irrigated for 7 weeks with wet sludges containing precipitated phosphates. Raw primary sludge and primary sludges from alum and ferric chloride addition to raw wastewater, each either limed or unlimed, were added at two different rates, corresponding to total application of 23 or 46 tonsihaiyear. Their results showed that barley plants irrigated with the three types of sludges grew as well as those supplied with inorganic fertilizer. Canadian workers have studied the effect of digested lime-, alum-, and irontreated sludges from Ontario on the growth of bromegrass (Bromus sp. L.), corn, and orchard grass (Dactylis glomeratu L.) (Bates er al. 1974; Chawla ef af., 1974b; Soon et al., 1980). The lime and iron sludges yielded more than the alum sludge at all but the lowest rate of application. In a field experiment with corn, however, the alum sludge produced as high yields as the lime and iron sludges (Bates et al., 1974; Chawla et al., 1974b; University of Guelph, 1975). The sludges increased extractable soil phosphorus, with the calcium sludge being much more effective than the other sludges (University of Guelph, 1976; Bates et al., 1977, 1978, 1979). Counts et al. (1975) also showed that lime does not tie

152

M. B. KIRKHAM

up phosphorus in soils treated with lime sludges. But, in Switzerland, Paccolat (1979) reported that the available phosphorus of soil can be much reduced after large applications of sludge flocculated with lime, and treatment of sludge with aluminum salts can prevent phosphorus uptake. Soil pH has been reduced by iron sludge application, slightly reduced by alum sludge, and increased by calcium sludge (Chawla et al, 1974b; Soon et al., 1980). Unlimed sludged soils may become acid, resulting in metallic toxicity to plants (Chawla et al., 1974b). Chawla et al. (1974b) found that concentrations of aluminum were slightly higher in the leachate from a silt loam compared to concentrations in the leachate from a loamy sand. All the concentrations were low and did not reveal any significant trend due to increasing rates of application or different types of sludges. Removals of phosphorus from the soils were entirely through plant uptake with nonsignificant contributions through leaching. They observed highest removals of phosphorus on silt loam soil as compared to sandy loam. Studies with lysimeters (30 cm in diameter, 1.8 m deep) exposed to atmospheric conditions showed that there was no evidence that aluminum, iron, or lime in the digested sludges was detrimental to production of orchard grass, wheat, and buckwheat (Fadopyrum esculentum Moench.) (Chawla et al., 1977; Bryant and Cohen, 1978; Cohen and Bryant, 1978). When air-dried sludge was used, phosphorus concentrations in leachates were always less than the maximum permissible concentration in effluent (1 .O mg/liter). But when fluid sludge was used, phosphorus in leachate from alum sludge exceeded 1 mg/liter and reached a value as high as 10 mglliter (see p. 83, Cohen and Bryant, 1978). Most of the studies concerning the effect of chemically treated sludges on agricultural land are of recent origin. Chemical and combined sludges, as well as organic sludge, however, have been spread on cropland for years (Bear and Prince, 1947; Trubnick and Mueller, 1958), even though the practice has not been formally recorded. The continued long-term use of the sludges with chemicals on land suggests that they do not adversely affect plant growth. Long-term sludge disposal sites that have received chemically treated sludges should be identified, and soil and plants analyzed for elemental composition. Recently, results of a 5-year study by the Canadian workers (Soon et al., 1980) have been published. They used three chemically treated sludges (calcium sludge, alum sludge, iron sludge) and found that uptake of copper, zinc, manganese, chromium, and lead by corn and bromegrass was dependent mainly on rate of sludge application rather than on sludge type. 2 . Runoff Studies

Curnoe (1974) measured runoff of an iron-precipitated sludge from the North Toronto, Canada, sewage treatment plant, which he spread in the fall, winter,

PHOSPHORUS IN SEWAGE SLUDGE

153

and spring on a poorly drained sandy clay loam soil with 2 and 6% slopes. Rates of sludge application, based on the nitrogen content of the sludge, were 200 and 800 kg/ha N. With one exception, phosphorus losses were low (less than 3 kg/ha P) with no significant difference between any sludge application time and the nosludge treatment. The exception occurred after a fall rainstorm, which immediately followed a sludge application. The amount of phosphorus lost on the 2% slope at the higher sludge application rate was 13 kg/ha. Phosphorus tied up in precipitated compounds in chemically treated sludges might present less of a hazard to surface drainage waters than phosphorus in organic sludges. The Canadian studies done since 1974 have shown that loss of phosphorus by runoff is low (generally less than 1 kg/ha P lost in 1 yr) (University of Guelph, 1976; Bates et al., 1977, 1978, 1979). C. COMPARISON OF ORGANIC AND CHEMICAL SLUDGES

Organic sludges have the advantage that they do not contain chemicals used to precipitate phosphorus. The use of chemicals may introduce trace quantities of impurities that result in water quality standard violations (Howe, 1973). Land spreading of potentially toxic materials is, of course, undesirable. Loehr (1972) concluded that it was better to use controlled land disposal for phosphorus removal from animal wastewaters rather than chemical treatment with alum, lime, or ferric chloride. Land disposal eliminated problems of chemical control and costs. Hinesly (Anonymous, 1971) pointed out that, while iron sludges should cause no problem because soils naturally contain 5 6 % iron, alum-treated sludges on cropland could be a problem, particularly if soil pH is low. Anderson and Hammer (1973) showed that lower organisms also are affected adversely by alum. Total microorganisms and protozoa in activated sludge decreased in number with increasing alum doses. Even small quantities of alum hydroxide can clog the soil, making it unfit for agriculture (Dean, 1968). Chemically treated sludges are watery and dewater slowly on land (Jansson, 1972; U.S. Department of Agriculture, 1972). Lime treatment of sludge adds calcium carbonate and phosphate, and sometimes magnesium hydroxide, to sludge. Alum and iron salts add the respective aluminum and ferric phosphates and the hydroxides (Dean and Smith, 1973). All three hydroxides are bulky and hard to filter. In addition, chemical treatment of wastewater adds to the mass of sludge produced and consequently the amount that must be disposed. These difficulties associated with land disposal of chemically treated sludges would not be encountered with organic sludge. The amount of phosphorus in sludges, even in those without precipitated phosphorus, apparently has not limited growth when sludges have been used as a fertilizer under field conditions. If quantities of sludge added to land are suffi-

154

M. B. KIRKHAM

cient to satisfy the nitrogen needs of the crop, phosphorus requirements for maximum yields will be ample. The phosphorus, however, would be beneficial if it were spread on phosphorus-deficient soils. Pierre (1953) said that many soils in the United States are deficient in phosphorus and that the use of phosphate fertilizers often increases crop yields. The area most deficient in phosphorus is the coastal plain of the South Atlantic and Gulf states (Parker, 1953). In many parts of the world, including the Sahel (Nabos et al., 1974), food production is limited by low concentrations of available phosphorus (Reeve and Sumner, 1970; Harre ef al., 1971). Phosphorus deficiency also is probably the most critical mineral deficiency in grazing livestock (Allaway, 1975). When phosphorus fertilizers are added to soils deficient in available forms of phosphorus, increased crop and pasture yields usually follow. Sometimes the phosphorus concentration in the crop is increased, and this increase may help prevent phosphorus deficiency in the animals eating the crop. As far as human nutrition is concerned, phosphorus deficiencies are not a serious problem. Therefore, phosphorus fertilizers are valuable, not because they increase the phorphorus content of food for humans, but because they increase total food production. Not only can phosphorus be used beneficially to fertilize hay and pasture land, but also it can be used to fertilize forest land. Many acres of forests are being fertllized profitably in Finland and Sweden, and there is a potential use of 15,580 metric tons of phosphorus on forests in North America (Harre et al., 1971; Valmari, 1974). When disposing of sludge on land, one often cannot choose the type of soil on which one puts the sludge. Transportation costs are high and can amount to as much as 44% of the cost of treatment and disposal of sludge (Bauer, 1973). Therefore, if one is putting sludge on land mainly for disposal rather than for fertilizing purposes, one is interested primarily in depositing it as inexpensively as possible. Whether or not the disposal site is phosphorus deficient is of secondary interest. One can, however, choose the type of crop planted on the sludgetreated soil. Plants known to have high phosphorus requirements could be grown on sludge-treated soil. Angiosperms (flowering plants), in general, have higher phosphorus concentrations than gymnosperms (e.g., pines, spruce, firs) (Gerloff et al., 1964). Peas, celery, spinach, peaches, and tomatoes, which are all angiosperms, are a few of the crops listed by Chapman (1973) that normally have high phosphorus concentrations (more than 0.5%).

IV. SUMMARY AND CONCLUSIONS The average concentration of phosphorus in organic sludges has increased since the mid-1920s when values were first published. Digested, organic sludges produced between 1931 and 1935, 1951 and 1955, and between 1971 and 1981

PHOSPHORUS IN SEWAGE SLUDGE

155

contained, on an average, about 0.96, I . 1, and 2.4% P, respectively. Activated sludges, which have more phosphorus than digested sludges, contained in 1931-1935, 1951-1955, and 1971-1981, 1.4, 2.5, and 2.5% P, respectively. Chemical sludges (no biological treatment) and combined sludges (chemical and biological treatment) vary in phosphorus concentrations, depending upon the characteristics of the wastewater, the type of chemical, the quantity of the chemical, and the place of addition of the chemical during the wastewater treatment process. The publications surveyed in this article showed chemical and combined sludges to contain the following percentages of phosphorus: Digested, lime Activated, lime Primary, alum Digested, alum Primary, iron Digested, iron Activated, alum Activated, iron

0.84 1S O 2.67 3.04 3.53 3.69 3.80 4.60

Studies done with organic, chemical, and combined sludges spread on cropland demonstrate that total and extractable concentrations of phosphorus in soils with sludge are higher than concentrations of phosphorus in fertilized, control soils even after one application of a small quantity (0.5 cm) of sludge. Large quantities spread for several years (e.g., 5 crdyear for 4 years) result in tremendous concentrations of phosphorus in the soil (e.g., 2994 ppm total P), and almost 40% of this can be in the extractable form. When these large quantities of phosphorus are applied, the phosphorus does not remain in the surface of the soil. Increases in extractable phosphorus concentrations occur at the 30- to 46-cm depth, and significant increases have been observed down to the 75-cm depth. The concentration of phosphorus in drainage and surface waters from land spread with large amounts of sludge contain phosphorus concentrations greater than those in drainage and surface waters from land without sludge. The phosphorus concentration of drainage and surface waters, however, does not increase as a result of modest (e.g., 1-2 crdyear) applications of sludge. The rapid increase of soil phosphorus may be a cause for concern. Alekseeva (1968), cited by Black (1970), found that plots in the Soviet Union receiving 20 kg/ha P as superphosphate each time sugar beets (Beta vulgaris L.) were grown in crop rotation, in addition to main fertilizer applications of 180 kg/ha P added presumably at annual intervals for 3 1 years, had an extractable phosphorus concentration at the 10-cm depth of 105 ppm and at the 90-cm depth of about 40 ppm. The control soil without fertilizer had a concentration of about 30 pprn. Hinesly et al. (1974) found that a silt loam soil fertilized with sludge for 3 years with a total phosphorus application of 5370 kg/ha had an extractable phosphorus

156

M. B . KIRKHAM

concentration of 376 ppm in the 0- to 15-cm depth and 49 ppm in the 30- to 46cm depth. The fertilized control soil at these two depths had extractable phosphorus concentrations of 16 and 15 ppm, respectively. Therefore, phosphorus concentrations built up quickly in soils treated with sludge. Despite the high phosphorus concentrations in sludge-treated soils, yields of plants grown on these soils apparently have not been reduced due to the large amounts of phosphorus, except in one case with soybeans (Lynam et af.,1975). The reactions of phosphorus in the soil are complex, and soil has a great capacity to immobilize phosphorus. Soils contain large amounts of aluminum, calcium, and iron which can react with phosphorus. High levels of phosphorus in the soil, therefore, should not necessarily be detrimental to growth. The phosphorus toxicity noted in the soybeans in Illinois probably was due to a micronutrient imbalance or to a high salt concentration caused by the excessive amounts of phosphorus. Phosphorus is an essential macroelement required by every living plant and animal cell. Consequently, one does not expect toxicity problems with it as one does with the large amounts of trace elements present in sludge (Page, 1974). In a study done on the buildup of heavy metals in corn plants grown on a silt loam soil that had received sludge for 35 years, larger than normal concentrations of cadmium and copper were found in the corn leaves (Kirkham, 1975). The phosphorus concentration in the treated soil also was elevated (0.07% P in surface of control soil, 0.57% P in surface of sludge-treated soil). The phosphorus concentrations of the plant parts, however, were within normal concentration ranges (Kirkham, 1976). Phosphorus cannot be recovered economically from sludge (Opferkuch er al., 1972). Therefore, recycling of sludge on agricultural land appears to be the only economical way to conserve the phosphorus wasted in sewage. Rising costs of commercial fertilizers and increasing demand to produce food make sludge a valuable nutrient source for crops. Conservation of phosphorus is especially needed because of limited long-term sources of this element. The value of phosphorus in sludge will increase if inorganic phosphorus fertilizers become unavailable. ACKNOWLEDGMENTS Partial support during the preparation of this article was provided by the National Science Foundation under grant numbers ENV77-04092 and ISP-8014715. All statements and opinions are the sole responsibility of the author. REFERENCES Adams, R. C., Maclean, F. S . , Dixon, J . K., Bennett, F. M . , Martin, G. I., and Lough, R. C. 1951. N. Z. Eng. 6, 396-424. Alekseeva, E. N. 1968. Agrokhirniya 8, 78-82.

PHOSPHORUS IN SEWAGE SLUDGE

157

Allaway, W. H. 1975. Cornell Univ. Agric. Inf. Bull. No. 378. Amarasin, S. L., and Olsen, S. R. 1973. Soil Sci. Soc. Am. Proc. 37, 716-721. Anderson, D. T., and Hammer, M. I. 1973. Water Sewage Works 120 (1). 63-67. Anderson, M. S. 1955. U.S. Dep. Agric. Circ. No. 972. Anderson, M. S. 1956. Sewage Ind. Wastes 28 (2), 132-135. Anderson, M. S. 1959. Sewage Ind. Wastes 31 (4). 678-682. Andrew, C. S., and Vanden Berg, P. J. 1973. Aust. J. Agric. Res. 24, 341-351. Andrew, C. S., Johnson, A. D., and Sandland, R. L. 1Y73. Ausr J. Agric. Res. 24, 324-339. Anonymous. 1971. “Sludge Disposal: A Case of Limited Alternatives,” pp. DI-D4. Water Pollution Control Fed. Deeds and Data, Washington, D.C. Anonymous. 1973. Water Pollut. Control 111 (3), 79-82. Backmeyer, D. 1948. Water Sewage Works 95, 353-356. Baldock, E. H. 1974. Annu. West. Can. Water Sewage Conf., 25th, Sept. 1973, Edmonton pp. 203-216. Bates, T. E. 1972. Min. Environ. Toronto Res. Rep. No. 1. Bates, T. E., Moyer, J. R., and Haq, A. 1974. Conf. Proc. Ontario Min. Environ. Polluf. Control Branch, Toronto pp. 150-158. Bates, T. E., Beauchamp, E. G . , Johnston, R. A,, Ketcheson, J. W., Protz, R., and Soon, Y. K. 1977. “Land Disposal of Sewage Sludge. Vol. IV (April, 1975-March, 1976),” Res. Rep. No. 60. Ministry of the Environment, Environment Canada, Toronto, Ontario, Canada. Bates, T. E., Beauchamp, E. G., Haq, A., Ketcheson, J. W., Protz, R., and Soon, Y. K. 1978. “Land Disposal of Sewage Sludge. Vol. V (April, 1976-March, 1977),” Res. Rep. No. 73. Ministry of the Environment, Environment Canada, Toronto, Ontario, Canada. Bates, T. E., Haq, A., Ketcheson, J. W., and Soon, Y.K. 1979. “Land Disposal of Sewage Sludge. Vol. VI (April 1, 1977-March 31, 1978);’ Res. Rep. No. 90. Ministry of the Environment, Environment Canada, Toronto, Ontario, Canada. Bauer, W. J. 1972. “Heavy Metals in Soils and Crops,” Vol. 4. Soil Enrichment Materials C o p . , Chicago, Illinois. Bauer, W. J. 1973. Proc. Joint Conf. Recyc. Municipal Sludges Efluents Land pp. 161-167. Bear, F. E., and Prince, A. L. 1947. N.J. Agric. Exp. Sta. Bull. No. 733. Bhatti, A. S., and Loneragan, J. F. 1970. Agron. J . 62, 288-290. Bingham, F. T. 1973. In “Diagnostic Criteria for Plants and Soils” (H. D. Chapman, ed.), pp. 324-361. Quality Printing, Abilene, Texas. Black, C. A. 1970. In “Agricultural Practices and Water Quality” (T. L. Willrich and G. E. Smith, eds.), pp. 72-93. Iowa State Univ. Press, Ames, Iowa. Black, S. A. 1974. Proc. Natl. Conf. Municipal Sludge Manage. pp. 107-113. Black, S. A,, and Schmidtke, N. W. 1974. Conf. Proc. Ontario Min. Environ. Pollut. Control Branch, Toronto pp. 1-9. Bouwer, H., and Chaney, R. L. 1974. Adv. Agron. 26, 133-176. Brewster, J. L., Bhat, K. K. S., and Nye, P. H. 1975. Plant Soil 42, 171-195. Brewster, J. L., Bhat, K. K. S., and Nye, P. H. 1976. Plant Soil 44,279-293. Brink, N. 1969. Grandfoerbaettring 22 (1-2), 33-41. (English transl. of abstr. in Chem. Abstr. 71, 317, 1969, No. 122994.) Brink, N. 1971. “Runoff from Fertilizing with Digested Sludge,” Publ. No. 7. Agricultural School, Uppsala, Sweden. Brown, B. A,, Hawkins, A., Rubins, E. J., King, A. V., and Munsell, R. I. 1950. Soil Sci. SOC.Am. Proc. 14, 240-243. Bryant, D. N., and Cohen, D. B. 1978. “Air-Dried Chemical Sewage Sludge Disposal on Agricultural Land. Vol. I,” Technol. Development Rep. EPA 4-WP-78-3. Water Pollution Control Directorate, Environment Canada, Ottawa,. Ontario, Canada.

158

M. B. KIRKHAM

Buckman, H. 0.. and Brady, N. C. 1969. “The Nature and Properties of Soils.” Macmillan, New York. Bunting, A. H. 1963. J . Agric. Sci. 60, 121-140. Burd, R. S. 1968. “A Study of Sludge Handling and Disposal,” Pub. WP-20-4. Fed. Water Pollution Control Admin., Office of Res. and Development, U.S. Dep. Interior, Washington, D.C. Carbeny, J. B., and Tenney, M. W. 1973. J . Water Pollut. Control Fed. 45, 2444-2462. Carroll, T. E., Maase, D. L., Genco, J. M., and Ifeadi, C. N. 1975. “Review of Landspreading of Liquid Municipal Sewage Sludge,” EPA-670/2-75-049. U.S. Environmental Protection Agency, Cincinnati, Ohio. Carter, L. J. 1980. Science 209, 372-374 & 376. Chae, Y. M., and Tabatabai, M. A. 1981. Soil Sci. SOC. Am. J . 45, 20-25. Chapman, H. D. 1973. “Diagnostic Criteria for Plants and Soils,” 2nd printing. Quality Printing, Abilene, Texas. Chawla, V. K., Stephenson, J. P., and Liu, D. 1974a. Conf. Proc. Ontario Min. Environ. Pollut. Control Branch, Toronto pp, 63-94. Chawla, V. K., Bryant, D. N., and Liu, D. 1974b. Conf. Proc. Ontario Min. Environ. Pollur. Control Branch, Toronto pp. 207-233. Chawla, V. K., Bryant, D. N., Liu, D., and Cohen, D. B. 1977. “Chemical Sewage Sludge Disposal on Land (Lysimeter Studies). Vol. I,” Res. Rep. No. 67. Ministry of the Environment, Environment Canada, Toronto, Ontario, Canada. Clark, L. J., and Hill, W. L. 1958. J . Assoc. 08.Agric. Chem. 41, 631-637. Cohen, D. B., and Bryant, D. N. 1978. “Chemical Sewage Sludge Disposal on Land (Lysimeter Studies). Vol. 11,” Res. Rep. No. 79. Ministry of the Environment, Environment Canada, Toronto, Ontario, Canada. Coker, E. G . 1965. J . Inst. Sewage Puri& 9, 419-426. Convery, J. J . 1970. “Treatment Techniques for Removing Phosphorus from Municipal Wastewater,” Water Quality Office, 17010-01/70, U.S. Environmental Protection Agency, Washington, D.C. Corbridge, D. E. C. 1980. “Phosphorus. An Outline of its Chemistry, Biochemistry and Technology,’’ 2nd ed. Elsevier, Amsterdam. Cosgrove, D. J. 1973. J . Environ. Qual. 2, 483-485. Council for Agricultural Science and Technology. 1976. “Application of Sewage Sludge to Cropland: Appraisal of Potential Hazards of the Heavy Metals to Plants and Animals,” pp. 18. EPA-430/9-76-013. U.S. Environmental Protection Agency, Washington, D.C. Counts, C. A., Shuckrow, A. J., and Smith, J. E., Jr. 1975. “Lime Stabilized Sludge: Its Stability and Effect on Agricultural Land,” EPA-670/2-75-012. U.S. Environmental Protection Agency, Cincinnati, Ohio. Cowlishaw, W. A., and Roland, F. J. 1973. In “Recycling Treated Municipal Wastewater and Sludge Through Forest and Cropland” (W. E. Sopper and L. T. Kardos, eds.), pp. 408-421. Pennsylvania State Univ. Press, Univ. Park, Pennsylvania. Cumoe, W. E. 1974. Conf. Proc. Ontario Min. Environ. Pollut. Control Branch, Toronto pp. 174- 184. David, J. D., Pinkerton, A., and Williams, C. H. 1978. J . Aust. Inst. Agric. Sci. 44, 132-137. Dean, R. B. 1968. Environ. Sci. Techno/. 2, 1079-1086. Dean, R. B., and Smith, J. E., Jr. 1973. Proc. Joint Conf. Recyc. Municipal Sludges EffZuents Land pp. 39-47. DeTurk, E. E. 1935. Sewage Works J . 7 , 597-610. Dick, R. I . , Simmons, D. L., Ball, R. O., Perlin, K., and O’Hara, M. W. 1978. “Process Selection for Optimal Management of Regional Wastewater Residuals,” Final Project Rep., Grant No.

PHOSPHORUS IN SEWAGE SLUDGE

159

ENV 75-15751-AO1, Nat. Sci. Found. Dep. Civil Eng., Univ. Delaware, Newark, Delaware, Dolar, S. G., Boyle, J. R., and Keeney, D. R. 1972. J . Environ. Qual. 1, 405-409. Ellis, B. G. 1973. In “Recycling Treated Municipal Wastewater Sludge Through Forest and Cropland” (W. E. Sopper and L. T. Kardos, eds.), pp. 46-70. Pennsylvania State Univ. Press, Univ. Park, Pennsylvania. Emmelin, L. 1973. “Sewage Treatment in Sweden, Environment. Planning and Conservation in Sweden.” Bull. No. 38, Swedish Information Service, New York. Farrell, J. B. 1974. Proc. Natl. Con$ Municipal Sludge Manage. pp. 5-10, Gerloff, G. C., and Krombholz, P. H. 1966. Limnol. Oceanogr. 11, 529-537. Gerloff, G. C., and Skoog, F. 1954. Ecolog.~35, 348-353. Gerloff, G. C., Moore, D. G . , and Curtis, J. T. 1964. Univ. Wisc. Agric. Exp. Sta. Res. Rep. No. 14. Gilkes, R. J., Young, R. C . , and Quirk, J . P. 1975. Aust. J . Soil Res. 13, 33-41. Goodroad, L. L., and Caldwell, A. C. 1979. J . Environ. Qual. 8, 493-496. Green, D. G., and Warder, F. G. 1973. Plant Soil 38, 567-572. Green, O., VanDam, D., LaBeau, B., Campbell, T. L . , and Daniels, S. L. 1974. “Ferric Chloride and Organic Polyelectrolytes for the Removal of Phosphorus,” EPA-67012-73-103. U.S. Environmental Protection Agency, Washington, D.C. Greenberg, A. E., and McGauhey, P. H. 1955. Soil Sci. 79, 33-39. Halderson, J. L., Lynam, B. T., and Rimkus, R. R. 1974. Proc. Natl. Cot$. Municipal Sludge Manage. pp. 129-133. Harnmond, L. C. 1973. Proc. Univ. Florida Workshop. Dep. Soil Sci. Hame, E. A,, Garman, W. H., and White, W. C. 1971. In “Fertilizer Technology and Use” (R. A . Olson, T. J. Army, J. J. Hanway, and V. J. Kilmer, eds.), p. 27-55. Soil Sci. SOC.Am., Madison, Wisconsin. Hecht, N. L., Duvall, D. S . , and Rashidi, A. S . 1975. “Characterization and Utilization of Municipal and Utility Sludges and Ashes,” EPA-67012-75-033b. U S . Environmental Protection Agency, Cincinnati, Ohio. Hinesly, T. D., and Jones, R. L. 1974. “Agricultural Benefits and Environmental Changes Resulting from Long-Term Use of Digested Sewage Sludge as Fertilizer for Field Crops and as an Amendment in Reclamation of Strip-Mine Spoil Banks.” Metropolitan Sanitary District of Greater Chicago, Dep. of Research and Development, Chicago. Illinois. Hinesly, T. C., Braids, 0. C., Dick, R. I., Jones, R. L., and Molina. J.-A.E. 1974. “Agricultural Benefits and Environmental Changes Resulting from the Use of Digested Sludge on Field Crops,” EPA15301SW-30d. I . U.S. Environmental Protection Agency, Washington, D.C. Hinesly, T. D., Jones, R. L., Ziegler, E. L., and Tyler, J. J . 1977. Environ. Sci. Techno/. 11, 182- 188. Hinesly, T. D., Ziegler, E. L., and Barrett, G. L. 1979a. J . Environ. Qua/. 8, 35-38. Hinesly, T. D., Sudarski-Hack, V., Alexander, D. E., Ziegler, E. L., and Barrett, G. L. 1979b. Cereal Chem. 56, 283-287. Hohla, G. N., Jones, R. L., and Hinesly, T. C. 1978. J . Environ. Qual. 7, 559-563. Howe, R. S . 1973. J . Soil Water Conserv. 28, 252-255. Hudgins, R. R., and Silveston, P. L. 1973. “Wet Air Oxidation of Chemical Sludges,” Rcs. Rep. No. 12 Ministry of the Environment, Environment Canada, Toronto, Ontario, Canada, Irving, C. E. 1960. Compost Sci. 1 (Spring), 18-20. Jansson, S. L. 1970. Siirtryck Hyg. Rev. 8, 350-356. Jansson, S . L. 1972. Stand Leistung Agrikulturchem. Agrarbiol. Forsch. 22, 61-66. Jones, J. B., Jr. 1967. Soil Sci. Soc. Am. Spec. Pub. No. 2, pp. 49-58. Jung, J., Isermann, K., and Henjes, G. 1979. Landwirtschafts Forschung 32, 262-274 (English summary).

160

M. B. KIRKHAM

Kadish, V. H. 1925. Trans. Inf. Conf. Sanit. Eng., London 1924 pp. 134-138. Keeney, D. R., Kelling, K. A., and Ryan, J. A. 1973. “Problems and Benefits of Land Disposal of Sewage Sludge,” Dep. Soil Sci., Univ. Wisconsin, Madison, Wisconsin. Kelling, K. A., Peterson, A. E., Walsh, L. M., Ryan, J . A. and Keeney, D. R. 1977a. J . Environ. Qual. 6, 339-345. Kelling, K . A., Walsh, L. M., Keeney, D. R., Ryan, J. A., and Peterson, A. E. 1977b. J . Environ. Qual. 6, 345-352. King, L. D., and Moms, H. D. 1973. J . Environ. Qual. 2, 411-414. Kirkham, M. B. 1975. Environ. Sci. Technol. 9, 765-768. Kirkham, M. B. 1976. Environ. Sci. Technol. 10, 285. Kirkham, M. B. 1980. Phyton Buenos Aires 38, 137-146. Kirkham, M. B., and Dotson, G. K. 1974. Proc. Natl. Conf. Municipal Sludge Manage. pp. 97- 106. Knezek, B. D., and Miller, R. H. (eds.). 1978. “Application of Sludges and Wastewaters on Agricultural Land: A Planning and Educational Guide,” p. 3.3, Gen. Services Admin. Rep. No. MCD-35. U S . Environmental Protection Agency, Washington, D.C. Lee, C. C., Soon, Y. K., and Bates, T. E. 1981. Can. J . Soil Sci. 61, 165-168. Lee, K. W., and Keeney, D. R. 1975. Water Air Soil Pollut. 5 , 109-1 12. Leeper, G. W. 1978. “Managing the Heavy Metals on the Land.” Dekker, New York. Levin, G. V . , and Shaheen, D. G. 1967. Biotechnol. Bioeng. 9, 457-470. Levin, G. V . , and Shapiro, J. 1965. J . Wafer Pollut. Control Fed. 37, 800-821. Levin, G . V . , Topol, G. J., Tarnay, A. G., and Samworth, R. B. 1972. 1. Water Pollut. Control Fed. 44, 1940-1954. Levin, G. V., Topol, G. J., and Tarnay, A. G. 1975. J . WaferPollut. Confrol Fed. 47, 577-590. Loehr, R. C. 1972. Proc. Cornell Agric. Waste Manage. Con$ pp. 41 1-427. Loehr, R. C., Jewell, W. J., Novak, J. D. Clarkson, W. W., and Friedman, G. S. 1979a. “Land Application of Wastes. Vol. I,” pp. 58, 76. Van Nostrand-Reinhold, New York. Loehr, R. C . , Jewell, W. J., Novak, J. D., Clarkson, W. W., and Friedman, G. S . 1979b. “Land Application of Wastes. Vol. 11,” pp. 24-39. Van Nostrand-Reinhold, New York. Loneragan, J. F. 1979. In “The Soil-Root Interface” (J. L. Harley and R. Scott Russell, eds.), pp. 351-367. Academic Press, New York. Lue-Hing, C., Lynam, B. T., and Peterson, J . R. 1974. “Digested Sludge Recycle to Land,” Rep. No. 74-21. Metropolitan Sanitary District of Greater Chicago, Dep. of Res. and Development, Chicago, Illinois. Lue-Hing, C., Lynam, B. T., Rimkus, R. R., and Peterson, J. R. 1975. “Digested Sludge Utilization in Agriculture and as a Soil Amendment,” Rep. No. 75-24. Metropolitan Sanitary District of Greater Chicago, Dep. of Res. and Development, Chicago, Illinois. Lund, L. J . , Page, A. L., and Nelson, C. 0. 1976. J . Environ. Qual. 5, 26-30. Lynam, B. T., Lue-Hing, C., Rimkus, R. R., and Neil, F. C. 1975. “The Utilization of Municipal Sludge in Agriculture.” The Metropolitan Sanitary District of Greater Chicago, Chicago, Illinois. Magdoff, F. R., Keeney, D. R., Bouma, J., and Ziebell, W. A. 1974. J . Environ. Qual. 3, 228-234. Malajczuk, N., McComb, A. J., and Loneragan, J . F. 1975. Aust. J . Bof. 23, 231-238. Martin, W. P., Fenster, W. E., and Hanson, L. D. 1970. In “Agricultural Practices and Water Quality” (T. L. Willrich and G. E. Smith, eds.), pp. 142-158. Iowa State Univ. Press, Ames, Iowa. Merz, R. C. 1956. In “Report on Continued Study of Wastewater Reclamation and Utilization,” Calif. State Water Pollution Control Board Pub. No. 15, San Diego County, California.

PHOSPHORUS IN SEWAGE SLUDGE

161

Milwaukee, Wisconsin. Sewerage Commission of the City of Milwaukee. 1971, “Phosphorus Removal with Pickle Liquor in an Activated Sludge Plant,” EPA 11010 FLQ 03/71. U.S. Environmental Protection Agency, Washington, D.C. Mitchell, M. J.. Hartenstein, R., Swift, B. L., Neuhauser, E. F., Abrams. B. I., Mulligan, R. M., Brown, B. A,, Craig, D., and Kaplan, D. 1978. J . Environ. Qual. 7 , 551-559. Mortvedt, J. J., Mays, D. A,, and Osborn, G. 1981. J . Environ. Qual. 10, 193-197. Nabos, J., Charoy, J., and Pichot, J. 1974. Agron. Trop. 29, I140-1l50. National Fertilizer Development Center. 1979. “1978 Fertilizer Summary Data,” Bull. Y-138. Natl. Fertilizer Development Center, Tennessee Valley Authority, Muscle Shoals, Alabama. Nishita, H., Haug, R. M., and Alexander, G. V. 1973. Plant Soil 39, 161-176. Olsen, S. R., and Barber, S. A. 1977. In “Soils for Management of Organic Wastes and Waste Waters,” pp. 196-215. Soil Sci. SOC.Am., Am. SOC.Agron., Crop Sci. SOC.Am., Madison, Wisconsin. Opferkuch, R. E., Ctvrtnicek, T., and Mehta, S. M. 1972. “Study of Utilization and Disposal of Lime Sludges Containing Phosphates,” Project No. 17070 FBE, Contract No. 14-12-199. Office of Res. and Monitoring, U.S. Environmental Protection Agency, Washington, D.C. O’Sullivan, J., Gabelman, W. H., and Gerloff, G. C. 1974. J . Am. SOC. Hort. Sci. 99, 543-547. Paccolat, B. 1979. Phosphore Agric. 75, 29-38. Page, A. L. 1974. “Fate and Effects of Trace Elements in Sewage Sludge When Applied to Agricultural Lands,” EPA-670I2-74-005. U.S. Environmental Protection Agency, Cincinnati, Ohio. Pammenter, N. W., and Woolhouse, H. W. 1975. Ann. Bor. 39, 337-346. Parker, F. W. 1953. In ‘‘Soil and Fertilizer Phosphorus in Crop Nutrition” (W. H. Pierre and A. G. Norman, eds.), pp. 401-426. Academic Press, New Yo&. Patrick, W. H., Jr., and Khalid, R. A. 1974. Science 186, 53-55. Pauly, H. 1973. Proc. Conf. Mineral. Inst., Denmark Inst. Technol.. 14rh, Lyngby Avfall 73. Pauly, H., and Simonsen, A. 1973. “22 Danish Sewage Plants’ Sludge Analysed for 29 Elements.” Mineralogical Inst., Technical Univ., Lyngby, Denmark. Peterson, J. R., and Gschwind, J. 1972. J . Environ. Qual. 1, 410-412. Peterson, J . R., Lue-Hing, C., and Zenz, D. R. 1973. In “Recycling Treated Municipal Wastewater and Sludge Through Forest and Cropland” (W. E. Sopper and L. T. Kardos, eds.), pp. 26-37. Pennsylvania State Univ. Press, University Park, Pennsylvania. Pierre, W. H. 1953. In “Soil and Fertilizer Phosphorus in Crop Nutrition” (W. H. Pierre and A. G. Norman, eds.), pp. ix-xi. Academic Press, New York. Pietz, R. I., Peterson, J. R., and Lue-Hing, C. 1974. “Groundwater Quality at a Strip-Mine Reclamation Area in West Central Illinois (Fulton County),” Rep. No. 74-18. The Metropolitan Sanitary District of Greater Chicago, Dep. of Res. and Development, Chicago, Illinois. Porcella, D. B., and Bishop, A. B. 1975. “Comprehensive Management of Phosphorus Water Pollution.” Ann Arbor Sci., Ann Arbor, Michigan. Pratt, P. F. 1977. In “Process Design Manual for Land Treatment of Municipal Wastewater,” Appendix B. EPA 62511-77-008 (COE EM1 110-1-501). U.S. Environmental Protection Agency, U.S. Army Corps Eng., U.S. Dep. Agric., Washington, D.C. Pratt, P. F., Bair, F. L., and McLean, G. W. 1964. Soil Sci. SOC.Am. Proc. 28, 363-365. Reeve, N. G., and Sumner, M. E. 1970. Soil Sci. SOC. Am. Proc. 34, 263-267. Regional Wastewater Solids Management Program. 1977. “LAIOMA Project. Sludge Processing and Disposal,” pp. 11-7. Regional Wastewater Solids Management Program, Los Angelesl Orange County Metropolitan Area, Whittier, California. Rehling, C. J., and Truog, E. 1939. Ind. Eng. Chem. 11, 281-283. Rudolfs, W. 1940. Water Works Sewage 87, 575-578.

162

M. B. KIRKHAM

Rudolfs, W., and Gehm, H. W. 1942. N.J. Agric. Exp. Sta. Bull. No. 699. Ryther, J. H., Dunstan, W. M., Tenore, K. R., and Huguenin, J. E. 1972. BioScience 22(3), 144- 152. Sabey, B. R., and Hart, W. E. 1973. Final Rep. Agron. Dep., Agric. Exp. Sru.,Cot. Srare Univ. Sabey, B. R., and Hart, W. E. 1975. J . Environ. Qual. 4, 252-256. Sawhney, B. L., and Norvell, W. A. 1980. Conn. Agric. Exp. Sta. Bull. No. 788. Schroeder, H. A., and Balassa, J. J. 1963. Science 140, 819-820. Scott, D. S., and Horlings, H. 1974. Con$ Proc. Ontario Min.Environ. Pollut. Control Brunch. Toronto pp. 413-443. Severson, R. C., and Gough, L. P. 1976. J . Environ. Qual. 5 , 476-482. Shapiro, J., Levin, G . V., and Zea, G. H. 1967. J . Water Pollut. Control Fed. 39, 1810-1818. Sheaffer, C. C., Decker, A. M., Chaney, R. L., and Douglas, L. W. 1979. J . Environ. Qrtal. 8, 450-454. Shipp, R. F., Wooding, N. H., and Baker, D. E. 1979. Pa. State Univ. Coop. Exr. Serv. Spec. Circ. No. 255, p. 3. Smilde, K. W. 1973. Plant Soil 39, 131-148. Smith, J. E., Jr., and Rosenkranz, W. A. 1975. “Municipal Sludge Management Research Program in the U.S.A.” US/USSR Seminar. Handling, Treatment and Disposal of Sludges, Moscow, USSR, May, 1975. U.S. Environmental Protection Agency, Cincinnati, Ohio. Sommers, L. E. 1977. J . Environ. Qual. 6, 225-232. Sommers, L. E., and Curtis, E. H. 1977. J . Water Pollut. Control Fed. 49, 2219-2225. Sommers, L. E., and Sutton, A. L. 1980. In “The Role of Phosphorus in Agriculture’’ (F. E. Khasawneh, E. C. Sample, E. J. Kamprath, eds.), pp. 515-544. Am. SOC.Agron., Crop Sci. SOC.Am., Soil Sci. SOC.Am., Madison, Wisconsin. Sommers, L. E., Nelson, D. W., Yahner, J. E., and Mannering, J. V. 1973. Proc. Indiana Acad. Sci. 82, 424-432. Sommers, L. E., Nelson, D. W., and Yost, K. J. 1976. J . Environ. Qual. 5 , 303-306. Soon, Y. K., Bates, T. E., and Moyer, J. R. 1978. J . Environ. Qual. 7, 269-273. Soon, Y. K., Bates, T . E., and Moyer, J. R. 1980. J . Environ. Qual. 9, 497-504. Stenstrom, T., and Vahter, M. 1974. Ambio 3 (2), 91-92. Taylor, J . M., Sikora, L. J., Tester, C. F., and Parr, J. F. 1978. J . Environ. Qual. 7, 119-123. Teletzke, G. H. 1966. Trans. Annu. Conf. Sanit. Eng. 16th, Bull. Eng. Architect. 56, 25-30. Tester, C. F., Sikora, L. J., Taylor, J. M., and Pam, J. F. 1979. 1. Environ. Qual. 8, 79-82. Touchton, J. T . , King, L. D., Bell, H., and Morris, H. D. 1976. J . Environ. Qual. 5 , 161-164. Trubnick, E. H., and Mueller, P. K. 1958. Sewage Ind. Wastes 30, 1364-1378. Truog, E. 1953. In “Soil and Fertilizer Phosphorus in Crop Nutrition” (W. H. Pierre and A. G. Norman, eds.), pp. 281-297. Academic Press, New York. U.S. Department of Agriculture. 1972. Interim Rep. by Agric. Res. Service for Maryland Environmental Service, District of Columbia, and U.S. Environmental Protection Agency. U.S. Department of Agriculture. 1978. “Improving Soils with Organic Wastes,” U.S. Government Printing Office Pub. No. 1979 0-623-484/770. U.S. Environmental Protection Agency. 1977. “Municipal Sludge Management: Environmental Factors,” Appendix VII, 2-3; Appendix VIII, 1. EPA 430/9-77-004. U.S. Environmental Protection Agency. 1978. “Municipal Sludge Agricultural Utilization Practices. An Environmental Assessment,” pp. 58, 98. SW-709. U. S. Environmental Protection Agency. 1979. “Process Design Manual for Sludge Treatment and Disposal,” pp. 4-8, 4-28. EPA 625/1-79-011. U.S. Salinity Laboratory Staff. 1954. Agric. Handb. No. 60. University of Guelph. 1975. “Land Disposal of Sewage Sludge. Vol. I1 (April, 1973-March,

PHOSPHORUS IN SEWAGE SLUDGE

163

1974),” Res. Rep. No. 24. Ministry of the Environment, Environment Canada, Toronto, Ontario, Canada. University of Guelph. 1976. “Land Disposal of Sewage Sludge. Volume 111 (April, 1974-March, 1975),” Res. Rep. No. 35. Ministry of the Environment, Environment Canada, Toronto, Ontario, Canada. University of Texas. 1971. “Phosphorus Removal and Disposal from Municipal Wastewater,” EPA 17010DYB02/71. U.S. Environmental Protection Agency, Washington, D.C. Valmari, A. 1974. Proc. Int. Symp. Forest Drain. Agric. Res. Center, Forest Res. Sta., Pelsonsuo, Finland pp. 361-364. Van Fleet, G. L., Barr, J. R., and Harris, A. J. 1974. J . Water Pollut. Control Fed. 46, 582-587. Vesilind, P. A. 1979. “Treatment and Disposal of Wastewater Sludges-Revised Edition.” Ann Arbor Sci., Ann Arbor, Michigan. Vlamis, J . , and Williams, D. E. 1972. Compost Sci. 13(1), 26-28. Wallace, A,, Romney, E. M., and Mueller, R. T. 1969. Phyton Buenos Aires 26, 151-154. Wallace, A,, EIGazzar, A , , and Alexander, G . V. 1973. Commun. SoilSci. PlantAnal. 4,343-345. Webber, L. R., and Hilliard, B. C. 1974. Conj: Proc. Ontario Min. Environ. Pollut. Control Branch, Toronto pp. 120-142. Webber, M. D. 1979. “Phosphate Fertilizer and Sewage Sludge Use on Agricultural Land-The Potential for Cadmium Uptake by Crops,” Technol. Development Rep. EPS 4-WP-79-2. Water Pollution Control Directorate, Environment Canada, Ottawa, Ontario, Canada. White, W. C. 1979. In “Situation 79,” pp. 36-39. Bull. Y-145. Natl. Fertilizer Development Center, Tennessee Valley Authority, Muscle Shoals, Alabama. Zenz, D., Peterson, J., Brooman, D., and Lue-Hing, C. 1974. “Digested Sludge Recycle to StripMine and Farmland-Effects on Public Health, Environmental Quality and Row Crops,” Rep. No. 74-22. The Metropolitan Sanitary District of Greater Chicago, Dep. of Res. and Development, Chicago, Illinois. Zwerman, P. J., and de Haan, F. A. M. 1973. Inst. Land Water Manage. Res. Wageningen Tech. Bull. No. 88.

This Page Intentionally Left Blank

ADVANCES IN AGRONOMY, VOL. 35

SUBTERRANEAN CLOVER IN THE UNITED STATES W. E. Knight,* C. Hagedorn,?V. H. Watson,?and D. L. Friesner? *Crop Science Research Laboratory, USDA-ARS, and ?Department of Agronomy, Mississippi State University and Mississippi Agricultural and Forestry Experiment Station, Mississippi State, Mississippi

I. Introduction. .

.........................

166

....................... .......................

169 170

.................................

171

V . Nitrogen Fixation . . . . . . . . . . . . A. Inoculation Responses. ........................................ B. Nitrogen-Fixation Assays ..........................

171

111. Breeding Subclover . . . . . . . . IV. Seed Characteristics . . . . . . . . A. Impermeability . . . . . . . . . . . . . . . . . . .

B. Sulfur.. ...............................

.................................

A. Soil Moisture.

183

A. Effects of Straw . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.......................................

186

..........................................

188

165

Copyright Q 1982 by Academic Press. Inc. All nghts of reproduction in any form reserved. ISBN 0-12-OOO735-5

166

W. E. KNIGHT ET AL.

I. INTRODUCTION Subterranean clover (Trifolium subterraneum L.) is a winter annual forage crop plant native to Mediterranean Europe, Asia, Africa, and southern England. Its leaves are similar in appearance to crimson clover (T. incarnatum L.). Flower stems develop in late winter and spring near the soil surface and 3-10 loosely attached, whitish florets are formed. A burr develops around the seed after pollination, and appendages that form on the burr pull the seed into the soil; hence the common name “subclover. ” It prefers a Mediterranean-type climate which features hot, dry summers and warm, moist winters, but excellent performance has been obtained in the humid southeastern United States (Fig. 1). Subterranean clover was introduced into Australia many years ago and then introduced into the United States, by way of Australia, in the early 1920s. The first seed of subclover grown in Texas was received by the Texas Agricultural Experimental Station in May 1921 (Leidigh, 1925). It was received with the request that it be tested in the field. Subclover gave good results for 3 years in succession at three different locations in southeastern Texas. The results of this

FIG. 1. Mixture of ryegrass and Mt. Barker subterranean clover on April 15, 1976, following the coldest winter on record and continuous heavy grazing by sheep at Mississippi State, Mississippi.

SUBTERRANEAN CLOVER IN THE UNITED STATES

167

field test suggested that it be sown during the month of September and that at least 224 kg acid phosphate/ha should be applied just before or after sowing. The study also showed that newly seeded areas should be properly cared for to ensure that the land is not grazed too heavily (Leidigh, 1925). The first planting in Oregon was made in 1922 at the Oregon Agricultural Experimental Station in Corvallis (Rampton, 1952). The field test showed that under Oregon climatic conditions, Mt. Barker, a midseason strain, and Tallarook, a medium-late strain, were well adapted (Rampton, 1952). In the early 193Os, subclover was introduced into California to improve both the quality and quantity of annual rangeland (Coats and Johnson, 1959b). At the same time, subclover was being tested in the southeastern states. Four years of testing at two Mississippi Agricultural Experiment Stations indicated that subclover was well adapted to the hilly sections of Mississippi (Coats and Johnson, 1959a). These tests suggested that subclover be planted between September 1 and October 15, fertilized with 60-100 Ib of phosphate, 25-50 lb of potash, and then limed to pH 6.0-6.5 (Coats and Johnson, 1959a). Varieties recommended for Mississippi are Mt. Barker, Woogenellup, Tallarook, and Nangeela. Other varieties with acceptable performance are Bacchus Marsh, Bassendean, and Howard (Knight et al., 1976). Mt. Barker and Bacchus Marsh were found to have greater disease resistance than other varieties tested (Coats and Johnson, 1959b). Another study in Mississippi compared five varieties of subclover grown with ryegrass, fescue, and alone, and indicated the great potential of this species in improving much of the grassland of Mississippi. In these studies, forage production and quality remained high until maturity, while reseeding was assured through both hard seed and seed dormancy (Knight and Watson, 1973).

II. POTENTIAL USE OF SUBCLOVER In recent years, a renewed interest has developed in the utilization of clovers in pastures. The energy crisis and subsequent high-priced mineral nitrogen has stimulated part of this interest. However, the greatest potential use of annual clovers is in grass-clover combinations as forage in pastuces, as a source of nitrogen, to improve forage quality, and to provide income from production of clover seed (Knight, 1974). It is estimated that Coastal and other bermudagrasses occupy 16 million ha in the southern region. Bermuda, dallis, bahia, and fescue are grown on an estimated 34 million ha. These grasses require a source of nitrogen for economic returns and suggest a tremendous potential use of legumes. Subclover is currently being used on approximately 60,000 ha of land in

168

W.E. KNIGHT E T A L .

western Oregon for livestock forage. Much of this is on hill land which does not have the potential to grow most other crops because of slope, soil depth, and lack of fertility (Mosher, 1976). Northwestern California has several thousand hectares planted to subclover for livestock feed. Currently, a small hectarage of subclover has been planted in the southeastern United States for cattle feed, and interest is increasing in the use of subclover in association with summer growing grasses in pasture areas (Fig. 2). Because of cold temperatures, there are only limited areas of the United States where subclover will survive, but within these areas it appears to have great potential. It will withstand heavy grazing and persists under very heavy utilization. In practical application, overgrazing is almost impossible, although heavy grazing seems to benefit the subclover at the expense of the grass. At the same time, utilization of the forage by livestock provides a significant return of nitrogen to the soil, which in turn benefits the associated grass. Some 800,000 ha in western Oregon could be converted from nonproductive land into excellent pastures for sheep and cattle, while in northwest California several million hectares of brushland and relatively unproductive native grasslands could benefit from the

FIG. 2. A volunteer stand of Mt. Barker subterranean clover overseeded on Coastal berrnudagrass (April 15, 1976).

SUBTERRANEAN CLOVER IN THE UNITED STATES

169

addition of a legume. Subclover appears to be an excellent choice for these areas (Mosher, 1976). Some research has also been done with subclover as a source of nitrogen in forest plantings. In a recent forestry study, sub- and crimson clovers were found to significantly increase sycamore seedling growth in northern Alabama (Haines et at., 1978). In this study, the purpose of the legume component was to provide weed control through good ground coverage, to maintain a low profile, to grow during the cool season so as not to compete with the sycamore seedlings, to provide cover for several years through the ability to reseed, and to make additional nitrogen available. Of the five legumes tested, sub and crimson gave good coverage, maintained a low profile, reseeded well, and increased the nitrogen content of the sycamore foliage in addition to improving height and volume of growth (Haines et al., 1978). Hagedorn et al. (1980) reported that Nangeela subclover produced 85-90% of unshaded yields under a 50% shade screen and was the most shade-tolerant of several legume species evaluated. A field study under a 25-year stand of longleaf pine in southern Mississippi yielded over 3200 kg clover dry matter/ha/year and nitrogen fixation rates of 250 kg N/ha/year. As well as an additional source of nitrogen, this forage increased the capacity for cool season livestock production and was also utilized by several different species of wildlife. Initial efforts have been started on “agroforestry” in Oregon, while current work in New Zealand and Australia gives us some promise in tree plantations in the Northeast and the Southeast. Subclover can provide nitrogen to the forage and the forest stand, and livestock can provide income between the initial tree planting and the closing of the canopy of the forest. In addition, putting the forage through animals can result in a much more rapid turnover of the nitrogen. Other uses of subterranean clover are in compliance with PL 500, Section 208 of the Clean Water Act, and in minimum tillage systems. With this act scheduled to go into effect in 1985, major use of subclover as a winter cover species for cropland could result. Because subclover dies in late spring, it might fit in a “no till” system, providing winter cover for the soil, supplying some nitrogen through fixation, and dying so as not to compete for moisture with summer crops (Fig. 3).

Ill. BREEDING SUBCLOVER In the Southeast, current breeding programs in Florida, Georgia, Mississippi, and Texas have emphasized plant introduction and evaluation for adaptation (Knight and Watson, 1977). At Mississippi State University, research to improve annual clovers by breeding began in 1958. Subterranean clover was added to the

170

W. E. KNIGHT E T A L .

FIG. 3. Subterranean clover is an ideal “no till” crop as shown in this vineyard. The clover suppresses weed growth, supports spray equipment in the winter, and provides nitrogen for the grapes.

improvement program in 1965. The objectives of this program were to select for good forage yield and distribution, reliable reseeding, and resistance to various diseases (Knight, 1974). However, five Australian cultivars appear to be adapted in their present state, and a number of the accessions evaluated offer a wide range of characteristics including morphology, maturity, and insect and disease resistance (Knight and Watson, 1977).

IV.

SEED CHARACTERISTICS A. IMPERMEABILITY

Impermeability exists when the seed coat prohibits imbibition of water that begins the process of seed germination. This is a desired characteristic in that it ensures reseeding stands of legumes after the initial seeding. Research was

SUBTERRANEAN CLOVER IN THE UNITED STATES

171

initiated at the University of California at Davis to study the amount of impermeable seed produced by subclover (Williams and Elliott, 1960). The impermeable seed percentage of subclover declined to a low level during the summer months after seed maturation at all locations except one site. This site had a marine influence, and a moderate amount of impermeable seed remained into the autumn months (Williams and Elliott, 1960). B. DORMANCY

Dormancy regulation in subclover was studied at Purdue University in an effort to determine whether the breaking of dormancy by carbon dioxide or by various types of poisons might be a result of enhanced ethylene formation following the treatments (Esashi and Leopold, 1969). This test established that carbon dioxide has a pronounced stimulatory effect on subclover seed germination, and that associated with the carbon dioxide stimulation of germination was a stimulation of ethylene production by the seed. The self-regulatory effect of ethylene may have an ecological benefit to the seed, encouraging germination when the seed is enclosed in soil medium, and depressing germination when it is exposed on the soil surface, even though adequate moisture is available (Esashi and Leopold, 1969). C. SEEDGERMINATION

Increasing legume seed germination by VHF and microwave dielectric heating was studied by the Nebraska Agricultural Experiment Station on Australiangrown legumes. The legume species were exposed to 39- and 2450-MHz electromagnetic fields and then examined for effects on germination and seed coat permeability (Nelson et al., 1976). Radiofrequency treatments appear to have practical potential for modifying hard-seed percentages in some species. In subclover, it appears likely that substantial amounts of hard seed can be rendered permeable by RF treatments.

V. NITROGEN FIXATION A. INOCULATIONRESPONSES

Failure of establishment or persistence of subclover is frequently due to lack of proper or effective nitrogen-fixing bacteria (Figs. 4 and 5 ) . Subclover requires

FIG. 4. Mississippi ecotype subterranean clover showing the importance of specific effective inoculum. Left to right, no inoculation, effective strain, and two ineffective strains (J. C . Burton photo, The Nitragin Company).

FIG. 5. Nodulation pattern from ineffective strain (left) and effective strain (right) (J. C. Burton photo, The Nitragin Company).

SUBTERRANEAN CLOVER IN THE UNITED STATES

173

inoculation with specific strains of Rhizobium trifolii, and much of the current microbiological research emphasizes the selection of more effective and dependable rhizobial strains. One study in California (Holland, 1970) was conducted to determine whether some failures with subclover were due to the inability of inoculum strains to form nodules in competition with naturalized rhizobia. It was found that the indigenous strains nodulated subclover but were of poor effectiveness with this host. Holland recommended that commercial inoculants be applied at rates greater than those suggested by the manufacturer to ensure effective nodulation at least in the establishment year. In California, it is often necessary to plant subclover under dry conditions, and the resulting frequent stand failures have been attributed to several factors. (1) The rhizobia die before there is adequate rainfall to insure seed germination. (2) Improper seed coating materials are used. (3) Effective nodulation requires large numbers of rhizobia (Jones et al., 1978). Field experiments demonstrated that subclover production increased as the amount of inoculum was increased, indicating a response to the quantity of rhizobia applied to the seed. The authors also demonstrated that 40% gum arabic in a peat inoculum with lime coating was superior to that applied as a water slurry. This was in agreement with an earlier report by Radcliffe ef a f . (1967) who found that R . trifolii survival was much greater in a peat suspension as compared to broth and that a wide variation in survival occurred among 13 seed coating materials. In Oregon, a survey was conducted on the effectiveness of native R . trifolii in an attempt to identify the causes of nodulation problems limiting subclover productivity (Hagedorn, 1978). At some locations, subclover became well established while at others the plants exhibited early nitrogen deficiency symptoms and died. Indigenous Rhizobium populations were found to nodulate subclover but were mostly ineffective with this host. Analysis of soil and rhizobia data yielded significant correlations between the size of the indigenous rhizobial populations and the soil texture and organic matter content, while effectiveness was associated with the level of base saturation. In a study on the nodulation of subclover by introduced rhizobia in western Oregon soils, Hagedorn (1979a) found that two highly effective strains successfully nodulated subclover at all field sites although there were differences in performance between sites. Fertilization (P + S + Mo) significantly increased the overall nodulation as well as the proportion of nodules occupied by the introduced strains. Both plant dry weights and total nitrogen levels reflected an additive effect of inoculation and fertilization. The two rhizobial strains were genetically tagged with antibiotic resistance markers so that in nodule occupancy tests they could be distinguished from each other and from indigenous strains. In a companion study (Hagedorn, 1979b), 50 R . trifolii strains recovered from Oregon soils were divided into two categories based on effectiveness and examined for resistance to 15 antibiotics. The strains were resistant to high concentrations of 1 1 antibiotics although resistance could

174

W. E. KNIGHT ET AL.

not be associated with any specific effectiveness category. The results indicated that there were several antibiotics which could serve in the selection of identifying markers for R. trifolii strains. B. NITROGEN-FIXATION ASSAYS

The acetylene reduction procedure has been widely used to determine rates of N, fixation although its application to intact systems in soil has been limited. Vaughn and Jones (1976) used acetylene reduction to evaluate N, fixation by annual rangeland species in California. They found that acetylene reduction assays of intact plants in potted soil could be performed with short incubations when the acetylene was forced through the soil. This method was not destructive, allowed for repeated samplings, and the results indicated that two naturalized clover species had equal potential for N, fixation as two of the widely planted introduced species. Assays on subclover nodules indicated large diurnal variation in acetylene reduction rates, and total N,-fixation estimates were in agreement with those reported from field studies. In a follow-up study, Williams et al. (1977) compared N,-fixation data from techniques based on the total N difference between clover and a grass species and the nitrogen A values in lysimeters. The relationship between the two methods was strongly linear although it appeared that total-N measurements were sufficient when N budgets were evaluated on a yearly production level. The A-value procedure seemed useful for more precise measurements of seasonal distribution in N,-fixation rates. Phillips and Bennett (1 978) compared acetylene reduction and '5N-enrichment procedures to measure N,-fixation rates in field plots of subclover and Bromus moflis L. The I5N technique was more accurate than acetylene reduction and more discriminating between the effects of various management schemes of N, fixation. It appeared that increasing the assay frequency would improve the performance of the acetylene reduction procedure, although this is not ideal for widely dispersed and poorly accessible rangeland plots. The 15N assay offered additional advantages in that site visits are required only for initial planting, fertilization, and one harvest. C . NITROGEN-FIXATION ENHANCEMENT

Recent attention has focused on the potential for improving legume production through selection of plant genotypes which are especially efficient in N, fixation and through development of a preferential plant selection for effective rhizobial strains. Such possibilities have been demonstrated in white clover (Hardason and

SUBTERRANEAN CLOVER IN THE UNITED STATES

175

FIG. 6 . Root system of a plant of Mississippi ecotype subterranean clover showing effective nodulation from an experimental Rhizobium strain.

Jones, 1979), alfalfa (Seetin and Barnes, 1977), and soybeans (Devine and Weber, 1977). Hagedorn and Caldwell (1981) reported that nodule occupancy differed from a 1:l ratio on 6 out of 12 subclover cultivars when two R . trifolii strains were used in equal proportions to inoculate the plants. For each cultivar that favored one of the strains, there were several plants that contained the strains in approximately a 1:l ratio. The deviation from this ratio was caused in each case by relatively few plants which exhibited a specific preference for one of the rhizobial strains. If such a feature could be developed on a cultivar basis, it would be very promising since many nodulation problems are caused by competition from poorly effective indigenous R . trifolii (Fig. 6).

VI. FERTILIZATION AND NUTRITION Much research has been done on the nutritive requirements of legumes. Because of the competition of grasses and weeds with legumes, a successful nutrition management plan must be developed for legumes where the legume is

176

W . E. KNIGHT ET AL.

favored and the associated plants do not become too competitive. A general investigation was conducted at the University of Kentucky on legume nutrition under some of the existing soil and climatic factors of the southern United States (Thomas, 1974). The conclusions were that (1) aluminum and manganese must be low for normal root growth; (2) molybdenum must be available for nitrogen fixation to occur; (3) a pound of fixed nitrogen makes the soil just as acid as a pound of applied nitrogen; (4) addition of nitrogen to a grass-legume mixture accomplishes nothing in total yield and in fact lowers legume yield; ( 5 ) grass and weeds compete strongly with legumes for potassium; (6) potassium is needed for reserve storage in the legume roots; (7) the critical level of potassium is not known, but in the early and young plants it is higher than in the later and older ones; (8) the critical level of phosphorus depends on whether the legume is cut and on its age; and (9) the sulfur requirement is about 1/15 of the nitrogen for both legumes and grasses (Thomas, 1974). A. NITROGEN

Research was conducted on the effect of nitrogen fertilizer on subclover stands at the Hopland Field Station in California. In 1967, several publications were released on the subject. The purpose of one of the investigations was to compare changes in botanical composition and protein levels of plants on an area fertilized with nitrogen with similar changes where subclover was established (Jones and Winans, 1967). It was found that the application of nitrogen increased the percentage of grass and depressed annual legumes. Subclover was very competitive with grass when the field was grazed or mowed. The establishment of subclover resulted in higher protein forage when the need was greatest during the dry season, compared with nitrogen fertilization which depressed protein levels during the same period (Jones and Winans, 1967). In another study, forage production and nitrogen uptake of a subclover-grass association were compared with those of annual-type grasslands fertilized with varying rates of nitrogen (Jones, 1967a). The results of this study indicated an advantage of subclover-grass pastures compared to nitrogen-fertilized annual grasslands where subclover can be established. Forage production and uptake of nitrogen were much more uniform throughout the growing season from subclover than from nitrogen fertilization. It appeared that where commercial nitrogen was applied, nitrogen w.as taken up rapidly, giving increased production and high concentrations of nitrogen within the plant early in the growing season. But when this was removed by grazing or mowing, little nitrogen remained for further growth. If the fertilized forage was left unclipped, a large increase in production was realized during the spring as well as the winter. However, this large volume of feed was low in quality since the percentage of protein in grass decreases with applied nitrogen (Jones, 1967a).

SUBTERRANEAN CLOVER IN THE UNITED STATES

177

A third study compared forage and protein and nitrogen uptake of a subclover-grass association with annual-type grasslands fertilized with varying rates of nitrogen (Jones, 1967b). Range-grass areas, including stands of subclover, produced forage yields equal to those from nitrogen-fertilized annual grasslands in a moisture-deficientyear, and more forage in a moisture-adequate year (Jones, 1967b) . Recently, additional studies have been conducted on the fate of applied nitrogen. In 1974, a study measured the effect of nitrogen applied to grass- subclover pastures on nitrogen leached into the drainage water, nitrogen uptake by the crop, and forage production (Jones et al., 1974). Nitrogen fertilization increased winter forage yields and nitrogen uptake. However, at the spring harvest, the subclover-grass pastures without applied nitrogen yielded just as much as the pastures with applied nitrogen. Apparent recovery of the applied nitrogen was 2% of the grass-subclover combination. The amount of apparent recovery of nitrogen fertilizer found in drainage water was 58%. Of total nitrogen leached, 94% occurred in the fall, 6% in the winter, and less than 1% in the spring (Jones et al., 1974). Results of another study published in 1977 dealt with the fate of applied nitrogen, especially that leached and that volatilized (Jones et al., 1977a). It was concluded that the recovery of applied nitrogen by subclover was poor, especially when applied in October. A single application of 500 kg/ha resulted in substantial leaching and gaseous losses (Jones et al., 1977a). A study of nitrogen-to-sulfur ratios in subclover tops published in 1977 compared six sulfur treatments, two nitrogen treatments, and three harvesting dates (Freney et al., 1977). While the total nitrogen:sulfur ratios varied widely from 4 to 33 1 depending on the plant part, age, nitrogen supply, and sulfur supply, there was no discernible trend in protein-nitr0gen:protein-sulfurratios attributable to any of these factors. The results also showed that it is difficult to extract all of the nonprotein nitrogen and sulfur from plant tissue. This could explain the variation in protein-nitrogen:protein-sulfurratios found by some workers (Freney et al., 1977). B. SULFUR

Many studies have investigated the mineral sulfur and its effect on subclover growth. In 1962, the sulfur requirements of subclover were studied at the Hopland Field Station (Jones, 1962a). Using a soil known to be sulfur deficient, it was concluded that the sulfate-sulfur concentrations can be used to identify sulfur-deficient plants but cannot be used to indicate the degree of deficiency (Jones, 1962a). In the same year, another publication dealt with the effects of sulfur supply on plant growth as related to total, organic, and sulfur concentrations within the plant parts at various growth stages of subclover (Jones, 1962b). The total sulfur and sulfate-sulfur concentrations in the plant did not change

178

W. E. KNIGHT E T A L

significantly over three growth stages. The first 20 lb/acre increased the yield and the organic sulfur concentration, but did not change the sulfate-sulfur concentration in the plant. More sulfate sulfur accumulated in the stems than petioles or leaflets at higher rates of applied sulfur, but at low rates there was little difference between the three parts (Jones, 1962b). A paper published in 1969 reported the effect of sulfur applied at different rates and particle sizes on the yield, sulfur uptake, and soil pH of subclover-grass pastures (Jones and Ruckman, 1969). Uptake of sulfur from the smallest-sized particles was highest during the first 2 years, indicating an annual decrease. With the 0.25- to 29-mm-sized particles, uptake was less than that from the smallest particles the first year, but the level remained relatively constant for 3 years, with only small decreases in subsequent years. Uptake of sulfur from the 0.30- to 0.59-mm-sized particles was less than half that from the smallest the first year after application, but there was no decrease with time. Increases in forage production with the application of 448 kg/ha sulfur did not follow the same pattern as sulfur uptake with regard to particle size. Where 45 kg sulfur/ha were applied, relative amounts of sulfur uptake from the various-sized particles were similar to that with the 448-kg application in the first year (Jones and Ruckman, 1969). In 1970, another report was published that compared the response of annual clover to soil additions of sulfur in combination with phosphorus (Jones et al., 1970). Ten varieties of subclover were used in the study and were grown on soil known to be deficient in sulfur and phosphorus. The response to sulfur alone was generally insignificant, but responses to phosphorus alone and to phosphorus and sulfur varied significantly. Differences in response to phosphorus and sulfur fertilization were as great among varieties within a species as between species. The highest-yielding clover varieties tended to have lowest percentages of phosphorus and sulfur. Differences in phosphorus and sulfur uptake by the clover varieties were significant; however, uptake responses to fertilization tended to be relatively smaller than yield response differences (Jones et al., 1970). In 1973, a study on the long-term effects of sulfur, phosphorus, and molybdenum fertilization on subclover was published (Jones and Ruckman, 1973). During the first 2 years of the study, highest forage yields occurred where only a combination of phosphorus and molybdenum was applied. The phosphorussulfur treatment produced less forage than did the phosphorus treatment alone. Responses to sulfur and molybdenum fertilization were small. In the following years, there was a definite response to sulfur when applied with phosphorus or with phosphorus and molybdenum. The yield response to all three elements diminished with time. The percentage of sulfur, phosphorus, or molybdenum in the subclover increased with the application of that element, but decreased when applied in combination with other elements, indicating a dilution effect (Jones and Ruckman, 1973).

SUBTERRANEAN CLOVER IN THE UNITED STATES

179

Additional research was conducted on the variations in total sulfur, sulfate sulfur, and total nitrogenkotal sulfur ratio, as related to plant age, nutrient status, and plant part in subclover when grown on a sulfur-deficient soil (Spencer et al., 1976). The concentrations of total sulfur and sulfate sulfur in the plant increased with increasing sulfur supply and decreased with plant age or with supplemental nitrogen. Overall, the laminae contained a higher concentration of total sulfur than the petioles when the optimal growth concentration was available. In general, the nitrogedsulfur ratio increased with the age of the plant or supplemental nitrogen and decreased with increasing sulfur supply. Critical concentrations of total sulfur and sulfate sulfur in the plant tissue were lowered by extra nitrogen; these fell markedly as the plant aged, but were not much affected by the maturity of the plant part (Spencer et al., 1976). Further investigations on sulfur and phosphorus fertilization evaluated the effect on yield, nutrient concentration, and critical phosphorus and sulfur levels in various growth stages of subclover (Drlica and Jackson, 1979). The form of sulphur applied had little effect on subclover; however, particle size greatly influenced uptake and critical levels. The phosphorus concentration of the grass-subclover forage at harvest was higher in plants that had received phosphorus fertilization. The concentration of phosphorus in the clover tissue remained relatively stable from mid-March to mid-April in both fertilized and unfertilized treatments. Sulfur fertilization increased yield 45% and more than doubled sulfur concentration of grass-subclover forage. Applied sulfur increased sulfur concentrations in clover and remained relatively constant from mid-March to mid-April. Addition of both phosphorus and sulfur was necessary for maximum nitrogen concentration in clover throughout the growing season (Drlica and Jackson, 1979). C. PHOSPHORUS In California, research was conducted on the effect of phosphorus on subclover where phosphorus deficiencies were widespread. In 1964, the effect of moisture stress on the concentration of phosphorus in subclover was studied (Wilson and Huffaker, 1964). In severely wilted plants, the concentration of phosphorus compounds decreased to less than half of that in plants with adequate moisture. The concentration of inorganic phosphorus, however, was not affected by moisture stress. Levels of most organic phosphorus compounds which had decreased as a result of moisture stress showed a marked increase during a 24-hr recovery period (Wilson and Huffaker, 1964). Another study concerning moisture stress and phosphorus was established under controlled conditions to better understand drought resistance in sub-, rose, and Spanish clovers and to determine the role that phosphorus might play in

180

W. E. KNIGHT ET AL.

response to soil moisture (Wilson et al., 1968). Applied phosphorus stimulated growth and decreased water use. Water use was higher in Spanish clover than in subclover, and therefore phosphorus does not contribute to its drought resistance (Wilson et al., 1968). In California, a study was designed to determine the changes in critical levels of phosphorus in subclover plant parts as influenced by clipping frequency and stage of growth (Jones et al., 1972). The plants were grown in phosphorusdeficient soils fertilized with increasing amounts of phosphorus until no additional increase in yield was obtained. Leaves were sampled from plants ranging from 48 to 151 days old. The critical level decreased from 0.61 to 0.11% phosphorus, and changes after 120 days were not significant. In each plant part, yields increased rapidly, and phosphorus percentage increased slowly up to the point of maximum yield (Jones et al., 1972). In another study, researchers investigated the relationship between microenvironment and phosphorus availability as it affects winter seedling growth of subclover on a soil known to be phosphorus deficient (Raguse and Evans, 1977). It was found that relatively small physical changes in the soil profile, the aspect of an exposed soil surface, phosphorus availability, the combination of seasonal climatic changes, and plant phenological development all interact to influence seedling survival, morphological development, dry matter accumulation, and phosphorus uptake and mobilization. It may be inferred that a brief midday period of favorable soil temperature permits near normal phosphorus uptake, even though the remainder of the day is unfavorably cold (Raguse and Evans, 1977). Subsequent research was designed to determine relationships between a known phosphorus-deficient soil, levels of applied phosphorus, and several parameters of plant growth in subclover (Osman et al., 1977). Shoot/root ratios, nodules per plant, and percent of crude protein in the shoot and root increased with available phosphorus. Marked differences associated with phosphorus levels were found in plant and leaf size and color, with optimal and vigorous growth at the highest phosphorus levels applied. Cotyledons of plants started on phosphorus-deficient soil remained small and yellowed early in development, while cotyledons of plants grown in adequate phosphorus increased in size and remained green (Osman et al., 1977). In 1979, another study was initiated to compare the establishment phase of growth of rose and subclovers in an environment with a common light and air temperature but with separate root-zone temperatures maintained in late fall and winter and with varying phosphorus levels (Raguse and Taggard, 1979). The leaf area of subclover was greater than that of rose clover and responded more to increasing phosphorus levels. Leaf areas tended to increase with higher soil temperatures. Nodule weight was positively correlated with leaf area, while leafarea regressions suggested that nodule development was adequate at 5°C and that

SUBTERRANEAN CLOVER IN THE UNITED STATES

181

nodule development was primarily related to time rather than to morphological development of the shoot. This suggests that while the practical soil threshold temperature for overall growth is near 5"C, some intraplant growth processes may proceed only if mineral nutrients are available (Raguse and Taggard, 1979). Certain reproductive anomalies in livestock seem to be associated with the estrogenic property of forage legumes. The estrogenic property of subclover is mainly related to its content of formononetin. A study was conducted to test the hypothesis that the nutrient supply of subclover may directly or indirectly affect the estrogen levels (Schoo and Rains, 1971). In this test, the effect of six rates of phosphorus and three of sulfur on the level of the estrogenic formononetin was evaluated. With increasing rates of fertilizer, phosphorus, and sulfur, the concentration of formononetin in leaf blades decreased significantly. The phosphorus and sulfur interaction was not statistically significant. It was concluded that the decrease in formononetin concentration may have resulted from a dilution effect and/or a decrease in isoflavone synthesis (Schoo and Rains, 1971).

D. MOLYBDENUM At the Oregon Agricultural Experiment Station, a study was done to determine the effect of molybdenum application on subclover in terms of yield and nutrient status and to examine an anion exchange resin as a means of assaying plant available molybdenum (Dawson and Bhella, 1972). Highly significant differences for yield and relative total nitrogen uptake were obtained following molybdenum fertilization. Generally, greatest yield as well as total nitrogen uptake from applied molybdenum occurred at the lower anion-exchangeable soil molybdenum levels. There was no significant change in plant copper concentrations following molybdenum fertilization. However, molybdenum concentration in subclover was higher in plants grown on treated soils than on untreated (Dawson and Bhella, 1972). E. CALCIUM

Serpentine soils are widespread in California and are characteristically deficient in nitrogen, phosphorus, potassium, calcium, and molybdenum as well as being gravelly and shallow to bedrock. Other problems include a high magnesium content and possible toxicities of nickel and chromium. Research was done in serpentine soils to determine the effect of growth stage, plant part, and defoliation treatment on the critical calcium concentrations and calciurdmagnesium ratios in subclover (Jones et al., 1976). It was found that percent calcium and calciudmagnesium ratios in new and old stems were insensitive as measures of

182

W.E. KNIGHT ET AL.

calcium status since these values changed less in stem than in leaves overa wide range of calcium levels. Percent calcium in young leaves was the best measure of calcium status because critical values varied less because of stage of growth or clipping than did percent calcium in old leaves or calciudmagnesium ratios in young or old leaves (Jones et al., 1976). Jones et al. (1977b) assayed the fertility status of serpentine soil and compared several criteria for diagnosing the calcium status of subclover. It was found that for the growth of subclover, the soils were very responsive to phosphorus and sulfur. Many others were also responsive to calcium, potassium, and molybdenum. The calcium status of subclover can be assayed by the percentage of calcium or calciudmagnesium ratio in clover tops, and by the exchangeable soil calcium expressed as exchangeable, or as percent saturation, or as calcium/ magnesium (Jones et al., 1977b). F. LIMING

It is well known that applying lime to acid soils increases the pH. Research shows that an acid soil can interfere with the absorption of essential minerals and formation of effective nodules, therefore, creating deficiency symptoms in the plant. Field experiments were established to determine the effect of limestone pelleting on nodulation (Williams and Kay, 1959). The heavier concentration of limestone resulted in a positive-yield response in subclover when associated with commercial inoculum (Williams and Kay, 1959). Researchers at Oregon State University studied the possible effects of lime with different materials on the establishment of subclover (Jackson, 1972). The general conclusions were: (1) the pH of an area immediately surrounding a granule of fertilizer may be critical in the establishmentof successful nodulation; (2) the application of phosphorus-improved seedling vigor; (3) low rates of nitrogen applied at planting gave marked increases in seedling size; (4) ammonium nitrate supplied at planting may reduce effective nodulation; (5) diammonium phosphate or magnesium ammonium phosphate gave a marked increase in good nodulation; (6) mixing lime and phosphorus together may be beneficial; and (7) increases in nodulation and yield were measured from applications of lime (Jackson, 1972). An experiment was done with Hawaiian Oxisol that tested the lime response in 22 legume varieties (Munns and Fox, 1976). Liming soils to a pH above 6.0 sometimes depressed plant growth, but most reported cases had been observed in experiments of short duration. Early growth or nodulation of certain species was depressed when soils were limed at rates above 6 tons/ha. In subclover, the depression was only temporary (Munns and Fox, 1976). In 1977, a report was published that described the lime requirements of tropi-

SUBTERRANEAN CLOVER IN THE UNITED STATES

I83

cal and temperate legumes (Munns and Fox, 1977). Lime response curves showed no distinct general difference between tropical and temperate legumes, but species varied within each group. Subclover was middle ranked according to the amount of lime needed for 90% of the maximum attained yield and lower ranked according to the magnitude of yield increase due to lime (Munns and Fox, 1977). A similar study compared the influence of lime on nitrogen fixation in tropical and temperate legumes (Munns et af., 1977). Again, no general difference could be found between them. Increased nodulation and plant nitrogen content were consistent for most species associated with liming (Munns et al., 1977). At the Oregon Agricultural Experiment Station, experiments were conducted on the effectiveness of seed pelleting, and on the use of small amounts of lime with certain fertilizers for promoting effective nodulation, and on the effect on yields when various lime levels in the presence of phosphate and molybdenum were used (McGuire et af., 1978). Effective nodulation in subclover was obtained by applying 2-3 tons of lime per acre or by drilling the inoculated seed with a lime-superphosphate mixture which served as a neutral medium for the survival of rhizobia. No evidence was obtained to suggest a response to the addition of molybdenum (McGuire et al., 1978).

VII. CLIMATIC VARIATIONS A. SOILMOISTURE

The effect of climatic variations on the establishment of subclover in California annual rangelands has been studied at the University of California at Davis. Often there are variations in the temperature and the amount and time of rainfall in this area during the fall season. The purpose of one experiment was to determine the effect of the planting date and variations in rainfall and temperature on the establishment of subclover (Jones et al., 1971). Healthy subclover plants were produced when the mean ambient air temperature in the 6 weeks following germination was between 49 and 62°F. Seed planted about 1 month before a rain produced good stands of subclover, indicating that the inoculum survived until the rains came. It was found that more vigorous clover grew from seed which was in the ground at the time of the first rain than from seed planted soon after the rain (Jones et al., 1971). Gerakis et al. (1975) determined the effects of moisture stress and population density on the tissue concentration of major nutrients and related these to principles of plant-water relations. Three winter annuals were tested, and subclover appeared to be the least sensitive to soil moisture stress. The reduction in total

184

W. E. KNIGHT ETAL.

biomass was significant when soil water potential was decreased from field capacity to - 3 bars. Increases in plant population density increased the total biomass of subclover significantly (Gerakis et al., 1975). Taggard et al. (1976) measured seedling development, forage yield potential, and management problems of winter annuals when supplemental irrigation was used to start growth in late summer. Forage yields from August and September plantings were similar but were greater than those from the October planting. Subclover planted and irrigated in August was subjected to daylengths which promoted fall flowering and increased the incidence of summer insects, primarily beet armyworm. It was concluded that early fall irrigation will ensure vigorous plant growth at temperatures under which maximum root development and nodulation can occur. During the summer of 1974, northern California experienced an unusual summer rain that provided an opportunity to observe germination responses of annual legumes resistant to summer germination. Subclover was found to be one of the few species that germinated in response to the summer rain, even though seedling densities were low compared to emergence under normal fall rains (Raguse et al., 1977). B. TEMPERATURE

In 1970, research was initiated on the effect of temperature on the establishment and response of subclover cultivars grown under varying temperature (Young et al., 1970). The germination of seeds of 11 cultivars of subclover grown in California was depressed by high temperatures, depressed by low temperature, or remained dormant or insensitive with temperatures of 0.5-20°C (Young et al., 1970). In 1972, another study reported the influence of ambient temperature on the early growth and development of annual legumes suitable for introduction into areas of a Mediterranean climate (Sumner et a!., 1972). Subclover was grown from seed to the 7- to 10-leaf stage using combinations of 10 and 20°C. Shoot ambient temperature appeared to be less important than root ambient temperature as a determinant of growth rate (Sumner et al., 1972). The effects of seed aggregation, availability of nutrients, and temperature on the rate of germination and emergence of subclover seeds were determined at the University of California at Davis (Caminos et al., 1973). Germination percentage was increased and emergence was accelerated when two or more subclover seeds were planted in close proximity instead of singly, and enhanced when a nutrient solution was applied. Raising the temperature from 10 to 15°C and from 15 to 20°C either reduced the effects of seed aggregation and nutrient availability or caused the effects to be evidenced during a shorter time interval. Since the influences of close seed-spacing arid of improved nutrient availability were more pronounced at cool temperatures, this may be an important condition for stand

SUBTERRANEAN CLOVER IN THE UNITED STATES

185

establishment from new seedings when rainfall adequate for germination occurs later than normal (Caminos et al., 1973). Guerrero and Williams (1975) evaluated the response of subclover to temperatures that reflect the conditions of the annual grassland type during the seedling stage. Subclover showed a linear increase in relative growth rate with temperatures up to 15 and 25°C. The results indicated that grasses were favored more than subclover when late fall rains occurred along with lower temperatures. Nitrogen-fixation response to variation in both light and temperature in subclover was determined by Eckart and Raguse (1980). Under constant temperature and either normal photoperiods or constant light, no significant diurnal fluctuation of acetylene reduction was measured. Under both constant temperature treatments, a gradual increase in acetylene reduction activity was observed. These results show that diurnal changes in acetylene reduction by subclover result more from fluctuations in temperature than from diurnal changes in light and suggest that nitrogen fixation by root nodules is buffered against short-term changes in photosynthate supply.

VIII. ESTABLISHMENT AND MANAGEMENT In areas where subclover is grown, early grazing and mowing are practiced to remove excessive growth and reduce lodging. Studies were initiated on the effect of mowing on the productivity of subclover at the University of California to determine the effect of clipping intervals on the botanical composition of a subclover-grass-forb pasture and lo obtain the maximum percentage of clover (Murphy et af., 1970). It was found that no consistent difference in the botanical composition of subclover resulted from clipping intervals after the first 2 years. The unclipped treatment consisted of mostly grass while the percentage of subclover diminished each year until none was present in the sixth year. Clipping reduced the percentage of grass, and the interval did not make any consistent difference. The yields taken in the sixth clipping year showed only minor differences due to clipping intervals. In a study in Mississippi, the influence of spring clipping on reseeding and the productivity of annual clovers was evaluated in a coastal bermudagrass sod (Knight, 1971). Subclover was found to reseed and produce well under three spring mowing treatments. A. EFFECTSOF STRAW

In 1953, a report was published discussing the problem of establishing subclover in the north-coast counties of California. Only when an application of unthreshed subclover straw was tried did a good stand of dark green virorous plants appear (Williams et al., 1953). Three characteristics of the straw were

186

W. E. KNIGHT E T A L .

considered to be potentially important in obtaining a stand of subclover: (1) the large amount of seed contained in the straw, (2) the mulching effect, and (3) the presence of legume bacteria. Another study was initiated to determine the beneficial effects of the subclover straw (Williams et al., 1954). The data showed that only 3-6% of the plants obtained from the plots receiving the seed only or the mulch appeared to have been effectively nodulated, while plants taken from plots receiving inoculation or straw were 58-67% effectively nodulated. Hence, the resident bacteria in the soil were ineffective or parasitic strains, whereas the organisms supplied in the inoculation culture or straw were highly effective (Williams et a/., 1954). B . EFFECTSOF FUNGICIDES

In 1960, the effect of six commercial fungicides on rhizobia and on its ability to inoculate subclover was reported (Williams et al., 1960). When a commercial subclover inoculum was applied to the seed, clover grew well on plots receiving no fungicide, but there was a significant decrease in yield with all six fungicide treatments. Until a fungicide that is nontoxic to rhizobia is found, it is recommended that the seed be inoculated but not treated with fungicide.

c.

EFFECTS OF HERBICIDES

Researchers have evaluated the effects of herbicides on pasture and rangeland vegetation almost since modem herbicides have been available. The large number of chemicals available and widely differing composition of swards in various climatic zones limit the value of generalized statements, and most often usefulness of a herbicide must be determined for specific situations. Since persistence and selectivity of herbicides is influenced by rate of application, climatic factors, and species present, the use of herbicides for pasture improvement should be treated in the context of the total management program. The concern with weeds in subclover production is focused on the two periods of germination and establishment and on the prevention of annual grass dominance. Campbell (1974) evaluated effects of glyphosphate on germination and establishment of alfalfa, subterranean clover, perennial ryegrass, and hardinggrass sown on bare soil before spraying, or surface-sown into an established pasture. The herbicide had little or no effect on germination of surface-sown seeds, but showed deleterious effects in subsequent establishment and growth, particularly on the legumes. They concluded that the seed and herbicide should not come in contact. In practice, this can be achieved by applying the herbicide prior to seeding, or if spraying and seeding simultaneously, by designing the equipment so that the seed and herbicide do not come in contact. In a legume herbicide-tolerance study, pure stands of white clover (Trifoliurn repens L.), crimson clover (T. incurnuturn L.), red clover (T. pratense L.),

SUBTERRANEAN CLOVER IN THE UNITED STATES

187

arrowleaf clover (T. vesicufosurn Savi .), subterranean clover (T. subterruneum L.), and vetch (Viciu sutivu L.) were treated with varying rates of 2,4-D, dicamba, or a 1:3 (w:w) mixture of 2,4-D and dicamba in the spring of 2 consecutive years (Watson et al., 1980). Phytotoxicity estimates from 2,4-D ranged from 0 for subterranean clover to 100% for vetch. Damage to stands from applications of dicamba or of 2,4-D and dicamba combinations was moderate (40%) to severe (100%).Species reactions to the various chemicals suggest their tolerances to the chemicals tested to be subterranean > white > crimson > red > arrowleaf > vetch. In Oregon, subclover has been shown to tolerate up to 1 lb MCPNacre (Mosher, 1976). Kerb has been used successfully at %-1 Ib/acre to control grasses. Several researchers have pointed out the need for controlled grazing or clipping following seeding on a renovated pasture or meadow (Norman and Green, 1957; Robinson and Cross, 1960). Such grazing or mowing would be needed following the use of herbicide such as paraquat which gives rapid burndown of the vegetation followed by complete or partial recovery, or with band spraying. Such regrowth would provide competition to the developing seedlings. These points are emphasized in a 1968 study that measured the effect of spraying on the immediate and future composition of the pasture (Kay, 1968). It was found that spraying with paraquat should be confined to the cold winter period of minimal growth and should not be sprayed after the period of rapid growth in the spring. Species composition was improved by all spraying treatments. Kay and Owen (1970) compared the effectiveness of band spraying and broadcast spraying to no weed control on establishment of a grass-subclover pasture. The pasture was established by seeding immediately after spraying the resident vegetation with paraquat. The broadcast spray system was found to be superior to the band spraying system and to no weed control since the grass forage did not establish itself without some weed control. Other considerations are pointed out by Warboys and Johnson (1966) who found that the fertilizer applied during renovation also enhanced the growth and competitiveness of the old sward vegetation that survived the renovation. It is also suggested that appropriate stock management must also follow pasture renovation if permanent improvement is to result (Robinson and Cross, 1960).

IX. MORPHOLOGICAL CHARACTERISTICS A. SEEDWEIGHT

A study was conducted in Australia, funded by a Fulbright Research Fellowship, on the effect of seed weight on development of the photosynthetic surface and the accumulation of dry matter in shoots during the vegetative phase

188

W. E. KNIGHT ET AL.

(Williams et al., 1968). The effect of seed weight on the competitive abilities of crimson and subclover was studied in simple and mixed communities. The yield of each component of the mixture was dependent mainly on its own seed weight and that of its associate. The effect of seed weight on growth was essentially linear (Williams et al., 1968). Raguse and Fianu (1971) determined the effect of seed weight on seedling development at the University of California at Davis and produced evidence that one parameter of seedling development was independent of seed size. The time-rate of appearance of trifoliolate leaves of subclover was studied over a wide range of seed weights. Seedling development of all seed sizes at 25°C could be expressed as a regression equation Y = - 1.96 - 0.34X, where Y equals the stage of plant.development and X equals the number of days from germination. This indicates that the resuIts for all seed sizes were similar. B . DEVELOPMENTAL MORPHOLOGY

A modification of the developmental morphology index proposed by Carlson in 1966 was proposed at the University of California at Davis, which provided a precise description of earlier stages of plant development (Raguse el al., 1970). Ten morphological stages of plant development from emergence to full expansion of the unifoliolate leaf were described for subclover. This system provides a precise morphological description of plant deveIopment when a high degree of experimental resolution is required. Another modification of the developmental morphology index was made dealing with the seedling leaves of subclover (Raguse er al., 1974). Ten morphologic stages of leaf development were described for seedling subclover leaves. This system has proved to be useful in field, greenhouse, and controIled-environment studies that require a rapid, nondestructive means of monitoring growth response to change in plant environment (Raguse et al., 1974).

X. SUMMARY Perhaps at no other time in the history of the United States has competition for land resources been so keen among the commodity crops. In many cases, this means that pasture crops will be grown on thinner, marginal soils which are undesirable for production of cotton, soybeans, corn, wheat, and peanuts. The greatest opportunity for success on these soils is to maximize forage production in the fall, winter, and spring, when rainfall is usually adequate, when drought is less likely to occur, and when temperatures are generally more favorable for succulent growth. In many cases, subterranean ciover is the ideal species for economical forage production.

SUBTERRANEAN CLOVER IN THE UNITED STATES

189

In recent years, interest has developed in the utilization of clovers in pastures. The energy crisis and subsequent high-priced mineral nitrogen have stimulated part of this interest. However, a desire to develop and use forage legumes was growing before the energy crisis. Producers and scientists have been emphasizing better quality forage with better seasonal distribution. They have been developing grazing systems in which the animal harvests most of the feed consumed with minimum amounts of feed harvested, stored, and fed back. There has also been an emphasis on better use of our land resources to produce food and fiber. Supporting this research, recent economic studies have stressed that yield alone does not necessarily make a practice economical, but instead it is the amount of quality forage (measured in energy units) consumed and converted into animal production that is important. An abundance of high-quality forage with good seasonal distribution is the foundation for livestock profits. Labor and machinery costs involved in the production, handling, storage, and preservation of food for livestock continually increase. Systems of year-around grazing that permit the animal to harvest most of the feed consumed should result in economical production of livestock products. In areas where subterranean clover is adapted, the proper use of this species in association with warm and cool season grasses can provide an economical system of grass-legume mixtures for high quality forage production. REFERENCES Caminos, J., Raguse, C. A,, and Sumner, D. C. 1973. Agron. J. 65, 1002-1003. Campbell, M. H. 1974. Aust. J . Exp. Agric. Anim. Husb. 14, 557-560. Coats, R. E., and Johnson, C. M. 1959a. Miss. Agric. Exp. Sta., Inf. Sheet 647. Coats, R. E., and Johnson, C. M. 1959b. Miss. Farm Res., Miss. Agric. Exp. Sta., October p. 7. Dawson, M. D., and Bhella, H. S. 1972. Agron. J . 64, 308-311. Devine, T. E., and Weber, D. F. 1977. Euphytica 26, 527-535. Drlica, D. M., and Jackson, T. L. 1979. Agron. 1.71, 824-828. Eckart, J. F., and Raguse, C. A. 1980. Agron. J. 72, 519-523. Esashi, Y., and Leopold, A. C. 1969. Plant Physiol. 44, 1470-1472. Freney, J. R . , Spencer, K., and Jones, M. B. 1977. Comrnun. Soil Sci. Plant Anal. 8, 241-249. Gerakis, P. A,, Guerrero, F. P., and Williams, W. A. 1975. J. Appl. Ecol. 12, 125-135. Guerrero, F. P., and Williams, W. A. 1975. Crop Sci. 15, 553-556. Hagedom, C. 1978. Soil Sci. SOC.Am. J . 42, 447-461. Hagedom, C. 1979a. SoilSci. Soc. Am. J . 43, 515-519. Hagedom, C. 1979b. Soil Sci. Soc. Am. J . 43, 921-925. Hagedom, C., and Caldwell, B. A. 1981. Soil Sci. Soc. Am. J . 45, 513-516. Hagedorn, C., Watson, V. H., and Knight, W. E. 1980. In “Southern Forest Range and Pasture Resources Symposium” (R. D. Child and E. K. Byington, eds.), pp. 143-145. Winrock, Morrilton, Arkansas. Haines, S. G., Haines, L. W., and White, G. 1978. Soil Sci. Soc. Am. J . 42, 130-132. Hardason, G., and Jones, D. G. 1979. Am. Appl. Riot. 92, 329-333. Holland, A. A. 1970. Plant Soil 32, 293-302. Jackson, T. L. 1972. Oreg. Agric. Exp. Sta. Circ. lnf. 634.

190

W. E. KNIGHT ET AL.

Jones, M. B. 1962a. Calif. Agric. 16, 4-5. Jones, M. B. 1962b. Soil Sci. Soc. Am. Proc. 26, 482-484. Jones, M. B. 1967a. Agron. J. 59, 209-214. Jones, M. B. 1967b. CaliJ Agric. 21, 4-7. Jones, M. B., and Ruckman, J. E. 1969. Agron. J. 61, 936-939. Jones, M. B., and Ruckman, J. E. 1973. Soil Sci. 115, 343-348. Jones, M. B., and Winans, S. S. 1967. J . Range Mange. 20, 8-12. Jones, M. B., Lawler, P. W., and Ruckman, J . E. 1970. Agron. J. 62, 439-442. Jones, M. B., Lawler, P. W., and Murphy, A. H. 1971. J . Range Manage. 24, 147-150. Jones, M. B., Ruckman, J. E., and Lawler, P. W. 1972. Agron. 1. 64, 695-698. Jones, M. B., Street, J. E., and Williams, W. A. 1974. Agron. J . 66, 256-258. Jones, M. B., Vaughn, C. E., and Harris, R. S. 1976. Agronomy 68 756-759. Jones, M. B., Delwiche, C. C., and Williams, W. A. 1977a. Agron. J . 69, 1019-1023. Jones, M. B., Williams, W. A., and Ruckman, J . E. 1977b. Soil Sci. SOC. Am. J . 41, 87-89. Jones, M. B., Burton, J. C., and Vaughn, C. E. 1978. Agron. J. 70, 1081-1085. Kay, B. L. 1968. Weed Sci. 16, 66-68. Kay, B. L., and Owen, R. E. 1970. Weed Sci. 18, 238-243. Knight, W. E. 1971. Agron. J. 63, 418-420. Knight, W. E. 1974. Proc. So. Pasture Forage Crop Improv. Conf.,31sr pp. 145-149. Knight, W. E., and Watson, V. H. 1973. Proc. Assoc. So. Agric. Workers p. 61. Knight, W. E., and Watson, V. H. 1977. Proc. Farm Seed Conf.,23rd pp. 8-26. Knight, W. E., Palmertree, H. D., and Watson, V. H. 1976. Miss. Agric. Forest. Exp. Sta. Inf. Sheet No. 1268. Leidigh, A. H. 1925. Tex. Agric. Exp. Sfa. Circ. No. 37, pp. 1-12. McGuire, W. S., Dawson, M. D., and Crofts, F. C. 1978. Oreg. Agric. Exp. Sru. Bull. pp. 1-34. Mosher, W. D. 1976. Proc. Hillands Int. Symp. Munns, D. N., and Fox, R. L. 1976. Plant Soil 45, 701-705. Munns, D. N., and Fox, R. L. 1977. Plant Soil 46, 533-548. Munns, D. N., Fox, R. L., and Koch, B. L. 1977. Plant Soil 46, 491-601. Murphy, A. H . , Jones, M. B., and Love, R. J. 1970. J. Range Manage. 23, 196-199. Nelson, S. O., Ballard, L. A. T., Stetson, L. E., and Buchwald, T. 1976. Trans. Am. SOC. Agric. Eng. 19, 369-371. Norman, M. J. T., and Green,J. 0. 1957. J . Br. Grussl. SOC.12, 74-80. Osman, A,, Raguse, C. A., and Sumner, D. C. 1977. Agron. J . 69, 26-29. Phillips, D. A., and Bennett, J. P. 1978. Agron. J . 70, 671-674. Radcliffe, J. C., McGuire, W. S., and Dawson, M. D. 1967. Agron. J . 59, 56-58. Raguse, C. A,, and Evans, R. A. 1977. Agron. J . 69, 21-25. Raguse, C. A,, and Fianu, F. K. 1971. J. Range Manage. 24, 385-387. Raguse, C. A., and Taggard, K. L. 1979. Agron. J . 71, 523-528. Raguse, C. A., Fianu, F. K., and Menke, J . W. 1970. Crop Sci. 10, 723-724. Raguse, C. A,, Menke, J . W., and Sumner, D. C. 1974. Crop Sci. 14, 333-334. Raguse, C. A., Young, J. A., and Evans, R. A. 1977. Agron. J . 69, 327-329. Rampton, H. H. 1952. Oreg. Agric. Exp. Sra. Bull. No. 432, pp. 3-12. Robinson, G. S . , and Cross, M. W. 1960. Proc. Int. Grassl. Conf. 8fh, Hurley, pp. 402-405. Schoo, H., and Rains, D. W. 1971. Crop Sci. 11, 716-718. Seetin, M. W., and Barnes, D. K. 1977. Crop Sci. 17, 783-787. Spencer, K., Jones, M. G., and Freney, J . R. 1976. Aust. J. Agric. Res. 28, 401-412. Sumner, D. C . , Raguse, C. A., and Taggard, K. L. 1972. Crop Sci. 12, 517-520. Taggard, K. L., Delmas, R. E., and Raguse, C. A. 1976. Agron. J . 68, 674-677. Thomas, G. W. 1974. Proc. So. Pasture Crop Imp. Conf.,31st pp. 234-247.

SUBTERRANEAN CLOVER IN THE UNITED STATES

191

Vaughn, C., and Jones, M. B. 1976. Agron. J. 68, 561-564. Warboys, I. B., and Johnson, R. J. 1966. Exp. Agric. 2, 309-316. Watson, V. H., Knight, W. E., and Strachan, W. F. 1980. Proc. So. Branch Am. Soc. Agron. 7, 5. Williams, W. A,, and Elliott, J. R. 1960. Ecology 1, 733-742. Williams, W. A , , and Kay, B. L. 1959. J . Range Manage. 12, 205-206. Williams, W. A , , Lenz, J. V., and Murphy, A. H. 1953. C a l f . Agric. 7 , 3, 16. Williams, W. A , , Lenz, J. V., and Murphy, A. H. 1954. Agron. J. 46, 95”6. Williams, W. A,, Hanvood, L. H., and Hills, F. J. 1960. Agron. J. 52, 363-365. Williams, W. A., Black, J. N., and Donald, C. M. 1968. Crop Sci. 8, 660-663. Williams, W. A , , Jones, M. B., and Delwiche, C. C. 1977. Agron. J. 69, 1023-1024. Wilson, A . M., and Huffaker, R. C. 1964. Plant Physiol. 39, 555-560. Wilson, A. M., McKell, C. M., and Williams, W. A. 1968. J. Range Manage. 21, 305-308. Young, J., Kay, B. L., and Evans, R. A. 1970. Agron. J. 62, 638-641.

This Page Intentionally Left Blank

ADVANCES IN AGRONOMY, VOL. 35

PREDICTING CROP PRODUCTION AS RELATED TO PLANT WATER STRESS R. J. Hanks and V. P. Rasmussen Department of Soil Science and Biometeorology, Utah State University, Logan, Utah

I. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

193

11. Review of the Literature 111.

IV. V. VI. VII. VIII.

................................

205

IX. Hill, Johnson, and Ryan Mo X. XI. XII.

I. INTRODUCTION

Crop production is usually limited by insufficient water at some time during the growing season. Even in humid parts of the world, periods of insufficient rainfall, and thus water stress, commonly occur. In subhumid-to-arid areas, insufficient water is the rule; therefore supplemental water, as irrigation, is often practiced. In many dry areas of the world, crop production without imgation is very low. Water supplies that can be used for irrigation are limited. Future irrigation, even in well-established areas, will probably face more limitations because of competing uses for water. Industrialization and urbanization have large water requirements and have the financial resources to buy the water from agricultural imgation. In the future, more "shared" use of water (with industries and municipalities) will undoubtedly occur with resulting problems of poorer water quality. More intensive use of water for irrigation and other uses naturally tends to 193

Copyright 6 1982 by Academic Press, Inc. All rights of reproduclion in any form reserved. ISBN 0-12Mx)735-5

194

R. J. HANKS AND V. P. RASMUSSEN

decrease the water quality and increase salinity. All of these factors will combine to make future availability of good water for irrigation more of a problem. Thus the possibility of dealing with water stress, even with irrigation, will become more of a reality in the future. This will be in contrast to the practice of much of the irrigated regions of the world where irrigation was supplied to meet maximum demands. Future population increases will further the trend to bring under production marginal lands that tend to have low water-holding capacity. Also, crop production in marginal climatic areas will become necessary to help feed the people of the earth. Thus, there will undoubtedly be much more demand in the future for knowledge about the influence of plant water stress on crop production. Fortunately, technology has progressed to the point where it is now possible to consider the many complicated and interrelated soil-climatic and crop factors that contribute to production. Predictions of crop production as related to plant water stress are now possible for many situations with reasonable accuracy. Present knowledge and predictive techniques have the capability of aiding water management decisions so that considerable improvements in crop production are possible.

II. REVIEW OF THE LITERATURE There is a vast amount of literature relating to plant water stress and crop production. Only those studies having rather direct relevance to predictability on a field scale will be emphasized here. The classic early work of Briggs and Shantz (summarized by Shantz and Piemesial, 1927) laid the foundation for predicting crop production as related to water stress. This early work demonstrated the close relationship between transpiration, T , and plant yield, Y, as illustrated in Fig. la-c. It is generally agreed that the processes of photosynthate production and transpiration are closely linked (Boyer and McPherson, 1975). The processes of photosynthesis become limited when water stress occurs due to closing of’the stoma and reduction in other activities in the plant. Because the transport processes which make CO, available for photosynthesis and allow water to evaporate from the stoma are so closely linked, yield can be reasonably estimated by analyzing the transpiration ratio. It was also recognized that the transpiration ratio (TIY) was not unique, but depended on climatic factors. It appears that these early workers were so discouraged with the variation in the TIY ratio that they found that these studies were not foIIowed up for many years. These early authors did not plot the overall relation between T and Y as shown in Fig. 1, but rather tabulated the ratios of

CROP PRODUCTION AND PLANT WATER STRESS

195

TIY. Adding to the discouragement about the applicability of these studies was the criticism that they were unrealistic because plants were grown in containeis under rather artificial conditions and that soil evaporation was not allowed or was minimized. Consequently, there was very little appreciation of the significance of these early studies for many years. Other agronomists, at about the same time, developed approximate relations to predict crops from water use. These estimates were based on the field-measured water use and thus include both soil evaporation ( E ) ,and transpiration (or evapotranspiration, ET). Cole and Mathews (1923) and Matthews and Brown (1938) indicate for wheat and sorghum in the Great Plains a relation of the form Y=a+bET,

(1)

where a is the y intercept and b is the slope of the line. The authors indicate a high correlation between yield and ET for most Great Plains locations but different values of a and b at different locations. The value of a was always found to be negative. Of significance is the value of ET when Y is zero, or of ET, = a/b.This is the amount of water used as ET before any yield is attained. Hanks (1974) indicates this is a good approximation of the amount of soil evaporation. Cole and Mathews (1923) indicated that ET, generally ranged from about 10 cm in the Northern Plains to 25 cm in the Southern Plains for spring wheat and that the value of b ranged from 66 kg/ha-cm (2.5 bu/acre-in.) for spring wheat at North Platte, Nebraska, to 132 kg/ha-cm (5.0 bu/acre-in.) at Edgeley, North Dakota. In more recent times, the value of b for wheat has generally ranged from about 66 to 160 kg/ha-cm (Leggett, 1959; Leggett et al., 1974), and the value of ET, has ranged from 10 to 12 cm. These relations have been widely used for management purposes in waterdeficient areas as a guide to planting (depending on stored soil moisture and probability of seasonal precipitation) as well as a means of assigning economic value to irrigation water or rainfall. The problem with Eq. (1) is that it is site specific. This means that the reaction found for one site does not apply to another site because of different climatic and soil conditions. Equation (1) does not account for the well-known effects of timing of water application or difference in climate at the same site from one year to the next. Nevertheless, the simplicity of the relation with its general applicability has led to its widespread use by farmers and others for management purposes. A major advance in the development of better predictive methods for relating crop production to water use was made by deWit (1958) who concluded that much of the earlier analysis was misleading because it emphasized the transpiration ratio T/Y rather than Y and T/Eo (Eo is potential evapotranspiration of free water during the measurement period). Few of the early publications showed a plot of Y versus T (Fig. l), nor did any of them plot Y versus T/Eo (Fig. 1). As a

a

8'

YIELD I N GRAMS P

0

0

0.0

0.0

100.0

100.0

Mo.0

W.0

300.0

400.0

m.0

500.0

so0.0

YIELD IN GRAMS 200.0

600.0

600.0

m

700.0

700.0

c a

CROP PRODUCTION AND PLANT WATER STRESS

197

0

.O

FIG. lc.

result of looking at the data from a different point of view, deWit (1958) concluded that the following equation holds in regions of the world similar to the Great Plains: Y = MT/Eo

(2)

where M is crop factor. This equation has the advantage that it takes account of some of the climatic influences as shown in Fig. 1 , where the variation of the yield data about T is greater than about T/Eo. DeWit (1958) indicated that values of M calculated from Eq. (2) from plants grown in containers, described by Shantz and Piemeisiel (1927), predicted field-grown crop conditions very closely. Hanks et al. (1969) later confirmed this conclusion. Recently Fisher and Turner (1978) concluded that Eq. (2) was more widely applicable than even deWit (1958) claimed. They showed data which indicates that the M value found for several C, crops gave the same value if root and top growth was considered. They also indicated that a different, higher value of M was found for C, crops if both above-ground and below-ground growth was included. The analysis by deWit (1958) had been for above-ground growth only. Equation (2) was not found to be appropriate for humid regions of the world. Arkley (1963) and Tanner and Sinclair (1981) have shown that an equation similar to Eq. (2) could be made more applicable to different climatic regions if the Eo term were replaced by a more appropriate climatic correction term.

198

R. J. HANKS AND V. P. RASMUSSEN

Tanner and Sinclair (198 1) give persuasive arguments why the better equation is

Y = kT/(e* - e)

(3)

where k is a crop factor and 7 - e is the vapor pressure deficit. Note that if the concern is for a given crop in a given season, the following simple equation results from either Eq. (2) or (3):

YIY, = TIT, where Y , is the maximum yield found where traspiration does not limit growth (defined as T,,,). It was felt that the transpiration ratio TIT, was more indicative of the interrelated responses of photosynthesis and transpiration rate than the T/ET, ratio. TIT, Is an indicator of the relative amount of carbon dioxide exchange and water exchange taking place at the leaf surface through the stomata. The problem still remains, however, of how to determine T separately from ET when ET is the only process measurable in the field. For many purposes, however, it may be sufficient to measure ET and estimate yield directly as proposed by Stewart et al. (1977) and Doorenbos and Kassam (1979). They propose a relative yield prediction like that of Eq. (3) which minimized some of the site-specific problems. The equation developed is Y/Y, = 1

- p(l

- ET/ET,)

(5)

where p is the slope of the relative ET plot. Like Eq. (I), this equation is particularly useful to predict yield influences at a given location and season due to changes in ET. Doorenbos and Kassam (1979) have used this approach and given ranges of values for p to use for many crops and even for deficits during different growth stages. Equation (5) is also useful because it can be broken down into other useful approximations. The ratio of ETIET,, where Y/Ym is zero, approximates the amount of ET that is soil evaporation (E) (Hanks, 1974) and is given by

E = [I - (l/P)]ET, Similarly, T can be derived as follows, since ET = E

T

=

ET - [ 1 - (l/P)]ET,,, = ET - ET,

(6)

+ T: =

ET,/p

(7)

and

T , = ETJP

(8)

Note that the value of P must be greater than one to satisfy the field observation that there is some ET value at which yield is zero. A value of p = 1.O would mean that E = 0 and ET = T . A value of P of 1.5 would mean that E = ET, 0.33, and T, = ET, 0.67. Thus, Eqs. (6) through (8) give a method for

199

CROP PRODUCTION AND PLANT WATER STRESS

Table I Values of p, Y,, ET,, and M for Corn Grown at Several Locations0

Location Davis, CA Ft. Collins, CO Logan, UT

Year

P

Eo (cm)

(kg/ha)

ET,

Variety

M

1974 1975 1974 1975 1974 1975

1.22 1.24 1.36 1.42 1.38 1.40

0.78 0.73 0.58 0.58 0.63 0.55

22,000 22,000 17,200 15,800 17,900 18,100

67.4 61.6 52.9 54.5 64.4 56.2

F4444 F4444 PX20 P3955 U4544A PX20

310 324 256 238 241 248

Ym

OASreported by Stewart et al. (1977). The data shown are adjustments of original published tables.

estimating T and E separately from field data, and can be used to estimate M from Eq. (2), or k from Eq. (3), provided data on Eo or e* - e are available. Table I shows this information for several varieties of corn grown at three different locations and different years. The data of Table I indicate values of p ranging from 1.22 at Davis, to 1.42 at Logan. The values of p at Logan and Ft. Collins are similar and are higher than at Davis. Figure 2 shows the ET yield data of these three locations. This divergence is probably due to higher soil evaporation at Logan and Ft. Collins than at Davis. The climate at Logan and Ft. Collins included “rainy” periods after corn planting but before rapid growth initiation (mostly bare soil), whereas at Davis there was almost no rain during this same period. However, the problem of estimating E and T separately from ET on a daily basis, needed for many predictive purposes, must be found by other methods to be discussed later. Nevertheless, the simplicity of Eq. ( 5 ) and the range of values of p which have been measured give a very useful equation for prediction. For corn, a value of p could be assumed and adjusted upward to about 1.40 if it is to be applied to locations where rainfall events are frequent soon after planting.

Ill. MEASURING ET The use of any of the equations discussed so far require a knowledge of ET or E or T. These data are not easily arrived at or measured because they are influenced by many factors. Measurements are usually based on the water balance equation

ET = I iR - R,

-+ AS - Dr

(9)

where I is irrigation, R is rainfall, R, is runoff (+) or runon (-), AS is water withdrawn from soil storage for root extraction (depletion), and Dr is drainage

200

R. J. HANKS AND V. P. RASMUSSEN

0

20.0

40.0

0

60.0

E l IN-Cfl

0

m

b

0

0

0

20 0

40 0

60 0

80

c

ET I N C V

FIG. 2. Yield as related to evapotranspiration ( E n for corn grown in a similar experiment at Logan, UT; Fort Collins, CO; and Davis, CA. Yield and ET were depressed by limited irrigation and/ or salinity.

20 I

CROP PRODUCTION AND PLANT WATER STRESS

0.0

20.0

40.0

60.0

80.0

ET I N CM

FIG. 2c. (+) or upward flow (-) below the root zone. Measurements of I , R , and AS are routinely made, but R , and Dr are almost always assumed to be zero and almost never measured. Under many situations, such as dryland conditions, R , and Dr may be close to zero, or measurements of soil water storage can be made after runoff to correct for errors in R,. Under irrigated conditions, especially where irrigation is applied to get maximum yield or for salt balance, R, and Dr are not usually zero. Upward flow is also significant where shallow water tables occur. Many irrigation systems have developed, such as basin, sprinkler, or drip, that eliminate or minimize runoff but the problem of evaluatingDr still exists. Spatial variability of soil properties have also contributed to problems of estimating Dr. One method using a simple water balance-a field-capacity approach-has been described by Stewart et al. (1977). This requires a knowledge of ET, throughout the season (accomplished in Davis, California by using lysimeters) and measurements of I , R , and AS (R, assumed zero).

IV. ESTIMATING ET Many methods have been developed over the years to estimate ET. Jensen (1973) and Doorenbos and Pruitt (1975) have indicated the present state of the art using the following crop-coefficient approach.

ET = K, ET,

(10)

202

R. J. HANKS AND V. P. RASMUSSEN

where K, is a crop coefficient and ET, is potential evapotranspiration using a reference crop or one of many climatic equations. The value of K , is dependent on the kind of crop as well as local climatic and imgation management conditions. Corrections are made for limited soil water. Hanks (1974), Nimah and Hanks (1973), and Childs and Hanks (1975) have developed models for systematically estimating E and T (and thus E7') from a knowledge of soil, climate, crop, and irrigation factors. In the simple approach used by Hanks (1974), the following equations are used in a day-by-day computation to estimate E and T: E = Em t - 1 / 2 (1 1)

T = T,,,

if

SWCIAW > 0.5

T = (T,/0.5) (SWUAW)

if SWCIAW < 0.5

(12) (13)

where Em is maximum evaporation for soil, t is time in days since the soil was last wet, SWC is soil water content in a given soil layer, AW is the total available water in the same layer. Water is added to a soil made up of layers each day that rainfall and irrigation occurred. If the water added to the top layer is more than enough to make SWC = AW (i.e., bring to field capacity), then the excess water is added to the next layer. This process continues for all layers, and drainage is said to occur if there is still excess water above that needed to bring all of the layers to field capacity (SWC = AW). The sum Em + T, is considered to be related to free water evaporation (from an evaporation pan properly adjusted) or reference crop and is thus closely related to ET, of Eq. (10). This procedure has been found to give good results for many situations (Stewart et al., 1977) and is easy to program on small or large computers. This model also computes TIT,, which allows computation of yield using Eq. (4). The method of estimating Emand T , is described by Retta and Hanks ( 1980) and Childs and Hanks (1975). It is believed that this method is much more transferrable than the values of K , published by Jensen (1973). The model of Nimah and Hanks (1973) is much more sophisticated than that of Hanks (1974) in that basic soil properties (hydraulic conductivity and matric potential functions with water content) are used. An estimation of runoff and drainage versus upward flow is made on an hourly basis throughout the season. Childs and Hanks (1975) have modified this earlier model to account for some salinity effects on T and E. The method used to determine Em and T,,, is the same as needed for the simple model of Hanks (1974). Kanemasu and co-workers have developed a method for evaluating Em and T,, separately from ET, using a micrometeorological approach with physically based relationships. Evapotranspiration models have been developed for sorghum and soybean (Kanemasu et al., 1976), corn (Rosenthal et ul., 1977), and wheat (Kanemasu et al., 1977). The approach has been similar in all cases.

CROP PRODUCTION AND PLANT WATER STRESS

203

ET, is estimated using the approach of Priestly and Taylor (1972), as modified by Jury and Tanner (1975): ET, = a [ s / ( s + y)] R ,

(14)

where a is crop- and location-related constant (1.35 for corn at Manhattan, KS), is the slope of the saturation vapor pressure curve at a weighted average temperature, y is the psychrometric constant, and R , is daily net radiation. R , is computed from empirically determined relationships of crop leaf area index (LAI) and stage of growth. Rasmussen (1979) showed that evaporation from the soil surface is computed by splitting the energy available for E, as dictated by Eq. (14), by a fraction, T , which represents the portion of energy reaching the soil surface (Ritchie, 1972) so that:

s

Em =

T

X

ET,

(15)

and

T , = ( 1 - 7) X ET,

(16)

is taken as an exponential function of LAI, or percent cover of the soil surface. Actual transpiration ( r ) is allowed to proceed at the maximum rate (T = T,) until a critical fraction of available water is left (near 65%), and is linearly decreased to zero at permanent wilting point. This is similar to the approach of Rasmussen and Hanks (1978). Actual evaporation from the soil surface ( E ) is assumed to proceed at the maximum rate ( E = Em) when the soil surface is wet, and then transfers to a falling-rate stage after some critical amount of available water is lost (Ritchie, 1972). Actual evaporation (E,) is then determined by the velocity of water flow to the soil surface and the time since last wetting as follows: T

E

= ~tl/Z- c(t -

1)IQ

(17)

where c is a constant dependent on soil hydraulic properties and t is the number of days since entering the falling-rate stage. This method is similar to that of Hanks and co-workers (Hanks, 1974; Childs and Hanks 1975), in that ET, is really a sum of separately determined E, and T,. This method does not rely on crop coefficients and thus accounts for uneven crop growth and cover, and rain/ irrigation timing,

V. ESTIMATING YIELD As discussed earlier, yields can be predicted once T or ET is predicted. These predictions do not account for any growth-stage effects that are not accounted for by effects on ET or T . This assumption of no growth-stage effects applies more to

Table I1 Predicted and Measured Grain Yields0 and Plant Growth for Five Irrigation Treatments for Mead, Nebraska, 197Sb Stover weight (dry weight)

Leaf area Irrigation applied (cdday)

Actual (dm2)

Predicted (dmz)

Actual

0.76 0.61 0.38 0.25 0.00

72.5 76.7 75.2 74.4 73.4

75.6 75.6 75.6 75.6 75.5

109.0 108.0 98.0 114.0 98.0

Wariety: Pioneer 3366. bSource: Childs ef al. (1977).

(!a

Ear weight (dry weight)

Grain yield (15.5% moisture)

Predicted

Actual

(8)

(g)

Predicted (9)

Actual (kg/ha)

Predicted (kglha)

104.0 104.0 104.0 104.0 104.0

163.0 141.0 160.0 113.0 54.0

158.0 158.0 157.0 120.0 55.0

7500 7260 7050 5230 2490

7480 7480 7410 5680 2620

CROP PRODUCTION AND PLANT WATER STRESS

205

total above-ground, dry-matter production than to the situation where only a plant part, such as grain, is produced. However, in many instances the ratio of grain to total dry matter (harvest index) is nearly constant. Hanks and Puckridge (1980) developed a modification of the Hanks (1974) model to account for plant development as related to water stress throughout the season. Both leaf area and total dry matter are predicted. The effects of limited water early in the season which cause limited T and growth are accounted for. This model predicted spring wheat yields well from several studies in South Australia where water was in excess as well as severely limited. In spite of a wide range of soil water availability in these studies, the harvest index was nearly constant. Childs er al. (1977) adapted the model of Childs and Hanks (1975) to predict plant leaf water potential and corn production. Photosynthesis, respiration, and available carbohydrates in the leaf are all estimated in this model. Yield predictions using this model have generally been good (Table 11).

VI. GROWTH STAGE EFFECTS There have been many studies indicating that water deficits during certain growth stages have more effect on grain yield than on ET or T. Jensen (1978) has found that delaying irrigation in grain sorghum reduced yields by 35% but reduced water use by only 20%. He developed the following equation to account for these growth stage effects:

Hanks (1974) provides for calculation of T/T, for various growth stages to estimate yield by an adaptation of the Jensen equation (T is used instead of ET). However, Stewart er al. (1977) found growth-stage correction was only slightly better in predicting corn grain yield than the more simple Eq. (4), when applied to the data collected in four states with many different irrigation regimes. The evaluation of values to use for hiis not straightforward because of interrelation, different time periods for growth stages, and so forth.

VII. RASMUSSEN AND HANKS SPRING WHEAT MODEL Rasmussen and Hanks (1978) used methods similar to Hanks (1974) to develop a transpiration-basedyield model for spring wheat. The model took a form of

206

R. J. HANKS AND V. P. RASMUSSEN

the Jensen (1968) yield equation, except that TITY, was used as a component of the model, rather than TIET,,,. It is generally agreed that the processes of photosynthate production and transpiration are closely linked (Boyer and McPherson, 1975). The processes of photosynthesis become limited when water stress occurs due to closing of the stomate and reduction in other activities in the plant. Because the transport processes which make CO, available for photosynthesis and allow water to evaporate from the stomates are so closely linked, yield can be reasonably estimated by analyzing the transpiration ratio. A growing root function was added, and phenological stages of development were clocked with a model specific for spring wheat, based upon growing degree days. The yield equation took the form:

Y = Y,

X

(TIT,,,):

X

(TIT,)$

X

(TIT,);

X

(TIT,):

(19)

where stage 1 was emergence to boot, stage 2 was boot to head, stage 3 was heading to soft dough, and stage 4 was soft dough to maturity and A = 0.25. This approach used empirically determined A values of 0.25 for each stage. This allows more mathematical “weighting” to fall upon the stage with fewest days, because a small change in TIT, can use a larger change in yield. Stress occurring in any growth stage multiplicatively lowers the yield-regardless of

10

-

g

9-

z

8-

8 .-0 L

7-

E

6 -

z

5-

+-

C

‘

Q)

5 4 .-C

-

:3 -

0,

.-t;

0

CALIBRATION VALIDATION

2-

-0

!t n

1I

I

I

1

I

I

1

I

1

I

FIG. 3. Observed versus predicted yields for Rasmussen and Hanks (1978) spring wheat model. Calibration (0); validation (0).

I

CROP PRODUCTION AND PLANT WATER STRESS

207

the status of other stages. This seems reasonable considering what we know of grain crops, in which stress in any stage can cause yield reduction. This method did not fully address the known fact of the “critical” growth stages by weighting them even more heavily with larger values. Figure 3 shows a good correlation of observed versus predicted line for this model (R2 = 0.66).

VIII. RASMUSSEN AND KANEMASU WINTER WHEAT MODEL Rasmussen and Kanemasu (1979) developed a model similar to the spring wheat model of Rasmussen and Hanks (1978), but used the ET modeling technique of Kanemasu et af. (1977) described earlier. The yield equation took a similar form, in that transpiration relationships were used. The model was developed for winter wheat from a large data set from the Great Plains coupled with remotely sensed satellite crop imagery. Dormancy of winter wheat has a confounding effect on yield when phenologically staged TIET, relationships are used. After considerable statistical analysis of many model forms, the following was found to be most logical and universally accurate:

Y

=

Y , x A[C(T/ET,)]:.” [z(T/ET,)]:.” [C(T/ETm)]g.6s

(20)

where Y , and A are really lumped into one term (2.856) when log transform regression is performed to solve for the best-fit values. Growth stages a, b, and c are planting to jointing, jointing to heading, and heading to soft dough. This model has a reasonable growth stage versus water stress approach. The summation of TIET, for any given stage is a gross measure of how much possible photosynthetic activity can occur in that stage, while still being weighted by a factor of water stress. Leaf area index (LAI) controls TIET, at its maximum, but water stress can always reduce that value for any given day. In winter, when green leaf area is low, although ET, can be high, the fraction TIET, remains low. A summation of these values for the winter period is low, but can still give a relative value for winter stress. An analysis of the exponential values in the equation shows clearly that this model gives much higher weight to the water relations and leaf area of the heading-to-soft-dough stage. This compares favorably with our knowledge of critical growth periods. This model has been used by Heilman et af. (1977) and Rasmussen and Kanemasu (1979) to predict yields over large areas of the Great Plains, as shown in Fig. 4. Brakke and Kanemasu (1979) extended the test to include data from the Pacific Northwest and compared the model to a photosynthesis model of Hodges and Kanemasu (1977). As shown in Fig. 5, the data show a good agreement in all

208

R. J . HANKS AND V. P. RASMUSSEN BU/ACRE (AT .10 Bw 8160 LB/BU)

I

0

9

n

- 40 - 30

2-

- 20 I-

.

PEARSON PRODUCT-MOMENT r .81 R2- .66 SL.m.61 INtm14.4

%\

I

I

I

BUIACRE I

ew

(AT.IO

20 I

-

I

2 3 OBSERVED GRAIN YIELD MTIHA)

10

b

- I0

30 I

EL 60 LBIBU) 40

50

I

0 BUSHLAND. TEXAS

RL

.91

SL.=.68

I

I

0

INT.=11.9

I

I

2 3 OBSERVED GRAIN YIELD (MTIHAI

Rasmussen and Kanemasu winter wheat yield model observed versus predicted data for (a) Kansas and (b) Texas.

CROP PRODUCTION AND PLANT WATER STRESS

209

OBSERVED GRAIN YIELD (BU/ACRE) 20 30 40 50 60 70

10

5000 -

-a a 4000-

ET YIELD MODEL

I

Y

w

F

3000-

w

a -10

0

2000 3000 4000 OBSERVED GRAIN YIELD (KG/HA)

1000

5000

FIG. 5. Observed versus predicted yield values for Rasrnussen and Kanemasu winter wheat model on data from Great Plains, Mountain States, and Pacific Northwest. Site (1975-1976): Saline Co., KS (0); Finney Co., KS (+); Rice Co., KS (0); Randall Co., TX (0); Oneida Co., ID (0); Whitman (2) Co., WA (A); Whitman (3) Co., WA ( X ) . ( r = .72) (from Brakke and Kanemasu, 1981).

respects except for the data from Washington State. R values are over 0.9 for the data without Washington sites, but fall to 0.72 when they are included. The relatively high yields associated with the Pacific Northwest (approaching 200 bu/ acre) were not included in the calibration of the model from Great Plains. Clearly, where varieties, husbandry practices, or other factors change the potential yield value, the model must be recalibrated. However, if potential yield values are similar in regions where the model is used (e.g., in the Soviet Union), it can be used with reliability.

IX. HILL, JOHNSON, AND RYAN MODEL Hill ef al. (1979) developed a T/T, model for the yield of soybeans. The approach was similar to that of Hanks (1974) in that grain was modeled with a Jensen (1968) multiplicative model that is further multiplied by a lodging factor and a seasonal yield factor. The seasonal yield factor accounts for differences in Y,,, from season to season so that a single Y , may be used for many sites and years. The observed versus predicted values are shown in Fig. 6 .

210

R. J. HANKS AND V. P. RASMUSSEN

70

rc

x 60 I

-

\

a

2 50 -

w%

1 40

-

b 4

J

30

-

J

f

2 20 -

0 Moiurliy Group II

P

Moturlty Group 111 r2s.989

-1 W

c

r*=.995

10-

0

/

A Moturlty Group IV

r2s.982

20 30 40 SO 60 10 MODEL PREDICTED RELATIVE YIELD y / y p - X

70

FIG. 6. Observed versus predicted yield for soybean model of Hill et a/. (1979). (8)Maturity Maturity Group IV, R2 = 0.982. Group 11, R* = 0.995; (m) Maturity Group 111, R2 = 0.989; (A)

The A values for weighting the T/T, fractions for each growth stage were changed with variety and were substantially determined with a log transform regression technique. The values are shown in Table 111. Of interest in this group of values is that most weighting is given to the TIT, of the last growth stage. This indicates that this model is performing more satisfactorily as an indicator of the negative pressure to yields by water deficits in the critical growth stages.

Table 111 Lambda Values for Different Soybean Maturity GroupscJ

Maturity group

Parameter values hl

A2

A3

A4

AS

Potential yield Y,, (quintalslha)

11 111 IV

0 0 0

0 0 0

0.380 0.075 0

0.05 0.03 0.30

0.410 0.475 0.400

51.8 60.5 63.9

OSource: Hill er al. (1979).

21 1

CROP PRODUCTION AND PLANT WATER STRESS

X. MORGAN, BIERE, AND KANEMASU MODEL FOR CORN Morgan et af. (1980) reported a model for corn that incorporates the ET model (discussed earlier) as reported by Rosenthal et al. (1977), and Kanemasu et af. (1976, 1977). The approach is to use the ET model to produce values for the fraction of available water in the soil profile. This fraction is then related to a moisture-growth response value as shown in Fig. 7. Yield is calculated from a general growth equation of the form:

where t is time in days, Y,,, represents potential yield at maturity, r(t)represents relative growth in potential yield in time period t when water is not limiting, and atis a function relating growth (relative to potential growth) to the level of available soil moisture. Dry matter accumulation occurs differently in the vegetative and reproductive stages of growth. During the vegetative stage, it is expressed as:

Dry matter accumulation I .o 0.9

1

I

I

= exp(- 1.7

1

I

1

I

I

1

+ 0.049t)

(22)

-

0.0

0.1

0.2

I

0.3 0.4 0.5 0.6 0.1 AVAILABLE SOIL MOISTURE

1

I

0.0

0.9

1.0

FIG. 7. Crop response function value versus available soil moisture of Morgan et a / . (1980).

212

R . J. HANKS AND V . P. RASMUSSEN

-

2 0,

\

140 Y

-

5 120 w,

n

-

0

-

Manhattan, 1974 0 Manhattan, 1975 A Manhattan, 1976 A Scandia, 1974 0 Scandia, 1975

0

20

40

60

80

100

120

140

160

180

ACTUAL GRAIN YIELD (Bu/Acre) FIG.8. Observed versus predicted yields for corn grain yields for several years and locations in Kansas: (0)Manhattan, 1974; (0) Manhattan, 1975; (A) Manhattan, 1976; (A) Scandia, 1974; (0) Scandia, 1975. (Morgan et al. (1980).)

During the reproductive stage, it is expressed as: Dry matter accumulation = exp(-3.573

+ 0.109(tr- tf/2t,,)

(23)

where rr is the time into the reproductive phase and t, is the total time of the reproductive phase. Predicted model values agreed generally with the observed values (Fig. 8). However, the model does not account as directly for growth stage effects as some others.

XI. OTHER MODELS WITH MOISTURE STRESS INCLUDED Several models use soil moisture status in determining yield, although not as a primary consideration. Arkin et al. (1976) and Vanderlip and Arkin (1977) developed a model for yield of grain sorghum that is based upon simulation of leaf area, photosynthesis, evapotranspiration, and temperature. Available soil water is assumed to have no affect on yield (photosynthesis) until the available soil water falls below 20%. The model did well in predicting dry matter production, but was not fully developed at the time of reporting to partition dry matter into grain. The strength

213

CROP PRODUCTION AND PLANT WATER STRESS OBSERVED GRAIN YIELD (BU/ACRE) 20 30 40 50

5 10

-a

5000

PHOTO MODEL 7.

I

a

F: w>

4000

- 70 W

K

0

-

-6o? 3 m

-

509 I *

- 40 a

n

- 30 K0

g2oM)-

2

w 0

n

- 20

w

K

P

1wo 0

0

I-

w

- 10 K 1000 2000 3000 4000 OBSERVED GRAIN YIELD (KG/HA)

5000

FIG. 9. Observed versus predicted yields from a large area test of the Hodges and Kanemasu model by Brakke and Kanemasu (1981). Site (1975-1976): Saline Co., KS (0); Finney Co., KS (+); Rice Co., KS (0); Randall Co., TX (0); Oneida Co., ID (0); Whitrnan (2) Co., WA Whitrnan (3) Co., WA ( X ) .

(a);

of this model is in its multifaceted approach that accounts for many variables besides soil moisture. Hodges and Kanemasu (1977) developed a complex photosynthetic model for wheat yield production. It relies heavily on LA1 data from ground measurements or remotely sensed images. It was compared with the Rasmussen and Kanemasu (1 979) ET yield model and was slightly less effective in predicting grain yields (Brakke and Kanemasu, 1979). However, it was very effective (R2 > 0.9) in predicting accumulation in dry matter. The only soil moisture input was a reduction in photosynthesis, as available soil water fell below 6576, with no partitioning of the effect of soil moisture during grain filling. Brakke and Kanemasu (1979) show an improved version of this model that performed well, compared to the ET/yield approach to Rasmussen and Kanemasu (1979), over a large area of the United States. Figure 9 shows the observed versus predicted values for this model.

XII. SUMMARY

The technology for evaluating the influence of plant water stress an crop production has advanced to the point that good prediction can be made for most purposes. Methods range from mechanistic prediction of details of growth of

214

R. J. HANKS AND V. P. RASMUSSEN

plant parts to statistical predictions of nationwide yields. Researchers have recently proposed many predictive models for many crop situations that account for many complex interactions to be managed in a practical manner. Thus, it is now possible for farmers to use these techniques to minimize the influence of water stress on yield. The simplest models require knowledge of total water available, such as irrigation, rain, and stored water, to predict yield using statistical methods. These methods are site and season specific, but give general guidelines and are widely used. The next step in complexity is to evaluate the relative amount of water actually used by a crop-not just available for use. These models are not so site specific but require information on maximum yield, maximum evapotranspiration, and drainage below the root zone. Both of these methods are based on field measurements. The next level of predictive models requires computers for solution. Soil, climatic, and crop information are used to predict water use as a function of time and can thus estimate stress as a function of time or growth stage. Some of these models do not differentiate between evaporation from the soil and transpiration by crops. The more general models base predictions on effects of stress on crop transpiration only. The disadvantage of this approach is in the difficulty of splitting up ET into E and 2“. There are several models of this type accounting for most of the important effects that are presently developed; thus they can be used for practical management decisions. Current research is being conducted along many facets of crop production and water stress. Our knowledge of these relations are being aided greatly by this research. It will probably be possible to predict, in the near future, how to manage water stress to harness the many components of yield to the benefit of man. REFERENCES Arkin, G. F., Vanderlip, R. L., and Ritchie, R. T. 1976. ASAE Trans. 19, 622-630. Arkley, R. J. 1963. Hilgardia 34, 559-584. Boyer, J. S., and McPherson, H. G. 1975. Adu. Agron. 27, 1-23. Brakke, T. W., and Kanemasu, E. T. 1979. Proc. Symp. Remor Sens. Enuiron., 13th. April. Ann Arbor pp. 629-641. Childs, S. W., and Hanks, R. J. 1975. Soil Sci. SOC. Am. Proc. 39, 617-622. Childs, S. W., Gilley, J. R., and Splinter, W. E. 1977. ASAE Trans. 20, 858-865. Cole, J. S., and Mathews, 0 . R. 1923. U.S.Dept. Agric. Bull. No. 1004. DeWit, C. T. 1958. “Transpiration and Crop Yields,” No. 64.6. Verslag van Lanbouwk, Donderzock. Doorenbos, J., and Kassam, A. H. 1979. F A 0 Irrig. Drain. Paper No. 33. Doorenbos, I., and Pruitt, W. 0. 1975. F A 0 Irrig. Drain. Paper No. 24. Fisher, R. A,, and Turner, N. C. 1978. Annu. Rev. Plant Physiol. 29, 277-317. Hanks, R. J. 1974. Agron. J . 66, 600-665.

CROP PRODUCTION AND PLANT WATER STRESS

215

Hanks, R. J., and Puckridge, D. W. 1980. Aust. J . Agric. Rec. 31, 1-11. Hanks, R. J., Gardner, H. R., and Florian, R. L. 1969. Agron. J . 61, 30-34. Heilman, J. L., Kanemasu, E. T., Bagley, J. O., and Rasmussen, V. P. 1977. Remote Sens. Environ. 6, 315-326. Hill, R. W., Hanks, R. J., Keller, J., and Rasmussen, V. P. 1974. “Predicting Corn Growth as Affected by Water Management: An Example,” p. 18. CUSUSWASH 221(d)-6. Utah State Univ. Hill, R. W., Johnson, D. R., and Ryan, K. H. 1979. Agron. J . 71, 251-256. Hodges, T., and Kanernasu, E. T. 1977. Agron. J . 69, 974-978. Jensen, M. E. 1968. In “Water Deficits and Plant Growth” (T. Kozlowski, ed.), Vol. 11, pp. 1-22. Academic Press, New York. Jensen, M. E., ed. 1973. ”Consumptive Use of Water and Irrigation Water Requirements.” Am. SOC.Cur. Eng., New York. Jury, W. A., and Tanner, C. B. 197.5. Agron. J . 67, 840-842. Kanemasu, E. T., Stone, L. R., and Powers, W. L. 1976. Agron. J . 68, 569-572. Kanemasu, E. T., Heilrnan, J. L., Bagley, J. O., and Powers, W. L. 1977. Using LANDSATdata to estimate evapotranspiration of winter wheat. Environ. Manage. 1, 5 15-520. Leggett, G . E. 1959. Wash. Agric. Exp. Stn. Bull. No. 609. Leggett, 0 . E., Ramig, R. E., Johnson, L. C . , and Massee, T. W. 1974. In “Summer Fallow in the Western United States,” Ch. 6. USDA Consumer Res. Dept. 17. Mathews, 0. R., and Brown, L. A. 1938. USDA Cric. No. 477. Morgan, T. H., Biere, A. W., and Kanemasu, E. T. 1980. Water Resour. Res. 16, 59-64. Nimah, M. N., and Hanks, R. J. 1973. Soil Sci. Soc. Am. Proc. 37, 522-532. Priestly, C. H. B., and Taylor, R. J. 1972. Mon. Weather Rev. 100, 81-92. Rasmussen, V. P. 1979. PhD dissertation, Kansas State University, Manhattan, Kansas. Ramussen, V. P., and Hanks, R. J. 1978. Agron. J . 70, 940-944. Rasmussen, V. P., and Kanemasu, E. T. 1979. Proc. Con$ Agric. Forest Meteorol., 14th; Conf. Biomet.. 4th. Am. Meteorol. Soc., Boston. Retta, A , , and Hanks, R. J. 1980. Utah Agric. Res. Dept. No. 48. Ritchie, J. T. 1972. Water Resour. Res. 8, 1204-1213. Rosenthal, W. D., Kanemasu, E. T., Raney, R. J., and Stone, L. R. 1977. Agron. J . 69,461-464. Shantz, H. L., and Piemeisel, L. N. 1927. J . Agric. Res. 34, 1093-1190. Stewart, J. I . , Danielson, R. E., Hanks, R. J., Jackson, E. B., Hagan, R. M., Pruitt, W. O., Franklin, W. T., and Riley, J. P. 1977. Utah Water Res. Lab. PRWG 151-1, 191. Tanner, C. B., and Sinclair, T. R. 1981. In “Limitations to Efficient Water Use in Crop Production.” Amer. Soc. Agron. (in press). Vanderlip, R. L., and Arkin, G . F. 1977. Agron. J . 69, 917-923.

This Page Intentionally Left Blank

ADVANCES IN AGRONOMY. VOL 35

IRON NUTRITION OF PLANTS IN CALCAREOUS SOILS Yona Chen and Phillip Barak The Seagram Centre for Soil and Water Sciences, The Hebrew University of Jerusalem, Rehovot, Israel

I. Introduction. . . . . . . . . . . . . . . A. Iron Oxide Minerals ........................

........................ ................. .................

................. ................. ......................... ................. ................. B . Mechanisms of Iron Uptake.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Causes of Iron Deficiency . . . . . . . . . . . . . ................. D. Indicators of Iron Deficiency . . . . . . . . . . . . . . . . . . . ................. IV. Correction of Iron Deficiency ...................... ................. ................. B . Soil Amendments ....................................... C. Genetic Selection of Cultivars . ........................ References . . . . . . . . . . . . ...................................... D. Soil Iron Extractants . . . . . . . . . . . . . . . . . .

217 218 218 219 219 221 222 222 225 227 230 230 23 1 237 238

I. INTRODUCTION

Many agricultural crops worldwide, especially in semiarid climates, suffer from iron deficiencies. Among plants sensitive to iron deficiency are apples, avocado, bananas, barley, beans, citrus, cotton, grapes, oats, peanuts, pecans, potatoes, sorghum, soybeans, and numerous greenhouse flowers. Deficiencies are usually recognized by chlorotic, or yellowed, intervein areas in new leaves and are typically found among sensitive crops grown in calcareous soils; calcareous soils cover over 30% of the earth’s land surface. Iron deficiency in extreme cases may lead to complete crop failure. Two principal methods of treating iron deficiencies are accepted practice at present. Spraying foliage with inorganic salts has been shown to be of benefit but often gives spotty results due to limited penetration of iron into leaves. Also, 217

Copyright Q 1982 by Academic Press, Inc. All rights of reproduction in any form reserved.

ISBN 0-12-0735-5

218

YONA CHEN AND PHILLIP BARAK

repeated treatments are required during the course of canopy development. Soil treatment with synthetic chelates, principally FeEDDHA (ethylenediaminedi-ohydroxyphenylacetic acid), has been found to be an almost unqualified success, but for the drawback of their high costs. Therefore, search continues for a lowcost, effective remedy to iron deficiency.

II. SOIL IRON COMPOUNDS AND METHODS FOR THEIR EXTRACTION A. IRONOXIDEMINERALS

Iron is the fourth most abundant element in the earth’s lithosphere, following oxygen, silicon, and aluminum. Most of the iron in the earth’s crust is in the form of ferromagnesium silicates. Weathering of such minerals in soil is usually accomplished by combined hydrolysis and oxidation due to reaction with water and air. Most of the iron released by weathering is precipitated as oxides or hydroxides; only a small part of the iron is incorporated into secondary silicate minerals or complexed by soil organic matter (Schwertmann and Taylor, 1977). Lindsay (1979) has summarized thermodynamic calculations that show that the solubility of Fe(II1) oxides decreases in the order: Fe(OH), (amorphous) > Fe(OH), (soil) > y-Fe,O, (maghemite) > y-FeOOH (lepidocrocite) > a-Fe,O, (hematite) > a-FeOOH (goethite). Goethite, because of its great stability, is found in almost every soil and climate and is the most abundant soil iron oxide. Hematite, the second most abundant soil iron oxide, is absent from recent soils in humid temperate climates. Lepidocrocite is found exclusively in noncalcareous hydromorphic soils, having had a Fe(I1) hydroxy compound precursor. Maghemite is common in highly weathered soils formed from basic igneous rocks in tropical and subtropical climates. Ferrihydrite (previously called “amorphous ferric hydroxide”) has been identified in numerous soil environments. Since Schivertmann and Taylor (1977) assess the solubility product Ksp (log Fe 3 log OH) of ferrihydrite as ranging from -37.0 for fresh ferric hydroxide to -39.4 {the value found by Norvell (1970) for Fe(OH), (soil) iron that regulated iron solubility in three near-neutral soils], it seems that Schwertmann and Taylor (1977) consider ferrihydrate to be the soil iron oxide that regulated iron solubility in Norvell’s experiments. For seven alkaline and calcareous soils, Sinha er nl. (1978) found that the Ksp of Fe(OH), averaged -39.0 +- 0.2. Using different techniques, Bohn (1967) found a K s p of -39.0 in acid soils, indicating that ferrihydrite is the most reactive iron oxide also in acid soils.

+

IRON NUTRITION IN CALCAREOUS SOIL

219

B. SOLIDPHASEORGANOIRON COMPLEXES

Relatively little is known of solid phase organoiron complexes in soils since investigation by necessity involves some type of extraction which alters the natural state of the complexes by making them soluble. The most common extractant of organoiron complexes is sodium pyrophosphate. In a study of 33 Ando soils, Wada and Higashi (1976) found that the iron in organoiron complexes ranged from 0.05 to 2.91% of soil weight and even exceeded the weight of iron in soil iron oxides in a number of soils. In general, they found that humus complexes more Fe and Al, the main complementary metal bound by soil organic complexes, as the soil ages and weathers. On the average, the Ando soil complexed 0.20 -t 0.03 (95% confidence) moles of pyrophosphate-extractableFe + A1 per pyrophosphate-extractable C. McKeague (1968) found Fe + AI/C ratios of 0.05-0.28 in various spodzol soils; pyrophosphate extractions of organoiron complexes are likely incomplete since McKeague et al. (1971) showed that artificial fulvic acid complexes with Fe/C = 0.2 were not completely pyrophosphate-extractable. Two additional types of organoiron interactions exist-the dissolution of iron from crystalline forms and the prevention of recrystallization. Schnitzer and Kodama (1977) have reviewed a number of experiments measuring the rate and extent of dissolution of iron-bearing silicates such as micas and chlorites, and iron oxides such as goethite and lepidocrocite by humic and fulvic acids. Schwertmann (1966) has presented data on the role of soil organic matter in preventing ferrihydrite from crystallizing into goethite, a less soluble iron compound. A sample of ferrihydrite was found to give no x-ray diffraction typical of iron oxides even after boiling in 1 N KOH for several hours, conditions favorable for goethite formation from laboratory-prepared ferrihydrite. Upon reduction of organic carbon content in the natural ferrihydrite from 10 to 1% with H,O,, samples boiled in 1 N KOH evidenced x-ray diffraction lines of goethite. Similarly, when humic acid was added to laboratory-prepared ferrihydrite, artificial aging produced very little goethite. Such evidence clearly suggests that soil organic matter inhibits thermodynamically favorable transformation of ferrihydrite to goethite and therefore prevents iron solubility from falling below that of ferrihydrite. C. SOLUBLE IRON SPECIES

Inorganic iron (111) species in solutions are Fe3 , FeOH2+, Fe(OH),t, Fe(OH)? and Fe(0H); . Calculations by Lindsay (1979) show that the three most +

220

YONA CHEN AND PHILLIP BARAK

abundant species in the pH range of 7-9 are Fe(OH),f, Fe(OH):, and Fe(0H)y ; the sum of inorganic species does not exceed moles Fe/liter, or 5.6 ppt (parts per trillion), in this pH range. Natural waters usually contain much higher concentrations, up to five orders of magnitude more. Bradford et al. (197 1) found that iron concentrations in saturated extracts of 68 soils representing 30 soil series in California ranged from 0.01 to 0.8 ppm, with a mean and median of 0.05 and 0.03 ppm, respectively. Such high iron concentrations are almost certainly due to either colloidal iron hydroxide or soluble organoiron complexes. Colloidal iron hydroxide has been found to pass lOOAmembrane filters, so that evidence of soluble organoiron complexes must be indirect. Perdue et al. (1976) have found a highly significant correlation between total Fe A1 concentrations and dissolved organic matter in water samples of the Satilla River system, “black waters” of the southeastern United States. The solubility of the organoiron complex was greater than that of the pyrophosphateextractable complexes of Wada and Higashi (1976) because the organic matter had a low molecular weight (1270) and because the F e S A K ratio was low (0.02). The addition of fresh organic matter to soils has been known to increase the water-soluble iron in soil solution. Data presented by Wallace and Lunt (1960) show that the soil solution of a calcareous soil incubated with 0- 10 tons/ acre of fresh clover for 11 days contained 0.21-1 S O pprn Fe; no data on soluble carbon were reported. Olomu et ul. (1973) incubated six calcareous soils at saturation for up to 70 days while continuously monitoring pH, Eh (soil redox potential), and watersoluble iron. The extracted soil solutions contained 1100- 1500 ppm C and up to 18 ppm Fe. The iron and organic matter passed virtually undiminished through Na-exchange resin and were both retained on C1-exchange resin, thus demonstrating that the iron was not present i n inorganic forms such as Fe2+, nor as colloidal ferric hydroxides, but rather as a negatively charged complex. The subject of soluble organoiron complexes has also been intensively studied in connection with podsolization, a process which forms highly weathered soils due to the decomposition of minerals and transport of weathering products by humic materials. As summarized by DeConninck (1980), mobile organic substances, e.g., humic acids, form during the decomposition of plant remains. If polyvalent cations, especially those of A1 and Fe, are present in the upper soil layer, the organic matter is immobilized as a gel which later becomes a solid organoiron complex. If insufficient A1 and Fe are present, then such quantities as are present are complexed and transported downward until immobilization occurs at some depth due to supplemental complexation of Fe or Al, desiccation, or change in ionic composition of soil solution causing coagulation. Rode (1937) has calculated the rate of iron mobilization by organic matter in a strongly

+

IRON NUTRITION IN CALCAREOUS SOIL

podzolic soil on varved clay to be 18 kg Fe/ha/year for the A1 (0.25 cm).

22 1

+ A2 horizons

D. SOILIRONEXTRACTANTS

The various extractants of soil iron may be divided into those used for selective dissolution of specific soil iron oxides and those used to dissolve some unknown fraction of soil iron that is empirically correlated with “availability” of soil iron to plants. Wada (1977) has summarized the efficiency of a number of common extractants upon the various soil iron oxides. Sodium pyrophosphate (0.1 M ) , either at pH 10 or adjusted to pH 7 with phosphoric acid, is a good extractant of organoiron complexes, a poor extractant of ferrihydrite, and does not extract iron from more crystalline iron oxides. The dithionite-citrate extraction process of Mehra and Jackson extracts crystalline iron oxides such as hematite and goethite as well as ferrihydrite and organoiron complexes by virtue of reduction of iron to Fe(I1) and chelation with citrate. Unlike the above extractants, the ones used for evaluating iron availability to plants do not selectively dissolve a particular iron compound. Instead, measures of iron availability are useful to the same degree that they are empirically correlated with the occurrence of iron chlorosis in sensitive crops. Cox and Kamprath (1972) report that although a number of methods have been devised to extract “available” iron, there are few reports on the evaluation of such tests. Olson and Carlson (1950) evaluated the use of 1 M ammonium acetate, adjusted to pH 4.8 with acetic acid, and found that the critical level dividing soils producing chlorotic and nonchlorotic plants was 2.0 ppm Fe. Johnson and Young (1968) found that soil iron extractable with 0.001 M EDDHA in 0.1 M NaNO, was correlated with chlorotic sudangrass. This method has to its credit a strongly colored Fe-chelate easily measurable colorimetrically. The method of choice today seems to be the DTPA (diethylenetriaminepentaacetic acid) extraction developed by Lindsay and Norvell (1969), which calls for a 1:2 soil/solution extraction with 0.005 M DTPA, 0.01 M CaCl,, and 0.1 M triethanolamine, adjusted to pH 7.3. The extractant is well buffered in order to standardize pH and Ca concentration and to prevent dissolution of occluded micronutrients in carbonate minerals. Lindsay (1979) reports that the test successfully separated 77 Colorado soils into deficiency and nondeficiency categories and that no response to iron fertilizer was obtained when DTPAextractable iron exceeded 4.5 ppm. Boer and Reisenaurer (1973) confirm the effacacy of the DTPA soil test but report the critical level to be 6 ppm for 11 of 13 field locations and 5 ppm for 13 of 14 greenhouse soils. In many cases,

222

YONA CHEN AND PHILLIP BARAK

however, the DTPA soil test results are not unequivocable but at present there is no better test for available iron.

Ill. IRON NUTRITION OF PLANTS A. METABOLIC FUNCTIONS OF IRON

Iron deficiency in plants typically causes chlorosis of leaf tissue because of inadequate chlorophyll synthesis. In a healthy plant, 60% of all leaf iron is concentrated in chloroplasts (Whatley et al., 1951). The exact role of iron in chlorophyll synthesis is not certain but according to DeKock (1971) there is evidence for the involvement of ferrous iron in the condensation of succinic acid and glycine to form y-aminolevulinic acid, which is condensed to form pyrrole groups, which in turn are condensed to form protoporphyrin 1X. Magnesium is then incorporated into the molecule to form chlorophyll, possibly with the catalytic action of iron. Possington (197 1) found that Fe-deficiency stress in spinach reduced both the number of chloroplasts per cell and chlorophyll content per chloroplast. Terry (1979) however, found that Fe stress in sugar beets caused no decline in chloroplasts per unit area (or per cell) or chloroplast volume but rather reduced chlorophyll per chloroplast. Spiller and Terry (1980) cite evidence that suggests that Fe deficiency retards not only chlorophyll synthesis but also the synthesis of the complete light-harvesting apparatus, including chloroplast membranes and the chlorophyll-protein complexes, carotenoids, reaction centers, and electron carriers associated with them. The experimental results regarding the effects of iron deficiency on the photosynthetic rate, expressed as CO, uptake per milligram chlorophyll per hour, are inconsistent because of differences in either plant species or experimental technique. For example, chlorotic sugar cane leaves were found by Naik and Joshi (1979) to contain 20% of the chlorophyll in Fe-treated and recovered leaves. On a leaf-weight basis, the chlorotic leaves could fix CO, at only 45% of the rate of nonchlorotic leaves; on a milligrams-of-chlorophyll basis, however, the chlorotic leaves fixed CO, at 237% of the rate of the nonchlorotic leaves. Naik and Joshi performed their experiment in natural sunlight in India (50,000 lux or more). Stocking (1975) found that Fe-deficient maize leaves, containing 38% of the chlorophyll in Fe-sufficient leaves, fixed CO, at a rate of 16% of the Fesufficient leaves on a dry matter basis and 42% on a milligrams-of-chlorophyll basis; Stocking did not specify illumination. On the basis of the most extensive experiments to date on the dependency of

IRON NUTRITION IN CALCAREOUS SOIL

223

photosynthetic rate and Fe deficiency, Terry (1979) reported that iron chlorosis in sugar beets did not reduce CO, uptake per milligram chlorophyll. Fe-deficient leaves containing 5 pg chlorophyll/cm2 absorbed 53% of the incident white light, while unstressed leaves containing 50 pg chlorophyll/cm2 absorbed 84%. Results of 25 lightKO, experiments demonstrated, however, that in general in chlorotic leaves, chlorophyll is the limiting factor, not because it limits light absorption, but rather because it limits photochemical capacity (the maximum rate at which a leaf can convert nonlimiting light energy to chemical products such as ATP and NADPH) associated with the electron carriers ancillary to chlorophyll. An increased photochemical capacity is particularly desirable at irradiance levels at which quanta may be absorbed faster than they can be converted to chemical energy. For photosynthetically active radiation (400-700 nm) at flux densities of 100 microeinsteins m-, sec - or 25% of full sunlight, the rate of photosynthesis was curvilinearly related to chlorophyll content, and thus limited by the photochemical capacity over the entire range of chlorophyll contents tested (5-65 pg/cm2). Buttery and Buzzel (1977) report concurring data showing that the photosynthetic rate per leaf area was linearly related to chlorophyll per leaf area for soybeans under field conditions. The result of photosynthesis of course is dry matter accumulation and, for agricultural plants, yields of fruit, tuber, and fiber. Sestak et al. (1971) have referred to 10 papers relating plant chlorophyll contents to growth and/or yield in various agricultural plants. Iron is also found in heme compounds, such as the various cytochromes, peroxidase, and catalase, and in phytoferritin and ferredoxin; these compounds may have reduced activities under conditions of iron deficiency (Bar-Akiva and Lavon, 1968; Aganvala and Sharma, 1961). Other enzymes may require Fe(I1) for activation (De Kock, 1971).

',

B.

MECHANISMS OF IRON

UPTAKE

Relatively little is known about the mechanism of iron uptake by plants in soil. Mengel and Barber (1974) calculated that iron intake for field-grown maize averages 5.7 pmol Fe/m root length/day for the first 20 days, declining rapidly until, at 70 days, no further iron is absorbed. O'Connor et al. (1971) pointed out that if mass flow alone were to be responsible for iron uptake of a plant transpiring at a rate of 500 g water/g dry matter and accumulating 100 pg Fe/g dry matter, the soil solution would have to contain 0.2 pprn Fe. Bradford et af. (197 I ) found that the iron content of saturated extracts of 68 soils averaged 0.05 ppm and thus perhaps 0.1 ppm at field capacity; these figures indicate that onehalf to three-fourths of the iron absorbed must come from sources other than mass flow. Oliver and Barber (1966) grew soybeans in pots under both high and

224

YONA CHEN AND PHILLIP BARAK

low transpiration rates in 1OO:O and 60:40 soi1:sand mixtures which had 0.02 ppm Fe in saturated extracts. Mass flow accounted for 3-9% of total Fe uptake. Another 23-56% of iron uptake was attributed by Oliver and Barber to root interception, i.e., assimilation of nutrients present in soil-pore solutions encountered and displaced by roots. Nye and Tinker (1977) argue against the concept of root interception, in which case iron uptake by means other than mass flow total led 91-97% of all iron adsorbed in Oliver and Barber’s experiment. Even less is known about iron uptake mechanisms other than mass flow. One possible mechanism is root contact with iron oxides, permitting a microenvironment favorable to iron uptake. Roots of sunflower, an Fe-efficient plant, have been shown by Kashirad and Marschner (1974) to cause a pH drop in nutrient solution when faced with iron stress, presumably in order to mobilize iron from iron oxides. By use of 59Fe, Azarabadi and Marschner (1979) demonstrated that iron mobilization from fresh ferric hydroxide by corn roots in sand culture was limited to the root-iron interface, and that the overall concentration of iron did not increase in the substrate. Chapman (1939) demonstrated the role of root contact with iron oxides by growing citrus seedling in quartz sand to which had been added 0.1% magnetite (Fe,O,), which contains two Fe(II1) and one Fe(I1) ions per molecule, and circulating nutrient solution maintained between pH 5.8 and 7.0. Seedlings in pots with magnetite grew and were not chlorotic; seedlings in pots without magnetite were stunted and chlorotic, even when nourished with the effluent of the magnetite pot nutrient solution. Chapman also found that addition of CaCO, to the sand culture with magnetite caused chlorosis unless the quantity of magnetite was increased. Mobilization of iron at the root-iron oxide interface seems to be achieved by either pH reduction or a drop in redox potential, or both, depending upon plant species and variety (Olsen and Brown, 1980). In a system in equilibrium with ferrihydrate, the pH of the ferrihydrite environment must be less than 3.3 in order to provide 0.2 ppm Fe in the solution adsorbed by plant roots (O’Connor er al., 1971), and less than 4.0 to supply 0.1 ppm Fe to supplement the 0.1 ppm that may be available by mass flow. These pH values at root-soil contact seem possible in view of the report by Oertli and Opoku (1974) that under favorable conditions maize roots dropped the pH of 2 liters of nutrient solution from 9.2 to 4.0 in 48 hr. The pH drop mechanism has been shown by Raju and Marschner (1973) to cease when the plant has regreened and to reoccur cyclically in response to Fe deficiency stress. A number of studies using nutrient culture demonstrate a drop in the redox potential of the solution. Calculations based on equilibrium with ferrihydrite at pH 7 show that redox patential (Eh) must drop to - 10 mV, or pe must be -0.2, in order to mobilize 0.2 ppm Fe as Fe2+. Interestingly, poor soil aeration in neutral and calcareous soils, which is favorable to the drop in redox, has nonetheless often been linked to the appearance of iron chlorosis (Wallace and Lunt,

IRON NUTRITION IN CALCAREOUS SOIL

225

1960). Evidence has accumulated that iron must be reduced to Fe2+ at the root surface in order to be adsorbed (Chaney et al., 1972). Natural soil chelates may contribute more to iron nutrition than their iron concentration in soil solution would seem to indicate by recharging after discharging iron at the root surface. Glauser and Jenny (1960) reported that the nutrient solution circulating between a pot of slightly acidic soil and a quartz sand culture had a pH of 6.1 and was able to relieve chlorosis in alfalfa in the sand culture. When the soil was amended with CaCO,, the dark-colored percolate had a pH of 8.0 and was unable to green alfalfa in sand culture; alfalfa plants growing directly in the calcareous soil were not chlorotic. Apparently the importance of soil chelates and of the recharging phenomenon is itself pH dependent. C. CAUSESOF IRON DEFICIENCY

Although iron chlorosis may occur in acid soils, iron chlorosis is most commonly identified with calcareous soils, hence the term “lime-induced chlorosis.” Even after decades of research, the exact soil-related cause of limeinduced chlorosis is not known with certainty. Complicating the understanding of iron deficiency is the observation that a particular plant may develop limeinduced chlorosis in a particular soil, yet another species or even another variety of the same species may be markedly more tolerant. Soil lime is in equilibrium with soluble calcium, bicarbonate, carbonate, pH, and CO, in the soil environment; soluble calcium and pH also govern the solubility of the various calcium phosphates which set the soluble phosphorus levels (Lindsay, 1979). Research indicative of the respective roles of soluble calcium, bicarbonate, soil CO,, and phosphorus in the induction of chlorosis is presented in the classic reviews by Wallace and Lunt (1960) and Miller (1960). In an attempt to predict the occurrence of chlorosis in rice based on soil data using statistical techniques, Place et al. (1969) measured pH, electrical conductivity, and concentrations of calcium, magnesium, potassium, and bicarbonate in the saturated extracts of 54 calcareous Arkansas soils, some of which induced chlorosis. The best single predictor, bicarbonate concentration, nonetheless misclassified 35% of chlorosis-producing soils and 42% of normal soils; all combinations of parameters, including all seven taken together, were unable to greatly improve the prediction based on soil solution data. A second approach to the detrimental effects of soil carbonate considers the reactivity of CaCO, rather than the solubility of its constituents. Drouineaux (1942) termed oxalate-precipitable calcium, expressed as calcium carbonate equivalent, “active lime”; active lime in soils was shown to be roughly equal to the lime content of the <20-p.m soil fraction. Secondary carbonates have been found to accumulate in the clay and fine silt fractions (Rostad and St. Arnaud,

226

YONA CHEN AND PHILLIP BARAK

1970); active lime therefore probably has a high specific surface area and may coat soil iron oxides. Yaalon (1957) reported that 10% active lime was the critical level in sensitive crops. Carter (1981) confirms this critical level for four tree species in Canada; no relationship between chlorosis and total CaCO, was found. If the quantity of CaCO, or active lime is indeed related to chlorosis, then calculations of soil mineral equilibrium of the bulk soil are not likely to be helpful since some aspect of the surface area of CaCO, is involved. Besides the strictly soil-related causes of iron deficiency, the soil solution composition may influence plant nutrition in such a way as to induce chlorosis. For example, the mobilization of iron by the export of H + out of the root is facilitated by a compensating high uptake of a permeant cation such as K + or NH,+ (Oertli and Opoku, 1974). On the other hand, uptake of more NO,, SO$-, and phosphate than K, Mg, and Ca necessitates export of OH-. Cunningham (1964) and others have shown that plants as a rule absorb more anions than cations; Cunningham’s data for 62 common plant species show that on the average 3.6 meq anions and 2.5 meq cations are absorbed per gram dry shoot material. To maintain electrical neutrality, the root must exude OH- in amounts equivalent to the excess of anions over cations; in the presence of water and CO,, OH- forms bicarbonate. As mentioned above, the precise manner in which bicarbonate is harmful to iron nutrition is uncertain (Miller, 1960), but clearly the pH-buffering effect of bicarbonate would be deleterious if a significant portion of plant iron is absorbed from iron oxides necessitating pH reduction at points of root contact. Since NO, is the principal anion absorbed by roots, nitrogen nutrition essentially governs the cation:anion balance and bicarbonate excretion into the rhizosphere. For example, Riley and Barber (1969) found that soil, shaken off the roots of soybean plants grown for 3 weeks in a soil fertilized with NO,, had accumulated bicarbonate at a rate of 35 meq/100 g roots, and the pH had risen one whole unit. High levels of NO, have been found to cause iron chlorosis in tobacco, avocado, macadamia, and soybeans by Steinberg et al. (1 949), North and Wallace (1959), and Aktas and van Egmond (1979), respectively. The role of NO, in iron chlorosis is also suggested by a number of reports on association of iron chlorosis with accumulation of organic anions in shoot material; the data of Kirkby and Knight (1977) show that organic anion accumulation increases linearly with increasing NO, concentration in nutrient solution up to a concentration of 6 meq NO, /liter. Kirby and Knight and others explain the organic anion accumulation as a result of reduction of NO, in plant shoots in the course of assimilation of NO,. Association of iron chlorosis with organic anion accumulation would not, therefore, suggest a cause-and-effect phenomenon but rather, that both are different effects of high NO, adsorption rates. Since nitrogen can alternatively be supplied as a cation, i.e., NH,+, the cation:anion balance may be altered, and the tendency to export bicarbonate is reduced or reversed. Esty et al. (1980) found that sorghum in Steinberg’s nu-

IRON NUTRITION IN CALCAREOUS SOIL

227

trient solution, which contains NH; , reduced the pH of the substrate from 6.5 to 4.0; in contrast, in Hoagland’s No. 1 solution, containing all nitrogen as NO,, the initial pH rose from 4.5 to 7.0. Nelson and Shelby (1974) reported that NO, was inferior to NH,f as a nitrogen source for spruce and pine seedlings in nutrient solution, reducing dry matter yield, increasing organic anion content, and causing chlorosis remediable with FeEDDHA. A similar phenomenon has been found by van den Driesshe (1978) for Douglas fir seedlings in nutrient solution and by Farrahi-Aschtiani (1972) for periwinkle plants in soil. Blonde1 (1970) has reported that iron chlorosis induced in field-grown sorghum by calciudmagnesium bicarbonate-rich irrigation water was alleviated by 800 kg ammonium sulfate/ha. NH; fertilizers in soils are in general, however, short lived due to nitrification and volatilization, particularly in calcareous soils. The significance of the cation:anion balance is summarized by Lucas and Knezek (1972) in photographs of petunias fertilized with calcium nitrate, potassium nitrate, ammonium nitrate, and ammonium sulfate, which range from severe chlorosis to no chlorosis. It may be noted that natural vegetation is rarely, if ever, chlorotic; therefore to a certain degree, modern agriculture, with its selection for high yield (i.e., Nefficient and therefore Fe-inefficient plants) and with its extensive use of nitrogen fertilizer, is responsible for the appearance of iron deficiencies. For example, Houba et al. (1971) found that sugar beets excreted bicarbonate until the nitrate source was either removed or depleted. Aktas and van Egmond (1979) investigated the effect of increasing NO, fertilizer on two soybean cultivars. The N-efficient cultivar had a cation:anion balance such that up to 1.3 meq bicarbonate were exported per gram dry matter, and the plants were chlorotic. The less N-efficient cultivar exported between 0.4 meq bicarbonate and 0.4 meq H + per gram dry matter, enjoyed a satisfactory iron supply, and had greater dry matter yields than the N-efficient cultivar. Because of the growing global dependence on modern agriculture and the use of nitrogen fertilizers to sustain it, an apparent increase in iron deficiencies has been identified and must be remedied by either proper iron fertilizers or genetic selection. Other causes of iron deficiency include poor aeration; interactions with other micronutrients such as manganese, copper, and zinc; low or high light intensities; certain organic matter additions to the soil; viruses; and root damage by nematodes or other organisms. Since these causes have been reviewed in detail by Wallace and Lunt (1960) and Olsen (1972), further discussion will be avoided here. D . INDICATORSOF IRONDEFICIENCY

Since lime-induced chlorosis was first demonstrated by Gris in 1843 to be due to iron deficiency, many researchers have performed chemical analysis for total

228

YONA CHEN AND PHILLIP BARAK

iron content of plant tissue with the objective of establishing the critical levels of iron below which deficiency symptoms occur. Jacobson (1943, Wallihan (1955), Gilfillan and Jones (1968), Terry (1980), and Harivandi and Butler (1980) have demonstrated a relationship between total iron content and degree of chlorosis, as measured by chlorophyll content in tobacco, citrus, macadamia, sugar beets, and Kentucky bluegrass, respectively. Wallihan’s values for iron content showed much overlap among sufficient, moderately deficient, and severely deficient leaves. thus precluding the determination of a unequivocable critical level of iron. The cubic correlation found by Terry for hydroponically grown sugar beets was such that at 1 pmol Fe/g dry matter, chlorophyll content of samples varied from 10 to 45 pg/cm2, while the total chlorophyll range was 2-60 pg/cm2. At 45 pg chlorophyll/cm2, leaf iron contents of samples ranged from 0.6 to 2.8 pmol Fe/g dry leaf matter, while the total iron content range was 0.3-3.8 pmol Fe/g. The variance in the iron-chlorophyll relationship is in general so great that the bulk of research indicates that chlorotic leaves may contain as much or more iron than healthy leaves and that total iron may show no relationship to the chlorotic or healthy appearance of the plant. The above conclusion is supported by Oserkowsky (1933), Twyman (1959), De Kock (1955), Agarwala and Sharma (1961), Brown and Jones (1962), Bar-Akiva (1964), Rio et al. (1978), Carter (1980), and Kaytal and Sharma (1980) for pear, tomato, mustard, barley, soybeans, citrus, peas, Scotch pine, and rice, respectively. Review of the several reports of success in establishing a connection between plant iron and chlorophyll indicates that plants need an adequate, timely supply of iron during chlorophyll synthesis; afterward, the plant may continue to accumulate iron without further altering the chlorophyll content. Jacobson and Oertli (1956), for example, modified the earlier findings of Jacobson (1945) that iron was related to chlorophyll by adding the proviso that the relationship is apparent only when iron concentration in the nutrient solution is held strictly constant. The significant correlation between iron and chlorophyll ( r = 0.90) for field-grown Kentucky bluegrass found by Harivandi and Butler (1980) seems to be due to harvesting the sample 3 days after a previous clipping, thus bringing iron uptake and chlorophyll synthesis into close conjunction. Such short growth periods are unusual for most crops and experiments. Because leaves continue to accumulate iron without further altering their chlorotic appearance, total iron content of plants is usually a poor indicator of iron nutrition status. Another approach is to fractionate total plant iron into either “active iron,” soluble in 1 N HCI over 6-24 hr, as described by Jacobson (1945), or ferrous iron, extracted at pH 3 with o-phenanthroline over 16 hr, as described by Kaytal and Sharma (1980). The search for a quantitative measure of iron deficiency has in general turned away ‘from total iron content of leaves to ratios of iron to other elements sus-

IRON NUTRITION IN CALCAREOUS SOIL

229

pected of inhibiting the absorption of iron or causing its internal inactivation. North and Wallace (1952) found that high nitrate absorption caused chlorosis in avocado and therefore suggested the use of Fe/N as a measure of iron deficiency. De Kock (1955) concluded that the degree of chlorosis was related to the ratio of P to Fe for mustard seedlings grown in nutrient solutions. The data of Aktas and van Egmond (1979) may be used to demonstrate the use of P/Fe and Fe/N ratios. Three nonchlorotic treatments of a Fe-efficient soybean cultivar had P/Fe ratios ranging from 32 to 35, whereas chlorotic treatments of a Fe-inefficient cultivar had P/Fe ratios of 59 and 68; the single Fe-inefficient cultivar that regreened after becoming chlorotic had an intermediate P/Fe ratio of 43. The Fe/P ratio would seem to separate the chlorotic and nonchlorotic treatments and perhaps indicates some type of interference of P in Fe uptake (De Kock, 1955). The three nonchlorotic treatments of the Fe-inefficient cultivar exhibited FeiN ratios (milligrams Fe per gram N) of 3.2-4.7, whereas chlorotic treatments of a Fe-inefficient cultivar showed Fe/N ratios of 1.7; the regreened chlorotic treatment had a ratio of 3.4. Both types of ratio, P/Fe and Fe/N, separated chlorotic plants from healthy plants but only one of the two ratios, or neither, is likely to reflect the cause of the iron deficiency. In fact, Aktas and van Egmond present evidence linking the cation:anion balance, and hence the exuding of bicarbonate into the rhizosphere, to the iron status of the plants. The reasons for the success of Fe/N and P/Fe in separating chlorotic plants from healthy plants may be speculated upon as follows: (1) increasing NO, uptake causes imbalance in the cation:anion ratio and exudation of bicarbonate into rhizosphere with concomitant difficulty in absorption of Fe, and thus of significance to Fe/N absorbed; and (2) exudation of bicarbonate may reduce soluble calcium in the rhizosphere, thereby increasing HPOZ- concentrations near the root surface and causing an increase in P uptake, and thus yielding an incidental relation between Fe/P absorbed and the iron status of plants. A third type of indicator of the iron status of plants is the activity of peroxidase, and iron hemoprotein, in leaf tissue. Bar-Akiva (1961) demonstrated that iron-deficiency symptoms in citrus can be distinguished from those of manganese by peroxidase-activity measurements. Bar-Akiva e l al. (1978) reported that maize cultivar Oh40B, recognized as an Fe-inefficient cultivar, had a peroxidase level half that of two other cultivars under equivalent iron nutrition (3 ppm Fe in nutrient solution) and similar uptake; since Oh40B also apparently had a normal appearance, varietal differences in peroxidase levels may exist independent of iron nutrition status. Perhaps the most straightforward approach to quantifying iron chlorosis is determination of chlorophyll content of leaf tissue. In nutrient solution cultures, Rio et al. (1978), Agarwala and Sharma (1961), and Bar-Akiva and Lavon (1968) have shown for peas, barley, and citrus, respectively, that leaf chlorophyll is strongly related to the iron supply in the nutrient solution. Chlorotic

230

YONA CHEN AND PHILLIP BARAK

leaves are readily discerned in the field, and the chlorophyll content is measurable by spectrometric measurements of acetone extracts of leaves; Rodriquez de Cianzio et al. (1979) have shown that visual ratings of chlorosis (0-5) by trained personnel are highly correlated ( r = 0.90) with laboratory analysis of chlorophyll content. The chlorophyll content of leaves is thus a suitable indicator of iron chlorosis since chlorophyll is related to both iron supply and chlorotic appearance of plants. However, determination of iron deficiency by chlorophyll measurements must be checked against the possibility that low values are due to leaf senescence or nutrient stress other than iron, by the use of a control treatment that greens upon use of effective remedies, such as FeEDDHA. This approach is similar to that of Gris in 1843 who first identified chlorosis of grape leaves on calcareous soils as an iron deficiency by observing greening of leaves after spraying with ferrous sulfate. A similar approach was applied by Barak and Chen (1982), who developed a bioassay-type test for the evaluation of iron deficiency.

IV. CORRECTION OF IRON DEFICIENCY A . FOLIARAPPLICATIONS

Foliar applications of inorganic iron salts has a long history dating to 1843, as mentioned above. Since then, many experiments have been performed, particularly with FeSO,, since use of synthetic iron chelates cannot be economically justified in some field crops. Spraying with iron salts alone has usually been found to be relatively ineffective, leaving green spots where drops of iron penetrated the leaf on a chlorotic background. Much research continues to identify effective surfactants to improve coverage of the spray. Uptake from foliar application of iron has been shown to be correlated with both total stornatal area and open stomata during spraying (Eddings and Brown, 1967; Wallihan et al., 1964). A major problem with foliar application is the poor translocation of applied iron within the plant. Eddings and Brown (1967) found that translocation rates were species dependent and did not exceed 50% of the iron applied to a given leaf or leaflet. In the field, respraying is often required, sometimes at 10-day intervals, to provide adequate iron for developing canopy, since Fe translocation from previously treated areas is insufficient. In practice, care must be exercised to avoid leaf burn due to either the iron salt or the surfactant. Foliar sprays other than iron salts include Fe lignosulfonates, Fe polyflavonoids, and Fe chelates. In experiments by Wallihan er al. (1964), these treatments did not surpass foliar applications of FeSO, with a surfactant at equal Fe rates.

23 1

IRON NUTRITION IN CALCAREOUS SOIL

B . SOILAMENDMENTS

1. Synthetic Iron Chelates

The most current method of treating plant iron deficiency by soil amendment is the use of synthetic iron chelates. Chaberek and Martell (1959) define metal complexes as the result of the replacement of water molecules surrounding the metal ion with other ions or molecules, called ligands. When the metal complex is both stable and soluble, the ligand is termed a metal-sequestering agent. When the ligands are attached to each other as well as to the metal, forming a ring, the metal complex is known as a metal chelate. Stability constants of proton-ligand and metal-Iigand complexes permit calculation of equilibrium concentrations of Fe-ligand in competition with Ca2 , Mg2 , Zn2 , Mn2+, A13 , etc., in soil solution as a function of pH. Examples of these calculations are given by Lindsay (1979). Reactions and stability constants are generally written as follows: +

Mn+ + LmK,,

+

+

+

% MLn-m

= [MLn-m]/[Mn+] [Lm-]

where M and L indicate metal and ligand, respectively, and brackets indicate molar concentration; the stability constant applies to a specific temperature, The rapid development of synthetic chelates has meant that agricultural testing has often preceded or been simultaneous with physicochemical investigations necessary to understand their behavior. The most important of the new class of synthetic amino acids developed in the 1930s, EDTA (ethylenediaminetetraacetic acid), was characterized by complexometric titrations by Schwarzenback between 1948 and 1954 and has come to be of great commercial importance in industry and food technology. FeEDTA was first used in plant nutrition by Jacobson (1951) who noted that FeEDTA was not stable in nutrient solution above pH 6. Stewart and Leonard (1952) in pioneering research reported the remedy of iron chlorosis in citrus trees growing in acid sandy soil by use of FeEDTA. Although subsequent work with FeEDTA in calcareous soils often showed noticeable improvement in the appearance of chlorotic plants, results usually fell short of a full remedy (Chaberek and Martell, 1959). Other synthetic amino acids were quickly tested for effectiveness in remedying lime-induced chlorosis. Wallace and Lunt (1960) report a large number of chelates tested in 1953. Among them was FeEDDHA (ethylenediaminedi-o-hydroxyphenylacetic acid), which by the early 1960s had emerged as the most effective chelate for relieving lime-induced chlorosis. Despite the high costs FeEDDHA is today the remedy of choice for correcting lime-induced iron deficiency. The success of synthetic metal chelates does not necessarily mean that mass flow of natural chelates is the primary mechanism by which plants normally absorb iron.

232

YONA CHEN AND PHILLIP BARAK

Two crucial questions relative to the agricultural cost-benefit equation are (1) recovery rate by plants, and (2) residual fertilizing effect. In a pot experiment, Burau et al. (1960) recovered 0.5-1.8% of 59Fe applied to a calcareous clay soil in the areolar portions of soybeans 30 days after fertilization with FeEDDHA; soybeans were 5 weeks old before treatment with 0.04, 1, and 5 mg Fe/kg (ppm Fe) soil as FeEDDHA. Plants were grown without drainage. Experiments by Salardini and Murphy (1978) demonstrated Fe recovery of 0.5-1.5% for sorghum grown for 30 days after emergence in soil which received 20, 40, and 80 ppm Fe as FeEDDHA. Two crops of sorghum grown in three soils by ElKhattari and Anderson (1979) extracted a maximum of l .2% of 59Feapplied as FeEDDHA; the second crop removed less 59Fe than the first. Residual effects of FeEDDHA are influenced by both simple leaching out of root zone by rainfall and irrigation and microbial decomposition of ligand and other iron-fixing processes. Follett and Lindsay (197 1) investigated iron status of soils which received 5 ppm Fe as FeEDDHA by the DTPA micronutrient soil test. For 11 soils an average of 87% of added Fe was recovered by DPTA extraction immediately after fertilization. After 14 weeks and two crops, DTPArecoverable Fe had decreased to 26% of added Fe. Salardini and Murphy (1977) monitored DTPA-extractable Fe in two calcareous silt-loam soils maintained at field capacity in pots and fertilized with 20,40, and 80 ppm Fe as FeEDDHA. In the Fe-deficient soil, recoverable Fe declined to 52% or less after 1 week and continued to decline slowly to 19-32% after 8 weeks. The exact fate of FeEDDHA was not established; it was uncertain as to whether FeEDDHA had been adsorbed on a solid surface or if Fe had been stripped from FeEDDHA, or if the EDDHA had decomposed. Since FeEDDHA exists at all soil pHs as a soluble anion, FeEDDHA in the field may also be removed from the root zone by leaching. Hochberg (1974) demonstrated that FeEDDHA passed through a column of sandy soil with the concentration of chelated iron unchanged. Hochberg also made field observations in a number of commercial agricultural fields, greenhouses, and citrus groves under sprinkler or drip irrigation regimes in which FeEDDHA has been used; analysis of soil cores revealed that more than half of the total water-soluble Fe, presumably as FeEDDHA, is 20 cm or more below the soil surface. Plant recovery, residual DTPA-extractable Fe, and translocated water-soluble Fe have been referred to previously as a problem of Fe applied as FeEDDHA, but the fact of the matter is, of course, that Fe is a negligible part of the cost of FeEDDHA.

2 . Organoiron Complexes Since soluble soil iron exists mainly as organoiron complexes, efforts have turned to the use of organic solid wastes, either enriched with iron or unenriched, as an iron source for sensitive crops in deficient soils. Among the organic wastes found effective in increasing iron uptake are animal manures of various sorts, composts of plant refuse, and extracts from forest by-products. Growth-promot-

233

IRON NUTRITION IN CALCAREOUS SOIL

ing effects of powdered charcoal, brown coal, and lignite have been found, in retrospect, also to be due to organoiron complexes. The effectiveness of these organoiron complexes is usually attributed to their similarity to soil organic matter, which is known to have a significant metalbinding ability. The binding ability of humic acids is often attributed to the presence of both carboxylic and phenolic groups which chelate in the manner of salicylic (0-hydroxybenzoic) acid. As with synthetic iron chelates, organoiron complexes have been tested for agricultural value without prior successful determination of the formula of the complex and its stability constant. Of the five methods of determining stability constants given by Chaberek and Martell (1959), none are suitable for a heterogenous substance of uncertain molarity. In addition, the method of determining the formula of the iron chelates by noting the appearance of the metal hydroxide, thus indicating saturation of the chelate, is invalid if, as has been shown, the humic acid peptizes ferric hydroxide sols (Deb, 1949), or if humic acids folocculate in the presence of metal ions (MacCarthy and Mark, 1976). 1. Manure and Compost. Several reports on the effectiveness of manure in solving problems of iron chlorosis have been published. For example, in a calcareous soil in Senegal which resulted in ivory-colored sorghum plants, Blonde1 (1970) increased dry matter yields from 520 to 970 kg/ha by adding 20 tons farmyard manure (FYM)/ha. Miller et al. (1969) have recommended the use of poultry manure for correction of Zn and Fe deficiencies in plants. Parsa and Wallace (1979) have shown that dog manure (1850 ppm Fe) added at a rate of I .5% to a calcareous loamy soil significantly increased dry matter yield of and iron uptake by sorghum plants in a pot experiment; ashed dog manure produced a lesser increase in dry matter and no increase in iron uptake. Parsa and Wallace found that the dog manure had an infrared spectrum similar to that of soil fulvic acids and had strong absorptions due to carboxylic and phenolic groups; complexing between these groups and iron was concluded to be the mechanism by which iron nutrition is improved by manure. Enrichment of manure with Fe as FeSO, has also been shown to be of interest. Thomas and Mathers (1979) grew three crops of sorghum in pots of calcareous sandy loam treated with combinations of up to 4% manure and 40 pprn Fe (as FeSO,). The first crop demonstrated a large dry yield increase with 2% manure and a small but significant increase with 20 ppm Fe. By the third crop, yields had fallen in general, but the yield of the optimal treatment, 2% manure 20 ppm Fe, was greater than the sum of yields of 2% manure and 20 ppm Fe alone, thereby revealing the presence and significance of an organoiron fertilizer. In a field experiment on a calcareous soil, Mann et al. (1978) investigated the interaction between the addition of 0 and 250 ppm Fe (as FeSO,) with 0 and 10 tons manure/ha on DTPA-extractable Fe and maize yield; statistical analysis verified that the Fe-manure treatment resulted in a greater yield increase than the sum of Fe and FYM alone. Further, the DTPA-extractable iron after maize harvest showed the same type of Fe-FYM interaction; values were 5.5,, 5.9, 7.0, and

+

234

YONA CHEN AND PHILLIP BARAK

10.8 ppm Fe for the control, +FYM, +Fe, and +FYM+Fe, respectively. Composts have been shown to be of use in alleviating iron shortages. Francis et al. (1979) reported on a field experiment where sorghum was grown in a calcareous soil to which had been added 0-45 ppm Fe (as FeSO,) composted with 0-20 tons cotton leavedha. Significant dry matter yield and Fe uptake increases were measured; the Fe and compost interaction was also statistically significant. Takkar (1969) found that 0-8% sugar trash added to a calcareous soil and incubated at field capacity initially caused a decline in ammonium acetate (pH 3)-extractable iron but after 35 days, extractable iron exceeded that present initially and rose to significantly higher values. 2. Sewage Sludge. Sewage sludge in some industrial nations may pose a disposal problem because of toxic levels of heavy metals; otherwise sewage sludge may find use as an iron source. Parsa and Wallace (1979) reported that activated sewage sludge (4200 pprn Fe) from Los Angeles mixed with a calcareous loam at a rate of 1.5% increased both dry matter of sorghum grown in pots and Fe uptake. Neither sewage sludge ash, nor 5 ppm Fe as FeSO, or as FeEDDHA, was able to match the yield and Fe uptake below those of l S % , indicating that a toxic level had been attained-perhaps with another microelement in the sludge. Parsa and Wallace present infrared spectrum analysis of the NaOH-extractable, pH 2-soluble fraction of the sewage sludge (termed fulvic acid fraction); Absorption due to carboxyl, N-bearing, phenolic OH, and sulfonic acid groups is strong. Baham et al. (1978) experimented with gel filtration of two fulvic acid fractions of sewage sludge (50 and 105 pprn Fe). Elution data showed multiple maxima in organic carbon concentration, a single maximum in C1- concentration, and multiple maxima of iron concentration, which coincide with both the organic carbon maxima and the chloride maximum. Analytical data showed that 13-24% of the iron was eluted together with the chloride maximum, indicating an inorganic complex, possibly hydroxides; 61-85% of the iron was distributed in the fractions collected near the carbon maxima, indicating organoiron complexes in sewage sludge. Abdou and El-Nennah (1980) have shown the cumulative effect of domestic sewage sludge disposed on calcareous loamy sand for up to 45 years; total iron was increased 6-fold, ammonium acetate (pH 2.8)-extractable iron was increased 10-fold, and water-soluble iron, 6-fold. Availability of manganese and zinc was also increased simultaneously. Though there is little doubt that urban sewage sludge may serve as an iron source, problems of toxicity and heavy metal contamination must be solved first. 3. Polyflavonoids and Lignosulfonates. Attention has turned to the use of polyflavonoids and lignosulfonates, both by-products of forest-product manufacturing, as potentially economical iron chelates. Although polyflavonoids and lignosulfonates have been tested as both foliar sprays and soil amendments, only the latter will be considered here. Rayonier Incorporated has marketed a polyflavonoid product called “Rayplex,” a chemical extract of western hemlock bark enriched to 11% Fe. Chesnin (1 968) found that maximum dry matter yield

IRON NUTRITION IN CALCAREOUS SOIL

235

and total Fe uptake were achieved with 2.5 mg Fe as Fe polyflavonoid/kg soil for sorghum in pots of a 10%CaCO, sandy loam. Richardson (1967) used 34 kgiha Rayplex by placing Rayplex granules 4 inches deep and 4 inches to one side of chlorotic sorghum plants in Arizona; apparently some effect was achieved, but the plants failed to yield harvestable amounts of mature grain. Walker and Smith (1967) applied 2.3 kg Rayplex (granular) per peach and pear tree in late May and reported a positive response. Salardini and Murphy (1977) measured DTPAextractable Fe in a normal and calcareous, Fe-deficient soil incubated at field capacity with 0, 20, 40, and 80 ppm Fe as Fe-polyflavonoid; iron levels fell quickly in the first week and by the end of 8 weeks, DTPA-extractable Fe ranged from 8 to 13 ppm for the normal soil, regardless of treatment, and from 3 to 6 ppm for the Fe-deficient soil. Lignin extracted from wood pulp by the sulfite process contains sulphonic groups and has been marketed as “Orzan” by the Crown Zellerbach Corporation, and as “Greenz 26” (Little, 1971). Using 59Fe for their study of soybeans in pots of calcareous loam, Wallace and Ashcroft (1956) found the relative effectiveness of Fe-lignosulfonate (4.6% Fe), FeEDTA, and FeEDDHA was 1:3:15 at a rate of 260 kg Felha. Chesnin (1968) compared a lignosulfonate with a Kraft-process lignin and polyflavonoids. The lignosulfonate and polyflavonoid were comparable to each other; the Kraft lignin seemed to have an advantage over the other treatments at 2.5 ppm Fe, presumably due to a lower molecular weight. After 8 weeks’ incubation in soil, lignosulfonate treatments had 9- I3 and 3-6 ppm DTPA-extractable iron in a normal and Fe-deficient soil, respectively. Overall polyflavonoids and lignosulfonates seem to evoke similar plant responses and have similar persistence in soil. 4. Coal, Lignite, and Peat. According to the summary presented by De Kock and Strmecki (1954), the growth-promoting effects of charcoal were recognized at the turn of the century and stirred interest in the fertilizing effects of brown coal and lignite as well. Research showed some brown coals in soil or water culture to be growth promoting while others were growth inhibiting. Theories based on hormonal or antibiotic characteristics of the coals and lignite were advanced in the 1930s; at the same time, evidence accumulated linking the stimulatory effect of coals to the humic acids they contain and linking the positive effect of humic acid to the maintenance of iron in a soluble form available to microorganisms and plants. De Kock and Strmecki (1954) investigated the growth promoting effects of lignite from Yugoslavia by comparing various size fractions of the lignite dust with a HCI extract, a 2% sodium oxalate extract (humic acid), and ash of both whole lignite and residues after extraction; mustard plants were grown in nutrient solution without iron. Of the various lignite size fractions, the finest fraction with a diameter less than 0.15 mm obtained the greatest growth response. Only treatments that combined humic acid with either the HCI extract which contained iron or 13 pprn Fe as FeCI, attained dry matter yields equal or surpassing those of the whole lignite. Extracts of lignite with ether, a nonpolar fluid intended to extract possible hormonal

236

YONA CHEN AND PHILLIP BARAK

stimulants, caused no growth increase at all. Although the growth-promoting effects were clearly due to the iron humates in the lignite, a separate experiment showed that the greening ability of the iron humate was lower than that of EDTA, though greater than citrate, on a per iron-added basis. Results of tests of the fertilizing effects of humic acids extracted by KOH or ammonia from a moderately to well-disintegrated reed sedge/sphagnum moss peat, a decomposed woody peat, and a partly decomposed reed sedge peat were reported by Burau ef al. (1960). Soybeans were grown in pots of silty clay soil containing 9% organic matter and 6%CaCO,. Liquid humic acids were enriched with iron by either 100 hr of contact with fresh ferric hydroxide, or addition of ferric sulfate or ammonium ferrous sulfate and titration to neutral pH; precipitates were removed by centrifugation. Iron humates were applied at rates of 10, 25, and 40 mg Fe/kg soil. In addition, ferric chloride at rates equal to those of iron humates was applied in 150 ml of solution brought to pH 2 with HCl. After 30 days’ growth, FeEDDHA at the 1-ppm-Fe rate was found to give plant iron concentrations equivalent to or slightly superior to iron humate treatments at the 10-ppm-Fe rate. No significant differences among the iron humates due to peat type, extracting base, or enrichment method were found. The iron humates were in fact equivalent or slightly inferior to acidic ferric chloride solutions. The conclusion of Burau et al. (1960) was that “further development of alkali extracts from peats as amendments to calcareous, chlorosis-producing soils for the purpose of increasing availability of iron does not appear warranted.” Little (1971)reports that iron sulfate was applied alone or mixed with peat (1/2 oz FeSO,/bu peat, type unspecified) at rates of 8.2 kg FeSO,/tree (type unspecified) in England with only partial success; the treatment was reported to be successful for greenhouse tomatoes, roses, and pot plants. No further details are given by Little, and no report on the subject is indexed in Soil and Fertilizers Index to date. Chen et al. (1982a) reported the remedy of lime-induced chlorosis by ironenriched peat in a pot experiment with peanuts grown on a rendzina soil containing 63%CaCO,. Chlorophyll levels reached those of FeEDDHA-treated plants. In a field experiment conducted by these investigators the application of ironenriched peat to peanuts grown on a soil containing 42% CaCO, resulted in yields amounting to 175 and 80% of those of control and FeEDDHA treatments, respectively. Iron-enriched peat was found to be superior to FeEDDHA and FeSO, in a field experiment with gladioli (Chen et al., 1982b). In this series of experiments, application rates of Fe in the form of iron-enriched peat were much higher than those of Fe as FeEDDHA. Other research in a similar direction has been carried out on ammonium nitrohumic acid, (NH,-NHA), which is produced by treating brown coal with nitric acid and neutralizing with ammonia. Aso and Sakai (1963) showed NH,NHA to be globular colloids with particle diameters between 0.01 and 0.1 pm; NH,-NHA was shown to complex with iron in nutrient culture of mulberry seedlings and to cause regreening in chlorotic barley. Overall, NH,-NHA

IRON NUTRITION IN CALCAREOUS SOIL

237

seemed equivalent or inferior to EDTA when both were used in nutrient solution at a rate of 2.5 ppm in the presence of 1 ppm FeCl,. Summarizing the research on the use of coal and peat as iron fertilizers, it may be said that whereas fertilizing effects of whole coal or peat have been noted in both soil and nutrient cultures, iron humates have been found to be relatively ineffective in soil culture when compared to synthetic chelates such as FeEDDHA on a per iron and per weight basis. It should be noted that the Commonwealth Bureau of Soils Bibliographies Nos. 997 and 1903 list 94 references for research on lignite and coal as fertilizers, published between 1950 and 1972. The bulk of the research has been performed in Eastern Europe and has not been translated into English. Abstracts given in the Bibliographies indicate that the research has proceeded in the direction of nitrogen or humic acid fertilization with little or no application to problems of iron deficiency. C. GENETICSELECTION OF CULTIVARS

It is not unusual for plants of the same species to display varying degrees of iron deficiency when grown in the same soil; such an occurrence demonstrates that different varieties of the same species respond differently to iron-deficiency stress. Varietal differences in iron nutrition have been noted in avocado, barley, bean, citrus, corn, cotton, eucalyptus, grape, Kentucky bluegrass, oats, peanuts, rice, sorghum, soybeans, and strawberries. In some cases, varietal differences are due to a single recessive gene which does not function properly. For example, the “yellow-stripe’’ gene ys-1 in corn, first reported in 1929, causes plants to be Fe inefficient because of an-inability to reduce Fe3+ to Fez+ at the root surface. Further work by Brown and Ambler (1970) showed that lines homozygous for ys-1 were also unable to reduce the pH of a nutrient solution. A single gene mutant, fer, of the tomato is unable to reduce redox potential or pH in the root environment, and has difficulties translocating iron from the roots to the shoot (Brown et al., 1971). Other varietal differences are apparently the result of multigenic combinations sometimes involving especially efficient uptake of some antagonistic nutrient, as in the case of nitrate-efficient T-203 soybeans (Aktas and van Egmond, 1979) or P-efficient Oh40B corn (Odurukwe and Maynard, 1969). Screening varieties in field tests for resistance to iron deficiency is a relatively straightforward procedure but time consuming. Brown (1976) recommends a quick screening procedure whereby seedlings grown in a modified Steinberg solution are transferred to a modified Hoagland and Arnon No. 1 solution containing 0.2 ppm Fe as FeHEDTA; after 6 days, plants may be graded for chlorosis. Additional Fe stress may be screened before field tests are conducted to measure yield potential. In the short run, cultivars should be selected for calcareous soils known to induce iron deficiency. Selection of resistant rootstock is already common for grapes and citrus and is possible also for avocado. In the long run, cultivars may be

238

YONA CHEN AND PHILLIP BARAK

specifically bred for problematic soils with iron efficiency in mind, with the aim of avoiding the financial loss of yield reduction due to uncorrected iron deficiency or the cost of iron fertilizers. ACKNOWLEDGMENT The authors gratefully acknowledge partial financial support for this research by the Israel Fertilizers Research Center.

REFERENCES Abdou, F. M., and El-Nennah, M. 1980. Planr Soil 56, 53-57. Agarwala, S. C., and Sharma, C. P. 1961. Physiol. Plant. 14, 275-283. Aktas, M., and van Egmond, F. 1979. Plant Soil 51, 257-274. Aso, S., and Sakai, I. 1963. Soil Sci. Plant Nutr (Tokyo) 9, 85-91. Azarabadi, S., and Marschner, H. 1979. 2. Pjlanzenernaehr. Bodenkd. 142, 751-764. Baham, J . , Ball, N. B., and Sposito, G. 1978. J . Environ. Qua/. 7, 124-127. Barak, P., and Chen, Y. 1982. Soil Sci. Soc. Am. J . (in press). Bar-Akiva, A. 1961. Nature (London) 190, 647-680. Bar-Akiva, A. 1964. In “Plant Analysis and Fertilizer Problems” (C. Bould, P. Prevot, and J. R. Magnes, eds.), Vol. 4, pp. 9-25. Geneva, New York. Bar-Akiva, A,, and Lavon, R. 1968. J . Agric. Res. 18, 145-153. Bar-Akiva, A,, Maynard, D. N., and English, J. E. 1978. Hortic. Sci. 13, 284-285. Blondel, D. 1970. Agron. Trop. Paris 25, 555-560. Boer, G. T., de, and Reisenauer, H. M. 1973. Commun. Soil Sci. Plant Anal. 4, 121-128. Bohn, H. L. 1967. Soil Sci. Soc. Am. Proc. 31, 641-644. Bradford, G. R., Bair, F. L., and Hunsaker, V. 1971. Soil Sci. 112, 225-230. Brown, J. C. 1976. Cornell Univ. Agric. Exp. Stn. Mem. pp. 355-358. Brown, J. C., and Ambler, J. E. 1970. Soil Sci. SOC.Am. Proc. 34, 249-252. Brown, J . C., and Jones, W. E. 1962. Soil Sci. 94, 173-179. Brown, J. C., Chaney, R. L., and Ambler, J. E. 1971. Physiol. Plant. 25, 48-53. Burau, R. G., White, R. G., and MacGregor, J. M. 1960. Congr. Soil Sci., 7th Madison. Wi.r. 3, 544-553. Buttery, B. R., and Buzzell, R. I. 1977. Can. J . Planr Sci. 57, 1-5. Carter, M. R. 1980. Planr Soil 56, 293-300. Carter, M. R. 1981. Can. J. SoilSci. 61, 173-175. Chaberek, S., and Martell, A. E. 1959. “Organic Sequestering Agents.” Wiley, New York Chaney, R. L., Brown, J. C., and Tiffin, L. 0. 1972. Plant Physiol. 50, 208-213. Chapman, H. D. 1939. Soil Sci. 48, 309-315. Chen, Y . , Navrot, J., and Barak, P. 1982a. J. Plant Nurr. (in press). Chen, Y., Steinitz, B., Cohen, A., and Elber, Y. 1982b. Sci. Hortic. (in press). Chesnin, L. 1968. In “Isotopes and Radiation in Soil Organic-Matter Studies,” pp. 411-419. IAEA, Vienna. Cox, F. R., and Kamprath, E. J. 1972. In “Micronutrients in Agriculture” (J. J. Mortvedt, P. M. Gordiano, and W. L. Lindsay, eds.), pp. 289-307. Soil Sci. SOC.Am., Madison, Wisconsin. Cunningham, R. K. 1964. J. Agric. Sci. 63, 109-1 11. Deb, B. C. 1949. J. Soil Sci. 1, I 12- 122. De Coninck, F. 1980. Geoderma 24, 101-128.

IRON NUTRITION IN CALCAREOUS SOIL

239

De Kock, P. C. 1955. SoilSci. 79, 167-175. De Kock, P. C. 1971. Tech. Bull. Min. Agric., Fish. Food No. 21. De Kock, P. C., and Strmecki, E. L. 1954. Physiol. Plant. 7, 503-512. Driessche, R., van den. 1978. Plant Soil 49, 607-623. Drouineau, G. 1942. Ann. Agron. 12, 441-450. Eddings, J. L., and Brown A. L. 1967. Plant Phvsiol. 42, 15-19. El-Khattari, S. K., and Anderson, W. B. 1979. Agron. Abstr. 71, 170. Esty, J. C., Onken, A. B., Hossner, L. R., and Matheson, R. 1980. Agron. J . 72, 589-592. Farrahi-Aschtiani. S. 1972. Z . Pflanzenernaehr. Bodenkd. 131, 190-196. Follett, R. H. and Lindsay, W . L. 1971. Soil Sci. Soc. Am. Proc. 35, 600-602. Francis, H. J., Rajagopal, C. K., and Krishnamoorthy, K. K. 1979. Mysore J . Agric. Sci. 13, 159-164. Gilfillan, I. N., and Jones, W . W. 1968. Proc. Am. Soc. Hortic. Sci. 93, 210-214. Glauser, R., and Jenny. H. 1960. Agrochimica 4, 263-278. Harivandi, M. A,, and Butler, J. D. 1980. Hortic. Sci. 15, 496-497. Hochberg, M. 1974. M. Sc. thesis, Fac. Agric., Hebrew University of Jerusalem, Rehovot. Houba, V. J. G., van Egmond, F., and Wittch, E. M. 1971. Neth. J . Agric. Sci. 19, 39-47. Jacobson, L. 1945. Plant Physiol. 20, 233-245. Jacobson, L. 1951. Plant Physiol. 26, 411-413. Jacobson, L., and Oertli, J. J. 1956. Plant Physiol. 31, 199-204. Johnson, G. V., and Young, R. A. 1968. Agron. Abstr. 60, 83. Kashirad, A,, and Marschner, H. 1974. Plant Soil 41, 91-101. Kaytal, J. C., and Sharma, B. D. 1980. Plant Soil 55, 105-119. Kirkby, E. A , , and Knight, A. H. 1977. Plant Physiol. 60, 349-353. Lindsay, W. L. 1979. “Chemical Equilibrium in Soils.” Wiley, New York. Lindsay, W. L., and Norvell, W . A. 1969. Agron. Abstr. 61, 84. Little, R. C. 1971. Tech. Bull. Min. Agric., Fish. Food No. 21. Lucas, R. E., and Knezek, B. D. 1972. In “Micronutrients in Agriculture’’ (J. J. Mortvedt, P. M. Giordano, and W. L. Lindsay, eds.), pp. 265-288. Soil Sci. SOC.Am., Madison, Wisconsin. MacCarthy, P., and Mark, H. B., Jr. 1976. Soil Sci. Soc. Am. J . 40, 267-276. McKeague, J. A. 1968. Can. J . SoilSci. 48, 27-35. McKeague, J. A., Brydon, J. E., and Miles, N. M. 1971. Soil Sci. Soc. Am. Proc. 35, 33-38. Mann, M. S., Takkar, T. S., Bansal, R. L., and Randhawa, N. S. 1978. J . Indian Soc. Soil Sci. 26, 208-2 14. Mengel, D. B., and Barber, S. A. 1974. Agron. J . 66, 399-402. Miller, B. F., Lindsay, W. L., and Parsa, A. A. 1969. Proc. Conf. Animal Waste Manage., Cornell Univ. Miller, G. W. 1960. Soil Sci. 89, 241-245. Naik, G. R., and Joshi, G. V. 1979. Plant Soil 53, 505-51 1 . Nelson, L. E., and Shelby, R. 1974. Plant Soil 41, 573-588. North, C. P., and Wallace, A. 1952. C a l f . Avocado Soc. Yeurh. 37, 177-186. North, C. P., and Wallace, A. 1959. C a l f . Macadamia Sac. Yearb. 5, 54-67. Norvell, W. A. 1970. Ph.D. thesis, Colorado State Univ., Ann Arbor, Michigan. (Univ. Microfilms; Diss. Abstr. 31-51 I IB). Nye, P. H., and Tinker, P. B. 1977. “Solute Movement in the Soilroot System.” Univ. of Calif. Press, Berkeley. O’Connor, G.A., Lindsay, W. L., and Olsen, S. R. 1971. Soil Sci. Soc. Am. Proc. 35,407-410. Odurukwe, S . 0.. and Maynard, D. N. 1969. Agron. J . 61, 694-697. Oertli, J . J., and Opoku, A. A. 1974. Soil Sci. Soc. Am. P r m . 38, 451-454. Oliver, S., and Barber, S. A. 1966. Soil Sci. Suc. Am. Proc. 30, 468-470. Olomu, M. 0.. Racz, G. J., and Cho, C. M. 1973. Soil Sci. Soc. Am. Proc. 37, 220-224.

240

YONA CHEN AND PHILLIP BARAK

Olsen, R. A,, and Brown, J. C. 1980.1. Plant Nutr. 2, 629-645. Olsen, S. R. 1972. In “Micronutrients in Agriculture” (J. J. Mortvedt, P. M. Giordano, and W. L. Lindsay, eds.), pp. 243-264. Soil Sci. SOC.Am., Madison, Wisconsin. Olson, R. V., and Carlson, C. W. 1950. Soil Sci. SOC. Am. Proc. 14, 109-112. Oserkowsky, J. 1933. Planr Physiol. 8, 449-468. Parsa, A. A,, and Wallace, A. 1979. Plant Soil 53,455-461. Perdue, E. M., Beck, K. C., and Reuter, J. H. 1976. Nature (London) 260, 418-420. Place, G. A., Dunn, J . E., and Hardy, G. W. 1969. Soil Sci. 107, 31-38. Possington, J . V. 1971. In “Recent Advances in Plant Nutrition” (R. M. Samish, ed.), pp. 155-165. Gordon & Breach, New York. Raju, K. V., and Marschner, H. 1973. 2. Pjlanzenernaehr. Bodenkd. 133, 227-241. Richardson, G . 1967. Farm Tech. 23, (6). Riley, D., and Barber, S. A. 1969. Soil Sci. SOC. Am. Proc. 33, 905-908. Rio, L. A., del, Gomez, M., Yanez, J., Leal, A., and George, J. L. 1978. Plant Soil 49, 343-353. Rode, A. A. 1937. “Podzol-forming Process.” Acad. Sci. USSR, Moscow; Isr. Prog. Sci. Trans/., Jerusalem, 1970. Rodriguez de Cianzio, S., Fehr, W. R., and Anderson, 1. C. 1979. Crop Sci. 19, 644-646. Rostad, H. P. W., and St. Arnaud, R. J. 1970. Can. .I. Soil Sci. 50, 65-70. Salardini, A. A,, and Murphy, L. S. 1977. Soil Sci. 124, 219-225. Salardini, A. A , , and Murphy, L. S. 1978. Plant Soil 49, 57-70. Schnitzer, M., and Kodama, H. 1977. p. 741-770. In “Micronutrients in Agriculture” (J. J. Mortvedt, P. M. Gordiano, and W. L. Lindsay, eds.), pp. 741-770. Soil Sci. Soc. A m . , Madison, Wisconsin. Schwertmann, U. 1966. Narure (London) 212, 645-646. Schwertmann, U., and Taylor, R. M. 1977. In “Minerals in Soil Environments” (J. B. Dixon, S. B. Weed, J. A. Kittrick, M. H. Milford, and J. L. White, eds.), pp. 145-180. Soil Sci. SOC.Am., Madison, Wisconsin. Sestak, Z., Catsky, J . , and Jarvis, P. G . 1971. “Plant Photosynthetic Production: Manual of Methods.” Junk N. V., The Hague. Sinha, M. K., Dhillon, S. K., Dhillon, K. S., and Dyanand, S. 1978. Aust. J . SoilRes. 16, 19-26. Spiller, S . , and Terry, N. 1980. Plant Physiol. 65, 121-125. Steinberg, R. A,, Bawling, J. D., and McMurtrey, J. E., Jr. 1949. Science 110, 714-715. Stewart, I . , and Leonard, C. D. 1952. Science 116, 564-566. Stocking, C. R. 1975. Plant Physiol. 55, 626-631. Takkar, P. N. 1969. Soil Sci. 108, 108-1 12. Terry, N. 1979. In “Photosynthesis and Plant Development” (R. Marcelle, H. Clijsters, and M. Van Pouke, eds.), pp. 151-160. Junk NV, The Hague. Terry, N. 1980. Plant Physiol. 65, 114-120. Thomas, J . D., and Mathers, A. C. 1979. Agron. J . 71, 792-794. Twyman, E. S. 1959. Plant Soil 10, 375-388. Wada, K. 1977. In “Minerals in Soil Environments” (J. B. Dixon and S. B. Weed, eds.), pp. 603-638. Soil Sci. SOC.Am., Madison, Wisconsin. Wada, K., and Higashi, T. 1976. J . Soil Sci. 27, 357-368. Walker, D. R., and Smith, R. L. 1967. Farm Tech. 23(6). Wallace, A., and Ashcroft, R. T. 1956. Soil Sci. 82, 233-236. Wallace, A , , and Lunt, 0. R. 1960. Am. SOC. Horric. Sci. 75, 819-841. Wallihan, E. F. 1955. Am. J. Bot. 42, 101-104. Wallihan, E. F., Embleton, T. W., and Sharpless, R. G. 1964. Am. Soc. Horr. Sci. 85, 210-217. Whatley, F. R., Ordin, L., and Arnon, D. I. 1951. Plant Physiol. 26, 414-418. Yaalon, D. H. 1957. Plant Soil 8, 275-288.

ADVANCES IN AGRONOMY, VOL. 35

NITROGEN TRANSFORMATIONS IN WETLAND RICE SOILS N. K. Savantland S. K. De Datta Department of Agronomy, International Rice Research Institute, Manila, Philippines

............................

I. Introduction

11. Chemical Nature of Soil Nitrogen.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

111, Physical and Physicochemical Processes Relevant to Nitrogen Transformations A. Transport.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Sorption-Desorption.. .................... C. Ammonia Volatilization . . . . . . . . . . . . . . . . . . . IV. Biochemical Nitrogen Transformations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Ammonification-Immobilization .............................. B. Nitrification-Denitrification.. . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . C. Urea Hydrolysis ......................... .... V. Fate of Fertilizer Nitrogen.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Fate of Basal Incorporated Nitrogen . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . B. Fate of Deep-Placed Nitrogen ............................ C. Fate of Topdressed Nitrogen . ............................ VI. Regulating Nitrogen Transformation Processes . . . . . . ........................

................ ...............................

241 244

26 I 261 274 285 286 286 289 290 29 1 293 294

I. INTRODUCTION Wetland soils differ markedly from well-drained dryland (upland) soils. For Asia, a wetland rice soil is a soil (or land) in a bunded area that is prepared by tillage such as plowing followed by puddling for transplanted rice. The area may be rainfed or irrigated or both; the crop has a high water requirement. The characteristic soil moisture regime vanes from near saturation to submergence (or flooding) resulting in several centimeters of standing floodwater for most of the crop growth period (De Datta and Barker, 1978). In other areas of the world-the United States, Australia, parts of Europe, and IPresent address: International Fertilizer Development Center, P. 0. Box 2040, Muscle Shoals, Alabama 35660. 24 1

Copyright 0 1982 by Academic Press. Inc. All rights of reproduction in any form reserved. ISBN 0-12-000735-5

242

N. K. SAVANT AND S. K. DE DATTA

in some Asian countries-rice land is prepared dry and flooded later. Therefore, neither puddling nor transplanting is an essential ingredient of a wetland rice soil. Many results discussed in this review are equally relevant to rice-growing soils that are flooded but not puddled. Flooding soil at the time of land preparation for wetland rice triggers several physicochemical and microbiological processes. The resultant soilwater-atmosphere system is highly complex and heterogeneous in nature. The characteristics of the wetland rice soils have been extensively studied and reported (Ponnamperuma, 1965, 1972, 1977, 1978a; Patrick and Reddy, 1978; Yamane, 1978; Tadano and Yoshida, 1978; Yoshida, 1978; De Datta, 1981). Some dynamic aspects of the wetland rice soils that are pertinent to nitrogen transformations can be summarized as retardation of gaseous exchange between flooded soil and atmosphere; decrease in soil redox potentials (Eh) or pe, which is the negative logarithm to the base 10 of the electron activity and is equal to Eh (volts)/O.0591; increase in pH of acid soils and decrease in pH of calcareous and sodic soils, to a near equilibrium pH range of 6.5-7.2; increase in specific conductance (EC) and ionic strength (p); and anaerobic decomposition of plant residues characterized by a slow rate with generation of an array of products such as CO,, CH,, fatty acids, mercaptans, H,S, and heterocyclic compounds such as purines and pyrimidines. Examples of some physicochemical properties of a wetland rice soil (Maahas clay) profile to a 20-cm depth are shown in Fig. 1. In the wetland rice soil a reduced plow layer is usually sandwiched between a thin oxidized surface soil layer and a partly oxidized subsurface soil layer. Within the reduced soil layer, in which most of rice roots proliferate and derive nitrogen, is a complex mosaic consisting of aerobic and anaerobic microsystems (Furusaka, 1977). The rice rhizosphere (and rhizoplane) is at relatively higher pe, compared with rest of the soil in the plow layer, due to diffusion of oxygen from the rice roots (van Raalte, 1941; Armstrong, 1967a,b, 1971; Naohara and Usami, 1976; Kimura et al., 1977). The behavior of nitrogen in wetland soil is markedly different from its behavior in dryland soil. Flooding the soil results in the accumulation of ammonium, the instability of nitrate, and a lowered nitrogen requirement for organic matter decomposition. Flooding also causes inefficient utilization of applied nitrogen (Patrick and Mahapatra, 1968). Ammoniacal nitrogen is subject to fixation by clays, loss by volatilization, leaching, runoff, seepage, and nitrification followed by loss through denitrification. Significant inorganic nitrogen is assimilated into the organic fraction of the wetland soil even though the magnitude of assimilation is lower than in the dryland soil. Leaching and these transformation processes have been studied primarily by laboratory incubation. However, most studies were limited to one specific transformation process. Some laboratory and field studies with I5N-labeled fertilizers

243

NITROGEN TRANSFORMATIONS

6

7

8

-

2

0

2

4

FIG. 1. Changes in volumetric water content (El), bulk density (p), pH, and pe with depth in a wetlant Maahas soil. From Savant and De Datta (1980).

have provided useful information regarding the transformation of soil nitrogen besides providing information on plant uptake and efficiency of fertilizer nitrogen in a wetland rice soil. Some of the studies on plant uptake are summarized by IAEA (1978), Craswell and Vlek (1979a), and Koyama (1981). A great body of literature exists on the management of fertilizer nitrogen in wetland rice and is covered in recent reviews (De Datta, 1978; Prasad and De Datta, 1979). A review of biological nitrogen fixation in wetland rice-growing soils is not included here because it is reviewed elsewhere (Balandreau et al., 1975; Yamaguchi, 1976; Watanabe, 1978; Matsuguchi, 1979; Dommergues and Rinaudo, 1979; Buresh et al., 1980). Nitrogen transformation processes in waterlogged soils were reviewed by Patrick and Mahapatra (1968) and Tusneem and Patrick (1 97 1). Recently, Broadbent (1978) highlighted the nitrogen transformation processes, and Hauck (1979) covered the methodology used by various workers for studying nitrogen transformations. Understanding the nitrogen transformation processes would greatly facilitate regulating nitrogen losses from wetland soils and increase the availability of nitrogen to rice. The primary focus of our review is on nitrogen transformation processes in wetland rice soils. Further, we summarize information on the physical, physicochemical, and biochemical processes relevant to nitrogen transformations and to improving fertilizer nitrogen efficiency in wetland soils. Although we summa-

244

N. K. SAVANT AND S. K. DE DATTA

rize papers published since 1970, necessary relevant references have been made to earlier work.

II. CHEMICAL NATURE OF SOIL NITROGEN

In wetland rice cultivation, the significance of native soil nitrogen is apparent from the fact that 60-80% of nitrogen absorbed by the crop is derived from the native nitrogen pool (Koyama et al., 1973; Broadbent, 1979). Knowledge about the nature of native nitrogen, therefore, should help in understanding the dynamics of nitrogen transformations and, in turn, its availability to the wetland rice plant. In rice soils of 10 countries studied by Kawaguchi and Kyuma (1977), the total nitrogen varied from 0.08 to 0.29%, with Indian soils containing least (0.08%) and Japanese soils containing exceptionally high amounts (0.29%) (Table 1). Table I Total Nitrogen and C/N Ratios of Rice Soils in Tropical Asia and Japan<‘

Soil

Average total Nh

Average

(n)

(%)

GIN"

Bangladesh

53

Burma

16

Kampuchea

16

India

73

Indonesia

44

Malaysia (West)

41

0.13 (0.05-0.52) 0.10 (0.04-0.18) 0.10 (0.03-0.29) 0.08 (0.02-0.20) 0.12 (0.05-0.59) 0.28 (0.05-0.92)

Philippines

54

8.7 (7.1-12.4) 11.5 ( 8 . 6I 3.4) 10.7 (7.7- 13.7) 11.7 (4.8-18.4) 12.6 (8.3-25.8) 11.8 (9.2-1 9.8) 11.7 (7.7-19.0) 10.5 (6.5- 15 . 6 ) 11.3 (6.Cb22.2) 11.6 (3.6-23.2)

Country

Sri Lanka

33

Thailand

80

Japan

84

“Kawaguchi and Kyuma (1977). ”Figures in parentheses indicate range.

0.15

(0.06-0.35) 0.13 (0.03-0.64) 0.09 (0.02-0.27) 0.29 (0.08-0.91)

245

NITROGEN TRANSFORMATIONS

As in dryland soils, a major portion of the nitrogen in wetland soils occurs in the organic pool (Fig. 2) whose chemical nature is not completely known. From the available literature, the distribution of nitrogen in the organic pool of some rice soils is given in Table 11, and the following generalizations can be made: (1) The water regime of the wetland rice soils does not seem to influence the distribution pattern of nitrogen within the organic pool. (2) The distribution pattern of nitrogen, especially the composition of acid-hydrolyzable nitrogen, is not influenced by soil type, climate, parent material, physiography, and age of soil. While extrapolating this information to the wetland rice soils, it should be remembered that air-dry oxidized soil samples, and not wetland soil samples, were used for the determination of these organic fractions in the rice soils studied. Various fractions present in the inorganic nitrogen pool of a wetland soil, shown in Fig. 3, represent a small part of the total soil nitrogen. These forms constitute an extremely dynamic nitrogen system in the soil, which is affected by physical, chemical, and microbiological reactions. The various reactions of inorganic forms of nitrogen are schematically depicted in Fig. 4. The predominant form of inorganic nitrogen in the wetland soil is NH4+. The concentration (or

U ORGANIC FRACTIONS OF N

Q HYDROLY 2 ABLE

I NONHYDROLYZABLE

HYOROLIZED SUGAR -N

Frc. 2. Fractionation of organic nitrogen in soils. Adapted from Cheng and Kurtz (1963) and Bremner (1965).

Table I1 Chemical Fractions of Organic Nitrogen in Some Rice Soils Fractions (% of acid-hydrolyzable N)

Soil Country Thailand0 Malaysia0 Japan0 Taiwanb

Total N

(fl)

(8)

6 3 6 20

0. I74 0.287 0.263 0.05-O. 22

Acidhydrolyzable N (8of total N) 53.3 57.8 72.7 75.6

aKawaguchi and Kyuma (1969b) used Bremner method (1965). bLin er al. (1973) used Bremner method (1965).

NH4 N

+

28.1 25.3 28.2 24.8

Amino acid N 47.0 46.2 45.8 35.7

Hexosamine or amino sugar

Serine and threonine

Acidnonhydrolyzable N

N

N

Unknown

(%)

6.5 5.9 7.0 4.8

7.2 8.4 7.9 6.9

18.4 22.6

46.7 42.2 27.3 24.4

19.0 10.5

247

NITROGEN TRANSFORMATIONS

I INORGANIC N

I

NH2-N STABLE AN0 PREDOMINANT IN THE REDUCED SOIL LAYER

I I

NO;-N IN REOUCED SOILLAYER

I

OF DENITRIFICATIONI IN SOIL AIR OR IN DISSOLVED PHASE

I

ORGANIC COLLOIDS

N03-N MOBILE PNO VERY UNSTABLE IN REDUCED SOIL SOLUTION ,STABLE IN FLOODWATER AND THIN OXIDIZED SOIL LAYER,AND MAY BE IN OXIDIZED RICE

FIG. 3. Inorganic fractions of nitrogen in a wetland rice soil.

activity) of NH, in soil solution (intensity parameter) is relatively small compared with the NH,+ in the sorbed phase (solid phase) in soils of high cation exchange capacity (CEC). No quantitative relationship, however, between total NH,+-N and NH,+-N in soil solution can be given because of interactions between CEC and ionic chemical equilibria, and chemical kinetics of the soils. However, the ratios of NH, +-N in solution to total NH, +-N may vary from nearly one-tenth for soils having high CEC to one-half for coarse textured soils (Ponnamperuma, 1965). Figure 5 shows the effect of nature and amount of clay on the relationship between total NH, and the NH, in soil solution for two Japanese soils. Nitrification-denitrification is considered to occur in the aerobic-anaerobic and soil layer of the wetland soil. However, the amounts of NO,--N NO, --N found are practically insignificant in a continuously flooded soil because of their high unstability. The equation for NO,--N, system +

+

pe = 21.06 - $pNO,-

+

+

&pN, - EpH

(1)

shows that the NO, - is extremely unstable in submerged soil having pe - 1 to 3 (Ponnamperuma, 1972). Similarly the equation pe

=

14.91 - $pNO,-

+ $pNH,+

-

$pH

(2)

248

N. K. SAVANT AND S. K. DE DATTA

FIG. 4. Schematic presentation of nitrogen transformations (biochemical) and relevant physical and physicochemical processes in the wetland rice soil.

indicates that at low potential of the wetland soils (pe = - 1 to 3), the concentration of NO,- that can be in equilibrium with NH4+ is infinitesimal (Ponnamperuma, 1972). Traces of NO,- and NO,- have, however, been reported to exist in floodwater or in the thin oxidized soil layer or both for a certain submergence period when the flooded soil system was fertilized (Racho and De Datta, 1968; Gotoh, 1973; Manguiat and Yoshida, 1973; Bilal, 1977; Singh et al., 1978). No NO,- was, however, detected in floodwater or soil solution collected at -20-cm soil depth following 10-cm-deep placement of supergranules of urea or prilled urea in mudballs, or basal incorporation of prilled urea or sulfur-coated urea (SCU-21), in a wetland Maahas clay (Savant and De Datta, 1978; IRRI, unpublished data).

249

NITROGEN TRANSFORMATIONS -T Soil SCL Odowora LiC

Takadale

‘

I

“

- I ~

CEC (meq/OOq) 20.8 40.8

I mowaro

rn

0

I 20

Clay miner01 Montmorillonitechloritic interprodo, ~lOhOllOy8it~ Montmorillonile

I

I

40

60

Total NH:-N

4

/

1

1

80

100

(mg/~proll)

FIG. 5. Effect of nature and amount of clay on relationship between total NH4+-N NH4+-N in soil solution for two Japanese soils. From Shoji (1976).

and

111. PHYSICAL AND PHYSICOCHEMICAL PROCESSES RELEVANT TO NITROGEN TRANSFORMATIONS Movement of nitrogen species, such as NH,(aq), NH4+, NO,-, and the urea molecule, occurs in soil through diffusion or mass flow or both. In the wetland soil, liquid phase diffusion, solid phase (adsorbed phase) diffusion, and mass flow may predominate and may play an important role in various pathways of nitrogen transformations (Fig. 4). A. TRANSPORT

General theoretical considerations and mathematical models of diffusion and mass flow of nitrogen compounds in soil have been reviewed by Gardner (1965). Other useful references are Crank (1956), Jost (1960), Jacobs (1967), Olsen and Kemper (1968), Hamaker (1972), and Letey and Oddson (1972). Realizing the significance of nitrogen movement in a flooded soil system, Reddy et al. (1976) determined apparent diffusion coefficients D (Philips and Brown, 1964), for NH4+ and NO,- in a flooded Crowley silt loam. The D

250

N. K. SAVANT AND S . K . DE DATTA

values for the NH4+ and NO,- were 0.216 and 1.33 cm2 day-', respectively (Fig. 16), suggesting about six times faster movement of NO,- than that of NH, . These apparent diffusion coefficients, however, may have limited relevance to the wetland rice soils, because it is not known whether Reddy et al. used nonaggregated or puddled soil having low pe. Using a technique similar to Pang et al. (1973) we measured D,,+ for deep-placed urea in a wetland Maahas clay (pe 0 to -2). The D,,,+ values were nearly 10 times smaller than those reported for the flooded Crowley silt loam (Savant and De Datta, 1980). Transport of nitrogen could be a significant factor in the transformations and subsequent availability of fertilizer nitrogen, especially after its deep placement. In the wetland soil, nitrogen release from placement sites of prilled urea, supergranules of urea, sulfur-coated urea (SCU-21), and urea placed in mudballs, appears to be a diffusion-controlled process (Savant and De Datta, 1979). With a deep-placement method of fertilizer application, the spatial distribution pattern of NH4+-N assumes importance. In a wetland soil (Maahas clay), we studied spatial distribution of NH, +-N in situ following 10-cm-deep placement of supergranules of urea (Fig. 6) in the presence of rice plants. Our findings are summarized as follows: (1) Despite a steep concentration gradient of NH, , it could hardly reach the soil-water interface in the presence of transplanted rice (IR42). ( 2 ) By and large, the movement of NH4+ was downward > lateral > upward from the deep placement site (Savant and De Datta, 1980; Savant et al., 1982). (3) The NH4+ concentration gradient disappeared earlier in dry season than in wet season, suggesting faster transport of NH, and/or more root-sink effect in dry season (Savant and De Datta, 1979; IRRI unpublished data). In a wetland field study at IRRI, Ventura and Yoshida (1978) also noticed very little horizontal movement of NH, +-N from the placement site; the distribution of I5N was restricted to the four adjacent hills of rice (IR26) transplanted 20 X 20 cm. Similar results are also reported by Lung-yie and Chin-cheng (1978). When Bilal (1977) studied transport of surface-applied ammonium sulfate in a flooded Sacramento clay, he noticed an appreciable concentration of nitrogen as NH, remaining in floodwater and up to a depth of 1.2 cm. In the absence of rice plants, we found downward movement of NH4+ 4 weeks after surface application of urea to be 12-14 cm in undisturbed flooded Maahas clay cores (Savant and De Datta, 1980). The leaching process may occur rapidly in coarse textured wetland soils or soils with appreciable amounts of hydrous oxides of aluminum and iron (Hsu et al., 1967; Pande and Adak, 1971; Chiang et al., 1972; Nair et al., 1976; Daftardar el al., 1979). Such soils generally have low CEC and there is more tendency for NH4+ to remain in soil solution. This may enhance downward convective transport of NH4+, probably below the root zone. Leaching loss of fertilizer nitrogen after deep placement of supergranules of urea in wetland soils having percolation rate >5 m d d a y could be very serious (Vlek et al., 1980a). +

+

+

+

25 1

NITROGEN TRANSFORMATIONS TRANSPLANTED 10

5

0

U

W

Distance from placement site (em) 5 10 10 5 0

5

10

20 I6 32

63

63

32

16

FIG.6 . Computed contours showing distribution patterns of NHd+-N (Fgicm' wet soil) following 10-cm-deep placement of 2 g supergranule urea in wetland soil. IRRI (1978), wet season. From Savant and De Datta (1980).

Loss of NH4+ through leaching may be accentuated in these soils as pe decreases, and probability of NH4+ desorption in solution by Fez+ or Mn2+ ion species, or both, is increased (Patrick and Mahapatra, 1968; Ponnamperuma, 1977). Using the perforated-tube (similar to a piezometer) technique for collection of in situ leachate from flooded soil (acidic alluvium with low CEC), Shinde and Vamadevan (1974) over a period of 8 weeks recorded an average loss of 45.64 and 31.78 kg N/ha during the monsoon and winter seasons, respectively. Puddling a soil decreases water movement through it and as a result may decrease leaching losses of the NH4+. This generalization is of particular significance for sandy soils or well-aggregated Oxisols and Inceptisols (Andepts) in the

252

N. K. SAVANT AND S. K . DE DATTA

tropics, where puddling is often essential for successful wetland rice cultivation (Sanchez, 1973, 1976). Addition of organic matter to a wetland soil may increase NH4+ in the leachate (Ponnamperuma, 1955; Raghupathy and Raj, 1971) or decrease it (Shinde and Chakravorty, 1975; Katyal, 1980). It appears that the effect of added organic matter on NH4+ leaching below the root zone depends on soil factors such as texture, nature and amount of clay, and on organic matter factors-for example, nature and amount of organic matter added, method of its incorporation in soil, etc. Split application of fertilizer nitrogen may reduce NH, leaching, but a single basal application of nitrogen during the early stage of transplanted rice may favor it (Daftardar et al., 1979). This is partly due to the absence of well-developed rice roots during early plant growth; roots act as potential sinks for nitrogen during the latter part of the growth period. Soil compaction to bulk densities of 1.72-1.75 g/cm3 has been reported to reduce leaching loss of nitrogen below the root zone by cutting down percolation rate, especially in light-textured soils (Varade and Patil, 1971; Ghildyal, 1978). +

B . SORPTION-DESORPTION

The various mechanisms of sorption and desorption of nitrogen compounds by the soil materials are extensively reviewed by Mortland and Wolcot (1965) and Nommik (1965). In most of the studies reported, however, oxidized, disturbed (<2 mm), and unpuddled soils have been used. The kinetics and mechanisms of sorption (or fixation) and subsequent desorption of nitrogen compounds in wetland soil are therefore subjects of some speculation. The NH4+ formed as a result of mineralization of soil organic matter, or added extraneously, interacts with the heterogeneous wetland soil under flooding. Besides microbial and phytic immobilization, the NH4+ is generally adsorbed as an exchangeable ion on the clays, and chemisorbed by humic substances, or fixed within the clay lattice (Gupta, 1964; Rodrigo, 1967; Bhattacharyya, 1969; Nommik, 1970; Manguiat and Yoshida, 1973; Yamada, 1975; Manguiat and Broadbent, 1977a,b; Sahrawat, 1979; and many others). It appears that these reactions may occur simultaneously. The rate and extent of these reactions are reported to be influenced by several factors, such as nature and amount of clays, alternate submergence and drying, nature and amount of soil organic matter, period of submergence, fertilizer additions, tillage, presence of crop, etc. The NH, +-N not extractable by KCI is generally deemed “fixed” in soil. In Korean soils, the amounts of NH, fixed in wet soils ranged from 6.3 to 23.5% of total adsorbed NH4+ (Kim and Chae, 1970). In a pot experiment, Manguiat +

NITROGEN TRANSFORMATIONS

253

and Broadbent (1977b) used Sacramento clay (pH 6.8) to study I5N recoveries under different management practices for wetland rice. They noticed 28.6-35.6% clay-fixed NH, + , which decreased with time. After equilibrating 12 tropical rice soils under flooding by continuous shaking for 2 hr, the amount of NH4+ fixed varied from 3.8 to 7.7 meq/100 g soil (Sahrawat, 1979). He reported a significant positive correlation (r = 0.61 *) between NH4+ fixing capacity and percent active iron in the soils studied. Under a wetland field, we (Savant and De Datta, 1979) noticed insignificant amounts of clay-fixed NH4+-N (<0.2% of total N added) at a deep placement site of urea after 45 days in a wetland Maahas clay. Using quantity/intensity (Q/ l) relationship, Pasricha (1976) studied the exchange equilibria of air-dried, and aggregated rice soils of the Philippines. The Q/I plots obtained for Maahas clay (pH 6.6) and Luisiana clay (pH 5.6) are shown in Fig. 7. They suggest that the NH4+-(Ca2+ + Mg2+) exchange equilibria obeyed the Schofield’s Ratio Law. The plots are straight lines at higher values of AR-NH4+ but curvilinear at lower values of AR-NH4+ . According to Pasricha, therefore, the NH4+ ions were adsorbed specifically (42-84%) and were associated with exchange sites from which they could not be easily desorbed by other cations such as Ca2+ or Mg2+, The level of such specifically held NH4+ was higher in mineral soils than in soils richer in organic matter. The dynamics of physicochemical properties of the wetland rice soils seem to influence NH,+ retention by them. When Chen and Chiang (1963) studied NH4+ adsorption from dilute NH4+ C1 solution by Chinese rice soils, they noticed a decrease in NH4+ adsorbed in waterlogged soils compared with that adsorbed by dry soils. It seems that the NH4+ in a reduced soil fails to replace

t 1.2

-

t 0.8

n

8 t 0.4

\

-z

o

Y

rO

$ -0.4

a

-0.8 LUtSlANA CLAY

FIG. 7. The NH4+-Q/l relationships of two paddy soils of the Philippines. AR-NH4+ means uNH4+/fa,-~+ + u ~ ; + ) ~From / ~ . Pasricha (1976).

254

N. K. SAVANT AND S. K. DE DATTA

some of sorbed Fe2+ or Mn2 , or both, ion species from the exchange complex. Soil samples collected from the oxidized layer, therefore, presumably adsorbed more NH4+ from aqueous (NH,),SO, solution than that collected from the underlying reduced layer in a wetland field (Savant and De Datta, 1979; IRRI unpublished data). On the other hand, Shimpi and Savant (1975) observed an increase in NH,-retention capacity of two tropical clayey soils continuously submerged for 45 days in a laboratory. In the process of adsorption of NH4+, the nature of clay minerals plays an important role. Marumoto and Yamada (1977) observed the order of adsorption power of NH4+ by clay minerals to be montmorillonite > kaolin minerals > allophanes. According to Thomas (1970), NO, - is weakly adsorbed in soils having appreciable contents of iron and aluminum oxides and pH <6. To some extent similar favorable conditions for adsorption of NO, -, formed as a result of nitrification, presumably exist in the thin oxidized soil layer. This type of NO,- adsorption in the aerobic layer may be of some significance, especially in highly weathered tropical acid soils containing appreciable amounts of aluminum and iron oxides. As information on this aspect is generally lacking, and since it may have some relevance to nitrogen losses from wetland soils, research is needed. Adsorption of urea by soils could be largely due to formation of H bonding arising from -0 or - - O H of broken edges of crystalline clays, -SiOH of groups of humic substances in soils amorphous clays, and --COOH or -C=O (Mitsui and Namioka, 1958; Mitsui and Mukai, 1959; Shiga, 1961; Mitsui and Takatoh, 1963; Hance, 1965; Dalal, 1974). According to Shiga (1961), urea sorption by soils can lead to formation of a double layer. Mitsui and Namioka (1958) reported urea adsorption by different Japanese rice soils to vary from 1 1 mg urea-N/100 g air-dry Yamaguichi sandy soil to 46 mg urea-N/100 g air-dry Tanashi soil of volcanic origin. Soil pe seems to influence urea sorption. The oxidized surface soil collected from a wetland field sorbed more urea from solution than the underlying reduced soil from the same field (Savant and De Datta, 1979; IRRI, unpublished data). However, on the whole, the sorption affinity of soils for urea molecules appears to be rather weak, especially in wetland soil (Chin and Kroontje, 1962). Information is lacking on the extent and energy of sorption of urea by the oxidized and reduced soil layers that would facilitate understanding of the fate of fertilizer urea immediately after its application. +

C. AMMONIA VOLATILIZATION

Conditions in the wetland rice field are conducive to ammonia volatilization loss. The following physicochemical and biological factors (or parameters) of a

NITROGEN TRANSFORMATIONS

255

flooded soil-water-atmosphere system and others simultaneously influence ammonia volatilization from wetland rice soils (Bouldin et al., 1974; Bouldin and Alimagno, 1976; IRRI, 1976, 1977, 1978, 1980; Wetselaar et a l . , 1977; Avnimelech and Laher, 1977; Vlek and Stumpe, 1978; Mikkelsen et a l . , 1978; Sahrawat, 1978; Feagly and Hossner, 1978; Mikkelsen and De Datta, 1979; Vlek and Craswell, 1979; Nor and Majid, 1979; Bouwmeester and Vlek, 1981; Freney et a l . , 1981; De Datta et a l . , 1981; Vlek and Craswell, 1981): 1. Soil Parameters Soil pH and pe Salinity (EC) and alkalinity (SAR) CaCO, content Cation exchange capacity (CEC) Predominant exchangeable ions 2. Floodwater Parameters PH ' pc,, Concentration of ammonia [NH,(aq) + NH4+ + NH40H] Total alkalinity pH-buffering capacity 3. Atmospheric Parameters Air temperature Solar radiation 4. Other Parameters Nitrogen management Water management Plant canopy

pH buffering capacity Partial pressure of carbon dioxide (Pcoz) Microbial activity

Temperature Water movement or turbulence Depth Algal growth or activity Concentration of phosphorus Use of pesticides Wind speed Partial pressure of NH, (PNH7)

As these factors presumably operate simultaneously, no attempt is made to discuss them separately. We discuss them in the following paragraphs where they are most relevant. The basic equilibrium reactions of ammonia volatilization from soil are given by Avnimelech and Laher (1977). The rate of ammonia volatilization from soil is a linear function of ammonia concentration: -d(NH,)ldt = k(NH,)

(3)

During the process of volatilization, gaseous ammonia is formed: NH4+ S NH, H + . That is, H + is released for each conversion of NH4+ += NH,(g). The pH and buffering capacity of the oxidized, as well as reduced, soil layers will therefore influence the volatilization process. In wetland soils, the kinetics of pH indicate that submergence causes pH values of most of the reduced acid and alkaline soils to converge between 6.5 and 7.2 (Ponnamperuma, 1972). If

+

256

N. K. SAVANT AND S. K . DE DATTA

other factors remain constant, the pH of the wetland soils per se may not be a dominant factor in ammonia volatilization (Vlek and Craswell, 1979). Further, soil pH is related to soil pe through an equation (Ponnamperuma, 1976b) such as: pe = 17.87

+ pFe2+ - 3 PH

(4)

It is, therefore, suggested that reference to the corresponding pe would help in the understanding of soil pH effect on ammonia volatilization from the wetland soil. For a nonflooded soil, Avnimelech and Laher (1977) suggest that pH-buffering capacity of soil is an important parameter in the process of ammonia volatilization. The pH-buffering capacity of the oxidized soil layer may be different from that of the reduced soil layer, presumably a result of differences in the redox systems of iron prevailing in them. In the stabilized pH range between 6.5 and 7.2, which is a characteristic of the wetland rice soils, the appearance of Fe,(OH),-Fez , or Fe(OH),-Fe,(OH), systems and FeCO, would increase pH-buffering capacity of the soil-water system (van Breemen and Wielemakar, 1974a,b; F. N. Ponnamperuma, IRRI,personal communication). In other words, absence of such redox systems in the oxidized soil layer may result in lowering of its pH-buffering capacity. In the wetland soil, the floodwater and atmospheric parameters probably assume more importance than soil parameters (Bouldin and Alimagno, 1976; Wetselaar et al., 1977; Vlek and Stumpe, 1978; Mikkelsen et al., 1978; Mikkelsen and De Datta, 1979; Vlek and Craswell, 1979). The floodwater-atmosphere system, therefore, warrants careful monitoring to understand and estimate the magnitude of ammonia volatilization loss from the wetland soil. Some physicochemical parameters of floodwater pertinent to ammonia loss have been reported by Vlek and Stumpe (1978), Mikkelsen and De Datta (1979), and Vlek and Craswell (1981). In floodwater, the overall inorganic ammonium nitrogen (equilibrium) system (N,) can be written as: +

N, = NH,(aq)

+ NH4+

(5)

The concentration of NH,(aq) changes in direct proportion with NH,' , and up to pH 9 it increases about 10-fold per unit increase in pH of an aqueous system. In the wetland rice soil, floodwater pH is markedly influenced by Pcoz and alkalinity, which in turn are determined by algal photosynthetic activity and respiration. By addition of 10 kg diuronlha to the wetland soil to inhibit algal growth, floodwater pH did not exceed 8.1 (IRRI, 1977). On the other hand, when algal bloom occurred in the same experiment due to addition of diammonium phosphate at 8.7 kg P/ha, floodwater pH increased to 9.7 within 3 days. As a result, 2.7 times more NH4+-N was found in floodwater of the diuron-treated

NITROGEN TRANSFORMATIONS

257

plot than that of the phosphate-treated plot 6 days after fertilizer application of 60 kg N/ha as ammonium sulfate (IRRI, 1977). The diurnal changes in floodwater pH as a function of carbon dioxide consumption and release due to algal activity are common. The rhythmic patterns of diurnal changes in floodwater pH overlying neutral Maahas clay (Aquic Tropudalf) and acid Luisiana clay (Typic Tropaquept) were markedly evident (Mikkelsen et al., 1978). For Maahas clay, daytime floodwater pH was as high as 10, but decreased by almost 2 units at night. In the case of Luisiana clay, less algal activity and consequently smaller diurnal pH changes in floodwater were noticed. The magnitude of ammonia volatilized from these soils was explainable on the basis of the pH differences. Mikkelsen et al. (1978) have also suggested that with floodwater pH <7, the NH, loss through volatilization may be of little agronomic significance. During 15N field studies, Wetselaar et al. (1977) observed significant correlation for the relationship: (pH) floodwater =

(NH,) volatilized [NH,(aq) + NH4+] floodwater

Their regression analyses, however, could not account for 44-57% of the variation in ammonia volatilized. This suggests that, under wetland conditions, other floodwater parameters besides floodwater pH and NH,(aq) + NH4+ in floodwater, and the atmospheric parameters need to be studied concurrently. The ammonia volatilization process can be described as an acid-base titration. The process of ammonia loss from floodwater may be written as (Vlek and Stumpe, 1978): k2

kl

NH4+ G NH,(aq)

+ H+ + NH, t

(6)

k3

This indicates two points: (1) the H formed after dissociation of NH4+ will consume alkalinity of the floodwater system, if present, or will otherwise lower its pH, and (2) the ammonia loss from floodwater will depend on its pH-buffering capacity. The floodwater overlying a calcareous soil will probably be more buffered and may, therefore, lose more nitrogen as ammonia than floodwater overlying noncalcareous soils. Although ammonia volatilization from a carbonate-free aqueous system seems to be a first-order kinetic process (Vlek and Stumpe, 1978), it may not hold for paddy floodwater, mainly because floodwater is a complex aqueous system and its pH shows marked fluctuations and rarely exceeds 10. Bouldin et ul. (1974) have proposed a model for ammonia volatilization loss +

258

N. K . SAVANT AND S. K. DE DATTA

from a pond:

where Ni = total amount of inorganic ammonium N [Ni = NH3(aq) + NH,+], No = the value of Ni when t = 0, and k = a constant, which includes Henry’s constant (A), volume (V) and surface area of water (S),and a constant representing rate of exchange between air and water system which also includes the integrated sine function (b’) and defined as k = A(S/V)b’.A near linear relationship between the natural log of inorganic NH, +-N in solution and time ( t )was noticed by Bouldin et al., suggesting the applicability and validity of the model. Although this model was originally developed for aquatic systems in ponds, it may be applicable to the wetland rice soil floodwater system during the first 1-2 weeks after transplanting. Further, Bouldin and Alimagno (1976) have suggested the following aerodynamic model for ammonia volatilization loss based on equilibrium between air and floodwater at the air-water interface: F ( N H ~ ) T= ( E H * ~ / A V“H3tg)ls )

(8)

where F(NH,), = flux of NH, loss (pg/cm2/hr), EHzO = estimated evaporation of water (mg/cm2/hr), AV = concentration of water vapor at the air-water interface (calculated from water temperature) minus the water vapor concentration at 2 m (calculated from air temperature and humidity (mg/liter), and [NH,(g)]s = calculated concentration of NH,(g) at the air-water interface. Bouwmeester and Vlek (1981), in an attempt to obtain a quantitative assignment of the chemical and environmental factors influencing NH, volatilization from the wetland rice, developed an ammonia diffusion model: Q

=

(AN/pt) Ief+zD‘erfc ~ v -EI + ( 2 i G ) p f i ~

(9)

For this application, Q = the ammonia volatilized per unit area; AN = the ammoniacal nitrogen concentration in the bulk liquid; p = k,kHiD( 1 H ik) where k, is the bulk transfer coefficient of ammonia in the air, k , is Henry’s constant, D is the molecular diffusivity of ammonia, and k is the equilibrium constant of the chemical reaction, NH4+ $ NH, + H + ; and t = time. Freney et al. (1981) used a combined chemical and micrometeorological method to measure the flux of ammonia loss in the atmosphere from the wetland paddy field at IRRI. Their relevant equation is

+

+

where F(NH3)g is the vertical flux density of NH,(g), R is the net radiation, G is the change in heat stored in floodwater, r is the psychrometric constant in terms of humidity, s is the slope of the curve relating specific humidity to temperature

NITROGEN TRANSFORMATIONS

259

at the average of the wet bulb temperatures concerned, p = density of air, C, is the specific heat of air at a constant pressure, ACNH, is the difference in the ammonia concentration in air at two different heights above the paddy, and ATw is the difference between wet bulb temperatures at the same two heights. In the absence of enough data and assumptions made, however, utility of these mathematical models for the wetland paddy field conditions cannot be fully appreciated. Data on ammonia volatilization from the wetland rice soils (field studies), summarized in Table 111, show large variation (0-60%) in the amounts of ammonia volatilized. Most of the studies, however, suggest ammonia volatilization to be less than 20% of the applied nitrogen. These differences in the magnitude of losses are ascribed to (a) the chemical nature of fertilizer, (b) the method of fertilizer application, and (c) the differences in the methods employed by the various workers. Nevertheless, the data do indicate that ammonia volatilization losses from wetland paddies could be serious. The merits and demerits of the various methods or techniques used in such studies are discussed elsewhere (Vlek and Stumpe, 1978; Mikkelsen and De Datta, 1979; Vlek and Craswell, 1981). The aerodynamic and micrometeorological models described earlier suggest that the environmental parameters such as air temperature, wind speed (directly), and solar radiation (indirectly) influence ammonia loss from the wetland paddies. In the typical temperature range in the tropics, ammonia volatilization may increase about 0.25% per degree Celsius increase (Vlek and Stumpe, 1978). Ammonia volatilization rate seems to follow the diurnal pattern of water temperature (Freney et al., 1981). High wind speed or rapid air exchange at the water-air interface results in a marked decrease in PNH3in air, which is in immediate contact with the body of floodwater and will therefore enhance ammonia volatilization (Bouldin et al., 1974). Vlek and Stumpe (1978) reported a curvilinear relation between the ammonia loss from a pH 8.9 aqueous solution containing 100 ppm N as (NH,),SO, and the rate of air exchange. The data presented by Bouwmeester and Vlek (1981) suggest that the effects of wind velocity, temperature, and pH on ammonia volatilization could be of the same order of magnitude. This envisages that the rice plant canopy, which modifies microclimate, will have a marked effect on ammonia volatilization from the wetland paddy. Two main effects of the rice plant canopy are visualized: (1) decreased transfer coefficient of NH, within the canopy due to the restricted air movement, and ( 2 ) reduction in pH and temperature of the floodwater presumably due to shading. These apparent effects would reduce ammonia volatilization. For example, the average nitrogen loss recorded by Bouldin and Alimagno (1976) for small canbut the opies (about 3 weeks after transplanting) was 6.1 p g NH,-N/cm*/hr, corresponding value for the large canopies (about 5 weeks after transplanting)

Table 111 Ammonia Volatilization Loss from Fertilizer-Amended Wetland Paddies Fertilizer N added NH3-N loss Soil

Season

Maahas clay loam (pH 7. I), Philippines

1973 wet season

Maahas clay, Philippines

1976 dry season 1977 wet and dry seasons

Maahas clay (pH 7), and Luisiana clay (pH 5.7), Phillippines Clay (pH 7.0-7.3, Thailand Black clay (pH 8.1, Vertisol), India Maahas clay (pH 6.7), Philippines

1971-1972 dry season 1976 wet season 1978 wet season

(%I

Methodology

4 8

Field study, closed system with opening for air pressure and acidified glass wool trap Field study, open and closed system

30-59

Greenhouse and field studies, closed system with air exchange and acid trap Field study, closed system with air exchange and acid trap Field study, static closed system and acid trap Field study, chemical and micrometeorological technique

Reference Ventura and Yoshida (1977)

1.0-20

Bouldin and Alimagno (1976); IRRI (1977) Mikkelsen and De Datta (1979); Mikkelsen el al. (1978)

0.25-6.8

("NH4)zS04

50-100

0.8-14.0

Wetselaar et al. (1977)

(I5NH2)2CO

I00

9.7

(NH4)2S04

4W30

5.1-10.6

Krishnappa and Shinde (1978a) Freney et al. (1981)

NITROGEN TRANSFORMATIONS

26 1

was 2.8 pg NH,-N/cm*/hr. In a field study using I5N-tagged (NH,)*SO,, Wetselaar et al. (1977) observed 3-4% less loss of ammonia from a planted rice plot (cultivar RD1, 25 X 25 cm hill spacing) than from unplanted plots. According to them, this could%e due to lowering of (1) floodwater pH in the presence of plants, (2) the NH, +-N content of the floodwater due to plant uptake, and (3) the NH, concentration in the air above the floodwater due to its absorption by plant tops.

IV. BIOCHEMICAL NITROGEN TRANSFORMATIONS

A. AMMONIFICATION-IMMOBILIZATION The interconversion of nitrogen from organic to inorganic and vice versa in the wetland soil are two concurrent opposing processes. They are summarized as shown in Fig. 8. With anaerobic conditions, such as those prevailing in the wetland rice soils, the conversion of organic nitrogen into inorganic nitrogen proceeds up to ammonia formation only. Such mineralization is, therefore, aptly described by the term ‘‘ammonification. Alexander (1977) has discussed microbiological aspects and biochemistry of a mineralization process in soils. A few comprehen”

Proteins, amino acids, purine and pyrimidine bases,

FIG. 8. Schematic presentation of ammonification and immobilization processes in the wetland soil.

262

N. K. SAVANT AND S. K. DE DATTA

sive reviews on these aspects of ammonification or mineralization in paddy soils are available (Patrick and Mahapatra, 1968; Tusneem and Patrick, 1971; Broadbent, 1978, 1979; Dei and Yamasaki, 1979). Our discussion mainly covers ammonification-immobilization in the wetland rice soil?.

I . Ammonification The ammonification process is essentially a catabolism of amino acids and presumably includes several types of deamination reactions. The oxidative deamination can be written as Amino acid -+ Imino acid -+ Keto acid

NH3

and may be operative in the oxidized soil layer. On the other hand, the reductive deamination (Rose, 1976) Amino acid --f Saturated acid + NH3

presumably takes place in the reduced soil layer. Although rate of decomposition of plant residues is slower in flooded soils than in nonflooded soils, the net ammonification may be greater under anaerobic than under aerobic incubation (Waring and Bremner, 1964; Borthakur and Muzumdar, 1968; Tusneem and Patrick, 1971; Yoneyama and Yoshida, 1977b). Presumably, this may also hold for the wetland rice soils. The ammonification process supplies significant amounts of nitrogen to rice in the absence or presence of fertilizer nitrogen, or both. In 1963, Ponnamperuma and his associates studied kinetics of ammonification or release of soluble ammonia (NH,-N) in submerged soils in the greenhouse. In the absence of rice plants, ammonia production followed a near asymptotic course. The kinetics of release of ammonia in submerged soils of the Philippines has been adequately described by the equation (Ponnamperuma, 1965): log(A - Y) = log A - ~t

( 1 1)

where A is the mean maximal NH, released from a dry soil, Y is the actual concentration of NH4+ after ? days of submergence, and c is the soil parameter or an apparent rate constant. For 31 mineral rice soils, A was highly correlated ( r = 0.81**) with organic matter (OM) content of the soils by the equation: A = 14.4

+ 38.44 OM

(12)

The equation suggests that the ammonification followed the first-order reaction kinetics. However, in the presence of rice plants, the kinetics of ammonification is modified (Fig. 9). Some of the kinetic data on ammonia release in soil solution collected at IRRI are summarized in Fig. 10. Note that (a) Initial

NITROGEN TRANSFORMATIONS

263

a

Doys after transplanting

FIG. 9. Changes in added '5N and soil nitrogen as NH4+ in (a) cropped and (b) uncropped flooded Sacramento clay system in the greenhouse. From Manguiat and Broadbent (1977a).

increase in NH4+-N in soil solution reached a maximum about 2 weeks after submergence and was then followed by a decrease; after 8 weeks, NH4+-N was < 1 ppm (Plot la). (b) Increased salinity created by amending Maahas clay with varying amounts of NaCl markedly increased NH4+-N in the quasiequilibrium soil solution (Plot lb). This could have been due to NH4+ replacement from the soil exchange complex by added Na+ . (c) Alkalinity created by the addition of NaHCO, to the Maahas clay did not show any significant effect on NH4+-N in soil solution (Plot lc). (d) Submergence, however, had a little effect on NH4+-N in solution of natural Bani saline soil (Plot 3), whereas it had a depressing effect in the Cotabato alkali soil (Plot 2). Boka (1973), studied the changes in concentration of water-soluble NH, +-N in the submerged acid sulfate soils in presence of pot-cultured plants. He also noticed an increase in water-soluble NH, , reaching peak value after 2 weeks of flooding and then declining rapidly (Plot 4). When the chemical kinetics of water soluble NH4+-N in wet organic soils (Histosols) of the Philippines were studied, very little release of NH4+-N was noticed (Plot 5) (IRRI, 1978). When 21 soils collected from different agroclimatic zones of Sri Lanka were kept submerged but not planted in a greenhouse, Rodrigo (1961, 1962) also noticed that the NH4+-N in the percolate reached a peak concentration followed by a decrease. Katsumi (1972) and Shoji e? al. (1971) attempted to understand the behavior of tagged fertilizer nitrogen in the wetland paddy. They observed a soilN mineralization pattern with two peaks of NH4+-N during the growing season. This presumably is the net consequence of rate of mineralization and other pathways of nitrogen losses from a flooded soil system such as nitrification-denitrification, immobilization-remineralization, and plant uptake. Lin et +

264

N.

K. SAVANT AND S. K.DE DATTA

1b

‘-.--O

No. -

SM

4

-%-

I

Maahas clay 6 . 6

1.2 2.0 1.9 0 . 3

2

Cotabalo olkali clay

8.9

3.6

27

0.5 0 I

3

Banisalins 5.9

26.9

2.6

05

4

Acid s u b l a clay 4.1

1 8 .Of

OrqanicIoil 5.5

-

5.9

5

730

clay

2

&

Aclive € 2 0 s Fa Mn

6 8 Weeks submerged

-

.01

-

12

10

FIG. 10. Kinetics of NH,+-N in submerged soil systems. From Boka (19731, Pasricha (19741, IRRI (1978), and Pasricha and Ponnamperuma (1978).

al. (1973) studied cumulative nitrogen mineralization rate in 20 surface soils from rice-growing areas in Taiwan. They reported that the cumulative nitrogen mineralization rate was linearly related to the square root of time (P2)through 12 weeks of incubation. Mizuno et al. (1978) studied the effects of the addition of organic matter to rice soils on concentration of cations in soil solution and noticed an increase in the NH4+ in soil solution with constancy of the NH, +/(Ca2 + Mg2 + K +) ratio. While working with some rice soils from West Bengal (India), with waterlogged conditions, Bhattacharyya (1971), also observed a marked increase in exchangeable NH4+-N in wet soils. According to him, the increase in exchangeable NH4+ during an initial short period of waterlogging was not due to microbial activity, but due to release of clay-fixed NH4+. This could be +

+

NITROGEN TRANSFORMATIONS

265

one cause, and more experimentation is warranted to support this hypothesis. The conversion of the NO, - to NH, through reduction is generally insignificant in wetland soils (Woldendorp, 1965; Gotoh, 1973) but may be noticed if a very high level of a soluble carbon source and NO, --N are available simultaneously (Stanford et al., 1975b; Buresh and Patrick, 1978). It is difficult to work out a mathematical model for the ammonification process as it occurs in the wetland soil, because the process is regulated by a variety of soil organisms and influenced by several factors. Nevertheless, Mohanty and Patnaik (1975) attempted to quantify NaC1-extractable NH, +-N in flooded soils at a given time. They used 20 soils from various rice-growing areas of India. Based on the patterns of ammonification observed, the following generalizations were suggested for the prediction of NH, content ( Y , ppm) in a flooded soil at a given time (t): +

+

(a) For soils with high organic carbon (OC) and nitrogen and medium clay and cation exchange capacity (CEC): Y = a + bt + ct2. (b) For soils with low organic carbon and nitrogen and high clay and CEC: Y = a + b ln(t + 1) + c [In@+ 1)]*. (c) For soils with medium organic carbon, nitrogen, clay, and CEC: In y = a + b ln(t + 1) + c [In@ + 1)12. The schematic presentations of ammonification patterns of organic nitrogen in a submerged soil are depicted in Fig. 1 1 . Ammonification pattern A is common when air-dried soil is waterlogged. Pattern B is observed in the case of continuously wet or flooded soil. Pattern C is similar to the pattern A but suggests slower ammonification of immobilized nitrogen. The sigmoid pattern D of ammonification may be common in the case of very fine-textured soil. Broadbent and Reyes (197 1) noticed a similar sigmoid pattern of mineralization as evident from nitrogen uptake by rice plants cultured in Luisiana clay well supplied with organic matter (Fig. 12). As evident from the patterns A and C of Fig. 1 1 , the time course of ammonification proceeds in two stages (Yoshino and Dei, 1974): 1 . Stage I is easily influenced by the soil moisture regime (dryness or wetness) prior to resubmergence. 2 . Stage 11, during which the ammonification increases steadily, appears to be a soil characteristic and not easily affected by pretreatment such as soil drying prior to resubmergence.

For interpretation of ammonification patterns obtained from anaerobic incubation of soils, there is need to understand regional differences in agroclimate (Shiga and Ventura, 1976; An and Kona, 1977; Kim and Kim, 1978). The ammonification process in the wetland rice soils is influenced by soil factors, organic-matter factors, and agronomic practices (Craswell et aZ., 1970; Asami, 1970, 1971a,b; Tusneem and Patrick, 1971; IRRI, 1973, 1976; Koyama

266

N. K. SAVANT AND S . K . DE DATTA

A 8C

= RECTILINEAR

B =LINEAR D =SIGMOID

T

z

2u !A

8

t

TIME

-

FIG. 11, Schematic presentation of some patterns of ammonification of organic nitrogen in a submerged soil. From Asami (1971a), Yoshino and Dei (1974), and Kai and Wada (1979).

et al., 1973; Ahmad et al., 1973; Takai et al., 1974; Yoshino and Dei, 1974; Mohanty and Patnaik, 1975; Hiura et al., 1976; Yoshino and Dei, 1977a,b; Cheng et al., 1977; Kawaguchi and Kyuma, 1977; Marumoto and Yamada, 1977; Furusaka, 1977; Marumoto and Higashi, 1977; Broadbent, 1978, 1979; Ventura and Watanabe, 1978; Maeda and Shiga, 1978; Dei and Yamasaki, 1979; Zhao-liang et al., 1979; Venkatakrishnan, 1980; Zhao-liang, 1980; Sahrawat, 1980b). These factors are grouped under soil, organic matter, and agronomic management.

1. Soil Factors Nature and amount of clay Cation exchange capacity pH, Salinity (EC), and alkalinity (SAR) Structure (aggregation and puddling) Soil water dynamics Phosphorus supply Free iron and manganese oxides Temperature Water percolation (soil drainage) Microbial activity

NITROGEN TRANSFORMATIONS

267

350 LUlSlANA CLAY

Days

FIG. 12. Nitrogen mineralization patterns as indicated by nitrogen uptake by rice plants in a greenhouse experiment. From Broadbent and Reyes (1971).

2. Organic Matter (Added) Factors C/N ratio moisture content 3. Agronomic Management Factors tillage (degree of pulverization) liming fertilizer phosphorus application application of pesticides

cropping pattern nature of fallow soil drying

The factors listed under agronomic management are discussed in Section VI. We discuss the effects of the important soil and organic matter factors on ammonification in this section. Ammonification is primarily an enzymatic decomposition of organic nitrogen (Alexander, 1977). Temperature of the soil is, therefore, the most important factor controlling kinetics and magnitude of ammonium production in the wet-

268

N. K . SAVANT AND S. K . DE DATTA

land soil (Fig. 13). Research on the influence of temperature on ammonification in flooded soils is reviewed by Ponnamperuma (1976a). The highlights are: (1) Ammonification rate increased with incubation temperature in the range of 15-45°C (IRRI, 1967, 1976; Cho and Ponnamperuma, 1971). (2) Apparent firstorder reaction rate constant ( k ) of ammonification showed a rectilinear relationship with the reciprocal of temperature (l/T), indicating that the temperature effect follows the Arrhenius law (IRRI, 1973). (3) The temperature coefficient values (Q,,) for the ammonification varied from 1.0 to 1.8 with soil type and temperature range. In peaty soils in alluvial areas of Hokkaido (Japan), organic matter decomposed and released ammonia at a lower rate in the months of lower temperature (June and July), the release of NH4+ being fast during the month of higher temperature (August) (Ishizuka et al., 1973). Yoshino and Dei (1974), therefore, critically studied the effect of temperature on release of ammonia during anaerobic incubation of Japanese paddy soils and introduced the following mathematical model based on the concept of summation of effective temperature [(T - To) D ]for predicting the amount of ammonia released (Y): Y = k [(T - To) 01"

(13)

where Y is the amount of NH4+-N released, k is the constant related to the N supplying potential of a soil, T is the incubation temperature, To is the minimum threshold temperature (usually 15"C), D is the incubation period, days, and n is the exponent (or constant) relating to the pattern of ammonification. Their subsequent experiments showed good agreement between the amount of

.,10

x)

?Q Temperature ("c)

FIG. 13. Ammonification as influenced by temperature following anaerobic incubation of nonair-dried Konosu paddy soil. From Yoshino and Dei (1974).

NITROGEN TRANSFORMATIONS

269

soil nitrogen mineralized and absorbed by rice plants grown in pots and the amount of nitrogen predicted by the above model (Yoshino and Dei, 1977a,b). There is, however, a need to test this model with field data (Kim and Kim, 1978; Yoshino and Onikura, 1980). The effects of changes in the soil moisture regime on ammonification, as a result of drying treatments prior to resubmergence, are well documented. As a result of the drying of soil prior to resubmergence-a situation common during dry fallow which is followed by the wetland rice in the tropics-a flush of ammonification occurs (Hayashi and Harada, 1969; Yoshino and Dei, 1974; Shiga and Ventura, 1976; Kai and Kawaguchi, 1977; Marumoto and Higashi, 1977; and many others). This kind of “drying effect” seems to give the most benefit when wetland rice is followed by a dryland crop in rotation rather than when followed by wetland rice (Dei and Yamasaki, 1979). As mentioned earlier, the unstable stage I of the ammonification pattern (Fig. 1 1 ) is easily influenced by drying of a soil prior to resubmergence. The drying effect is observed not only in the case of mineral soils, but also in Histosols. For example, continuously wet Histosols of the Philippines released very little ammonia during anaerobic incubation at 3WC, but there was a marked increase in the release of ammonia from those soils as a result of drying prior to resubmergence (IRRI, 1978). The amino acid-N and amino sugar-N appear to be the labile-N fractions of soil organic matter that are readily mineralized as a result of drying (Marumoto and Yamada, 1977; Marumoto and Higashi, 1977), the latter being less labile than the former (Kai and Kawaguchi, 1977; Marumoto, 1977). During an initial period after submergence, Kato (1972) noticed rapid decrease in water-soluble organic matter, probably soluble amino acid-N, as a result of decomposition by soil microorganisms. Recent findings of Kai et al. (1976) clearly show that a significant fraction of the soil organic matter derived from microbial cell wall constituents, such as mucopeptides containing D-isomers of amino acids, is an important source of the so-called easily mineralizable nitrogen (labile N) in soils. These results explain why nitrogen-supplying power of some soils was found to be related to total acid-hydrolyzable N fractions, especially amino acid-N or amino sugar-N of the soil organic nitrogen pool (Lin el af., 1973; Aggarwal and Ramamoorthy, 1975; An and Kona, 1977). The results of Craswell et al. (1970) suggest that small soil aggregates contain a larger proportion of readily mineralizable organic nitrogen than do larger aggregates. In the wetland rice soils, however, the effect of different aggregate sizes may not be noticed because puddling the soil destroys aggregates. The role of the C/N ratio in the process of ammonification is well known. In general, the greater the C/N ratio or the amount of organic matter added, the slower the ammonification process becomes. This is probably because the concurrent immobilization process overshadows the ammonification process. Kai and Kawaguchi (1977) reported the influence of the C/N ratio of a mixture of rice

270

N. K. SAVANT AND S . K . DE DATTA

straw and 15N-tagged fertilizer nitrogen on the mineralization-immobilization process. The rate of ammonification was maximum immediately after the immobilization peak was reached (Fig. 14). When the C/N ratio was 84, the ammonification was not apparent after 20 weeks of incubation. They also recorded differences in the rates of ammonification at different incubation temperatures; the rate differences were more pronounced at the higher CIN ratios. The low percentage of ammonification for Indian and Burmese soils has been attributed to higher resistance to microbial decomposition of the small reserve of organic matter in these soils (Kawaguchi and Kyuma, 1977). Ammonification is also influenced by the soil colloidal properties. As noted earlier, the time versus ammonification curve may be sigmoidal in fine-textured soils, whereas it is linear or curvilinear in light-textured soils (Yoshino and Dei, 1974). Onikura et al. (1975) reported a positive relationship between NH,+-N production and CEC of soils. In the Malaysian paddy soils, the regional differences in the ammonification are explainable on the basis of climatic differences and clay mineralogy of these soils (Kawaguchi and Kyuma, 1969a). When Boka (1973) did a greenhouse incubation experiment on the kinetics of

I

Weeks incubated

FIG. 14. Immobilization and release of nitrogen in a paddy soil (pH 5.6) amended with rice straw and tagged (NH4)2S04 at different C/N ratios at 30°C. From Kai and Kawaguchi (1977).

NITROGEN TRANSFORMATIONS

27 1

NH, release in soil solution as influenced by soil texture, he found that release of NH4+ was lower in clayey soils than in sandy soils. If other factors remain constant, the ammonification process in wetland soils generally appears to be rapid in soils dominant in 1:l-type or allophane clay minerals, and slow but continuous in the soils dominant in 2: l-type clay minerals (Yamashita, 1969). The relationship between the ammonification potential and some chemical properties of 18 mineral alluvial soils was studied by Yoshino and Dei (1977a). The multiple regression analysis of their data showed high multiple coefficient of determination (R2 = 87%) when total nitrogen, CEC, exchangeable calcium, and free iron were independent variables, and amount of NH,’ released was a dependent variable. The effect of added fertilizer nitrogen on the ammonification of soil nitrogen has been described as a “priming effect” (Broadbent, 1965). This effect also seems to operate in the wetland rice soils (Maeda and Shiga, 1978). Onikura er af. (1975) reported accelerated NH4+-N release from a flooded soil following application of fertilizer nitrogen. In a field study with “N-tagged ammonium sulfate, the increments of soil nitrogen uptake induced by the addition of fertilizer nitrogen in Bangkhen dark heavy clay (Thailand, pH 5.0) constituted as much as 15-29% of the total nitrogen contained in plants (Koyama et al., 1973). Similar stimulating effects of added nitrogen on uptake of soil nitrogen by the wetland rice plant are reported by other workers (Wada et al., 1971a; IRRI, 1975; Khind and Datta, 1975; Reddy and Patrick, 1976; Krishnappa and Shinde, 1978b). A similar priming effect on the ammonification of native soil nitrogen results from added organic nitrogen. Kai and Kawaguchi (1977) reported a stimulating effect of added organic nitrogen as 4% ISN-tagged rice straw on ammonification of native soil nitrogen in three Japanese soils. They noticed a significant positive correlation ( r = 0.93) between the magnitudes of priming action and the mineralized amounts of rice-straw nitrogen. Negative priming effects of fertilizer nitrogen on the wetland soil have not been reported. Because the pH values of acid and alkaline wetland soils converge to the pH range of 6.5-7.2 following 4-6 weeks of submergence, reduced soil pH per se may not have marked effect on ammonification. However, the initial aerobic soil pH may influence the ammonification process during the early period of submergence, or until the pH reaches a stabilized value. In other words, the extent of change in pH (ApH) and rate of change in pH (dpH/dt) due to submergence may influence the ammonification in freshly submerged soil. This is supported by the findings of Kawaguchi and Kyuma (1969a). They observed significant correlation between change in pH (ApH) as a result of waterlogging and ammonia produced in anaerobically incubated Malaysian soils. The soils with pH >6 or ApH > 1 unit produced significantly higher amounts of ammonia than those with either pH <6 or ApH <1 unit. +

272

N. K. SAVANT AND S. K. DE DATTA

A paucity of phosphorus and low pH were considered as major factors causing low production of ammonia in acid sulfate soils of Thailand (Kawaguchi and Kyuma, 1969b; Seirayosakol, 1971; Motomura er al., 1975). In 12 Taiwan soils, Hsu and Wen (1969) noticed faster ammonification in neutral soils than in acidic soils. In acid sulfate soils of Thailand, the formation of H,SO, through oxidation of pyrite or FeSO, near the soil-water interface is probably common (van Breemen, 1976). In such soils, liming appears necessary to promote ammonification. From the foregoing, it is quite apparent that ammonification in the wetland rice soils is a function of several factors and, therefore, regional and seasonal

Table IV Average Ammoniacal-N Production and Ammonification Percentage in Anaerobically Incubated Paddy Soilsc1

Region or country

Samples (n)

Tropical Asia

410

Bangladesh

53

Burma

16

Kampuchea

16

India

73

Indonesia

44

Malaysia (West)

41

Philippines

54

Sri Lanka

33

Average ammoniacal N producedh (mg N/IOO g air-dry soil)

Average ammonification percentageh (% of total N)

8.5 (0.3-63.0) 6.1 ( 1 . 6 3 1.8) 2.0 (0.4-6.0)

6.8 (0.3-26.5) 4.5 ( 1 ,610.8) I .8 (0.3-5 .O) 5. I (0.7-10.0) 3.9

4.0 (0.5-7.8) 2.7 (0.5-7.3) 14.1 (3.6-36.6) 14.9 (3.8-40.3) 17.2 (3.3-63 .O)

8.4 ( I . 1-21.9)

( I .O-14.8) 12.4 (4.4-25.4) 6.3

(1.5-14.7) 10.8 (2.8-26.5)

7.0

80

5.2 (0.3- 16.4)

(1.8- 15.7) 6.6 (0.8- 18.3)

Temperate Asia Japan

84

17.5 (0.637.0)

6.5 (0.7-1 1.8)

Mediterranean countries

62

Thailand

7.6

( I . 1-23.6) UKawaguchi and Kyuma (1977). hFigures in parentheses indicate range.

4.9 (0.7-9 .O)

NITROGEN TRANSFORMATIONS

273

differences in the amounts and the patterns of ammonia production are often observed. These observations are amply supported by the ammonification data given in Table IV for the paddy soils of tropical Asia (Kawaguchi and Kyuma, 1977). 2. Immobilization

The conversion of inorganic forms of nitrogen (native or added) into organic nitrogen is largely a process of assimilation by the soil microorganisms. Immobilization is important because it affects the availability and supply of nitrogen to rice. It has been studied extensively, and was recently reviewed by Kai and Wada (1979). The highlights of some of the latest 15N studies, elucidating the immobilization process, are cited here. 1. The chemical composition, mainly the C/N ratio, of plant residues added to the soil has a marked effect on the rate and pattern of net immobilization (Fig. 14). The wider the CIN ratio, the larger the amount of nitrogen immobilized (Yoshida and Padre, 1975; Kawaguchi and Kyuma, 1977; Yoneyama and Yoshida, 1977a; Higuchi and Kurihara, 1978b). It ranged from 0 to 80% of the 15N added, depending upon the chemical composition of organic matter added (Higuchi and Kurihara, 1978a). 2. Although the contents of crude fat, crude fiber, lignin, and N-free extracts had no significant correlation with the 15N immobilization, the total amount of these components was related to I5N immobilization. The crude protein content was inversely correlated with the lSN immobilization (Higuchi and Kurihara, 1978b). 3. The immobilization of 15N added as ammonium was temperature dependent and the temperature effect followed the order 37 > 30 > 265°C (Asami, 1970). 4. In the presence of an easily decomposable organic carbon source, the I5N immobilization was rapid (Asami, 1970). 5. The presence of amorphous iron oxides ( 1 4 % ) or manganese oxides (0.5-3%) resulted in an apparent increase in the 15N immobilization (Asami, 1970). 6. The immobilization of added inorganic I5N was rapid during the first 1-2 weeks of incubation and decreased thereafter (Broadbent and Nakashima, 1970; Broadbent and Tusneem, 1971; Tusneem and Patrick, 1971; Asami, 1971a; Higuchi and Kurihara, 1978a). 7. As a result of immobilization of added IsN and rice straw, the amino acidN or amino sugar-N fractions, or both, were considerably enriched with 15N (Tusneem and Patrick, 1971; Asami, 1971b; Krishnappa and Shinde, 1978b). 8. In lowland rice soils, amended with (*sNH,),SO, and incubated anaerobically, the ratio of mineralization to immobilization (M/I) was found to be relatively constant, with a value of approximately 2 (Maeda and Shiga, 1978).

274

N. K . SAVANT AND S. K . DE DATTA

3. Ammonification of Recently immobilized Nitrogen The ammonification pattern of immobilized nitrogen shows deviation from that of native soil nitrogen. According to Asami (1970), the rate of ammonification of immobilized nitrogen is initially rapid and almost ceases at an early stage of incubation, whereas in the case of native nitrogen, ammonification may continue at a reasonably steady rate for a longer period (see Plots A and C in Fig. 11). This is presumably because the immobilized nitrogen enriches amino acid-N and amino sugar-N, which are highly labile forms and are ammonified rapidly (Kai and Kawaguchi, 1977), whereas the native soil nitrogen is relatively resistant to microbial decomposition and is, therefore, slowly ammonified. The decrease in the rate of 15N immobilization after about 1 week of incubation has been attributed to reammonification of I5N incorporated in the labile organic nitrogen pool (Higuchi and Kurihara, 1978a). In an anaerobic incubation experiment at 30°C, immobilized lSN was reammonified within 2 weeks in the absence of glucose (1%) and within 1 week in the presence of glucose (Asami, 1971a; Tusneem and Patrick, 1971). The plantuptake data suggest, however, that the reammonification of the immobilized I5N is rather slow. Krishnappa and Shinde (1978a), and Ventura and Watanabe (1978), reported low recovery of the immobilized fertilizer 15N by the second crop. Zhao-liang et al. (1979) observed that 14-22% of the immobilized "N was remineralized during the growing period of early rice. According to Kai and Wada (1979), immobilized nitrogen probably becomes progressively less available as a function of time. This has been attributed to the conversion of organic nitrogen compounds into those having higher molecular weight. McGill and Paul (1976) showed that the amorphous iron and aluminum compounds on the clay surface are probably active agents in bonding organic nitrogen to inorganic colloids, thus making it less labile to ammonification. Rice plants heavily depend on soil nitrogen for their nitrogen requirement and, therefore, the ammonification-immobilization process assumes practical significance. Most of the data on these vital processes were from anaerobic incubation studies in vitro or in greenhouse experiments using nonpuddled soils. This seriously limits their utility in the wetland rice field (Hawk, 1979). B . NITRIFICATION-DEN~TRIFICATION

The wetland soil system as it exists in the field is a complex matrix consisting of aerobic and anaerobic micro- and macrosites (Fig. 4). As discussed earlier, the two aerobic sites are the oxidized surface soil layer (site I) and the rhizosphere (site 11). These both favor nitrification. The reduced soil is highly conducive to biological denitrification, during which the NO, - is readily reduced mainly to

275

NITROGEN TRANSFORMATIONS

nitrous oxide (N,O) or N,, or both. In other words, concurrent nitrification and denitrification processes are inevitable in the wetland rice soil.

I . Nitrification Biological oxidation of NH, to NO, - in the wetland soil is common but it is undesirable because it leads to loss of nitrogen. Nitrification, strictly an aerobic process, occurs in the floodwater and in the thin, oxidized, surface soil layer (Figs. 15 and 16) where oxygen is not a limiting factor (Broadbent and Tusneem, 1971; Tusneem and Patrick, 1971; Takai and Uehara, 1973; Patrick and Gotoh, 1974; Yoshida and Padre, 1974; De Datta et a l . , 1974; Alexander, 1977; Bilal, 1977; Watanabe et al., 1981). Sakai (1970) studied nitrification in a simulated oxidized layer of a flooded soil and noticed varying lag periods until NO,formation began. Yoshida and Padre (1974), also with similar conditions, reported an increase in NO,- 6 days after submergence of Maahas clay soil. In another laboratory study, a Sacramento clay column was topdressed with (NH,),SO, at the rate of 112 kg N/ha in floodwater, and high nitrification of +

1.25

-

/

-5 1.000 t z

on soil surface

-z

z

b

0

(NH&S04

placed at 2-cm depth

075-

c

z I

g0.50-

z

0.25

-

0 0

3

6

9

12

Davs

FIG. 15. Effect of depth of placement of (NH&S04 on N03--N flooded soil. From De Datta et al. (1974).

formation in a simulated

276

N. K. SAVANT AND S. K . DE DATTA OVERALL NlTRlFtCATlON- DENITRIFICATION REACTION :

+

24 NH$ t 48 02 24NO3 t 24H20 t 48H' 2 4 N 0 3 t 5C6H1206t24Ht +12N~C30C02+42H20

OXYGEN

DIFFUSION

A

7"

FLOODWATER

2-IOcm

FIG. 16. Nitrification-denitrification reaction and kinetics of the related processes controlling nitrogen loss from the aerobic-anaerobic layer (site I) of a flooded soil system. From Patrick and Reddy (1977).

applied NH, +-N occurred in the floodwater as well as in the oxidized top soil layer (Bilal, 1977). Peak concentrations of NO,--N were observed about 7 days and of NO,--N about 10 days after topdressing. According to Patrick and Reddy (1977), nitrification in the oxidized layer (site I) is a rather slow reaction and seems to follow zero-order reaction kinetics (Reddy et al., 1976). Although we detected NO,--N in the floodwater in incubation studies, no NO,--N was detected in the floodwater (4-8 cm) in the field (IRRI, unpublished data). Various studies conducted under controlled conditions indicate that nitrification of fertilizer NH, +-N in the floodwater-aerobic soil layer system is influenced by several factors (Sakai, 1970; Yoshida and Padre, 1974; Alexander, 1977; Watanabe et al., 1981). Those factors include soil pH; soil organic matter content; soil temperature; floodwater depth; amount, form, and method of apgrowth stage of rice plants; and use of plication of fertilizer NH,+-N; pesticides. Typically, the nitrification rate is reduced markedly below pH 6 and becomes

277

NITROGEN TRANSFORMATIONS

negligible below pH 5 (Dancer et al., 1973). In other words, soil pH >6 favors nitrification. A soil pH up to 8.3 is reported to support a high population of nitrifying bacteria in the oxidized soil layer (Sakai, 1970). The presence of fresh plant residues in a flooded soil promotes immobilization of NH, and thereby retards nitrification of added NH, (Manguiat and Broadbent, 1977a). The effect of floodwater depth on nitrification in the oxidized soil layer was investigated in laboratory studies by Yoshida and Padre (1974). Their data suggest that adding depth to floodwater probably limits diffusion of oxygen to the soil-water interface, and therefore nitrifying activity in the oxidized soil layer is retarded. Sakai (1970) showed that the nitrifying power of the surface layer of the submerged soil changes with the growth of rice plants. During early growth, the method of fertilizer nitrogen application seemed to influence nitrifying power of the oxidized soil layer markedly. There was a considerable lag period for nitrification observed in the oxidized soil layer of nonfertilized plots. Broadcast application of fertilizer nitrogen, however, reduced the lag period. At the advanced growth stages of rice, the nitrifying power of the oxidized soil layer may be slightly higher than that at the early growth stage. This is attributed to the buildup of the population of nitrifying microorganisms and an increase in the thickness of the oxidized soil layer with submergence time (Tangcham and Sarutanontana, 1971; Patrick and Delaune, 1972; Takai and Uehara, 1973; Watanabe et al., 1981). In pot-culture studies, nitrifying bacteria showed slower adaptation to urea incorporation than to ammonium sulfate incorporation (at the rate of 100 kg N/ha), but these differences disappeared 2 weeks after the fertilizer application (Sakai, 1970). In the wetland rice field, the effects of transformations of urea would be rapid and of a milder nature because of environmental factors, and nitrifying bacterial activity should be restored in less than 2 weeks. Frequent use of pesticides is common on high-yielding wetland rices. The surface-applied pesticides presumably influence nitrification in the oxidized soil (Vlassak and Livens, 1975; Vojnova et al., 1975; Sahrawat, 1976; Tu, 1978; Turner, 1979). Algicide-treated plots showed a noticeable nitrification lag, indicating some inhibitory effect on ammonia formation in the oxidized soil layer (Sakai, 1970). Heterotropic bacteria that induce nitrification appear to be more tolerant to the action of the most commonly used fungicides unless they are applied at a high rate (Wainwright, 1977; Vlassak et al., 1977). In the simulated oxidized layer of a flooded alluvial clayey soil (pH 6.0; organic matter, 1.35%) addition of benomyl [methyl 1-(butylcarbamoyl)-2-benzimidazolecarbamate]at the rate of 100 ppm inhibited autotropic nitrification by Nitrosomonas and Nitrobacrer spp., but at 5000 ppm it stimulated heterotropic nitrification by Pseudo+

+

278

N. K. SAVANT AND S. K. DE DATTA

monas sp. (Gowda et al., 1977). In flooded alluvial soil (pH 6) amended with carbofuran up to 100 ppm and incubated for 40 days at ambient temperature, oxidation of NH, was not adversely affected, but NO, formation at the 1000ppm level was partly inhibited (Ramakrishna et al., 1978). To minimize nitrification in the oxidized soil layer while controlling rice pests, it may be worthwhile to investigate further such inhibitory or retarding effects of various pesticides used for wetland rice. Watanabe et al. (1981) observed depressing effect of nitrapyrin on nitrification in the oxidized surface soil layer of the wetland rice field on the IRRI farm. There is a strong evidence that the oxidized rhizosphere of wetland rice is another site for nitrification (site 11, Fig. 4).The oxidized state of the rhizosphere can be due to enzymatic oxidizing activity of the roots and the diffusive flux of oxygen from the roots to the surrounding soil (Jensen et al., 1967; Armstrong, 1967a,b, 1971). Armstrong (1971) recorded oxygen diffusion flux from the live roots of rice (cultivar Norin 36) as (15.7 1.38) X g 02/cm2root surface per minute at 23°C. The diffusion of oxygen alters the soil chemical and microbiological environment in the rhizosphere, as evident from the change in soil color and pH (Naohara and Usami, 1976). The deposition of Fe(II1) in the close vicinity of, and on, the rice roots (rhizoplane) is a common phenomenon (Kawata and Ishihara, 1965). Armstrong (1967~)attempted to study the relationship between oxygen diffusion rate (ODR) and Eh in some waterlogged organic soils. For ODR equal to that found at the rice root surface, he recorded soil Eh values ranging from 350 to 450 mV. The rice roots were also found to maintain a higher redox potential of the soil solution in a submerged soil (Tadano and Tanaka, 1970). The apparent potentials (pe) of the rhizosphere of 5- to 6-week-old rice (IR36) grown in a reduced Maahas clay (pe = 0 to -3) ranged from + 2 to $ 5 (Savant and De Datta, 1978; IRRI unpublished data). The availability of oxygen and the concomitant higher soil pe in the rhizosphere certainly induce some alternations in the microbial flora from anaerobic to aerobic (Armstrong, 1967c; Furusaka, 1977). It is, therefore, apparent that the rhizosphere of live rice plant may be a favorable site-I1 for nitrification to occur in the wetland rice soil. However, no soil analytical data on the amounts of NO,--N in the rice rhizosphere are found in the literature to date. This is presumably dueto two factors. (a) The NO,--N is highly mobile. Immediately after formation, it is likely to move into the surrounding reduced soil through diffusive transport or convective transport, or both. The reduced soil acts as a sink wherein it is readily denitrified and lost (see Fig. 4). (b) The rice roots (Tanaka et d., behave like sinks in that they also can readily absorb NO, --N 1959). This partly concurs with the observation made by Manguiat and Broadbent (1977a). From observations by Naohara and Usami (1976) it appears that nitrifying power of the oxidized rhizosphere changes with the physiological growth stages +

*

NITROGEN TRANSFORMATIONS

279

of rice. Therefore, it can be hypothesized that the nitrification in the rice rhizosphere should follow a similar decreasing trend with time after the flowering stage. Studies are essential to understand the extent and kinetics of nitrification that may be occurring in the rice rhizosphere. 2 . Denitrification Rice agronomists and environmental chemists are now very much concerned about denitrification occumng in the wetland rice soil because it adversely affects the efficiency of fertilizer nitrogen, and because the flux of nitrous oxide into the atmosphere plays an important role in the stratospheric ozone regulatory cycle (CAST, 1976). Nitrogen loss via the denitrification pathway occurring in the wetland paddies is probably significant at times. For 280 Philippine rice soils, the nitrogen loss between harvest of one crop and flooding for the next crop was reported as 10-78 kg Nlha per season with a mean of 26 (IRRI, 1973). The generally accepted pathway of the denitrification is: NO, - + NO, - + N,O -+N,. Some workers list NO between NO,- and N,O. However, St. John and Hollocher (1977) do not consider that NO is an obligatory intermediate of nitrite reduction. Innumerable papers report nitrification-denitrification leading to serious nitrogen losses from flooded soils and natural water-sediment systems. Some useful references to the understanding of nitrification-denitrification in the wetland rice soils are given in Table V.

Table V Some Useful References to Understanding Nitrification-Denitrification in the Wetland Rice Soils Reference Broadbent and Clark (1965) Painter (1970) Ponnamperuma ( 1972) Patrick er al. (1976) Focht and Vestraete (1977) Mitsui (1 977) Patrick and Reddy (1977) van Kessel (1977) Alexander ( 1977) Hauck (1979) Focht (1979) Broadbent (1979) Cho and Mills (1979) Watanabe and Mitsui (1979)

Published information Literature review on denitrification in soil Inorganic N metabolism in microorganisms Literature review on denitrification in submerged soil Mechanism of NO3 removal at water-mud interface Biochemical ecology of nitrification-denitrification Japanese work on denitrification in flooded soil systems Nitrification-denitrification in flooded soil systems Factors affecting denitrification in water-sediment systems Introduction to nitrification and denitrification in soils Methods for studying denitrification in soils Microbial kinetics of nitrification-denitrification Literature review on nitrification and denitrification in rice soils Kinetic formulation of denitrification in soil Historical review of denitrification in Japanese paddy soil

280

N. K. SAVANT AND S . K. DE DATTA

The nitrification-denitrification occurs in the oxidized and reduced soil layers (site I) and presumably in the rice rhizosphere-reduced soil system (site 11) of the wetland soil (see Fig. 4). For the sake of convenience, nitrification- denitrification at sites I and I1 is discussed separately.

3. Site I Nitrijkation-Denitrification Patrick and his associates (1972, 1974, 1976, 1977, 1978) have extensively studied the mechanisms and kinetics of the nitrification-denitrification process at the simulated site I in controlled laboratory experiments without plants. The highlights of their work are summarized in Fig. 16 and discussed here. The biological oxidation of NH,+ to NO,- ( k = 3.18 kg/cm3/day) takes place in the surface aerobic layer. This results in the development of a concentration gradient of NH4+-N across the aerobic and anaerobic layers, and causes the NH4+ present in the anaerobic layer to diffuse (d = 0.216 cm2/day) into the aerobic layer where it is nitrified. The NO,--N formed diffuses back into the anaerobic layer (d = 1.33 cm2/day) where it is readily denitrified ( k = 15.O Fg/ cm3/day). The slow rate of upward diffusion of NH4+ and also the slow rate of nitrification are probably the two limiting factors in nitrogen loss through the process of sequential nitrification-denitrification in the flooded surface soil. These studies were, however, conducted in experimental systems without any water movement. Natural field systems may be much more dynamic. Various incubation and other studies conducted to date envisage that many factors directly or indirectly influence the denitrification process in the anaerobic surface soil layer (site I) (IRRI, 1965, 1969, 1972, 1973; Azad and Khan, 1968; Chiang and Yang, 1969; Broadbent and Tusneem, 1971; Datta et al., 1971; Tusneem and Patrick, 1971; Ponnamperuma, 1972, 1976a, 1978a; Takai and Uehara, 1973; Gotoh, 1973; Patrick and Gotoh, 1974; Burford and Bremner, 1975; Mandai and Datta, 1975; Stanford e t a l . , 1975a,b; Bremner, 1977; Mitsui, 1977; Reddy and Patrick, 1977; Manguiat and Broadbent, 1977a; Reddy et al., 1978; Komatsu et al., 1978; IAEA, 1978; Denmead et al., 1979; Ryzhova, 1979; Letey et al., 1980; Firestone et al., 1980). Those factors that affect the denitrification process are: 1. Soil Factors pH and pe temperature organic matter (nature and amount) nitrate-N content and nitrification rate

9

submergence period degree of puddling or aggregation activity of denitrifiers Po, soil fertility

NITROGEN TRANSFORMATIONS

2. Other Factors floodwater regime fertilizer nitrogen management

28 1

pesticide use presence of plants

Although various reports on the effects of the environment and other factors on nitrification-denitrification in the flooded soil are published elsewhere (Table V), for this article we give only the highlights of a few recent reports. Stanford et al. (1975b) found no correlation between initial aerobic soil pH (prior to submergence) within the range of pH 5-8 and potential denitrification. This is explainable on the basis of the kinetics of pH in submerged soils (Ponnamperuma, 1972). The stabilized pH (6.5-7.2) and pe (- 1 to 3) values of the wetland soils are generally highly conducive to denitrification (Ponnampentma, 1972, 1977). The group of soil microorganisms mediating denitrification seems to function best at near pH 7 (IRRI, 1973). Komatsu et al. (1978) have obtained data that indicate a possibility of participation of a Fe3 - Fe2 system in denitrification in a waterlogged soil system. The IRRI results indicate that the denitrification reaction rate in anaerobically incubated rice soils varied with soil type (Fig. 17) and increased markedly as the temperature rose from 5 to 45°C (Ponnamperuma, 1977). Because denitrifying bacteria need organic matter as a source of carbon (and nitrogen) for their activity, the nature and amount of organic matter in the wetland soils largely determine denitrification. The capacity of Iowa surface soils for denitrification when waterlogged was highly and significantly correlated (r = 0.99" *) with their content of water-soluble or readily decomposable organic matter (Burford and Bremner, 1975). Some of the recent studies of Bremner's group (Blackmer and Bremner, unpublished work as quoted by Bremner, 1977) indicate that the ratio of N,O to N, in the gases produced during denitrification was influenced by the molecular weight and water solubility of organic substances added to soil. As expected, the water-soluble organic substances with low molecular weight had much greater effect on denitrification than highmolecular-weight substances with low water solubility. However, the field studies on denitrification in Australia (Denmead et al., 1979) and in the Philippines (IRRI, 1980; Freney et al., 1981) indicate that N,O is by far the minor product of denitrification occurring in flooded rice soils. In general, loss of nitrogen through denitrification at site I may not be a serious problem in the tropical wetland rice fields such as those in India or Burma, presumably because of their low organic matter contents (Kawaguchi and Kyuma, 1977) and their eventually weak development of reducing conditions (Yamada, 1975). Earlier work of Bremner and Shaw (1958) indicated little or no gaseous loss of nitrogen from incubated waterlogged soils containing < 1% organic carbon. +

+

282

N. K. SAVANT AND S . K. DE DATTA

“ “ I

0

2

4

6 8 1 Incubation (days)

0

FIG. 17. Kinetics of denitrification in six submerged rice soils of the Philippines at 35°C. From Ponnampemma (1977).

A few workers have considered that the rate of denitrification was independent concentration, suggesting that it followed zero-order reaction of the NO, --N kinetics (Patrick, 1960; Broadbent and Clark, 1965; Keeney, 1973). Others have reported denitrification following first-order reaction kinetics, indicating its dependence on NO, - concentration (IRRI, 1965; Ponnamperuma, 1972; Bowman and Focht, 1974; Stanford et al., 1975b; Kohl et al., 1976). Reddy et al. (1978) have attempted to elucidate this phenomenon on the basis of the NO, - diffusion from the overlying floodwater layer to the soil layer. They have concluded that where nitrogen loss due to denitrification involved NO,- diffusion, the denitrification appeared to be a first-order reaction rather than a zero-order reaction. In a water-sediment system, van Kessel (1977) found that the denitrification rate was dependent on NO,- concentration in the overlying water, approximating a first-order reaction at lower concentration, and gradually becoming independent of the NO, - concentration (zero-order reaction) at higher NO, - contents. According to Ryzhova (1979) dependence of denitrification in anaerobic soil on NO, - concentration can be described with adequate accuracy by the Michaelis-Menten equation. Takai and Uehara (1 973) studied changes in bacterial flora in the surface and underlying layers of incubated flooded soil system with progressing sub-

NITROGEN TRANSFORMATIONS

283

mergence time. In the early stage of flooding, the NO,- disappeared rapidly and was accompanied by a pronounced increase in denitrifiers. There was no marked increase in nitrifiers in the aerobic layer. At the middle stage, as the oxidized layer developed, a considerable increase in nitrifiers in the surface layer occurred. The denitrifier population was maintained at a high level, and as a result NO,--N did not accumulate. At the later stage, the number of nitrifiers was sufficiently high but the number of denitrifiers dropped, presumably because of a decrease in substrate concentration. Using ISN-tagged NH,+ , Patrick and Reddy (1976a), observed that the IsNH, content of a flooded Crowley silt loam decreased and a buildup of N, became evident after 15 days. They noticed rapid conversion of NH4+ to N, after 30 days. It is apparent, therefore, that nitrogen loss from the flooded soil due to sequential nitrification-denitrification would become significant only after 1-2 weeks of flooding. This period is normally required for the differentiation of the discernible oxidized (or aerobic) layer overlying the reduced (or anaerobic) layer. In order to avoid repetition, the effects of other factors, such as management of water, fertilizer nitrogen, and pesticides, are discussed in Section V.

Whenever extrapolating the results discussed so far to the wetland rice field, the following points should be borne in mind: (a) Most of the studies were incubation experiments conducted in a controlled environment using air-dried, nonpuddled, and < 2 mm soil. (b) The NO,--N concentrations used in the experiments were rather high. (c) Most of these studies were done in soils without rice plants. 4 . Site I1 Nitrification-Denitrification

Plant effects on concurrent nitrification-denitrification occurring at site I may be of small magnitude. Plant roots proliferating in the surface layers (site I) may decrease NO,--N concentration directly or indirectly, or both, and as a result, denitrification may be decreased. But the oxidized rice-rhizosphere-reduced soil-microsystems (site 11, Fig. 4) could be the potential site for denitrification to occur. As indicated earlier, as a result of nitrification, NO,- may be produced at site 11. If the plant root fails to trap and immediately absorb NO,- immediately after its formation, it may readily move into the surrounding reduced soil (outside the oxidized rhizosphere) where it will presumably be denitrified. Further, the denitrification may also occur within the rhizosphere itself. A positive rhizosphere effect of rice on denitrification has been demonstrated by comparing induced activity of N,O-reductase in planted and nonplanted soil (Garcia, 1975a,b). A study of the influence of the rice rhizosphere on the potential denitrifying activity and the number of denitrifiers in wetland rice soils from Senegal has shown that the denitrifying effect of the rice rhizosphere was more

284

N. K . SAVANT AND S. K. DE DA'ITA

evident in soils that were poorer in organic matter content (Garcia, 1973). The potential activity of denitrifying bacteria seems to be supported by the root exudates (Alexander, 1977; Russell, 1977; Kimura et al., 1979). This stimulation of denitrification would, however, be restricted to the thin layer of soil adhering the rice roots. But considering the total root surface that is in contact with soil, such denitrification could be tremendous and would lead to a significant loss of nitrogen. As the root activity changes with time, one would also expect the rhizosphere effect on denitrification to change. Garcia (1977) observed the maximum denitrifying effect of rice rhizosphere during the first stages of growth. However, it decreased progressively with the age of the plant presumably because of root senescence. Mandal and Datta (1975) did 15N tracer studies (Fig. 18) that indicated a marked decrease in N, loss by denitrification with the advancement of growth of rice (IR8). Although critical studies to quantify the magnitude of nitrogen loss through concurrent nitrification-denitrification at site I1 have not been attempted, there is considerable indirect evidence that envisages that de-

First month

Second mnth

4 c

I"\

01 0.6

Last month

/

*

*

1

I

I

I

0.7

0.8

0.9

1.0

NO>/NH+~ ratio

FIG. 18. Loss of fertilizer nitrogen as N2 during different growth stages of the rice plant under controlled pot-culture conditions: soil, alluvial; pH 6.8, OC, 0.69%; CEC, 20.59 meqilOO g. For 0.63, 0.78,0.95, and 1.0 N O J - / N H ~ + ratios, fertilizers of high, medium, and low water-solubility (tagged nitrophosphate, and ammonium nitrate, and superphosphate fertilizers, respectively) were used. From Mandal and Datta (1975).

NITROGEN TRANSFORMATIONS

285

nitrification losses may operate with more intensity at site I1 than at site I. This issue is further discussed in Section V. C. UREAHYDROLYSIS

Urease activity is common in wetland soils. In a 20-day incubation study, Delaune and Patrick (1970) found nearly the same urea hydrolysis for 0.33-bar moisture and waterlogged soil conditions. This was probably due to the presence of oxidized (aerobic), as well as reduced (anaerobic), soil in the incubated samples. In a recent study, we found that the oxidized surface soil collected from a wetland field hydrolyzed urea faster than the underlying reduced soil from the same field (Savant and De Datta, 1979; and unpublished data). In wetland soils, urease activity may be influenced by soil pH, temperature, and organic matter content. In 12 wetland rice soils of Taiwan, urea hydrolysis was faster in neutral soils than in acid soils (Hsu and Wen, 1969). Delaune and Patrick (1970) recorded urea hydrolysis over a pH range of 6-9, with a maximum at about pH 8. In the wetland field, urea should be incorporated before transplanting or topdressed, when soil is normally reduced and when the pH is stabilized in the 6.5-7.2 range. The pH values of the wetland soils per se, except those of acid sulfate or sodic soils may not, therefore, exert a marked effect on urea hydrolysis. Added or native organic matter shows a positive effect on rate or urea hydrolysis (Delaune and Patrick, 1970; Suthipradit, 1973). The optimum temperature for conversion of urea to ammonium seems to be in the 30-40°C range (Hsu and Wen, 1969; Magnaye, 1971). Sahrawat (1980a) has reported 2- to 3times-increased urease activity in Maahas clay amended with 1.3% Na,CO, but no effect on urease activity of added 0.5% NaCl. The level of urease activity is lower in floodwater than in the underlying soil (Suthipradit, 1973; IRRI, 1978). Savant and De Datta (IRRI, unpublished data, 1978), therefore, probably noticed urea-N in floodwater up to 5 days after topdressing IR36 rice at panicle initiation stage. However, urease activity in the floodwater of some rice soils may be enough to hydrolyze urea remaining in floodwater after broadcast application or incorporation, eventually resulting in NH, loss (Sahrawat, 1980a; Vlek et al., 1980b). Waterlogged soils seem to hydrolyze urea rapidly even when it is point placed. Craswell and Vlek (IFDC, unpublished data, 1977) found in laboratory studies that 460 mg urea-N (as supergranule) placed in 100 cm3 of waterlogged soil was completely hydrolyzed within 3 days. Savant and De Datta (1979) noticed < 10% of urea-N remaining at a placement site I day after application of supergranule urea to a wetland rice soil. They found, however, that a part of applied urea moved from its placement site through diffusive transport or convective transport, or both, before it was converted to ammonium (Savant and De Datta, 1980).

286

N. K. SAVANT AND S. K. DE DATTA

Hydrolysis of urea derivatives has been studied in the anaerobic incubation experiment by Islam and Parsons (1979). Isobutylidine diurea (IBDU), ureaform, and glycoluril hydrolyzed at much slower rates than urea and in the case of glycoluril, a lag period of 8-16 days occurred before hydrolysis began.

V. FATE OF FERTILIZER NITROGEN The fate of fertilizer nitrogen in wetland soil is governed by the various, often concurrent transformations discussed so far (Fig. 4) and reviewed earlier (Craswell and Vlek, 1979a; De Datta, 1981). In general, it is observed that these transformations of added nitrogen are nearly complete within 30-45 days after its application in a cropped wetland soil (Wada et al., 1971a,b; Shoji et al., 1971; Katsumi, 1972; Takahashi et al., 1973; Wetselaar, 1975; Maeda and Onikura, 1976; Patrick and Reddy, 1976b; Manguiat and Broadbent, 1977a). For a given agroclimate, the extent of effects of various transformation processes on the fate of fertilizer nitrogen will largely depend on the form, method, and time of its application. Therefore, we shall discuss the following: fate of basal-incorporated, deep-placed, and topdressed nitrogen. A. FATEOF BASALINCORPORATEDNITROGEN

In tropical Asia, basal fertilizer nitrogen is sometimes mixed in a puddled soil just before transplanting. This results in an increase in the NH4+-N in floodwater and surface soil (10- to 20-cm depth). The relative distribution of added nitrogen in floodwater and in underlying soil will depend on the method of incorporation and water management. In a field experiment, Takamura et al. (1976) recorded 63 ppm nitrogen in paddy water following surface application of fertilizer nitrogen. They probably, therefore, recorded a large loss of added nitrogen (-66%) due to runoff from the field. In the Philippines, Singh et al. (1978) reported runoff loss equal to 3.6 kg N as NH, /ha per season following application of ammonium sulfate at the rate of 60 kg N/ha to wetland rice fields. When 60 kg N/ha as urea were incorporated in a puddled Maahas clay just before transplanting IR36, the concentrations of urea-N and NH4+-N in static -2 cm floodwater were -54 and 8 ppm, respectively (Savant and De Datta, IRRI, unpublished data, 1978). These contents of nitrogen in floodwater disappeared, however, within a week after application, which may have been due mainly to ammonia volatilization. Aquatic weeds and algae growing in floodwater may immobilize a marked proportion of basal-incorporated fertilizer nitrogen (Craswell and Vlek, 1979b; Craswell et al., 1981; Savant et al., 1982). Chemical analyses of wet soil samples from fertilized transplanted fields reveal that with time, NH4+-N in soil solution, as well as exchangeable +

287

NITROGEN TRANSFORMATIONS

NH,+-N, decreases very rapidly. This is evident from the data on NH4+-N, measured using an electroultrafiltration (EUF) technique (Nemeth, 1976) (Fig. 19). The EUF NH4+-N fractions I, 11, and 111, representing NH4+-N in soil solution (intensity parameter), readily exchangeable, and with difficulty exchangeable (capacity factors) NH, +-N, respectively, reached peak values following basal incorporation of 40 kg N/ha as ammonium sulfate and thereafter declined rapidly to near the NH4+-N concentration of a check soil within -40 days after application. Similar trends for KC1-extractable NH, +-N were observed (Wanasuria et al., 1979, IRRI unpublished data). This decline in the

Soil: pH4.6, O.C.1.85%, clay 51.4% (8icol, Philippines) Rice: 1R36 Season : wet, 1978

'

0 kg N/he lNW2SO4

1

0

BASAL

I

30kgN/ha INW2SQ TOPDRESSING

60 Days after transplanting

30

I

90

FIG. 19. Changes in NH4+-N determined by the electro-ultrafiltration(EUF) technique in wet soil samples collected from wetland paddy field. Prior to transplanting (day 0) and at 40 days after transplanting, soil samples were collected before application of fertilizer N. From S . Wanasuria, S. K . De Datta, and K . Mengel, IRRI (unpublished data, 1979).

N. K. SAVANT AND S . K. DE DATTA

288

NH, +-N in the plow layer of the wetland soil can be attributed to (a) absorption by the rice plant (root-sink effect), (b) concurrent nitrification-denitrification resulting in nitrogen loss, (c) immobilization and fixation, and (d) deep percolation below the root zone. Katsumi (1972) observed absorption of basal-applied nitrogen by rice plants until 40 days after transplanting (i.e., until maximum tillering 'stage); its availability was 30% of the total nitrogen applied. Plant recovery of basal fertilizer nitrogen varied from 17.5 to 27% (Wada et a / . , 1971a; Shoji et at., 1971; Koyama et al., 1973; Khind and Datta, 1975; Krishnappa and Shinde, 1978a). Krishnappa and Shinde (1978a) reported that about 27% of applied nitrogen remained in a Vertisol (with 50% clay, pH 8.1) after harvest of the first rice crop. As an example, the overall recovery of basal-applied 15N as urea to a wetland rice soil is shown in Table VI. If a soil is well puddled, as expected in the wetland rice cultivation, deep-percolation loss of fertilizer nitrogen should be small to neglible because permeability is reduced. However, percolation losses may be high in light-textured rice soils. After the first week of incorporation of nitrogen, NH, volatilization loss may not be critical (IRRI, 1976, 1977, 1978). The unaccounted portion of nitrogen, however, shows a gradual increase with time (Wetselaar, 1975), which has been mainly attributed to concurrent nitrification-denitrification. It is our belief that such nitrogen loss may be more due to denitrification occurring at site I1 (the oxidized rhizosphere-reduced soil system) than at site I (the oxidized and reduced surface soil layers, see Fig. 4).This is explainable as follows: First, at site I, the nitrification-denitrification loss would be rather limited to the first 1 or 2 weeks, that is, as long as there is incorporated NH4+ (substrate) in the oxidized layer. The slow diffusion of the NH,+ would limit its upward movement to site I (from lower depth to the surface layer) and would also limit loss through denitrification. But at site 11, the nitrification-denitrification leading to nitrogen loss may Table VI Fate of Fertilizer Nitrogen Following Basal Application of Tagged Urea to the Wetland Paddy under Tropical Field Conditions" Recovery or loss of IsN Nitrogen fraction

(%)

24.27 9.73 7.51

Crop uptake Ammonia volatilization Leaching Soil (0-20 cm) As Kjeldahl nitrogen Clay-fixed NH,+-N Unaccounted nitrogen

26.49 0.87 31.13 ~~

UKrishnappa and Shinde (1978a)

289

NITROGEN TRANSFORMATIONS

continue for a longer period (beyond 2 weeks) because the roots proliferate and explore more soil volume and eventually more NH,+-N. In other words, the slow diffusion coefficient of NH4+ may not be a limiting factor in this process. Second, it is commonly observed that the apparent pattern of nitrogen loss of incorporated fertilizer nitrogen, presumably through nitrification-denitrification (as the unaccounted nitrogen) (Patrick and Reddy, 1976a), also follows a course analogous to that of growth and oxidizing-reducing activity of rice plant roots (Mandal and Datta, 1975; Naohara and Usami, 1976). B. FATEOF DEEP-PLACED NITROGEN

The fate of nitrogen deep-placed in the wetland paddy is shown in Figs. 20 and 21. It is apparent that ammonia volatilization loss following deep placement of nitrogen is almost negligible (<1%). Ammonia volatilization studies at IRRI confirmed this observation (IRRI, 1977, 1978; Mikkelsen and De Datta, 1979; De Datta, 1981). The probable N loss through concurrent nitrification-denitrification at site I would be also negligible because fertilizer nitrogen placed at 10-cm depth or

-

100kg/ha -fertilizer - N ( '5N-labeled )

7 Applied 7 DEEP

ON SURFACE

Total loss : 59 96 Recovery in tops: 19 I % Grain yield : 4.7 tans/ha

Total loss: 45.3 % Recwery in tops: 30.2 % Gain yield : 5.4 tcns/ha

290

N. K. SAVANT AND S. K. DE DATTA

Days after planting

FIG. 21. Changes in unaccounted IJN fraction of deep-placed (NH4)2S04 in flooded rice (cultivar, Vista) culture. Field experiment adapted from Partick and Reddy (1976b).

more may not get a chance to diffuse up to the soil-floodwater surface in the transplanted wetland rice field (Savant and De Datta, 1980). Thus, one should be able to account for nearly all deep-placed nitrogen as (a) nitrogen absorbed by the rice plants, (b) nitrogen remaining in a soil, and (c) nitrogen lost by leaching. In general, deep-placed fertilizer nitrogen recovered in rice plants has been consistently greater than surface-applied and/or incorporated nitrogen (Craswell et al., 1981; Savant et al., 1982). However, the gross recovery of deep-placed nitrogen in a soil-plant system showed a deficit of 36-45% (Fig. 21; Wetselaar, 1975). This suggests that this apparent loss of fertilizer nitrogen may have occurred via nitrification-denitrification at site 11. C . FATEOF TOPDRESSED NITROGEN

Topdressed nitrogen is usually applied at the panicle initiation stage, when the root system (sink) is well developed. The plant canopy, which is also large enough to modify considerably microclimate near the floodwater-air interface, thereby may help lessen ammonia volatilization loss. Only a small portion of topdressed nitrogen (3-10%) was found remaining in the plow layer (Wada et al., 1971b). Plant recovery of topdressed nitrogen has been always higher (Fig. 22) than that of basal-applied nitrogen (Wada et al., 1971b; Koyama et al., 1973; Khind and Datta, 1975; IAEA, 1978; Zhao-hang et al., 1979). Nevertheless, the unaccounted portion of topdressed nitrogen is still significant (Patrick and Red-

29 1

NITROGEN TRANSFORMATIONS

0 Koyarna et al (1973)

o Khind and Datta (1975) A

Fried (1976) (not to scale)

tillerinp

Pantie primordia initiation

Flowering

FIG.22. Recovery of '5N as (NH&SO4 separately added at various stages of growth by the wetland paddy. From Koyama el al. (1973), Khind and Datta (1975), and Fried (1976).

dy, 1976b; Reddy and Patrick, 1978). Katsumi (1972) reported that a substantial amount of the topdressed nitrogen was effective for about 2 weeks, its availability being only 30%. These findings lead us to suggest that the portion of topdressed nitrogen unaccounted for is probably lost largely through nitrification-denitrification, percolation, and/or runoff.

VI. REGULATING NITROGEN TRANSFORMATION PROCESSES It is important to regulate nitrogen transformation processes in the wetland rice soil in order to (a) check transformations (or pathways) that lead to nitrogen loss from the soil-plant system, and (b) regulate transformations that directly or indirectly aid in better nitrogen supply to the rice plant. Ammonia volatilization, denitrification, leaching, and runoff are the principal transformations or pathways that normally lead to nitrogen losses from the wetland rice soil. Ammonia volatilization is markedly influenced by the method of fertilizer nitrogen application. The extent of ammonia losses occur as broadcast applica-

292

N. K . SAVANT AND S. K . DE DATTA

tion > incorporation > deep placement (Rajaratnam and Purushothaman, 1973; Wetselaar, 1975; IRRI, 1976, 1977, 1978; Wetselaar et al., 1977; Ventura and Yoshida, 1977; Mikkelsen and De Datta, 1979; Vlek and Craswell, 1979). Topdressing at the later stages, such as panicle initiation stage can also help reduce ammonia volatilization because of a well developed root system (sink') in the surface soil and a larger plant canopy, which has a moderating effect on floodwater pH and the microclimate near the floodwater-air interface (IRRI, 1975; Bouldin and Alimagno, 1976; Wetselaar et al., 1977). Use of modified forms of urea such as sulfur-coated urea (SCU) and isobutylidene diurea (IBDU), show promise in minimizing ammonia volatilization loss (IRRI, 1978; Mikkelsen and De Datta, 1979; Vlek and Craswell, 1979). Addition of an algicide, such as diuron, to the floodwater, and thorough incorporation of basal-applied phosphatic fertilizer into the soil may help checking algal bloom and concomitant rise in daytime pH of floodwater (IRRI, 1976). Such practices, which check rise in floodwater pH, would certainly aid in reducing ammonia loss from the wetland rice soil. Denitrification at site I, i.e., occurring in the oxidized and reduced surface soil layers, can be effectively controlled by deep placement of fertilizer nitrogen. However, the magnitude and importance of denitrification at site I1 must be critically evaluated before any corrective measures are formulated. Attempts are made to check or retard nitrification in order to limit substrate concentration for denitrification. The patented synthetic compounds, for example, Nitrapyrin [2-chloro-6-(trichloromethy1)pyridine],AM (2-amino-4-chloro6-methylpyrimidine), ST (2-sulfanilamidothiazole), and some synthetic and natural products such as extract of neem (Azadirachta indica Juss) kernels and karanjin (the major furanoflavonoid from Pongamia glabra seeds) have been tried as nitrification inhibitors or retarders in wetland soils with varying degree of success (Patrick et al., 1968; Sakai, 1970; Lakhdive and Prasad, 1970; IAEA, 1970; Broadbent and Tusneem, 1971; Noguchi and Shinhara, 1971; Sarma, 1972; Manguiat and Yoshida, 1973; Rajale and Prasad, 1973, 1974; Sahrawat, 1973; Arunachalam et al., 1974; Yoshida and Padre, 1974; Ketkar, 1974; Narain and Datta, 1974; Bazilevich and Sidorenko, 1975; Reddy and Prasad, 1977; Rao and Shinde, 1977; Xian-fang et al., 1979; and many others). By and large, the nitrification retarders or inhibitors were found less effective in field studies than in laboratory or greenhouse studies. These results suggest that there is an apparent need for a more potent nitrification inhibitor or retarder. N-lignin is one of the slow-release nitrogen fertilizers that has the property to regenerate highly oxidative products to inhibit nitrification of fertilizer nitrogen. This advantage was apparent from the field studies of wetland rice soil under varying agroclimate conditions in India (Subbiah et al., 1977). Condensation products of urea and aldehydes apparently reduce the rate of nitrification-denitrification in wetland soil. Therefore, nitrogen loss through

NITROGEN TRANSFORMATIONS

293

denitrification of slow-acting condensation products of urea may be of the following order (Chiang, 1970): ureaform (UF) > isobutylidene diurea (IBDU) > crotonylidene diurea (CDU). Immobilization of fertilizer nitrogen via biological and chemical processes is an unavoidable process but may not be considered a loss because the nitrogen is not removed from the soil system. It may become available to rice plants in the course of time as a result of release if chemically fixed within the clay lattice, or after ammonification if assimilated by microorganism. By ameliorating adverse soil conditions, such as that of low pH and aluminum toxicity of acid sulfate soils, by liming (Seirayosakol, 1971; Motomura el al., 1975), by removing excess sodium from sodic soils by improving drainage conditions, or by adjusting the C/N ratio with added soil organic matter, ammonification can be improved. Point or band placement of fertilizer nitrogen may also help in reducing NH,+-N fixation by soil (Savant and De Datta, 1979; Craswell and Vlek, 1979b). Proper puddling will reduce downward movement of water, which in turn will reduce percolation loss of added nitrogen (Sanchez, 1973, 1976). Because the sulfur-coated urea fertilizers release nitrogen at a slow rate, their use may reduce leaching losses of nitrogen in the wetland soils (Savant and De Datta, 1979, 1980). Rao and Shinde (1977) prepared a slow-release, ball-type fertilizer material from rice straw or husks, wet soil, and fertilizer nitrogen (urea or ammonium sulfate). When these ball-type fertilizers, air-dried, and with a C/N ratio of 12:I , were placed at 8-cm depth between rows of flooded rice at planting time, a substantial decrease in 15N loss through leaching was observed. Incorporation of carbonaceous plant residues such as rice straw may immobilize fertilizer nitrogen, thus reducing nitrogen loss through leaching (Shinde and Chakravorty, 1975). This may result in better nitrogen turnover under the wetland conditions (Krishnappa and Shinde, 1978b). Runoff loss of nitrogen can be minimized by thorough incorporation of fertilizer nitrogen in a wetland soil and subsequent impounding of floodwater for at least for 5-7 days (Singh, 1978). A similar practice of impounding water for 1 week after topdressing is also suggested wherever possible to minimize runoff loss of topdressed nitrogen. To minimize the mobility of urea, Patnaik and Nanda (1967) suggested mixing urea with 2-5 times its weight of soil followed by 48-hr incubation before application in the field.

VII. UNRESOLVED CHALLENGES A great deal of research data are available and have increased the knowledge of nitrogen transformations in wetland soils. However, there are many unanswered questions, and research efforts are needed to provide greater under-

294

N. K. SAVANT AND S. K. DE DATTA

standing of the various transformation processes associated with a wetland rice crop. Greater understanding of transformation processes would lead to development of better management practices, which could increase fertilizer nitrogen efficiency in rice. The following are examples of the areas we believe need attention. 1. Information is needed on exchange equilibria, and sorption-desorption of NH4+, and kinetics of urea hydrolysis. Such studies should involve relatively undisturbed soil samples collected separately from the oxidized and reduced soil layers, or simulated wetland soil samples. 2. Detailed studies are needed for a better understanding of leaching, movement, and transport of fertilizer nitrogen. Again, wherever possible, undisturbed soil core samples should be used. 3. There is a need to study the mechanism and extent of nitrification-denitrification processes occurring in the rhizosphere in the reduced soil systems, especially as they are influenced by soil and plant factors. 4. In order to minimize nitrogen loss through denitrification, especially at site 11, more potential nitrification inhibitors should be identified. The following characteristics are needed for a nitrification inhibitor or retarder to be effective in wetland rice soils in the tropics: (a) it should be stable under oxidized as well as under reduced soil conditions at least for a period of 2 months; (b) it should not adversely affect rice roots; and (c) it should not have any adverse residual effects on soil ecology. 5. Nitrogen losses by various processes can be considerably minimized by deep placement of fertilizer nitrogen. More extensive research data are needed on release, distribution, and movement of nitrogen following deep placement of fertilizer nitrogen in wetland soils. 6. Information is needed on the nitrogen transformations in adverse soilshigh salinity, alkalinity, strong acidity, and high organic matter (Histosols). In South and Southeast Asia alone, these soils cover about 100 million hectares of potential rice lands (Ponnamperuma, 1978b). 7. Very little is known on the effects of placement of pesticides on the nitrogen transformation processes of deep-placed nitrogen. 8. Nitrogen transformation processes including nitrogen-balance studies should focus on using urea including 15N-labeled urea if these studies are to be relevant at the farm level. Urea is the most important source of fertilizer nitrogen for wetland rice and its importance will grow in the coming years.

REFERENCES Aggarwal, R. K., and Ramamoorthy, B. 1975. R i m 24, 127-134. Ahmad, Z., Yahiro, Y., Kai, H . , and Harada, T. 1973. Soil Sci. Plant Nutr. 19, 287-298.

NITROGEN TRANSFORMATIONS

295

Alexander, M. 1977. “Introduction to Soil Microbiology,” 2nd ed. Wiley, New York. An, S. B., and Kona, S. 1977. J . Kor. Soc. Soil Sci. Fert. 10, 29-37. Aomine, S. 1978. Soil Sci. Plant Nutr. 24, 571-580. Armstrong, W. 1967a. Physiol. Plant. 20, 540-553. Armstrong, W. 1967b. Physiol. Plant. 20, 920-926. Armstrong, W. 1967c. J . Soil Sci. 18, 27-34. Armstrong, W. 1971. Physiol. Plant. 25, 192-197. Arunachalam, N., Morachan, Y. B., and Rajagopal, K. 1974. Madras Agric. J . 61, 677-679. Asami, T. 1970. J. Sci. Soil Manure (Jpn.) 41, 433-437. Asami, T. 1971a. J . Sci. Soil Manure (Jpn.) 42, 22-25. Asami, T. 1971b. J . Sci. Soil Manure (Jpn.) 42, 74-80. Avnimelech, Y., and Laher, M. 1977. Soil Sci. Soc. Am. J . 41, 1080-1084. Azad, M. I., and Khan, A. A. 1968. Wesr Pak. J . Agric. Res. 6, 128-133. Balandreau, J., Rinaudo, G., Fares-Hamad, I . , and Dommergues, Y . 1975. In “Nitrogen Fixation by Free Living Microorganisms” (W. D. P. Stewart, ed.), pp. 57-70. Cambridge Univ. Press, London and New York. Bazilevich, S. D., and Sidorenko, 0. D. 1975. Dokl. TSKhA. 213, 22-25. Bhattacharyya, A. K. 1969. J . Indian SOC. Soil Sci. 17, 51 1-515. Bhattacharyya, A. K. 1971. J . Indian Soc. Soil Sci. 19, 209-213. Bilal, I. M. 1977. 2. Pfanzenernaehr. Bodenkd. 140, 199-207. Boka, C. K. 1973. “Chemical and Electrochemical Changes in Two Flooded Acid Sulfate Soils and Interrelationships between Iron and Manganese in Rice Plant as Affected by Added Manganese in Relation to the Growth and Yield of Rice” (Terminal Report). Int. Rice Res. Inst. (IRRI), Los Bafios, Philippines. Borthakur, H. P., and Muzumdar, N. N. 1968. J . Indian Soc. Soil Sci. 16, 143-147. Bouldin, D. R., and Alimagno, B. V. 1976. “NH, Volatilization from IRRI Paddies Following Broadcast Applications of Fertilizer Nitrogen” (Terminal Report). Int. Rice Res. Inst. (IRRI), Los Bafios, Philippines. Bouldin, D. R., Johnson, R. L., Burda, C., and Kao, C. W. 1974. J . Environ. Qual. 3, 107-1 14. Bouwmeester, R. J. B., and Vlek, P. L. G. 1981. Atmos. Environ. 15, 131-140. Bowman, R. A,, and Focht, D. D. 1974. Soil Biol. Biochrm. 6, 297-301. Bremner, J. M. 1965. Agrononrj 9, 1238-1255. Bremner, J. M. 1977. In “Soils Organic Matter Studies,” Vol. 11, pp. 229-239. IAEAIFAO. Bremner, J. M., and Shaw, K. 1958. J . Agric. Sci. 51, 22-39. Broadbent, F. E. 1965. Soil Sci. Soc. Am. Proc. 29, 692-696. Broadbent, F. E. 1978. In “Soils and Rice,” pp. 543-559. Int. Rice Res. Inst. (IRRI), Los BaRos, Philippines. Broadbent, F. E. 1979. I n “Nitrogen and Rice,” pp. 105-1 18. Int. Rice Res. Inst. (IRRI), Los Bafios, Philippines. Broadbent, F. E., and Clark, F. E. 1965. Agronomy 10, 344-359. Broadbent, F. E., and Nakashima, T. 1970. Soil Sci. Soc. Am. Proc. 34, 218-221. Broadbent, F. E., and Reyes, 0. C. 1971. Soil Sci. 112, 200-205. Broadbent, F. E., and Tusneem, M. E. 1971. Soil Sci. Soc. Am. Proc. 35, 922-926. Buresh, R. J., and Patrick, W. H., Jr. 1978. SoilSci. Soc. Am. J . 42, 913-918. Buresh, R. J., Casselman, M. E., and Patrick, W. H., Jr. 1980. Adv. Agron. 33, 149-192. Burford, J. R., and Bremner, J. M. 1975. Soil B i d . Biochem. 7, 389-394. Chen, C. F., and Chiang, P. S. 1963. Acra Pedol. Sin. 11, 171-183. Cheng, H. H., and Kurtz, L. T. 1963. Soil Sci. Soc. Am. Proc. 27, 312-316. Cheng, J . H., Chen, F. I., and Su, N. 2. 1977. Taiwan Agric. Q. 13, 69-80. Chiang, C. T. 1970. Soils Fert. Taiwan. pp. 1-27.

296

N. K. SAVANT AND S. K. DE D A V A

Chiang, C. T., and Yang, C. C. 1969. Tech. Bull. Taiwan Fert. 34, 1-54. Chiang, C. T., Hou, T. T., and Lee, W. H. 1972. Tech. Bull. Taiwan Ferr. 40, 1-30. Chin, W. T., and Kroontje, W. 1962. Soil Sci. Soc. Am. Proc. 26, 479-481. Cho, C. M., and Mills, J. G. 1979. Can. J. Soil Sci. 59, 249-257. Cho, D. Y., and Ponnamperuma, F. N. 1971. Soil Sci. 112, 184-194. (CAST) Council for Agricultural Science and Technology. 1976. Report No. 53, p. 33. Dept. of Agron., Iowa State University, Ames, Iowa. Crank, J. 1956. “Mathematics of Diffusion.” Clarendon, London. Craswell, E. T., and Vlek, P. L. C . 1979a. In “Nitrogen and Rice,” pp. 175-192. Int. Rice Res. Inst. (IRRI), Los Batios, Philippines. Craswell, E. T., and Vlek, P. L. G. 1979b. Soil Sci. Soc. Am. J . 43, 1184-1 188. Craswell, E. T., Saffigna, P. G., and Waring, S. A. 1970. Plant Soil 33, 383-392. Craswell, E. T., De Datta, S. K., Obcemea, W. N., and Hartantyo, M. 1981. Fert. R t s . 2, 247-259. Daftardar, S. Y., Deb, D. L., and Datta, N. P. 1979. J. Nucl. Agric. B i d . 8, 94-97. Dalal, R. C. 1974. Trop. Agric. (Trinidad) 51, 264-271. Dancer, W. S., Peterson, L. A,, and Chesters, G. 1973. Soil Sci. Soc. Am. Proc. 37, 67-69. Datta. N. P., Banerjee, N. K., and Prasada Rao, D. M. V. 1971. I n t . Symp. Soil Ferril. Eval. , New Delhi 1, 631-638. De Datta, S. K. 1978. I n “Soils and Rice,” pp. 671-701. Int. Rice Res. Inst. (IRRI). Los Batios, Philippines. De Datta, S. K . 1981. “Principles and Practices of Rice Production.” Wiley, New York. De Datta, S. K., and Barker, R. 1978. In “Soils and Rice,” pp. 623-648. Int. Rice Res. Inst. (IRRI), Los Batios, Philippines. De Datta, S. K., Saladaga, F. A,, Obcemea, W. N., and Yoshida, T. 1974. Proc. FAI-FA0 Semin.. New Delhi pp. 265-288. De Datta, S. K., Stangel, P. J . , and Craswell, E. T. 1981. I n “Proceedings of Symposium on Paddy Soil,” pp. 171-206. Science Press, Beijing, China. Dei, Y., and Yamasaki, S. 1979. In “Nitrogen and Rice,” pp. 451-463. Int. Rice Res. Inst. (IRRI), Los Batios, Philippines. Delaune, R. D., and Patrick, W. H., Jr. 1970. Soil Sri. Soc. Am. Proc. 34, 603-607. Denmead, 0. T., Freney, J . R., and Simpson, J . R. 1979. Soil Sci. Soc. Am. J . 43, 716-718. Dommergues, Y. R., and Rinaudo, G. 1979. In “Nitrogen and Rice,” pp. 241-260. Int. Rice Res. Inst. (IRRI), Los Bafios, Philippines. Feagley, S. E., and Hossner, L. R . 1978. Soil Sci. Soc. Am. J . 42, 364-367. Firestone, M. K., Firestone, R . D., and Tiedje, J. M. 1980. Science 208, 749-751. Focht, D. D. 1979. In “Nitrogen and Rice,” pp. 119-134. Int. Rice Res. Inst. (IRRI), Los Batios, Philippines. Focht, D. D., and Vestraete, W . 1977. Adv. Microbiol. Ecol. 1, 135-214. Freney, J. R., Denmead, 0 . T., Watanabe, I . , and Craswell, E. T. 1981. Aust. J. Agric. Res. 32, 37-45. Fried, M. 1976. Proc. Rev. Meet. Incr. Prod. under Tight Suppl. Trials Proj., Ist East-West Center. Honolulu. pp. 133-138. Furusaka, C. 1977. Proc. Int. Semin. Soil Environ. Fertil. Manage. Intens. Agric.. Tokyo pp. 769-776. Garcia, J . L. 1973. Oecol. Plant. 8 , 315-323. Garcia, J. L. 1975a Soil Biol. Biochem. 7 , 139-141. Garcia, J. L. 1975b. Soil Biol. Biochem. 7 , 251-256. Garcia, J. L. 1977. Cah. OSTROM, Ser. Biol. 12, 83-87. Gardner, W. R. 1965. Agronomy 10, 550-572.

NITROGEN TRANSFORMATIONS

297

Ghildyal, B. P. 1978. In “Soils and Rice,” pp. 317-336. Int. Rice Res. Inst. (IRRI), Los BaAos, Philippines. Gotoh, S. 1973. Bull, Kyushu Agric. Exp. Sra. 16, 669-714. Gowda, S. T. K . , Rao, V. R., and Sethunathan, N. 1977. Soil Sci. 123, 171-175. Gupta, S. K. 1964. Ph.D. thesis, Calcutta University. Hamaker, J. W. 1972. In “Organic Chemicals in the Soil Environment” (C.A.I. Goring and J . W. Hamaker, eds.), pp. 341-397. Dekker, New York. Hance, R. J. 1965. Weed Res. 5, 98-107. Hauck, R. D. 1979. In “Nitrogen and Rice,” pp. 73-93. Int. Rice Res. Inst. (IRRI), Los Banos, Philippines. Hayashi, R., AND Harada, T. 1969. Soil Sci. Plant Nutr. 15, 226-234. Higuchi, M., and Kurihara, K. 1978a. J . Sci. Soil Manure (Jpn.) 49, 58-64. Higuchi, M., and Kurihara, K. 1978b. J . Sci. Soil Manure, (Jpn.) 49, 65-70. Hiura, K., Hattori, T., and Furusaka, C. 1976. Soil Sci. Plant Nutr. 22, 459-465. Hsu, S. C . , and Wen, C. P. 1969. J. Taiwan Agric. Res. 18, 49-54. Hsu, S. C., Lin, M. F., Chiu, T. F., Lee, L. T., and Huang, W. L. 1967. J. Taiwan Agric. Res. 16, 33-43. IAEA (International Atomic Energy Agency). 1970. Tech. Rep. Ser. (log), 1-177. IAEA (International Atomic Energy Agency). 1978. Tech. Rep. Ser. (181), 1-134. IRRI (International Rice Research Institute), Los Baios, Philippines. 1965. Annu. Rep. 1964. IRRI , Los BaAos, Philippines. 1967. Annu. Rep. 1966. IRRI , Los Bados, Philippines. 1969. Annu. Rep. 1968. IRRl, Los BaAos, Philippines. 1972. Annu. Rep. 1971. IRRl , Los BaAos, Philippines. 1973. Annu. Rep. 1972. IRRI, Los BaAos, Philippines. 1975. Annu. Rep. 1974. IRRI , Los BaAos, Philippines. 1976. Annu. Rep. 1975. IRRI, Los BaAos, Philippines. 1977. Annu. Rep. 1976. IRRI, Los Banos, Philippines. 1978. Annu. Rep. 1977. IRRI, Los BaAos, Philippines. 1979. Annu. Rep. 1978. lRRl . 1980. IRRI Res. Highlights 1979. Ishizuka, Y.,Shimazaki, Y., Tanaka, A,, Satake, T . , and Nakayama, T. 1973. “Rice Growing in a Cool Environment,” p. 98. Food and Fertilizer Technol. Center, Taipei, Taiwan. Islam, M. S., and Parsons, J. W. 1979. Plant Soil 51, 319-330. Jacobs, M. H. 1967. “Diffusion Processes.” Springer-Verlag, Berlin and New York. Jensen, C. R., Stolzy, L. H., and Letey, J. 1967. Soil Sci. 103, 23-29. Jost, W. 1960. “Diffusion.” Academic Press, New York. Kai, H . , and Kawaguchi, S. 1977. Proc. Ini. Semin. Soil Environ. Soil Ferril. Manage. Intensive Agric., Tokyo pp. 3 15-323. Kai, H., and Wada, K. 1979. In “Nitrogen and Rice,” pp. 157-174. Int. Rice Res. Inst. (IRRI), Los Barios, Philippines. Kai, H., Kawaguchi, S., and Marumoto, T. 1976. Tsuchi Biseibursu 18, 27-41. Kato, T. 1972. J. Sci. Soil Manure (Jpn.) 43, 357-363. Katsumi, F. 1972. Bull. Fukui Agric. Exp. Sta. 9, 1-13. Katyal, J . C. 1980. Field Crops Res. 3, 137-145. Kawaguchi, K., and Kyuma, K. 1969a. “Lowland Rice Soils in Malaya,” pp. 1-154. Center for Southeast Asian Studies, Kyoto University. Kawaguchi, K., and Kyuma, K. 1969b. “Lowland Rice Soils in Thailand,” pp. 1-270. Center for Southeast Asian Studies, Kyoto University. Kawaguchi, K., and Kyuma, K. 1977. “Paddy Soils in Tropical Asia.” Univ. Press of Hawaii, Honolulu.

298

N. K. SAVANT AND S. K. DE DAlTA

Kawata, S., and Ishihara, K. 1965. Proc. Crop Sci. SOC. (Jpn.) 33, 168-173. Keeney, D. R. 1973. J . Environ. Qual. 2, 15-29. Ketkar, C. M. 1974. Fert. News 19, 25-26. Khind, C. S., and Datta, N. P. 1975. J . Indian SOC. Soil Sci. 23, 442-446. Kim, B. T., and Chae, D. U. 1970. Res. Rep. Oflice Rural Dev. (Korea) 13, 49-56. Kim, K. S., and Kim, Y. U. 1978. Theses Chonnam Univ. 24, 55-61. Kimura, M., Wada, H., and Takai, Y. 1977. J . Sci. Soil Manure 48, 85-90. Kimura, M., Wada, H., and Takai, Y. 1979. Soil Sci. Plant Nufr. 25, 145-153. Kohl, D. H., Vithayathil, F., Whitlow, P., Shearer, G., and Chien, S. H. 1976. Soil Sci. SOC.Am. J . 40, 249-253. Komatsu, Y., Takagi, M., and Yamaguchi, M. 1978. Soil Biol. Biochem. 10, 21-26. Koyama, T. 1981. Fert. Res. 2, 261-278. Koyama, T., Chammek, C., and Niamsrichand, N. 1973. Trop. Agric. Res. Cent. Tokyo, Tech. Bull. 3, 1-79. Krishnappa, A. M., and Shinde, J . E. 1978a. Working Paper No. 41 presented at the 4th Res. Coordination Meeting of the Joint FAOIIAEAIGSF Coordinated Program on N Residues held at Piracicaba, Brazil. Krishnappa, A. M., and Shinde, J. E. 1978b. FAONAEA Inr. Symp. Colombo. Sri Lanka. Lakhdive, B. A., and Prasad, R. 1970. J . Agric. Sci. Cambridge 75, 375-379. Letey, J., and Oddson, J. K. 1972. In “Organic Chemicals in the Soil Environment” (C.A.I. Goring and J. W. Hamaker, eds.), pp. 399-440. Dekker, New York. Letey, J., Hadas, A., Valoras, N., and Focht, D. D. 1980. J . Environ. Qual. 9, 232-235. Lin, C. F., Chang, A. H., and Tseng, C. C. 1973. J. Taiwan Agric. Res. 22, 186-203. Lung-yie, C., and Chin-cheng, F. 1978. Acta Pedol. Sin. 15, 75-82. McGill, W. B., and Paul, E. A. 1976. Can. J. Soil Sci. 56, 203-212. Maeda, K., and Onikura, Y. 1976. J. Sci. Soil Manure (Jpn.) 47, 99-105. Maeda, K., and Shiga, H. 1978. Soil Sci. Plant Nutr. 24, 515-524. Magnaye, C. P. 1971. M. S. thesis, College of Agriculture, University of the Philippines, Los Bafios, Laguna, Philippines. Mandal, S. R., and Datta, N. P. 1975. Indian Agric. 19, 127-134. Manguiat, I. J., and Broadbent, F. E. 1977a. Philipp. Agric. 60, 354-366. Manguiat, I. J., and Broadbent, F. E. 1977b. Philipp. Agric. 60, 367-377. Manguiat, I. J., and Yoshida, T. 1973. Soil Sci. Plant Nutr. 19, 95-102. Marumoto, T. 1977. J. Sci. Soil Manure (Jpn) 48, 391-395. Marumoto, T., and Higashi, T. 1977. Proc. Int. Semin. Soil Environ. Ferril. Manage. Intensive Agric., Tokyo. pp. 303-314. Marumoto, T., and Yamada, Y. 1977. Bull. Faculry Agric. Yamaguri Univ. 28, 71-81. Matsuguchi, T. 1979. In “Nitrogen and Rice,” pp. 207-222. Int. Rice Res. Inst. (IRRI), Los BaRos, Philippines. Mikkelsen, D. S., and De Datta, S. K. 1979. In “Nitrogen and Rice,” pp. 135-156. Int. Rice Res. Inst. (IRRI), Los Bafios, Philippines. Mikkelsen, D. S., De Datta, S. K., and Obcemea, W. N. 1978. Soil Sci. Soc. Am. 1. 42,725-730. Mitsui, S. 1977. Proc. Inr. Semin. Soil Environ. Ferril. Manage. Intensive Agric., Tokyo. pp, 259-268. Mitsui, S., and Mukai, N. 1959. Res. Rep. Urea Res. Organ. No. 118. Mitsui, S., and Namioka, H. 1958. Res. Rep. Urea Res. Organ. No. 54. Mitsui, S., and Takatoh, H. 1963. Soil Sci. Plant Nutr. 9, 103-1 10. Mizuno, N., Miyamori, Y., and Kamada, K. 1978. J . Sci. Soil Manure (Jpn.) 49, 85-89. Mohanty, S. K., and Patnaik, S. 1975. Acta Agron. Acad. Sci. Hung. 24, 446-451. Mortland, M. M., and Wolcott, A. R. 1965. Agronomy 10, 150-197.

NITROGEN TRANSFORMATIONS

299

Motomura, S., Seirayosakol, A , , and Cholitkul, W. 1975. Paper presented at the Int. Rice Res. Conference, Int. Rice Res. Inst. (IRRI), Los Barios, Philippines. Nair, R. R., Pisharody, P. N., and Gopalakrishnan, R. 1976. Agric. Res. J . Kerala 14, 1-4. Naohara, T . , and Usami, S. 1976. J . Sci. Soil Manure (Jpn.) 47, 391-397. Narain, P., and Datta, N. P. 1974. fndian J . Agric. Sci. 44,339-344. Nemeth, K. 1976. Appl. Sci. Dev. 8, 89-1 11. Noguchi, H., and Shinhara, K. 1971. Res. Rep. Fukuoka Agric. Exp. Sra. 9, 17-22. Nommik, H. 1965. Agronomy 10, 198-258. Nommik, H. 1970. PIanr Soil 33, 581-595. Nor, Y. M., and Majid, N. M. 1979. Workshop held at Univ. of Chiang Mai, Thailand. Olsen, S. R., and Kemper, W. D. 1968. Adv. Agron. 20, 91-151. Onikura, Y., Yoshino, T., and Maeda, K. 1975. J . Sci. Soil Manure (Jpn.) 46, 255-259. Painter, H. A. 1970. Water Res. 4, 393-450. Pal, S. S., Misra, A. K., and Sethunathan, N. 1980. Soil Sci. 129, 54-57. Pande, H. K., and Adak, N. K. 1971. Exp. Agric. 7, 329-336. Pang, P. C., Hedlin, R. A , , and Cho, C. M. 1973. Can. J . Soil Sci. 53, 331-341. Pasricha, N. S. 1974. Terminal Report. Int. Rice Res. Inst. (IRRI), Los Barios, Philippines. Pasricha, N. S. 1976. Soil Sci. 121, 267-271. Pasricha, N. S., and Ponnamperuma, F. N. 1978. Riso 27, 13-18. Patnaik, S., and Nanda, B. B. 1967. Oryza 4, 78-79. Patrick, W. H., Jr. 1960. Trans. Int. Congr. Soil Sci., 7th 2, 494-500. Patrick, W. H., Jr., and Delaune, R. D. 1972. Soil Sci. Soc. Am. Proc. 36, 573-576. Patrick, W. H . , Jr., and Gotoh, S. 1974. Soil Sci. 118, 78-81. Patrick, W. H., Jr., and Mahapatra, I. C. 1968. Adv. Agron. 20, 323-359. Patrick, W. H., Jr., and Reddy, K. R. 1976a. J . Environ. Qual. 5 , 469-472. Patrick, W. H . , Jr., and Reddy, K. R. 1976b. Soil Sci. SOC.Am. J . 40, 678-681. Patrick, W. H., Jr., and Reddy, K. R. 1977. Proc. Int. Semin. Soil Environ. Soil Fertil. Manage. Infensive Agric. Tokyo. pp. 275-281. Patrick, W. H., Jr., and Reddy, C. N. 1978. In “Soils and Rice,” pp. 361-379. Int. Rice Res. Inst. (IRRI), Los BaAos, Philippines. Patrick, W. H., Jr., Peterson, F. J., and Turner, F. T. 1968. Soil Sci. 105, 103-105. Patrick, W. H., Jr., Delaune, R. D., Engler, R. M., and Gotoh, S. 1976. Proj. Rep. EPA-600/3-76042. Louisiana Agric. Exp. Sta., Louisiana State University, Baton Rouge. Philips, R. E., and Brown, D. A. 1964. Soil Sci. Soc. Am. Proc. 28, 758-763. Ponnamperuma, F. N. 1955. Ph.D. thesis, Cornell University, Ithaca, New York. Ponnamperuma, F. N. 1965. In “The Mineral Nutrition of the Rice Plant,’’ pp. 295-328. Johns Hopkins Press, Baltimore, Maryland. Ponnamperuma, F. N. 1972. Adv. Agron. 24, 29-96. Ponnamperuma, F. N. 1976a. In “Climate and Rice,” pp. 249-263. Int. Rice Res. Inst. (IRRI), Los Barios, Philippines. Ponnamperuma, F. N. 1976b. IRRI Res. Paper Ser. 2 , 1-18. Ponnamperuma, F. N. 1977. IRRI Res. Paper Ser. 5 , 1-32. Ponnamperuma, F. N. 1978a. In “Soils and Rice,” pp. 421-441. Int. Rice Res. Inst. (IRRI), Los Bafios, Philippines. Ponnamperuma, F. N. 1978b. Agron. Abstr. Am. Soc. Agron. Annu. Meer. Chicago p. 43. Prasad, R., and De Datta, S. K. 1979. In “Nitrogen and Rice,” pp. 465-484. Int. Rice Res. Inst. (IRRI), Los Bafios, Philippines. Racho, V. V . , and De Datta, S. K. 1968. Soil Sci. 105, 419-427. Raghupathy, B., and Raj, D. 1971. Annamalia Univ. Agric. Res. Annu. 3, 30-34. Rajale, G . B., and Prasad, R. 1973. J. Agric. Sci. Cambridge 80, 479-487.

300

N. K. SAVANT AND S. K. DE DATTA

Rajale, G. B., and Prasad, R. 1974. J . Agric. Sci. Cambridge 83, 303-307. Rajaratnam, J . A., and F’urushothaman, V. 1973. Malays. Agric. Res. 2, 59-64. Ramakrishna, C., Rao, V. R., and Sethunathan, N. 1978. Soil Biol. Eiochem. 10, 555-556. Rao, C. U. M., and Shinde, J. E. 1977. J . Nucl. Agric. Biol. 6, 139-141. Reddy, K. R., and Patrick, W. H., Jr. 1976. Agron. J. 68, 965-969. Reddy, K. R., and Patrick, W. H . , Jr. 1977. Soil Sci. 123, 142-148. Reddy, K. R., and Patrick, W. H., Jr. 1978. Agron. J. 70, 465-467. Reddy, K. R., Patrick, W. H., Jr., and Philips, R. E. 1976. Soil Sci. Soc. Am. J . 40, 528-533. Reddy, K. R., Patrick, W. H . , Jr., and Philips, R. E. 1978. Soil Sci. Soc. Am. J. 42, 268-272. Reddy, R. N. S., and Prasad, R. 1977. R i m 26, 217-224. Rodrigo, D. M. 1961. Trop. Agric.(Cey/on) 117, 245-272. Rodrigo, D. M. 1962. Trop. Agrir. (Ceylon) 118, 139-218. Rodrigo, D. M. 1967. Inr. Rice Comm. Newsl. 16, 7-16. Rose, A. H. 1976. “Chemical Microbiology.” Plenum, New York. Russell, R. S. 1977. “Plant Root Systems, Their Function and Interaction with the Soil.” McGrawHill, New York. Ryzhova, I. M. 1979. Sov. SoilSci. 11, 168-171. Sahrawat, K. L. 1973. Ph.D. thesis, Indian Agric. Res. Inst., New Delhi. Sahrawat, K. L. 1976. Pesticides 10, 15-18. Sahrawat, K. L. 1978. In?. Rice Res. Newsl. 3, 16. Sahrawat, K. L. 1979. Commun. Soil Sci. Plant Anal. 10, 1015-1023 Sahrawat, K. L. 1980a. Soil Biol. Biochem. 12, 195-196. Sahrawat, K. L. 1980b. Plant Soil 55, 225-233. Sakai, H. 1970. Progress Report of the All-India Coordinated Rice Improvement Project, Vol. 2. Sanchez, P. A. 1973. Soil Sci. 115, 149-158. Sanchez, P. A . 1976. “Properties and Management of Soils in the Tropics.” Wiley, New York. Sarma, A. D. 1972. Andhra Agric. J. 19, 139-143. Savant, N. K., and De Datta, S. K . 1979. Soil Sci. Soc. Am. J . 43, 131-134. Savant, N. K . , and De Datta, S. K . 1980. Soil Sci. Soc. Am. J. 44, 559-565. Savant, N. K., De Datta, S. K., and Craswell, E. T. 1982. Soil Sci. Soc. Am. J . 46, 567-573. Seirdyosakoi, A. 1971. Annu. Res. Rep. Rice Dept.. Ministty Agric., Bangkok. pp. 119-128. Shiga, Y. 1961. Soil Sci. Plant Nutr. 7 , 119-124. Shiga, H., and Ventura, W. 1976. Soil Sci. Plant Nurr. 22, 387-399. Shimpi, S. S., and Savant, N. K . 1975. Soil Sci. Sac. Am. Proc. 39, 153-154. Shinde, J . E . , and Chakravorty, S. P. 1975. Acta Agron. Acad. Sci. Hung. 24, 419-423. Shinde, J. E., and Vamadevan, V. K. 1974. In “Effects of Agricultural Production on Nitrates in Food and Water with Particular Reference to Isotope Studies,” pp. 45-52. International Atomic Energy Agency, Vienna. Shoji, S. 1976. In “The Fertility of Paddy Soils and Fertilizer Applications for Rice,” pp. 36-48. Food and Fertilizer Technol. Center, Taipei, Taiwan. Shoji, S., Wada, G., Saito, K . , Shinbo, I., and Takahashi, J . 1971. Proc. Crop. Sci. Soc., Jpn. 40, 28 1-286. Singh, V. P. 1978. Ph.D. thesis, University of the Philippines at Los Bafios, Laguna, Philippines. Singh, V. P., Wickham, T. H., and Corpuz, I. T. 1978. Tech. Paper No. 11, Vol. 3, 144-160. Int. Comm. on Imgation and Drainage, Manila, Philippines. Stanford, G . , Legg, J. O., Dzienia, S., and Simpson, E. C., Jr. 1975a. Soil Sci. 120, 147-152. Stanford, G., Vander Pol, R. A,, and Dzienia, S. 1975b. Soil Sci. Soc. Am. Proc. 39, 284-289. St. John, R. T., and Hollocher, T. C. 1977. J . Eiol. Chem. 252, 212-218. Subbiah, B . V., Joshi, 0. P., and Sachdev, M. S. 1977. Proc. Symp. IAEAIFAO 1, 365-370. Suthipradit, S. 1973. J . Aust. Inst. Agric. Sci. 39, 272-273. Tadano, T., and Tanaka, A. 1970. J . Sci. Soil Manure (Jpn.) 41, 142-148.

NITROGEN TRANSFORMATIONS

30 1

Tadano, T., and Yoshida, S. 1978. In “Soils and Rice,” pp. 399-420. Int. Rice Res. Inst. (IRRI), Los Batios, Philippines. Takahashi, J., Shoji, S . , Nogi, T., and Wada, G. 1973. Crop Sci. SOC. Jpn. Proc. 42, 91-96. Takai, Y., and Uehara, Y. 1973. J . Sci. Soil Manure (Jpn.) 44, 463-470. Takai, Y., Wada, H., Kagawa, H., and Kobo, K. 1974. Soil Sci. Plant Nutr. 20, 33-45. Takamura, Y., Tabuchi, T., Suzuki, S., Harigae, Y., Ueno, T., and Kubota, H. 1976. J . Sci. Soil Manure Jpn. 47, 398-405. Tanaka, A., Patnaik, S., and Abichandani, C. T. 1959. Proc. IndianAcad. Sci. Sec. B 49,386-396. Tangcham, B., and Sarutanontana, S. 1971. Annu. Res. Rep., RiceDept., MinistryAgric., Bangkok pp. 150-156. Thomas, G. W. 1970. In “Nutrient Mobility in Soils: Accumulation and Losses” (0.P. Englestad, ed.), pp. 1-20. Soil Sci. SOC.Am., Madison, Wisconsin. Tu, C. M. 1978. Commun. Soil Sci. Plant Anal. 9, 620-636. Turner, F. T. 1979. SoilSci. SOC. Am. J . 43, 955-957. Tusneem, M. E., and Patrick, W. H . , Jr. 1971. La. Agric. Exp. Sta. Bull. 657, 1-75. Van Breemen, N. 1976. Agric. Res. Rep. 848. Center for Agric. Publ. Documentation, Wageningen. Van Breemen, N., and Wielemaker, W. G. 1972a. Soil Sci. SOC.Am. Proc. 38, 55-60. Van Breemen, N., and Wielemaker, W. G. 1974b. Soil Sci. SOC.Am. Proc. 38, 61-66. Van Kessel, J. F. 1977. Water Res. 11, 259-267. Van Raalte, M. H. 1941. Ann. Bot. Card. Buitenzorg 51, 43-57. Varade, S. B., and Patil, E. A. 1971. Riso 20, 219-223. Venkatakrishnan, S. 1980. Plant Soil 54, 149-152. Ventura, W., and Watanabe, I. 1978. Soil Sci. Plant Nutr. 24, 535-545. Ventura, W . B., and Yoshida, T. 1977. Plant Soil 46, 521-531. Ventura, W., and Yoshida, T. 1978. Soil Sci. Plant Nutr. 24, 473-479. Vlassak, K., and Livens, J. 1975. Sci. Total Environ. 3, 363-372 [Fert. Abst. 9, 95, 19761. Vlassak, K., Govindaraju, K., and Verstraeten, L. M. J. 1977. Proc. Int. Semin. Soil Environ. Fertil. Manage. Intensive Agric., Tokyo, pp. 750-754. Vlek, P. L. G., and Craswell, E. T. 1979. Soil Sci. SOC. Am. J . 43, 352-358. Vlek, P. L. G., and Craswell, E. T. 1981. Ferr. Res. 2, 227-245. Vlek, P. L. G., and Stumpe, J. M. 1978. Soil Sci. SOC. Am. J . 42, 416-421. Vlek, P. L. G., Byrnes, B. H., and Craswell, E. T. 1980a. Plant Soil 54, 441-449. Vlek, P. L. G.,Stumpe, J. M., and Bymes, B. H . 1980b. Ferr. Res. I(3), 191-202. Vojnova, Z., Petkova, P., and Kostov, 0. 1975. Rocz. Glebozn 26, 207-21 1 [Fert. Abst. 9, 337, 19761. Wada, G., Shoji, S., and Takahashi, J. 1971a. Crop Sci. SOC. Jpn. Proc. 40, 275-280. Wada, G., Shoji, S., Takahashi, J., Saito, K., and Shinbo, I . 1971b. CropSci. SOC. Jpn. Proc. 40, 287-293. Wainwright, M. 1977. Z. Pflanzenernaehr. Bodenkd. 140, 587-603. Waring, S. A., and Bremner, J. M. 1964. Nature (London) 201, 951-952. Watanabe, I. 1978. In “Soils and Rice,” pp. 465-478. Int. Rice Res. Inst. (IRRI), Los Baiios, Philippines. Watanabe, I., and Mitsui, S. 1979. IRRI Res. Paper Ser. 37, 1-10, Watanabe, I., Padre, B. C., Jr., and Santiago, S. T. 1981. Soil Sci. Plant Nutr. 27, 373-382. Wetselaar, R. 1975. Proc. Conf. Thai-Ausr. ChaoPhyaRes. Proj., Chainat 1966-1975pp. 91-100. Wetselaar, R., Shaw, T., Firth, P., Oupathum, J., and Thitipoca, H . 1977. Proc. Int. Semin. Soil Environ. Fertil. Manage. Intensive Agric. Tokyo pp. 282-288. Woldendrop, J. W. 1965. Ann. Inst. Pasteur, Paris, Suppl. 3, 316-329. Xian-fang, W., Bao-zhong, W., Fu-jun, W . , Gen-yuan, P., and Hong-kui, M., 1979. Acta Pedol. Sin. 16, 380-386. Yamada, Y . 1975. Proc. Ferr. INPUTS Proj., 1974, East-West Food Inst., Honolulu, pp. 116-124.

302

N. K. SAVANT AND S . K. DE DATTA

Yamaguchi, M. 1976. In Seminar paper “Fertility of Paddy Soils,” pp. 60-75. Taipei, Taiwan. Yamane, I. 1978. In “Soils and Rice,” pp. 381-398. Int. Rice Res. Inst. (IRRI), Los Banos, Philippines. Yamashita, K. 1969. In “New Techniques in Relation to Soil and Fertilizer” (in Japanese). pp. 84- 100. Gihodo, Tokyo. Yoneyama, T., and Yoshida, T. 1977a. Soil Sci. Plant Nirtr. 23, 41-48. Yoneyama, T., and Yoshida, T. 1977b. Soil Sci. Plant Nutr. 23, 175-183. Yoshida, T. 1978. In “Soils and Rice,” pp. 445-463. Int. Rice Res. Inst. (IRRI), Los Batios, Philippines. Yoshida, T., and Padre, B. C., Jr. 1974. Soil Sci. Plant Nutr. 20, 241-247. Yoshida, T., and Padre, B. C., Jr. 1975. Soil Sci. Plant Nutr. 21, 281-292. Yoshino, T., and Dei, Y. 1974. Japan Agric. Res. Quart. 8, 137-141. Yoshino, T., and Dei, Y. 1977a. J . Cent. Agric. Exp. Sta. Jpn. 25, 1-62. Yoshino, T . , and Dei, Y. 1977b. Proc. Int. Semin. Soil Environ. Ferti/. Munagc. Inreiisilu Agric.. , TO~YO.pp. 297-302. Yoshino, T., and Onikura, Y. 1980. J. Cent. Agrir. Exp. Sta. 31, 73-86. Zhao-liang, 2 . 1980. Nitrogen Advisory Committee Meeting, Int. Rice Res. Inst., Los Batios. Philippines, April 1980 (unpubl. mimeo). Zhao-hang, Z . , Rong-ye, C . , Yong-fu, X., Yin-hua, X., and Shao-lin, 2. 1979. A m Prdol. Sin. 16, 218-233.

Index A

Crop production, predicting, 193-215 Cynodon ducqdon, 149

Aphelenchoides besseyi, 60 Alfalfa, 145, 175 Aluminum, 142, 147, 152 Ammonia volatization, 254-261, 288, 291 Ammonification, 262-2’74 Avena saliva, 151 Avocado, 226

D 2,4-D, 187 Dactylis glomeraiu, 15 I Denjtrification, 279-285 Dicamba, 187 Dormancy, seed, 171 Douglas fir, 227

B Barley, 151, 229 Beet, sugar, 222, 227, 228 Bermudagrass, 149 Blast, 58 Bluegrass, Kentucky, 228 Boron, 142 Breeding technique, innovative, 82-84 Bromus. 151 Bromegrass, 151 Brown leaf spot, 59 Buckwheat, 152

E Ethylenediaminetetraacetic acid, 23 1 Ethylenediaminedi-o-hydroxyphenylaceticacid, 231-232 Evapotranspiration, 195-205 estimating, 20 1-205 measuring, 199-201

F Fagopyrurn esculentum, I52 Fungicide, 186

C Cadmium, 142 Calcium, 181- 183 Canary grass, reed, 145 Charcoal, 235 Chilo plejudellus, 61 Chlorosis, 223-230 temperature effect, 97, 99- 100 Citrus. 228, 229 Clover, arrowleaf, 187 crimson, 186 subterranean, 165- 191 white, 174, 186 Compost, 233-134 Copper, 142, 147 Corn, 28, 145, 146, 151, 199-201, 211-212, 224, 237, see also Maize; Zea mays

G Genetics, iron utilization, 237-238 rice resource, 37-91 Germ plasm evaluation, 58-68 preservation, 68-71 use, 71-80 Glwine mux. 145 Glyphosphate, 186 Grain moth, angoumois, 61 Growth study, sludge use, 144-148

H Helminthosporium oryzae, 59

303

304

INDEX

Herbicide, effect, 186-187 Hoja blanca, 60

I Iron, 142, 147 calcareous soil, 217-240

K Kernel smut, rice, 60

L Leaching, 250-252 Legume fertilization, 175-183 nodulation, 146 Lignite, 235-236 Lignosulfonate, 234-235 Liming, 182-183 Lissorhoptrus oryzophilus, 6 I

M Macadamia, 226, 228 Magnaporthe salvanii, 59 Maize, 223, see also Corn; Zea mays Male sterility, 102 Manganese, 142 Manure, 233-234 Medicago sativa, 145 Mineral nutrition, temperature, 110-1 12 Millet, 148 Molybdenum, 181

N Neovossia barclayana, 60 Nickel, 142, 147 Nitrification, 274-285 Nitrogen fixation, 171-174, 185 nitrate, 226 transformation, 241-302

0 Oat, 151 Oebalus pugnax, 61 Orchard grass, 151, 152 Oryza barthii, 44 Oryza fatua. 42 Oryza glabberima, 42, 44 Oryza longistaminata, 44 Oryza minuta, 42 Oryza izivara, 43, 75 0,yza officinalis, 42 Oryza perennis, 42 Oryza rufpogon, 43, 59 Oryza sativa, 42-45 Oryza sativa f. spontanea, 74 Oryza stapfii, 44

P

Panirum miliocium, 148 Paraquat, 187 Paspalum, 98 Pea, 229 Peanut, 18 Petunia, 227 Phalaris arundinaceae, 145 Phaseolus vulgaris, 147 Phosphate fertilization, 11 1-1 12 Phosphorus, 178-181 sewage sludge, 129-163 Photosynthesis, 222-223 low temperature, 98, 113-1 14 Pine, 227 Pyricularia oryzae, 58

R Rhizobium trifolii, 173-174 Rhizoctonia solani, 59 Rhizotron, 1-35 Rice, 102, 114, 225 genetic resource, 37-91 soil, nitrogen, 241-302 Root illumination, 18 research, rhizotron, 1-35 Rye, 146, 149

INDEX

S Secult. cereule, I46 Sheath blight, rice, 59 Sitotroga cerealella, 61 Sludge, 234 phosphorus in, 129-163 Snapbean, 147 Sogatodes orizicola, 60 Soil, calcareous, iron in, 217-240 Sorghum, 212, 226, 233 Sorghum, 98, 102 Sorghum vulgare sudanense, 145 Soybean, 18, 28, 98, 114, 145, 146, 175, 209-210, 223, 226, 229, 237 Spinach, 222 Spruce, 227 Stem borer, 61 rot fungus, 59 Stink bug, 61 Straighthead, 60 Sudan grass, 145 Sulfur, 177-179 Sunflower, 224 Sweet potato, 114

T Temperature, 184- 185 adaptation breeding, 119-124 chilling injury, 96-99 freezing injury, 94-95 growth, 104-109 low nonfreezing, 95-103 mineral nutrition, 110-1 12 morphogenesis, 114-1 17 rice, 61-62 water relations, 109-1 10

305

Tilletia barclaj~ana.60 Tobacco, 226, 228 Tomato, 237 Tr{folium incarnaium, 186 Trijolium repens, 186 Trijolium subterranean. 166, I87 Trifolium vesiculosum, .I 87 Triticum destivum, 147

U Urea, 254, 285-286

V Vetch, 187 Vicea sativa. 187

w Water, drainage, 149-153 relation, temperature, 109- 110 stress, crop production, 193-215 Water weevil, rice, 61 Wheat, 147, 152 transpiration model, 205-209 White tip disease, rice, 60-61

Z Zea mays, 145, see also Corn; Maize low temperature effect, 93- 128 Zinc, 142, 147

This Page Intentionally Left Blank