Advances in Agronomy, Volume 16

ADVANCES I N AGRONOMY VOLUME 16 CONTRIBUTORS TO THIS VOLUME R. W. ALLARD LOWELL E. ALLISON R. J. BULA KENNETHL. DAV...

Author: Arthur Geoffrey Norman

20 downloads 981 Views 22MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

ADVANCES I N

AGRONOMY VOLUME 16

CONTRIBUTORS TO THIS VOLUME

R. W. ALLARD LOWELL E. ALLISON R. J. BULA KENNETHL. DAVXSON V. C. FARMER P. E. HANSCHE W. M. HOFFMAN J. LETEY D. E. MCCLOUD W. J. MCHARDY B. D. MIMIELL NORMANJ. ROSENBERC R. H. SHAW

L. H. STOLZY G. H. STRLNCFIELD G. L. TERMAN B. C. WRIGHT MADISON J. WRIGHT

ADVANCES IN

AGRONOMY Prepared under the Auspices of the AMERICANSOCIETY

OF

AGRONOMY

VOLUME 16 Edited by A.

G. NORMAN

The University of Michigan, Ann Arbor, Michigan

ADVISORY BOARD H. D. M o m F. L. PATIXRSON G. M. VOLK

C. 0. GARDNER C. L. HAMILTON E. G. HEYNE

1964

ACADEMIC PRESS

New York and London

COPYRIGHT @ 1964, BY ACADEMIC PRESSINC. ALL RIGHTS RESERVED. NO PART OF THIS BOOK MAY BE REPRODUCED IN ANY FORM, BY PHOTOSTAT, MICROFILM, OR ANY OTHER MEANS, WITHOUT WRITTEN PERMISSION FROM THE PUBLISHERS.

ACADEMIC PRESS INC. 111 Fifth Avenue, New York, New York 10003

United Kingdom Edition published by ACADEMIC PRESS INC. (LONDON) LTD. Berkeley Square House, London W.l

LIBRARY OF CONGRESS CATALOG CAR^ NUMBER: 50-5598

PRINTED IN THE UNITED STATES OF AMERICA

CONTRIBUTORS TO VOLUME 16 Numbers in parentheses indicate the pages on which the authors’ contributions begin.

ALLARD,R. W. (281), Professor of Agronomy, Department of Agronomy, University of California, Davis, California

ALLISON,LOWELL E. (139), Research Soil Scientist, United States Salinity Ldoratory, Agricultural Research Seruice, United States Department of Agriculture, Riverside, California BULA,R. J. ( l ) ,Research Agronomist, United States Department of Agriculture and Agronomy Department, Purdue University, Lafayette, Indiana DAVISON, KENNETHL. (197), Research Specialist in Plant and Animal Nutrition, Departments of Agronomy and Animal Husband y,Cornell University, lthaca, N e w York FARMER, V. C. (327), Senior Research Oficer, Department of Spectrochemistry, Macaulay Institute for Soil Research, Aberdeen, Scotlund HANSCHE, P. E. (281), Assistant Pomologist, Department of Pomology, University of California, Davis, California HOFFMAN, W. M . (59), Chemist, United States Fertilizer Laboratoy , Soil and Water Conservation Research Division, United States Department of Agriculture, Beltsville, Maryland LETEY, J . (249), Assistant Professor of Soil Physics, Department of Soils and Plant Nutrition, University of California Citrus Research Center and Agricultural Experiment Station, Riverside, California MCCLOUD, D. E. ( l ) ,Research Leader, Humid Pasture and Range Investigations, Crops Research Division, Agricultural Research Service, United States Department of Agriculture, Beltsville, Ma yland MCHARDY, W. J. (3271, Research Officer, Department of Pedology, Mamulay Institute for Soil Research, Aberhen, Scotland

MITCHELL, B. D. (327))Senior Research Officer, Department of Pedology, Macaulay Institute for Soil Research, Aberdeen, Scotland V

vi

CONTRIBUTORS

ROSENBERG,NORMAN J. (181), Assistant Professor, Department of H d i culture and Forestry, The University of Nebraska, Lincoln, Nebraska SHAW,R. H. ( l ) ,Professor of Agricultural Climatology, Department of Agronomy, Iowa State University, Ames, Iowa STOLZY,L. H . (249), Associate Soil Physicist, Department of Soils and Plant Nutrition, University of California Citrus Research Center and Agricultural Experiment Station, Riuerside, Calif omia STRINGFIELD,G. H . (101), Seniw Research Agronomist, DeKaZb Agricultural Association, Incorporated, DeKalb, Illinois

TERMAN, G. L. (59), Agronomist, Soils and Fertilizer Research Branch, Tennessee Valley Authority, Muscle Shoals, Alabama WRIGHT,B. C. (59),* Associate Professor and Associate Agronomist, Department of Agronomy, Mississippi State University, State College, Mississippi

WRIGHT, MADISON J. ( 197), Associate Professor of Agronomy, Department of Agronomy, Cornell Uniuersity, Ithaca, N e w York

* Present address: Associate Soil Scientist, The Rockefeller Foundation, Chanakyapuri, New Delhi, India

PREFACE The nine chapters in this volume illustrate well the diversity of research activities in soil and crop science that contribute to advances in the broad field of agronomy. In the preface to earlier volumes the Editor has defended the position that it is appropriate to include in this series any topics, basic or applied, scientific or technological, that relate to the soil, its productive use, and to the characteristics and improvement of crop plants. Indeed the primary test applied to topics considered for inclusion is that the information be helpful or useful to agronomists. The reader will find in this volume a scholarly review of the nature of amorphous inorganic soil components by B. D. Mitchell and colleagues from the Macaulay Institute, and a discussion by Stolzy and Letey of the value of the platinum electrode for characterizing soil oxygen conditions or following oxygen diffusion rates. More applied soil problems are represented by L. E. Allison’s analysis of salinity in relation to irrigation, a matter not to be overlooked in many areas of the world where ambitious land development projects are being formulated. Crop responses to the physical effects of soil compaction or compression by vehicular tr&c are considered by Rosenberg; Terman and colleagues review the procedures that may be used to evaluate the crop availability of phosphorus fertilizers, particularly those with new or unconventional chemical structures. The range of topics from basic to applied is equally great in the field of crop science. D. E. McCloud and colleagues review in depth the developing techniques of field physiology, which strives to relate crop growth to the components of the environment and thereby to identify the factors limiting yield. In a different vein Allard and Hansche explore the relative importance of population variability in plant breeding and crop improvement. Their paper is unique in one respect because it reports the result of computer simulations of highly complex genetic sequences, and is the first example in these pages of the application of the computer to an agronomic problem. From his rich experience as a corn breeder G . H. Stringfield provides a critical reappraisal of the objectives in corn improvement in the mid-West, and points out the need for continual alertness to changing circumstances. The threat posed by nitrate accumulation in crops to animals consuming them is authoritatively surveyed

viii

PREFACE

by Wright and Davison, a timely review because of heavier use of nitrogenous fertilizers to increase forage yields. Through their labors in preparing these critical surveys the contributors to these volumes perform a valuable service to their colleagues and their profession. A. G. NORMAN Ann Arbor, Michigan August, 1964

CONTENTS

CONTRIBUTORS TO VOLUME16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Page v

PREFACE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

vii

FIELD PLANT PHYSIOLOGY BY D. E. MCCLOUD, R. J. BULA,AND R. H. SHAW I. 11. 111. IV.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Characterization of the Environment . . . . . , . . . . . . . . . . . . . . . . . . . . . . . Environmental Elements and Plant Growth . . . , . . . . .. . . . . . . . . . . . . . Controlled Environment Facilities as a Supplement to Field Research . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 2 15 47 54

CROP RESPONSE TO FERTILIZERS IN RELATION TO CONTENT OF "AVAILABLE" PHOSPHORUS BY G. L. TERMAN, W. M. HOFFMAN,AND B. C. WRIGHT 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Status of Chemical Methods in the United States and Other Countries 111. Chemical and Physical Nature of Fertilizers Marketed in the United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV . Crop Response Results Prior to 1950 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Recent Crop Response Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Problems Concerned with Nonorthophosphates and Other Fertilizers . . VII. In Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

59 60

66 73 77 93 96 98

OBJECTIVES IN CORN IMPROVEMENT BY G. H. STRINGFIELD I. Introduction

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

.

11. Hybrid Corn and Yield . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. The Offense and the Defense . . . . . . . , . . . . . . . . . . . . . . . . . . . . . , . . .

.

.

IV. Culture and Improvement . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . v. Breeding for Industrial Uses and Nutritive Value . . . . . . . . . . . . . . . .. . ix

102 103 108 114 119

X

CONTENTS

.

VI VII . VIII. IX . X.

Parent Stocks ................................................ Exotic Germ Plasm ............................................ The Cytoplasm ............................................... Tetraploid Corn .............................................. Summary and Conclusions ...................................... References ...................................................

122 132 133 133 134 136

SALINITY IN RELATION TO IRRIGATION

BY LOWELLE . ALLISON I . Introduction ................................................. Salinity of Irrigation Waters .................................... 111 Effect of Salts on Soils ......................................... IV . Effect of Salts on Crops ........................................ V. Reclamation of Salt-Affected Lands .............................. VI . Management Practices for Salt-Affected Land ..................... VII. Conclusions .................................................. References ...................................................

. .

I1

139 140 146 156 164 171 177 178

RESPONSE OF PLANTS TO THE PHYSICAL EFFECTS OF SOIL COMPACTION

BY NORMAN J. ROSENBERG I. I1 I11. IV. V. VI .

.

Causes of Soil Compaction ..................................... Compaction Effects on Soil Productivity .......................... Plant Response to Soil Compaction ............................. Experimental Difficulties ....................................... A Mechanistic Study of Compaction Effects on Plant Growth ......... Outlook ..................................................... References ..................................................

181 182 185 191 192 194 195

NITRATE ACCUMULATION IN CROPS AND NITRATE POISONING IN ANIMALS

BY MADISON J . WRJCHT AND KENNETH L . DAVISON I . Introduction .................................................. Recognition of Nitrate as a Toxic Agent .......................... 111. Accumulation of Nitrate by Plants ............................... IV Postharvest Losses ............................................ V Toxicity of Nitrate to Animals .................................. VI . Conclusions .................................................. References ....................................................

. . .

I1

197 198 201 220 221 240 241

xi

CON?-ENTS

CHARACTERIZING SOIL OXYGEN CONDITIONS WITH A PLATINUM MICROELECTRODE BY L . H . STOLZY AND J . LETEY I. I1. I11 IV

. . V. VI .

Introduction ................................................. Polarography ................................................ Problems Associated with the Use of Platinum Microelectrodes in Soils Relationships between Oxygen Diffusion Rates and Biological Responses Results of Field Measurements ................................. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References ..................................................

249 250

253 258 272 275 277

SOME PARAMETERS OF POPULATION VARIABILITY AND THEIR IMPLICATIONS IN PLANT BREEDING BY R . w . ALLARD AND P . E . HANSCHE

I. I1. I11. IV. V.

Introduction ................................................. The Genetics of Predominantly Self-pollinated Populations . . . . . . . . . . The Exploitation of Exotic Variability . . . . . . . . . . . . . . . . . . . . . . . . . . . Variability within Agricultural Varieties .......................... Summary ................................................... References ..................................................

281 282 302 313 323 324

AMORPHOUS INORGANIC MATERIALS IN SOILS BY B. D . MITCHELL.v . c. FARMER.

I. 11. I11. IV . V. VI .

AND

w. J . MCHARDY

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327 Nature and Occurrence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328 Methods of Detection and Estimation ........................... 338 Origin of Amorphous Material in Soil ........................... 364 Relationships between Amorphous Inorganic Material and Specific Physical and Chemical Properties ................................. 372 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 References .................................................. 375

AUTHORINDEX ......................................................

385

SUBJECTINDEX .....................................................

400

INDEX OF CONTRIBUTORS, VOLUMES 1-15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . KEYWORDTNDEX,VOLUMES1-15 . . , .. , , . , ..............................

405

411

This Page Intentionally Left Blank

FIELD PLANT PHYSIOLOGY D. E. McCloud, R. J. Bula, and R. H. Shaw United States Deportment of Agriculture and Iowa State University, Beltsville, Maryland, Lafayette, Indiana, and Ames, Iowa

I.

Introduction

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

...........

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

B. Phenological Observations C. Agricultural Seasons

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

Page 1

7 15

B. Temperature

B. Control of Plant Environment Factors in the Growth Room . . . . C. Program-Controlled Environmental Conditions References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

48 54

1. Introduction

In Physiological Researches, almost half a century ago, McLean (1917) wrote, “The dependence of plants upon climate conditions is almost selfevident, but the quantitative aspect of the relation between plant activities and climate presents an exceedingly complex problem, the solution of which can not be expected for a very long time.” In the ensuing years many investigators have attacked this problem by attempting to measure plant production or crop yield in terms of the climatic conditions observed during the growth period. This type of research has usually resolved itself into attempts to correlate plant growth with one or more climatic factors, such as temperature or rainfall. However, not only do environmental conditions around the plant change, but the plant responds differently to the same external condition at different times in the life cycle. Thus, expression of plant growth in terms of the environmental factors cannot be accomplished except by an exceedingly complex formula. Each advance in plant physiological knowledge seems only to add to the complexity. 1

2

D. E. MCCLOUD, R. J. BULA, AND R. H. SHAW

During the last two or three decades another approach has been taken, that of attempting to delineate the relationship of a particular environmental factor to the growth and development of the plant. This approach is not without pitfalls since the interactions of environmental factors often is quite important in field studies. However, some progress is evident from the formulation of biophysical models which have furthered comprehension of the relation of the plant to environmental factors. The objective of this paper is to review briefly some of the more recent findings from field plant physiological studies which may be helpful in clarifying the relation of the plant to its environment. II. Characterization of the Environment

A. THEAGRICULTURALWEATHER STATION

1. Requirements for a Station Although the purpose of an agricultural weather station is to define the environment of the biological object, opinions differ widely on the observations necessary to define that environment. To cite only one comparison, Brooks and Kelly (1951) list a large number of observations to be recorded hourly, whereas Newman et al. (1959) list fewer observations, only maximum and minimum, day or night, or total daily values, being recorded. Both these types of stations have a proper place in research activities. Measuring the environment to explain a physical process may require instrumentation more precise than that required to measure the same factor when relating it to a biological process. Since micrometeorological stations are generally established to measure or evaluate specific physical processes, standardization of these stations is difficult, and such stations will not be discussed here. An attempt will be made to describe the objectives of an agricultural weather station, with one suggested station arrangement presented. This will be a “minimum” station to which additional, more extensive, and specialized equipment could be added, and, for many problems, would need to be added. Gol’tsberg (1963) has stated: ‘What we must have are special climatic descriptions indicating the natural climatic potential of the territory from the view of agricultural production, as well as all the adverse climatic factors, together with specification of the necessary measures to counteract them.” The investigator must know what factors are needed to evaluate the particular biological process being examined, then take such measurements as will define the weather factors. Too often the

FIELD PLANT PHYSIOLOGY

3

approach has been to collect all data possible with the hope that some will prove useful. An accurate description of the environment is difficult to obtain. Some factors are d a c u l t to measure accurately, and analysis of the data may be complex. It is essential that the researcher examine his problem carefully to decide which factors are necessary. What factor or factors in the physical environment exert major controls on the life cycle of the organism? Are there critical limits to these factors? What are the optimal conditions? How do these vary in the different stages of the life cycle of the organism? With these questions in mind, a more intelligent decision can be made on what measurements are needed. A further fact of considerable importance is that the environmental effects are frequently cumulative. The sampling interval used and method of accumulating the data may be very important. In many problems, one biological measurement, such as yield, is related to the many environmental observations being recorded. In such a problem, simplicity of expression of the environmental data is of paramount importance. Gol'tsberg (1963) has stated that the formulation of new indexes between environmental factors and crops has been retarded for two reasons. There has been confusion as to whether the agrometeorologist or agrobiologist should perform the task, and there have been inadequate existing methods and empirical data available for a concise solution to this complex problem. The latter is extremely important. At present, our techniques for relating extensive data over a period of time to a biological response are frequently a limiting factor in application of these data. Many of the environmental-plant relationships are not quantitatively understood, and until this is accomplished only a Iimited solution to the problem can be obtained. Integration of the whole problem is difficult until the component parts are understood. Only the intensive study of specific, limited-scope problems can provide many of the answers required. The weather station, to be briefly described, will provide data to solve many biological problems and provide data for application to local areas, but it does not provide the detailed data necessary for solution of many special problems, particularly those that require a detailed evaluation of a physical process in the atmosphere. The station to be described would require a minimum of time for observation, limited skill for maintenance and operation, yet would provide valuable agricultural data more extensive than that now available. Such station would be the base station to which additional special observations could be added. Only the minimum basic observations can be standardized, as pointed out by Ventskevich ( 1958) and each agrometeorological station

4

D. E. MCCLOUD, R. J. BULA, AND R. H. SHAW

should be specialized to fit the agricultural character of its locale or the particular problems being studied. According to Newman et al. (1959) the essential requirements for such a station are: 1. That the site be representative of general crop-climate conditions. 2. That observations be of moderate accuracy but highly reliable. 3. That instrumentation be simple and rugged in design. 4. That observations be easy to obtain and record. 5. That data be expressed in a simple gross manner on a daily, weekly, monthly, or seasonal basis. 6. That routine observations be held to a daily, weekly, monthly, or seasonal basis. 7. That any special attention necessary for instruments be on a daily, weekly, monthly, or seasonal basis. The selection of the time interval and the sampling techniques used are as important as precision of the measurement in most physical environmental-biological response relationships,

2. Selection of Site and Location of Instruments Since the main objective of an agricultural weather station is to study the influence of various measurable environmental parameters on the biota, the site must be representative of the biosphere. The station should be in a level area, or near a crest in sloping or hilly areas. It should never be in small depressional areas unless that location represents the particular problem being studied. A grass sod is generally considered best for the location of such a station, but in arid regions maintenance of a sod cover will introduce unnatural climatic conditions. (For tall vegetation it may be necessary to locate the station within the crop cover.) Instruments should be isolated from obstructions for a minimum horizontal distance of approximately 2 to 3 times the height of these obstructions, and for major obstructions the distance should be even greater (Schaal and Newman, 1958). Examples of good instrument and equipment exposures can be found in the U. S. Weather Bureau (1955) Instructions for Climatological Observers. Recommendations on instruments for certain uses can be found in the World Meteorological Publication (1961), “Guide to Meteorological Instrument and Observing Practices.” 3. Instruments T o Be Used Maximum and minimum thermometers can be used to obtain daily temperature data. Air temperature thermometers should be exposed in standard weather shelters, painted white, mounted over grass sod when-

FIELD PLANT PHYSIOLOGY

5

ever possible. Sensing elements should be located at approximately 150 cm. (60 inches) above the soil surface. The shelter door should always open to the north. Although artificial ventilation of the shelter is recommended (World Meteorological Organization, 1961 ), at the present time this is seldom done. If duration of temperatures is desired, a recording thermograph can be installed. Another observation that may be desired is the minimum temperature near the surface. To obtain this a minimum thermometer is exposed just above the vegetation surface. This is called the “grass minimum thermometer” or terrestrial radiation observation. This observation gives information about ground frost at night and records approximately the minimum vegetation temperature. Daily measurements of atmospheric moisture are extremely important, but in many respects are difficult to obtain. A psychrometer can be used for wet-bulb and dry-bulb readings from which relative humidity or the dew point can be calculated. Unfortunately this nonrecording instrument makes observations of maximum and minimum values quite difficult. Recording hair hygrographs can be used but require frequent calibration. DaiIy maximum-minimum dew-point observations can be made using a lithium chloride dewcel addition to a dial-indicator type of maximumminimum thermometer, A description of such an instrument is given by Tanner and Suomi (1956, 1958), but this instrument is not available commercially. Precipitation is measured using a standard type rain and snow gauge. Intensity and duration can be observed using a standard recording gauge. These “standard” gauges differ in different countries, but they should have a receiving area of 200 to 500 cm.2. Instructions for using these are given by the U.S.Weather Bureau (1955) and by the World Meteorological Organization ( 1961). Soil temperature measurements are preferably recorded as maximum and minimum values, although indicating thermometers can be used. Depths recommended by the World Meteorological Organization ( 1981) are 5,10, 20,50, and 100 cm. Soil temperatures are generally taken under sod or bare ground. A sensing element which samples over some horizontal distance (as a small tube, for example) is preferred to a point sample as it helps overcome area variations in soil temperature. Heat flow and energy balance problems require a greater accuracy in measuring soil temperature and special instruments, such as a space integrating thermometer (Suomi, 1957) or a flat-plate heat flux meter (Gurevich, 1958) should be used. Deacon (1950) has described equipment for recording heat flux in the soil. The measurement of wind into two time periods, daylight and dark-

6

D. E. M C CLOUD, R. J. BULA, AND R. H. SHAW

ness, has been proposed (Newman et al., 1959). Wind measurements are recommended at a standard 10 meter height (World Meteorological Organization, 1961). This is a high level for most agricultural uses, and a height of 1meter has been proposed in the United States (Newman et al., 1959). For wind measurements in conjunction with an evaporation pan, a height of 6 inches above the rim of the pan is recommended by the U. S. Weather Bureau (1955). Most commercial anemometers are satisfactory and could be adapted to record day and night total wind movement, through the use of time clock-operated counters. If wind profiles are to be measured, the installation of commercially available systems designed specifically for this purpose is recommended. Radiation is important in most field physiology problems and can be measured with respect to duration, intensity, and quality. In the agricultural station it should be measured as to quantity-both shortwave and net radiation. Net radiation is important since it provides a measure of the radiation balance. Present commercial instruments are expensive and generally operate on a continuous recorder, More economical types nonrecording (Fritschen, 1963; Suomi and Kuhn, 1958) are available, but if converted to automatic recording, the recording device may be the major cost. Solar radiation can be recorded with an Eppley pyrheliometer, or in other countries with instruments recommended by the Weather Service. An integrating device which gives hourly, or daily, totals facilitates data processing. All radiation instruments should be placed to alleviate any shading effects. Dew supplies some water and provides a wet surface for pathogens (Stone, 1957; Lloyd, 1961). No known dew intensity recorders are available commercially, but experimental recorders have been described by Hirst (1954) and Jennings and Monteith (1954). The Duodevani dew gauge (Newman et aE., 1959), a specially treated wood surface, can be used to obtain an estimate of dew intensity. The World Meterological Organization, Commission on Agricultural Meteorology ( 1958) has a working group on this problem. A duration recorder for dew is commercially available (Newman et al., 1959). It operates with a lamb gut strip which, when wet, allows a pen to mark on a chart. When not wet, no mark is made. Another type, using an indelible pencil to mark on a ground glass plate, has been used experimentally (Taylor, 1956; Thesis and Calpouzos, 1957). Both types of instruments were tested in the field by Shaw (1954), who reported relatively good results compared to visual observations of dew duration. Some measure of the evaporating potential of the atmosphere is desirable. Several types of instruments have been used. In the United

FIELD PLANT PHYSIOLOGY

7

States, the U. S. Weather Bureau (1955) recommends a pan 4 feet in diameter. This pan was adopted as the reference instrument by World Meteorological Organization ( 1961) for the International Geophysical year measurements. Several types of atmometers ( Livingston, 1908, 1915, 1935) have been used. The Bellani plate atmometer has been widely used in Canada (Robertson, 1955; and Robertson and Holmes, 1956, 1958). Rider (1958) has discussed several types of these instruments in a World Meteorological Organization Technical Note. Each of the many available types has certain advantages or disadvantages, but all are designed to measure potential evaporation, where water is not limiting. Soil moisture data are an essential part of agricultural weather station observations. Very few stations in the United States now systematically record such data, but numerous soil moisture stations have been established in the Soviet Union, Soil moisture observations generally are taken under sod in the instrument area, but can also be taken in nearby areas under crops representative of the area. Data should be recorded by 6or 12-inch increments, but no generally accepted standards have been set for these observations. No attempt will be made here to discuss the different methods that can be used in soil moisture sampling. An example of good arrangement for a proposed agricultural weather station is shown in Fig. 1. Not all the observations previously discussed are included in this station, and any station should be large enough so that additional equipment can be added. The important consideration to be used in locating different instruments is that they do not interfere with each other. Those that offer the most obstruction to wind movement should be located downwind from the prevailing wind direction for the area. Stations somewhat similar to those shown in Fig. 1 are commonplace in the Soviet Union (Gurevich, 1958). In the Ukraine alone there are over 2000 agricultural weather stations at state and collective farms, and in the Kazakhstan over 4000 are planned. In addition, at the scientific and experimental agricultural institutions, more extensive observations are taken. It should be stressed that a network of agricultural weather stations, plus additional meteorological instrumentation as required for special problems, would provide invaluable weather data for field physiological studies, and this aspect is too often overlooked.

B. PHENOLOCICAL OBSERVATTONS Although phenological observations are generally not taken within the physical confines of the agricultural weather station, these observations should be included as part of the agrometeorological observations.

8

D. E. MC CLOUD,R. J. BULA, AND R. H. SHAW

oll weather

-4

m o l l gate wwd wolk ways, never large oreos of grovel, stone, or cement.

0

10 foot gote

instrument shelter door to North (cotton belt type)

electric service poles ond pmer line to the North

c

60' North

w

6' x 8'

pa

maintenance shelter with electric power, heot , water

mox- min oir thermometers rnox- rnin dew point thermometers max-min soil thermometers slondard 8 inch rohgouge grass minimum c-

@

I 9 0'

recording 8 inch roin gouge

hp water supply

wind observations z

5?

0

0 0

evoporotion observations

15' x 15' mointained under grass for soil moisture observations dew observations

T

1

10' x 10' mointoined I bore soil

I 1

Area shculd be fenced with o wire fence of a two inch mesh and at leost four feet high.

FIG. 1. A plan for placement of instruments in an agrometeorological weather station. Newman et al. (1959).

FIELD PLANT PHYSIOLOGY

9

Webster’s Third New International Dictionary defines phenology as “a branch of science concerned with the relations between climate and periodic biological phenomena.” It involves a study of the relationship between physical factors in the environment and seasonal changes in growth and development during the life cycle of plants and animals. These phenological observations can be related to crop plants for a better understanding of the relationship between various meteorological variables of the environment and the associated biological responses. In many cases, these observations provide essential information for evaluating the effect on the crop of extreme weather conditions of short duration. Indicator species can be used as seasonal integrators of the weather. Newman and Beard (1962) have suggested four methods of approach in utilization of phenological observations. First is morphological or physical changes in the structure of the plant or animal, such as flowering or change of hair coat. This method is particularly applicable to plants. Second is changes of mass or rate of growth of an organism under observation. Third is the changing of activities, such as bird migration. Fourth is chemical analysis, either qualitative or quantitative, related to some biological change within the organism. All these kinds of biological events recorded on a periodic basis are phenological observations. The observation of phenological events dates to biblical times. Such events were also reported by ancient Chinese civilizations. More modern studies may be dated from Reaumur (1735), who developed a thermal constant concept in postulating that all plants have a certain minimal air temperature necessary for growth. In the United States early work was done by Merriam (1898) on the biotic distributions of North America. Probably the best known work done in the United States was that of Hopkins (1938), who studied the periodic responses of plants and animals to climatic factors and formulated a bioclimatic law to describe the relationship, Schnelle ( 1955) presented isophene contour maps of equal vegetative development for central European areas. An extensive review of the literature on phenology is presented by Wang and Barger (1962), in which over 400 references are given. Most European countries, and many countries on other continents, maintain phenological departments as part of their government meteorological services (Schnelle and Volkert, 1957). Most of the observations are now taken on native plants. In the USSR these observations are taken by the Hydrometeorological Service at the meteorological stations and by the Commission for Plant Breeding at agricultural experimental stations (Ponomarev, 1958). Their observations on crop plants include dates of basic stages of plant development, such as emergence from soil,

10

D. E. MCCLOUD, R. J. BULA, AND R. H. SHAW

first leaves, secondary shoots, stems, flower buds, flowering, opening of seeds. In addition the condition of crops is estimated periodically. After the occurrence of unfavorable meteorological phenomena (frost, drought, dry or strong wind, heavy precipitation, hail), the condition of field crops is examined and damage estimated (Gurevich, 1958). Many phenological observations are taken in the United States, but these are not collected in such a manner that they can be compiled. No formal program of phenological observation has been maintained in the United States, although Abbe (1905) reported some very early work done by the U. S. Weather Bureau, then a part of the U. S. Department of Agriculture. Recently two regional research projects in the United States include some phenological observations. Also, data collected by the Statistical Reporting Service of the U. S. Department of Agriculture provide some general phenological information, but the program in the United States lacks the detail of the formal phenological programs of European countries. In taking phenological observations, as in agrometeorological observations, certain questions need to be raised regarding the observations. Newman and Beard (1962) proposed five questions:

1. Can the observation be expressed quantitatively both with respect to time and state of the organism? 2. How often with respect to time and state of organism change is it necessary to repeat the observation? 3. What are the possible causal physical factors within the environment? 4. How should each of these factors be measured with respect to time and space? 5. What skills are necessary on the part of the observer? Of particular importance is the first question. The observation must be one which can be expressed quantitatively, and in many cases must be an easily discernible morphological change. However, this should not exclude more detailed observations for special programs such as biochemical or internal observations, but the readily visible observations should become an integrated part of the agrometeorological station. In addition to taking observations on crop plants on experimental plots, phenological observations on selected noncrop plants can serve as seasonal indicators. These observations measure the variation of biological events between seasons as well as within seasons. To be a desirable seasonal indicator species, a plant must meet certain essential requirements. These have been listed by Newman and Beard (1962) as: (1) that the species be easy to identify; ( 2 ) that events.be

FIELD PLANT PHYSIOLOGY

11

easy to observe; ( 3 ) that events be rather uniformly spaced throughout the season; ( 4 ) that the occurrence of the events be widespread geographically; (5) that the species be stable ecotypically, with a minimum of genetic variation. By recording both weather factors and phenological observations, a more complete description of agricultural weather patterns can be obtained which will provide information in field physiological studies.

SEASONS AND PLANT-CLIMATE ZONES C. AGRICULTURAL In addition to characterizing the weather by meteorological observations, or by phenological observations, the annual timetable of climatic change may be classified into natural agricultural seasons. The four astronomic seasons provide a means of doing this, but in themselves provide little information on the agricultural climate. Many climatologists have recognized the inadequacy of this system, particularly the seasonal classification or the middle latitude regions. In Italy, Azzi (1914) attempted to improve seasonal time scales through the use of phenological events. Angots (1914), in France, used the number of days with a given mean temperature level to define the severity of weather for certain seasons. In the United States, Alciatore (1915))classified summer conditions by the use of daily mean maximum and minimum temperatures. Although not strictly a means of defining a plant-climate zone, the heat unit approach (Went, 1957; Boughner and Kendall, 1959; Wang, 1960) can be useful in field physiology studies. Extensive references in this area are listed by Wang and Barger (1962). In this approach, temperatures ( generally daily mean temperatures) above a selected threshold are accumulated during the growing season. If a plant has a base or threshold temperature of 50"F., and the mean temperature for the day is 70"F., 2Q degree-days or heat units have accumulated for that day. It has often been assumed that a certain sum of heat units are necessary for a particular crop variety to reach a particular stage of development. Wang (1960) has pointed out that plants respond differently to the same environment at different stages in their life cycle. As a result of this, threshold temperatures may also change. Responses may not always be linear with temperature change. Medcalf (Personal communication. This material is in the Hawaiian Pineapple Planters' Record, 1952.) has avoided this problem by devising temperature-weighted growth rates. This approach involves using thermograph records and counting the number of hours specified temperature ranges occurred. This number of hours is multiplied by the growth rate appropriate for

12

D. E. MC CLOUD,

R. J. BULA, AND R. H. SHAW

that particular temperature and stage of development. By adding these temperature-weighted growth rates excellent results were obtained. Within recent years, Newman (1956),Newman and Wang (1959), and Baker and Strub (1983) have described methods of classifying the annual timetable of climatic change into natural agricultural seasons. In the latter two studies, maximum and minimum temperatures were used. The advantage of this approach is the dense network of stations from which data can be used. In many areas of the world this network provides a sufficient density of stations for detailed climatic study. In studying climatic variations one should remember that minimum temperatures reflect strongly the local conditions, or the microclimate, while maximum temperatures reflect the macroclimate of the region. In using climatic data in crop physiology studies lethal as well as optimal temperature limits must be considered for many crops. These limits, and temperature growth responses are then used to subdivide the calendar year into seasons that represent gross crop responses. These subdivisions are called agricultural seasons. They give the average date on which a season begins, the length of the season, and extremes or optima of interest can be expressed in terms of the probability of occurrence. These statistics should prove most useful in considering the adaptation of new plant species. The criteria used by Baker and Strub ( 1964) to characterize agricultural seasons were : "1. Early spring begins when 20% or less of the minimum temperatures are 16°F. or lower. In early spring cool season perennial crops, such as bluegrass, begin to grow, and cool season annuals, such as spring oats, are planted. 2. Late spring begins when less than 20% of the minimum temperatures are 32°F. or lower. In late spring warm season crops, such as dent corn and soybeans, are planted, and cool season crops grow rapidly. 3. Summer begins when less than 10% of the minimum temperatures are 40°F. or lower. In summer warm season crops, such as soybeans, grow rapidly, and cool season annuals, such as small grains, are harvested. [Newman and Wang included the condition that 20% or less of the daily maximums are as low as 70"F.I 4. Early fall begins when more than 20% of the minimum temperatures are 40°F. or lower. In early fall cool season crops, such as winter grains, are planted and warm season crops, such as dent corn, mature rapidly. [Newman and Wang included the condition that more than 20% of the maximums are 70°F. or below.] 5. Late fall begins when more than 10% of the minimum are 32°F. or lower. In late fall cool season crops, such as winter grains, grow

FIELD PLANT PHYSIOLOGY

13

rapidly, and warm season annuals, such as dent corn and soybeans, are harvested. 6. Winter begins when more than 20% of the minimum temperatures are 16°F. or lower. In winter, crop plants are dormant.” The arrival of a season, or its duration, can be influenced by largescale factors, such as air mass movement, local topography, and the kind of soil surface. Air drainage, poor heat conductivity, altitude, marine

FIG. 2. Average duration in days of summer-less than 10 per cent of the minimum temperatures are 40°F. or lower; small grains are harvested and warm season crops such as dent corn and soybeans grow rapidly. Baker and Strub (1963).

effects also affect these seasons. All these factors can be integrated into the agricultural seasons. The pattern of duration of agricultural summer days in Minnesota is shown in Fig. 2. The range from 10 to 90 days is striking and points out the usefulness of this system in defining the growing season for an area. As stated by Newman and Wang: “If this same definition was applied to the southern parts of the Gulf Coastal states, winter would cease to exist, since cool season crops grow all through the so-called winter dormant season there. The same can be said of the summer at northern

14

D. E. M C CLOUD, R. J. BULA, AND R. H. SHAW

latitudes. When temperature regimes are cool and short enough to eliminate the use of warm season crops from agricultural cropping systems, summer as an agricultural season does not exist. In the continental middle latitude regions, characterized by a cold winter season and a hot summer season, the seasonal change concepts described in this paper exist rather consistently.” Went (1957) devised another system using maximum and minimum temperatures as the basic data, but dividing them into effective day vs. night regimes. Effective nyctotemperature (night) as: mean minimum

+ 0.25 (mean maximum - mean minimum)

Effective phototemperature (day) as: mean maximum - 0.25 (mean maximum - mean minimum)

These effective day and night temperatures can be used to characterize the temperature regime at which specific plants grow best (Fig. 3). Then for any region when the climatic ellipse passes through the optimal growing conditions for a plant, that is the time when that plant grows best in that particular climate. Similar type relations might well be established for different field crops.

0

60

70

80

Effective Day Temperatures, O F

FIG. 3. Effective day-night temperature conditions for optimal growth and flower production. Kimball and Brooks ( 1059).

FIELD PLANT PHYSIOLOGY

15

Kimball and Brooks (1959) have used these definitions to develop a detailed map delineating the plant climate zones of California. They state that knowledge of effective day and night temperatures may be a valuable tool in determining the most favorable environment for plants and animals. The same knowledge may also be used to relate controlled laboratory studies to field conditions. Went (1957) has also used a multidimensional representation of climate, for example, phototemperature, nyctotemperature, and photoperiod. Climographs of this type, although somewhat complex to comprehend, are useful in explaining the interrelationships between a number of factors and a specified crop response. Nuttonson (1947) has used the climatic analog aproach for defining comparable zones of climate. He subsequently published a series of comparisons for various areas with areas in North America. Elements of comparison are mean monthly and yearly temperatures, maximum and minimum temperatures, average monthly, seasonal, and yearly precipitation, precipitation effectivity indexes and ratios, length of frostless periods, and latitudes. After analysis of these factors, the environment in North America found to resemble most closely that of another area was recorded as climatically analogous to it. Although it would be desirable to compare distributions of weather elements in addition to mean values, and to use shorter intervals, his climatic analogs provide a means of comparing climates in widely different geographic areas. As yet, there are few, if any, generally accepted techniques for making detailed plant-climate zones. However, such comparisons are extremely useful, particularly when considering the introduction of new crops into an area. 111. Environmental Elements and Plant Growth

A. RADIANT ENERGY 1. Characterization of Radiant Energy Radiant energy is that form of energy which is impelled through space as electromagnetic waves. The electromagnetic spectrum is a sequential arrangement of radiant energy according to wavelength and frequency extending from the long wave, low energy photons of the radio region to the extremely high energy particles of the short wave cosmic rays; this is presented graphically in Fig. 4.Of this broad radiant energy spectrum, solar radiation spans only a minute section, yet it is this section which is of prime importance in plant growth and development. The various spectral regions have been somewhat arbitrarily delineated by certain of their most evident properties. The visible spectrum

16

D. E. MC CLOUD, R. J. BULA, AND R. H. SHAW

extends from about 400 millimicrons (mp) in the violet to 700 mp in the red and is determined by the limits of spectral sensitivity of the average human eye. The shorter wavelengths down to about 300 mp encompasses the ultraviolet of the solar spectrum. On the long wavelength side of the visible is the infrared, which is a very broad band. For the solar spectrum most of the energy is contained in the region below 15OOmp. COSMIC PAYS FAMMA RAY? X RAYS I---------I

SOLAR SPECTRUM

RADAR, TELEVISION, RADIO I

I

l IO-IP

l

l 10-10

l

I l l l l l

l 10-8

10-6

lo-*

I

10-p

l

l

100

l I O ~

l

l 104

l

l

l

106

WAVE LENGTH, .cm

FIG.4. Radiant energy spectrum.

According to Withrow and Withrow (1956), sources of radiant energy in the visible and adjacent spectral regions may be divided arbitrarily into three general classes: the thermal radiator, the electrical discharge or electron-excited source, and the fluorescent lamp. The sun and incandescent lamps are thermal radiators. The electrical discharge sources include the carbon arc with a discontinuous spectrum consisting of lines characteristic of the elements present in the discharge along with a background of thermal radiation from the incandescent electrodes. In the fluorescent lamp, ultraviolet energy is absorbed by a phosphor coating on the inside of the lamp and then reemitted by fluorescence at longer wavelengths. Only about 40 per cent of the energy of solar radiation falls within

FIELD PLANT PHYSIOLOGY

17

the range 400 to 700 mp. Conversely the radiation from an incandescent light source is largely (about 80 per cent) in the infrared above the visible spectrum. Fluorescent lamps on the other hand have the reverse pattern, 75 to 80 per cent of their radiant energy being in the visible with a small fraction in the ultraviolet and the balance in the infrared just beyond the visible. The upper limit of radiance which plant and animal tissues can tolerate is often determined by the heating effect of the absorbed energy (Withrow and Withrow, 1956). Thus, in studies of photosynthesis infrared is an undesirable component. At noon under summer sunlight in tropical and temperate zones, plants and animals are exposed to a maximum irradiance of up to 1.5 calories/cm.2/min. This approach is the maximum flux tolerated by most organisms. An incandescent lamp energy of 2000 foot-candles at 20 lumens per watt will produce the same total radiance and nearly the same heating effect as noon sunlight. However, if water is used to absorb the infrared from an incandescent source, the visible radiation flux can be increased close to that of the sun (Gordon, 1930). Van der Veen and Meijer (1959) propose that for the purpose of determining the spectral composition of light from a plant standpoint, the radiant flux can be divided into a number of wave bands or zones, each having a specific physiological effect on the plant: 1. All radiation of wavelength longer than 800 mp: No specific effect is known to be caused by such radiation. It may be assumed, therefore, that the absorbed portion of this radiation is converted to heat. 2. Radiation between 800 and 700mp: This region encompasses the radiation having specific elongating effects on plants. Included is the far-red effect of the phytochrome system. 3. Radiation between 700 and 610mp: This is the spectral zone of peak chlorophyll absorption and maximum photosynthetic activity. The “night-break and other effects are also most marked in this band. 4. Radiation between 610 and 510mp: This is the spectral zone of minimal photosynthesis and for most plants reduced formative influences. 5. Radiation between 510 and 400mp: In this region absorption by the yellow pigments takes place and a secondary chlorophyll absorption peak occurs. The yellow pigments induce various important reactions, such as phototropism, the streaming of protoplasts, and the movement of chloroplasts. 6. Radiation between 400 and 315 mp: This band produces formative effects; plants become shorter and leaves thicker. 7. Radiation between 315 and 280mp: Radiation in this zone is detrimental to most plants.

18

D. E. MCCLOUD, R. J. BULA, AND R. H. SHAW

8. Radiation of wavelengths shorter than 280 mp: These wavelengths rapidly kill plants.

2. Light Intensity and Photosynthesis a. Growth analysis. Of the major environmental factors that influence the growth and development of higher plants, light has been the most obscure and least studied. Today’s concept of light relations within plant communities is based on research contributions over the last four or five decades. The foundation for this work was the development of methods of growth analysis. Gregory (1917) developed the concept of net assimilation rate, defined as “rate of increase in total plant weight per unit of assimilating material.” Any attribute of the plant which is primarily concerned in carbon assimilation and thus has some claim to be taken as a measure of the internal factor for growth was taken to be assimilating material. It remained for Briggs et a,!. (1920) to formulate appropriate methods of growth analysis by combining net assimilation rate and leaf ratio into a product, the relative growth rate. This was equivalent to the efficiency index developed by Blackman (1920). From these studies the formula for net assimilation rate (NAR) evolved as: lnL2-lnLl W2 - W1 NAR = -- X L I -L , ta - tl Williams (1946) pointed out that this equation applies only provided the total plant weight is linearly related to the total leaf area, that is, AW remains constant. NAR then involves measurement of dry weight changes with time and it involves measurement of the leaf area. Early workers realized that it was not the leaf area per plant which was important, but the leaf area supported over a given ground area. This ratio of leaf area to ground area has been termed leaf area ratio or more recently leaf area index (LAI). Later contributors in the NAR field followed this form of analysis, but bases other than leaf area have been used in determining relative growth rate. Because of the difficulty in accurately measuring leaf area, Crowther ( 1934), Ballard and Petri ( 1936), Williams ( 1936), Heath (1937), and others have substituted leaf weight for leaf area. Williams (1939) indicated that leaf area and leaf weight were satisfactory indicators only during early vegetative stages and suggested leaf protein nitrogen as a better measure of the active growth substance over extended growth periods. Brougham (1960) showed a highly significant correlation between the maximal growth rates of various forage species and the amount of chlorophyll per unit land area. The first intensive field investigation of the LAI-NAR relation was

FIELD PLANT PHYSIOLOGY

19

made by Watson (1946). Work previous to this time had been usually with pot cultures. Watson made numerous observations on the relationship of LA1 to NAR on field crops. Watson and co-workers also developed detailed sampling procedures for determining fresh dry weight, plant and shoot numbers, and leaf area. b. Light interception. Analysis of the light relationships in plant communities is complicated. Under field conditions the intensity of solar illumination is not static but varies erratically, frequently from minute to minute. In addition, the diurnal light cycle as well as the yearly seasonal light rhythm both impart predictable aberrations that are related to latitude. The intensity and wavelength components of light environment are markedly altered by these factors, but within the plant canopy even more important are the factors such as the quantity and quality of plant parts which absorb, reflect, and transmit the incident radiation. Ecologists for many years have been aware of the effects of heavy shade under forest canopies. Boysen-Jensen (1918, 1932) in early work elucidated the importance of the light factor in plant communities and its relation to dry matter production. However, the development of the leaf area index concept fostered increased attention in the light-leaf area relationships, and during the last decade intensive interest has developed in this field. Blackman (1938) showed that variations in light intensity had a direct effect on clover growth and also acted indirectly, affecting grass competition in mixed swards. The foundation for much of the recent work on light and plant growth was deveIoped by Blackman and his collaborators. Mitchell ( 1953) and Black (1955) have delineated the interactions of light with temperature and other environmental and management factors as influences on growth and development of forage plants. Work in this area has been reviewed by Black (1957). Monsi and Saeki (1953) found that herbaceous communities shade the ground as effectively as deep forest canopies. The steep light gradiant in these plant communities resulted in intense competition of plant parts for light. Based on many plant communities the logarithm of relative light intensity at one height in a homogeneous plant community was shown to decrease linearly with increasing leaf area, according to the formula I =e-"L

I0

I is the light intensity beneath the leaf canopy, lo is the light intensity above the crop, L is the leaf area index, and K is the extinction coefficient.

20

D. E. MC CLOUD,R. J. BULA, AM) R. H. SHAW

Davidson and Phillip (1958) used a similar equation derived by analogy with Beer’s law. This equation gives a steep decline in light intensity from the surface of the crop downward within the plant community. In addition to leaf area index many other factors also influence light interception. Transmissibility of leaves is important in determining the degree of light penetration in the plant community. Kasanaga and Monsi (1954) measured transmissibility in some 80 plant species and found that for most species the transmission values ranged between 5 and 10 per cent. From Beer’s law the light intensity beneath one leaf with 10 per cent transmissibility would be 10 per cent of full daylight, beneath t w o layers of leafage the light would be 1 per cent daylight, and beneath three layers of leafage only 0.1 per cent of the daylight intensity. However, although there are few actual measurements of this aspect, the decline in light intensity within the crop community is much less marked. Brougham ( 1958) determined the light interception capacity of perennial ryegrass equivalent to 74 per cent per unit LA1 while for white clover the transmissibility was 50 per cent per unit LAI. The reason for these differences in light transmission through the plant canopy is to be found in leaf arrangement. Kasanaga and Monsi (1954) have derived the theoretical relationship of light intensity, transmissibility, and density of leaves as related to growth rate. At low light intensity a continuous leaf layer is more efficient whereas at high light intensities discontinuous leaf layers give a crop growth rate 40 per cent greater than with the same amount of leafage in an unbroken layer. Warren Wilson (1960) examined the theoretical importance of leaf dispersion using the inclined point quadrat concept. This work suggests that the more uniform the leaf distribution the greater will be the crop growth rate because of reduction in overlap of leaves and less light penetration to the ground surface. The work of Watson and Witts (1959) in the field suggests that leaf arrangement and leaf angle exert a large influence on the net assimilation rate of sugar beets. With a leaf area index of 1, the net assimilation rate of cultivated varieties of sugar beet was about the same as that of three types of wild sugar beet. With closer plant spacings and leaf area indexes greater than 1, the net assimilation rate of the cultivated sugar beet was higher than that of the wild types at equal leaf area indexes. This suggests that the photosynthetic efficiency of these species is similar and that the differences are related to the leaf arrangement. The wild beet had a more prostrate growth habit with more overlapping of leaves. Thus, in the wild beet a much smaller fraction of the total leaf area is exposed to high light intensities. Another leaf character influencing light interception is leaf angle. Monsi and Saeki (1953) calculated the relative light interception by

21

FIELD PLANT PHYSIOLOGY

horizontal leaves as 1.0 contrasted to vertically oriented leaves as 0.44. Warren Wilson (1960) emphasized the theoretical reason for the photosynthetic effectiveness of leaves not displayed perpendicular to the incident light: horizontal leaves will be exposed to a light intensity far above that needed for maximum photosynthesis, while leaves acutely angled to the incident light rays will receive a reduced light intensity distributed over a much greater area. It follows that the optimum leaf to light inclination will increase with increasing light intensity. c. Maximum photosynthesis. The photosynthesis of crop plants as influenced by light, carbon dioxide, temperature, and other factors has been the subject of several recent comprehensive reviews: Gaastra (1959), Talling ( 1961 ), and Evans ( 1963). Several workers have developed formulas for calculating the theoretical maximum photosynthetic rates

20 f=

t

K.07

K4.0

1 O0O

2

6

4

K.2.0

8

10

LA1

FIG.5. Relationship of LA1 and extinction coefficients ( K ) to net photosynthesis. Saeki (1980).

of plant communities. Based on the light photosynthesis relation of isolated leaves, deWit (1959) devised a formula for the calculation of potential photosynthesis of a closed crop surface. This formula is derived from a simplified curve of the light intensity-photosynthesis relationship with linear increase in photosynthesis up to a maximum at light saturation. Leaf arrangement was simplified by assuming that the leaves of the crop surface are arranged randomly with respect to incident light. From the resultant formula a potential photosynthetic rate of 5 g. CH2O/ m.”day in December and 29 g./m.2/day in June €or the average radiation received in the Netherlands. Saeki (1960) gives comparable values up to about 40 g. dry weight/m.2/day (Fig. 5). Loomis and Williams (1963) has estimated maximum levels of crop productivity, by evaluating the photosynthetic energy potential solar radiation on a quantum basis, as 71 g. CH20/m.2/day. This estimated maximum compares with the highest values reported from field studies by Vicenti-Chandler et QZ. (1959) of 26 g. dry matter/ma2/day. Maximum values reported for most crop plants

22

D. E. MC CLOUD,R. J. BULA, AND R. H. SHAW

do not exceed 20 g. dry matter/m.2/day. Alberda (1962) determined maximum rates for sugar beet in the Netherlands of 2Q g./m.2/day. While considerable effort has been devoted to determination of maximum potential yields for plant communities both from the theoretical standpoint and the determination of actual yields in the field, little attention as yet has been directed toward contribution of the various 1 nfoldIng

Optimum leaf area index

0

p

R

1 2 3 4

12

2 2

10 6

3 31

-

N e t assim

-

2 2 -

0 -

10 8 4 1 23 -

-

Ceiling leaf area index

'Y

6 7

0 0

2

1 2 1 3

R

-2

-1

N e t asrim of crop

Ceiling yield

Total net assimilation

a

FIG.6. Idealized plant community showing relationship of increasing leafage to net assimilation of the foliage. ( L ) leaf area index, ( P ) photosynthesis, ( R ) respiration. (From Donald, 1961, p. 280.)

FIELD PLANT PHYSIOLOGY

23

layers of leafage within the plant community. Donald (1961) presented a hypothetical example to portray the development of an idealized leaf canopy through three successive stages (Fig. 6 ) . First he defines the optimum leaf area index as the amount of leafage necessary to maximize net assimilation. At this stage all leaves are making a positive contribution to dry weight increase though the contribution by the lowest leaves may be quite small. Second, as the leafage continues to increase above the optimum, the leaf area index ceiling or maximum leaf area index is reached. At this point, the rate of death of leaves at the base of the canopy as a result of aging or low light intensity equals the rate of appearance of new young leaves. The net assimilation rate by the foliage at the ceiling leaf area index is below that at optimum leaf area index. Third, as the plant canopy continues to develop, a ceiling yield is reached where the nonphotosynthetic tissues have increased until the respiratory losses by the crop equal the photosynthetic gain. At this stage, the dry weight of living material per unit area is static and total net assimilation is zero. If the crop canopy continued to develop, respiration losses would exceed photosynthetic gains and the total yield would diminish below the ceiling yield. Verhagen et al. ( 1963) from mathematical and physical considerations of hypothetical foliages have concluded that a foliage in which the bottom leaves are at the compensation point is not always at the optimum LAI. This happens because of compensatory changes in leaf arrangement while leaf area continues to increase so that the light received by the bottom leaves does not diminish further. Emecz (1962) has postulated a larger optimum leaf area for spring versus autumn based on increasing versus decreasing solar angles. Models for photosynthesis of leaves within plant communities have been based on measured or computed light at the leaf surfaces. Until recently no experimental information has been available on the assimilation of leaves within the plant canopy. The contribution of various layers of leafage in situ has recently been shown experimentally by McCloud ( 1964). Measurements of net photosynthesis by successive layers of leafage at different light intensities shows that photosynthesis is related to light interception leaf area and leaf angle (Fig. 7 ) . At low light intensities, lower layers of leafage contribute negatively to the total dry matter production of the plant community while with increasing light intensities, light penetration raises these lower leaves above the compensation point and maximizes net photosynthesis. Thus, during the natural diurnal fluctuation of light intensity the optimum leaf area changes from zero at night to a maximum at midday. d. Carbon dioxide and photosynthesis. The fixation of radiant energy

24

D. E. MCCLOUD, R. J, BULA, AND R. H. SHAW

in the photosynthetic process can be estimated from the net carbon dioxide exchange between the plant and the natural environment. The flux of carbon dioxide above a crop in the field can be estimated from the vertical gradient of gas concentration and an appropriate transfer coefficient. This method was first used by Thornthwaite and Holzman (1942) to measure water vapor flux and was later refined by Pasquill (1950). Application of the theory to photosynthesis came later because carbon dioxide gradients are more difficult to measure. The development of the infrared gas analyzer has facilitated the determination of carbon dioxide fluxes over growing crops. Van Oorschot and Belksma (1961) have given a detailed description of an assembly for the continuous Leaf Area Light & Angle Intercepted

Net Photosynthesis

' 7 6 cm

30

16 0 L24

'-30-'

LA1 & La I

-%

-

5

0

5

-

5

0

5

-

5

0

5

M g C02/m2/min

FIG. 7. Relation of leaf area, leaf angle, and light interception to net photosynthesis by successive layers of leafage, Pearlmillet (Pennketum typhoides) at a density of 12 plants per square meter. McCloud (1984).

recording of carbon dioxide exchange in transpiration of plants. Lemon ( 1960) applied this aerodynamic method for determining carbon dioxide exchange between the atmosphere above a corn field. He showed that turbulence may be a limiting factor in supply of carbon dioxide to an active crop under conditions of high incident radiation. Monteith and Szeicz (1960) showed that meteorological estimates of carbon dioxide transfer over a field of sugar beet gave good agreement with conventional estimates from dry matter production. Monteith (1963) lists the resistances to carbon dioxide difusion as: the external resistance above the canopy, the external and internal resistances within the canopy, including stomata1 resistance, and the intercellular resistances of the chloroplast. From this comprehensive review it is concluded that the aerodynamic resistance to carbon dioxide diffusion is not large compared with internal resistances of the plant.

FIELD PLANT PHYSIOLOGY

25

The soil system also serves as a source of carbon dioxide. Monteith (1962) on the basis of measurements over grass and beans at Rothamsted concluded that soil carbon dioxide production was of the order of 0.03 mg./cm.2/hour while the maximum rates of grass photosynthesis were of the order of 0.2 mg. of C02/cm.2/hour. Monteith stressed the importance of determining the respiratory flux and its contribution to dry matter production, since the conventionally assumed respiration rates are inadequate and field measurements of this aspect are urgently needed.

B. TEMPERATURE 1. The Role of Temperuture in Crop Production The effects of temperature on plants have been extensively studied and the major generaIity which can be drawn from the extensive literature is that the effects of temperature are strongly interrelated with the other factors of the environment. However, the role of temperature in crop production is apparent from the specificity of the cultivated species grown in the subarctic, temperate, and tropical regions. Surprisingly, the wide range of cultivated species found among these regions and their definite seasonal relations are brought about by a relatively narrow temperature range. Growth of higher plants is largely restricted to temperatures between 0°C. and 60°C. (Spector, 1956). Crop production is confined to an even narrower range, 10 to 40°C. Temperature regimes within this range plays a dominant role in delineating species adaptation. Crop production research on plant-temperature relations falls into three general categories. First is the optimum temperature regime for the various crop species. Second is the maximum temperature endured by the crop species without either reduced dry matter accumulation or death of the plant. Third is the lowest temperature tolerated by the crop species, which for some species may be much below freezing and for others well above freezing. However, Parker (1946) points out that attempts to establish the fixed cardinal temperatures and particularly the optimum temperature have been inconclusive. The physiological complexity of the plant as an organism may preclude the definition of these cardinal points because different physiological processes within the plant may have different temperature coefficients. Although the definition of these cardinal temperature points may remain empirical, there is presumably an optimum, maximum, and minimum for each crop variety grown under a given set of environmental conditions. Agronomic research relative to temperature-plant response can be categorically related to these three cardinal temperature regimes.

26

D. E. M C CLOUD, R. J. BULA, AND R. H. SHAW

2. Optimum Temperature Regimes The cardinal temperature most pertinent to crop production is the optimum regime. For example, such crops as corn, sorghum, and soybeans need much warmer temperatures for maximum yields than do such crops as oats, peas, and potatoes. However, since the effects of temperature are conditioned by other environmental factors, research conducted under field conditions to determine optimum temperature regimes for given crop species is difficult to interpret. Recently, studies have been conducted under controlled environmental conditions which have reduced the number of interacting variables. Friend et al. (1962a,b) report that for MARQUIS wheat the optimum temperature for the maximum relative growth rate was 15 to 20°C. The optimum temperature represented a compromise between two opposing physiological processes, As temperature increased the leaf area (photosynthetic surface) increased, but the rate of respiration also increased, Further as the temperature increased, over the range of 10 to 30"C., the rate of development of the wheat plant increased. Thus, the relative growth rate of plants grown at high temperatures decreased because of the effect of temperature on the rate of development of the plant. With time, both net assimilation rate and leaf area per unit plant weight declined. This anomaly illustrates another difficulty involved in attempting to establish the cardinal temperatures-cardinal temperature regimes are modified by the stage of development or physiological condition of the plant. Optimum temperatures of 20°C. for perennial ryegrass (Lolium perenne), orchardgrass ( Dactylis glomerata), Agrostis tends, and Holcus lanutus, have been reported by Mitchell (1956) as shown in Fig. 8. Optimum temperature for Paspalurn dilutatum, however, was near 30°C. and for white clover near 25°C. Figure 9 shows that over a relatively wide range of temperatures, from 12°C. to 28"C., the rate of development of individual tillers was relatively constant. Brown ( 1939) reported similar results for bluegrass and orchardgrass. However, he showed that optimum temperatures for top growth were higher than optimum temperatures for root and rhizome growth. Sullivan and Sprague (1949), likewise, concluded that optimum temperature regimes for ryegrass were near 20°C. based on maximum growth and accumulation of reserve carbohydrates. Temperature can affect the chemical composition of the plant as well as the overall growth rate. Howell and Cartter (1953) reported positive correlations between the oil content of soybeans and maximum temperatures on the basis of field data. It was assumed that the temperature

27

FIELD PLANT PHYSIOLOGY I

1

1

-

TEM PERATURE

20

rZ-i----

C

-----

TEM PERATURE

FIG.8. Percentage daily increase in dry weight of the total shoot of 10 forage species grown at various temperature regimes. Ba, perennial ryegrass; Bc, orchardgrass; Bt, bentgrass; By, velvetgrass. Mitchell ( 1956).

28

D. E. MC CLOUD, R. J. BULA, AND R. H. SHAW

2

i

1

l5-

W fr U

45'

55'

I

I

65' 75' TEMPERATURE

- 00

20-

-------a-

I-

I

85'

I

95'

I

0

8c BY 00

-

15-

w U P

lo-

II ' Ba 20 - - - - _ _Ac I

---

-a-

c z

I

I

I

Ak Al

15-

W

v

I

45O

55*

65.

75'

I

1

85.

95'

1

TEMPERATURE

FIG.9. Percentage increase per day in tiller numbers of 10 forage species grown at various temperature regimes. Ba, perennial ryegrass; Bc, orchardgrass; Bt, bentgrass; By, velvetgrass. Mitchell ( 1956).

29

FIELD PLANT PHYSIOLOGY

coefficient of the oil synthesis reactions was about 2, which would mean marked oil formation rate changes concurrent with higher temperature. The highest correIations were obtained for the periods of 20 to 40 days before maturity, which coincided with periods of maximum rates of oil accumulation in the soybean seeds (Table I ) . TABLE I Mean Correlation Coe5cients of 5 Soybean Varieties between Oil Percentage and Maximum or Minimum Temperatures during Various Periods before Maturitya Number of days before maturity

0

Oil and maximum temperature

0.74 50 to 40 0.75 40 to 30 30 to 20 0.83 20 to 10 0.67 10to 0 0.52 From data of Howell and Cartter (1953).

Oil and minimum temperature

0.43 0.70 0.66 0.64 0.43

For some crops, the optimum temperature regime may, of necessity, be a compromise between temperature regimes optimal for growth rate and temperature regimes optimal for the desired chemical composition (Brown, 1939; Sullivan and Sprague, 1949). Another important consideration related to optimum temperature regimes is the genotype or strain and temperature interaction. Morley (1958) found markedly different temperature response patterns for five strains of subterranean clover. Optimum temperatures were high for two strains, low for two other strains, and one strain showed little difference over the temperature range studied, which was 17 to 24°C. Grafius (1956) has suggested that night temperature may be an important factor in the relative yields of varieties of oats and barley. The varieties were most sensitive to night temperatures during heading, illustrating again the importance of stage of development in the temperature response. 3. Maximum Temperature Regimes The effects of high temperatures on plant growth have recently received considerable attention. Bonner (1957) proposed that at the temperature extreme, either maximal or minimal, plant growth is depressed by the inactivation of temperature-sensitive reactions producing essential metabolites and that the growth depression can be alleviated by introducing these metabolites into the plant from an external source. Sizable increases in dry weight of pea plants were noted by Ketellapper and Bonner (1961) when plants were treated with

D. E. MC CLOUD, R. J. BULA, AND R. H. SHAW

30

vitamin B or ribosides while growing at above optimum 30°C. day and 23°C. night temperatures. They interpret these data as establishing that high temperatures depress growth by affecting specific biochemical events. Ketellapper (1963) reported further data which showed that in a number of cases the reduction in growth resulting from unfavorable temperature exposure was prevented, either partially or completely, by applying chemically well-defined substances to plants growing under such unfavorable temperature conditions. The effective chemical appears to be specific with species and temperature regimes involved (Table 11). TABLE I1 Dry Weight Means (in Grams) for 5 Species Treated with Metabolites at 2 Temperature Levelsa Temperature regime “C. Species

Pisum satiuum L. Vicia faba L. Lupinus nanus Dougl. Cosmos bipinnata Cav.

Solanum melongenu L. a

Treatment

17/10

20/14

23/17

30/23

Control Vitamin C Control Vitamin C Control Vitamin B Control Vitamin B Control Ribosides

-

-

0.45 0.47 -

0.22 0.31 1.59 1.96 0.78 1.10 2.09 1.93 2.03 1.54

-

-

2.50 2.42

-

1.32 1.38

0.57 0.72

-

-

-

0.56 0.82

-

-

-

-

-

From data of Ketellapper ( 19fX3).

This would support the suggestion that these “climatic lesions” can be considered to be the temperature-dependent requirement for the product of a chemical reaction which is more sensitive to temperature than all others. Langridge and Griffing (1959) reported that biotin prevented deleterious high temperature effects in two races of Arabidopsis thalianu, while cytidine partly alleviated the effects in a third race, and two races did not respond to the chemical supplements. Daday (1963) described a high temperature-sensitive mutant in DU PUITS alfalfa that appeared to be a result of the high temperature sensitivity of the nitrate-reductase system. It is quite likely that many similar yet undiscovered temperature sensitive systems exist in crop plants. Their discovery could have considerable economic importance since the deleterious effects could be overcome either by selection for specific temperature-tolerant strains or by direct application of the metabolite which would alleviate the high

31

FIELD PLANT PHYSIOLOGY

temperature depressive effect. A recent review by Langridge (1963) gives five causes for the high-temperature deleterious effects, Each cause ultimately blocks the synthesis or accelerates the breakdown of some essential metabolite. A considerably different analysis of high-temperature effects has been proposed by Oppenheimer and Drost-Hansen ( 1960). They reported maximum growth of Clostridium sp. at 13, 23, and 37"C., which corresponds to temperatures at which the structural stability of water is at a maximum. Minimum growth rates were observed at 16, 31, and 43"C., which corresponds closely to temperatures at which rather abrupt changes in the viscosity of water occur. These water structural changes may relate to the hydration shells around the protein macromolecules (Klotz, 1958). Thus, stability of proteins (enzymes) could be postulated as maximum for the temperature at which water structure is highly stable. 4. Minimum Temperature Regimes

The reduced growth rate of plants below optimum temperatures is generally considered a result of decreased rates of chemical reactions. TABLE I11 Effects of Lowering Day or Night Temperatures on the Growth of Forage Speciesn Relative growth (per cent) 60°F. day temperatures Species

45°F. night

Perrenial ryegrass Orchardgrass White clover Subterranean clover

100 100 100 100

0

35°F. night &I

57 61 86

45°F. day temperatures 45°F. night

35°F. night

58 37 35 58

45 22 27 37

From data of Mitchell and Lucanus ( 1960).

However, no reasonable explanation has been advanced for the observed range of growth of different crop species at temperatures near 10°C. For example, corn and tomatoes cannot survive such conditions whereas the cool season grasses, some strains of alfalfa, and small grains are able to maintain substantial net assimilation rates. McCloud ( 1963) observed that the subtropical grasses develop severe chlorosis when exposed to night temperatures of 1O"C., and that the lack of chlorophyll undoubtedly results in a reduced photosynthetic rate. Mitchell and Lucanus (1960) found that low day temperatures gave a much greater relative reduction in growth than did lower night temperatures (Table 111). Presumably, this is a reflection of lower photosynthetic rates as a result of the lower day temperatures. Morley

32

D. E. M C CLOUD, R. J. BULA, AND R. H. SHAW

et al. (1957) have reported considerable genetic variation within alfalfa varieties for winter growth indicating that varieties capable of producing more dry matter at near freezing temperatures could be selected. Survival of plants exposed to subfreezing conditions or winter hardiness has been extensively investigated by agronomists and plant physiologists. As Levitt ( 1962) points out, no inclusive hypothesis has evolved to account for all experimental observations. Levitt proposes that frost resistance is a resistance toward sulfhydryl oxidation and its interchange with disulfide to resist the formation of intermolecular disulfide bonds. Levitt maintains that this is the only hypothesis compatible with the accumulated information on frost injury. Whether the theoretical considerations proposed by Levitt regarding the molecular bonding-low temperature relationships of plants will be substantiated or whether the previously suggested protective property classification remains to be determined. The available evidence establishes that tolerance to subfreezing temperatures involves the protoplasmic protein components, not merely a physical depression of the freezing point by sugars or other cellular constituents. Specifically how the protein components effect cold tolerance has not been elucidated. This aspect undoubtedly will be the focal point of considerable research in the immediate future. A companion aspect of low temperature effects on plants is vernalization. This phenomenon, like cold resistance, has been the subject of extensive research. Vernalization was first described in winter cereals but has since been shown to be important in most crop species adapted to temperate or subarctic regions. The classical research by Purvis and Gregory (1937) is accepted as having contributed a major portion of our knowledge on vernalization as it relates to cereals. Cooper (1957) studied extensively vernalization in forage grass species. The observation that gibberellin promoted flowering among species which were known to require cold treatment promoted many investigators to suggest that gibberellin replaces the cold requirement and thus was the active chemical (vernalin) involved ( Lang, 1956). Later work indicates that only long-day plants with a rosette habit of growth can be induced to bolt and flower with gibberellin treatment. Furthermore, in the naturally vernalized plant development the floral primordia differentiate prior to bolting whereas in gibberellin-treated plants bolting may occui prior to differentiation of flower primordia. ' Peterson and Bendixen (1963) working with Lolium temulentum L. have concluded that the main effects of exposure to low temperature cannot be replaced by gibberellins. As evident in Table IV, concentrations

33

FIELD PLANT PHYSIOLOGY

of 100 to 10,000 p.p.m. of gibberellic acid in the water used to moisten the seeds during the vernalization or germination period had very little effect on leaf number at heading. Where no cold or vernalization period was imposed, the average leaf number was near 22 at heading. A vernalization period of 28 days reduced the leaf number at heading to near 17-still much higher than the number considered optimal for this species. A vernalization period of 56 days reduced the leaf number at heading to 8, and gibberellic acid treatment in conjunction with this vernalization reduced the leaf number to near 6. As pointed out by Peterson and Bendixen (1963) flowering probably involves a balance of promotive and inhibitory substances together with an adequate supply of energy substances. GibberelIins may be involved in this balance, but the main effects of exposure to low temperature in the vernalization process appear to be more profound and evasive. TABLE IV Effect of Increments of Cold and Gibberellic Acid (GA) on Leaf Number at Heading of Loliurn terndenturn L.a ~

Days cold treatment None 28 56 a

~

~

Leaf number at heading for various p.p,m. GA: 0

100

22.3 17.3 8.0

21.5 15.4 6.8

1000 21.9 17.9 6.1

10,000 21.1 16.3 5.9

Mean 21.7 16.7 6.7

From data of Peterson and Bendixen ( 1963).

Whyte (1960) comments that even though recommendations of early Russian workers stressed the desirability of practical application of the vernalization technique on a large scale in the cultivation of cereals and other crops, this practice has not been used routinely in crop production.

5. Thermoperiodicity Diurnal changes in temperature normally associated with the light and dark period produce pronounced effects on plant growth and development. Thermoperiodic effects have been extensively investigated, and Went (1953, 1957) has reviewed the literature on this subject. The general basis presented for the response of plants to diurnal thermoperiodicity is the relationship between the Qlo of the predominant processes of photosynthesis and respiration. Grafius ( 1956) suggested that the effect of night temperature on the yield of barley and oats may be a reflection of its effect on respiration during the period when no photosynthates are being accumulated. Other physiological reactions are also affected by such diurnal temperature changes. Robertson et aE. (1962) reported that diurnal temperature changes have a pronounced

34

D. E. MCCLOUD, R. J. BULA, A N D R. H. SHAW

c: 3

5

E v L

0

60

40-

30-

m 3

* 200

c

10-

0

11111111111 10

20

30

40

50

Days from flowering FIG. 10. Total sugar content of developing pea seeds as affected by various temperature regimes. Robertson et a2. ( 1962).

FIG. 11. Starch content of developing pea seeds as affected by various temperature regimes. Robertson et a2. ( 1962).

FIELD PLANT PHYSIOLOGY

35

effect on the carbohydrate composition of developing pea seeds. At low night temperatures the conversion of sugars to starch in developing seeds was delayed whereas at high temperatures the sugars were rapidly converted to starch (Figs. 10 and 11). Went ( 1957) pointed out that thermoperiodic effects quite probably involve much more complex physiological relationships than the Qro of various physiological processes. Certain species are very sensitive to high night temperatures and die when exposed to repeated high night temperatures. Still other species, such as potatoes, require cool nights for the development of tubers and accumulation of organic reserves. Thermoperiodic responses are undoubtedly an expression of many internal processes. In crop species, natural selection as well as breeding have in all probability resulted in the evaluation of varieties best adapted to the thermoperiodicity existing under field conditions. This is further illustrated by the fact that, through selection, varieties selected for tropical regions produce comparable yields to those varieties of the same species selected for adaptation to temperate regions. 6. Heat-Unit Accumulation

An interesting agronomic application of the temperature effect on plants, which has received considerable attention recently, is the “heatunit theory.” Actually, this theory integrates phenology, physiology, and climatology as a tool for predicting plant growth, development, maturity, and yield. Katz ( 1952) reported an essentially linear relationship between heat-unit accumulation and tenderometer readings of two varieties of canning peas. Heat unit accumulation data along with controlled planting and harvesting are used in planning cannery operations. Chinoy (1956) proposed a photothermic quanta theory for predicting the development of wheat, which takes into account both the number of light-hours and the degree-days during the vegetative period. Chinoy reports different quanta requirements for different varieties. Wiggans ( 1956) found small variations from year to year in the number of heat units required for a specific variety of oats to reach maturity. Considerable use has been made of the heat-unit approach in Canada. Holmes and Robertson (1959) present considerable information on this concept. Boughner and Kendall (1959) have summarized growing degree day normals for a large number of stations in Canada, and present maps showing the pattern of growing degree days for selected periods. Went (1953) is somewhat critical of the use of heat units because no consideration is given to the effects of thermoperiodicity and also because optimum temperatures change during the development of the plant.

36

D. E. MC CLOUD, R. J. BUM, AND R. H. SHAW

Wang (1960) has also pointed out limitations of this approach. These limitations of the heat-unit theory can be overcome if more detailed climatic data become available and the plant responses are more precisely delineated. Gilmore and Rogers (1958) proposed a modification of the degreedays calculation to take into account temperature above or below the optimum temperature range for growth of corn. Corrections were made for temperature exposures below 50" and above 86°F. A comparison of a number of methods of calculating degree-days is shown in Table V. The use of the means of 3-hour interval temperatures (8 for each day) gave TABLE V Comparative Coefficients of Variation, as Per Cent, of Heat Units Required for Development of Corn from Planting to Silking As Calculated by 12 Methods5 Method of calculation

Coefficient of variation (per cent)

Maximum and minimum temperatures, OF 2-50

X -50, base 50 X -50, optimum 90 X - 50, optimum 86 X - 50, bases 50 and 90 X -50, bases 50 and 86

6.08

3.65 4.49 2.74 2.05 1.63

3-Hour interval temperatures, "F 2-50 X 50, base 50 X 50, optimum 90 X -50, optimum 86 X -50, bases 50 and 90 X 50, bases 50 and 86

-

-

5

4.63 3.74 3.77 2.46 2.85 1.55

From data of Gilmore and Rogers ( 1958).

lower coefficients of variation than the corresponding method using only daily maximum and minimum temperatures. Gilmore and Rogers (1958) point out that because of its simplicity the maximum and minimum method is preferred. However, where diurnal variation of temperatures might be rather wide, use of the 3-hour interval method would be justified. Thus, as our knowledge of temperature responses of crop species increases, the calculation of heat units or degree-days can be refined to include diurnal temperature variations and other considerations, such as weighting of seasonal temperatures to correspond to the sensitivity of the appropriate developmental stage of the plant, thus increasing the precision of the method.

FIELD PLANT PHYSIOLOGY

37

C. WATERAND HUMIDITY Biologists are inevitably concerned with water. In crop production, plant-soil-water relations are of paramount importance and a vast amount of literature has accumulated on this topic. Russell (1959) presented an extensive review of water and its relation to crops and soils. An equally comprehensive review on plant water deficits and physiological processes was presented by Vaadia et aZ. ( 1961). Thus, the present discussion will be limited to some specific aspects of this broad subject.

1. Characterization of the Plunt-Soil System Under field conditions water loss is due both to evaporation from the soil surface and transpiration from plant surfaces. Fritschen and Shaw (1961) attempted to determine the loss of water due to the two components by growing corn through a plastic film and comparing the water loss under these conditions with water loss under normal field conditions. They found that 70 to 90 per cent of the water loss during the period when corn provided good ground cover was due to transpiration. The relative water loss due to transpiration changed during the development of the corn plant. Early in the growing season transpiration accounted for a much smaller portion of the total water loss than that when the corn had attained maximum size. As the plant matured and changed color, the per cent lost by transpiration decreased. Near maturity transpiration was the major source of water loss. Peters (1960) reported that transpiration accounted for only 50 per cent of the total water loss, particularly, if soil moisture was frequently replenished and the soil surface was wet. He found only minor variations in transpiration related to soil moisture supply. 2. Internal Plant Water Status The internal water status of a plant is not specifically related to soil moisture. Kramer (1959) emphasized that it is the internal plant water status which is relevant to plant growth. A plant may undergo water stress even when soil moisture is near field capacity and likewise encounter little moisture stress even though soil moisture is low, depending on atmospheric conditions. The internal water status of a plant is a function of water absorption versus transpiration. Transpiration and absorption in turn are basically a function of the radiant energy input at the leaf surfaces. The plant is seldom in a thermodynamic equilibrium with its environment. Energy gradients existing between the plant and its environment depend largely on the presence or absence of radiation and the vapor pressure differences between the

38

D. E. MC CLOUD, R. J. BULA, AND R. H. SHAW

plant and the atmosphere. The temperature of the plant plays a major role in the water vapor exchange between the plant and its environment because it determines the saturation vapor pressure of the moistureemitting surface (Raschke, 1958). When the stomata of a leaf are fully opened and the plant is carrying on normal metabolic functions, such as photosynthesis and respiration, transpiration is proportional to leaf temperature and the vapor pressure difference between the plant and the atmosphere. Any increase in radiant flux causes an increase in transpiration. However, if the internal water status of the plant is such that additional increments of water loss cannot be maintained, stornatal closure is induced. Normally stomatal closure is associated with a loss of turgor of the plant cells or wilting. Transpiration has generally been regarded as a necessary evil with no useful function in plant growth or development. However, Winneberger (1958) observed that growth was reduced or stopped when the plants were grown under high (almost 100 per cent) relative humidities and transpiration was completely stopped. Winneberger suggests that transpiration may be the energy source for all translocation except that resulting from diffusion. Particular emphasis is placed on translocation of materials between adjacent cells as compared to mass translocation from root to shoot. Another beneficial effect of transpiration has been suggested by Kinbacher (1963). His data indicate that transpiration may be an effectiveplant mechanism for alleviating high temperature injury. Plants exposed to high temperatures (44 to 45°C.) and 100 per cent relative humidities were damaged more severely than plants exposed to these same temperatures and 50 per cent relative humidities. At least under these experimental conditions, the cooling effects of high transpiration rates were of considerable benefit particularly with respect to plant survival. On the other hand, Russell and Barber (1960) have minimized any functional aspects of transpiration. They suggested that only under some specific circumstances does transpiration affect salt uptake and consequently plant growth. From an agronomic viewpoint, there appears to be no question that crop production could be enhanced if some means were devised to reduce transpiration and/or evaporation. It has been suggested that this could be accomplished by spraying the aerial portions of the plant with materials resistant to water vapor transfer or the application to the soil of highly hydrophobic chemicals which ultimately would either be transported to the plant-atmosphere interface in the stomata and there influence water vapor transfer, or reduce evaporation from the soil surface. Extensive use has been made of plastic sprays to reduce water loss

39

FIELD PLANT PHYSIOLOGY

from nursery stock during transplantation. Gale ( 1961) has evaluated a number of chemicals for possible use as antitranspirants on crop plants. One specific compound, a vinyl acetate-acrylate ester, sprayed on bean plants reduced the amount of water used per unit of dry matter production under field conditions (Fig. 12). On another report, Gale (1962) described a further effect of this same chemical on water vapor transfer. In this case an antipathogenic effect was noted presumably because the fungal hyphae were not able to penetrate the plastic film or nutrients and moisture was unable to diffuse to the fungus. 12

-

D

,....' .' _.-'

10 8 -

..

6

'

i

.'

./' ....

......._..

/'

./','

3

' 0 .............. ....I

I

.*'

6-

/' I'

...c . _----_--,.

.-.*'-._.-.-.

.'

.. .

/'*

0<...................

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

d...'

2.

& 4 -/'

#;

p-

&$

- ,

A

. I .

2,/st

,Znd

I

,3rd treatmenf

FIG. 12. Effect of antitranspirant spray on growth of bean plants in the field. Gale ( 1961). D = adequate irrigation and anti-transpirant spray; C = adequate irrigation, no spray; B = irrigation one-half of adequate and anti-transpirant spray; A = irrigation one-half of adequate, no spray.

Another method of inducing reduced transpiration rates has been reported by Roberts (1961). He applied cetyl and steryl alcohols to the plant root system and reported sizable reduction of transpiration. These results have not been confirmed by other investigators. Oertli (1963) added high molecular weight fatty alcohols and related compounds to nutrient solutions of barley and bean plants. High concentrations ( u p to 10 per cent) added to the nutrient solution resulted in some reduction in transpiration. (Table VI). However, these high concentrations also reduced plant growth, which in turn may have resulted in lower transpiration rates, Olsen et al. (1962) added hexadecanol and octadecanol to soil

40

D. E. MC CLOUD, R. J. BULA, AND R. H. SHAW

in which corn was grown without moisture conserving effects. These tests were conducted under greenhouse conditions, In their studies the amount of water transpired per unit of dry matter increased under some treatments. They postulated that the results, when extrapolated, meant that these fatty alcohols would not affect water loss under field conditions. Possible interrelationships between moisture stress, the use of antitranspirants, and plant temperatures has been discussed by Tanner TABLE VI Effects of Antitranspirants Added to the Nutrient Solution on the Growth and Water Use of Barley Seedlings5

Antitrans pirant Steric acid

Oleyl alcohol

Stearyl alcohol

Cetyl alcohol

a

Concentration ( % of nutrient sol. )

Increase in top dry weight (mg./plant )

(ml./g. dry wt.)

0.0 0.1 1.0 10.0 0.0 0.1 1.0 10.0 0.0 0.1 1.o 10.0 0.0 0.1 1.o 10.1

10.6 9.8 9.6 8.0 11.8 10.9 9.2 5.3 10.6 10.8 9.8 9.7 12.3 10.7 12.2 7.6

63.3 63.3 68.7 55.7 68.7 60.0 68.3 68.8 80.0 66.5 67.6 66.8 67.3 78.7 72.3 72.4

Water use

From data of Oertli (1963).

(1963). Preliminary data shown in Fig. 13 indicate that temperatures of the plant material sprayed with an antitranspirant were slightly higher than unsprayed plant material. Likewise, an abrupt increase in plant temperatures were observed shortly after noon without concurrent increases in air temperature. The implication drawn by Tanner (1963) is that such increases in plant temperature may reflect a decrease in transpiration due either to moisture stress or stomata1 closure. Whether antitranspirants increase plant temperatures above critical levels has not been demonstrated. It is obvious that a primary requirement of antitranspirants is imperviousness to water vapor. Equally important, however, are the requirements that diffusion of other gases, particularly carbon dioxide, be unimpeded and that the chemical have no phytotoxic effects. Presently,

'

PRLINGTON, WISC. GRASS SURFPCE

-

27

211 1200

'

1

'

I

'

'

1 ' 1400 HWRS, SEPTEMBER 19,1961

I300

'

'

'

I ' 1500

'

FIG. 13. Temperatures of grass, of grass treated with evaporation suppressor, and of air. Tanner ( 1963).

critical stages could minimize deleterious effects of even limited periods of drought. An interesting plant response that apparently circumvents the deleterious effects of transpiration has been described for plants that are able to fix appreciable amounts of carbon dioxide during the dark period. Nishida ( 1963) reports observations on some Crassulacean species which show nocturnal opening of stomata. This stomata1 behavior appears to coincide with a daily periodicity of the accumulation of organic acids produced by the absorption of carbon dioxide during the day. Thus, during the day when vapor pressure gradients are the steepest the stomata are closed, thereby preventing water loss. No decrease in photosynthetic activity results because the carbon dioxide fixed in the organic

42

D. E. MCCLOUD, R. J. BULA, AND R. H. SHAW

acids during the previous dark period is available for reduction by the energy absorbed by the chloroplasts. This particular type of metabolism has been suggested as the explanation for the high yields of pineapple even under limited moisture conditions (Ekern, 1959). It is generally assumed that dark fixation of carbon dioxide in any appreciable quantities is restricted to crassulacean species. However, a more thorough survey of such metabolism particularly among tropical and xerophytic species may make some contribution to crop production. 3. Water Deficits and Plunt Growth

The physiological effects of water stresses resulting in permanent wilting of the plant are obvious. Water stresses which are not as severe are more generally encountered in crop production. What effects such moderate water stresses impose on crop yields have not been determined critically. This lack of information on moderate water stresses may be an indirect result of the general presumption that transpiration (and other plant processes) remains at normal rates until soil moisture nears the permanent wilting percentage. It has been suggested by Hagan et ul. (1957) that increasing soil moisture stress does not have a uniform effect on the plant. Plant processes appear to vary in their sensitivity to water deficits. Hagan et al. (1957) found that in Ladino clover chemical composition, flowering and seed production were affected progressively within most of the range of available soil moisture from field capacity to permanent wilting percentage. On the other hand, photosynthesis and dry matter production were not affected until the soil moisture content approached the permanent wilting percentage. The curves in Fig. 14 show little change in photosynthetic rates until soil moisture neared the wilting point. Whereas in Fig. 15, seed yields were affected over the entire range of moisture conditions. Likewise, Owen (1958) concluded that dry matter production and net assimilation rate of sugar beets, broad beans, and lettuce were not affected by repeated short periods of water stress. As shown in Table VII however, leaf area of the plants subjected to water stresses was found to be lower under some conditions. In these experiments the plants were grown in containers and there remains some question whether the results obtained would be applicable to field responses. Gates (1955a,b) reported significant reductions in the growth of tomato plants for most of the period that water stress was imposed even though the soil moisture was not below permanent wilting percentage. He suggested that because of the magnitude of the growth depression the effects of moisture stress were imposed on the plant soon after the soil

43

FIELD PLANT PHYSIOLOGY

moisture content dropped below field capacity. The younger leaves of the plant were most severely affected by moisture stress and apparently were the major source of the observed growth depression (Gates, 1955b). The older leaves did not display moisture stress effects as readily as the

C

2.

In

,o 0

L

a

10

AI

0 I

C I

D I

Harvest date FIG.15. Seed yields at various harvest dates of Ladino clover grown at various levels of soil moisture. B, moisture depleted to 50%; D, moisture depletion to wilting point; E, moisture depletion until plants wilted overnight. Hagan et al.

(1957).

44

D. E. MC CLOUD, R. J. BULA, AND R. H. SHAW

younger leaves, but the younger leaves recovered more rapidly from the moisture stress than did the older leaves. Another interesting observation was that upon rewatering plants that had been visibly affected by moisture stress the subsequent growth was more characteristic of physiologically young tissue. TABLE VII Dry Weight and Leaf Area of Sugar Beets Grown under Different Watering Regimes& Water regimes Field capacity

Rewatered when pF reached 4.0

Samplings

Dry weight ( g./plant)

Leaf area (dm.Z/plant)

Dry weight (g./plant)

First (May 8 ) Second Third Fourth Fifth Tune 3 )

7.5 25.4 38.9 46.1 72.8

11.0 15.2b 14.8b 20.6b 21.9

7.6 25.0 35.0 47.7 73.9

Leaf area (dm.z/plant)

10.4 14.1b 13.5b 18.9b 19.5

From data of Owen (1958). Difference between watering treatments exceeds at least the 5 per cent level of probability. a

b

Under moisture stress conditions, Gates and Bonner (1959) found that destruction of ribonucleic acid (RNA) was accelerated although the ability to synthesize RNA was not lost. West (1962) reported that corn seedlings germinated under water stress, as simulated by high osmotic concentration of the nutrient solution, accumulated RNA but proteins and nucleotides were quantitatively reduced (Fig. 16). Thus, water stress on germinating seedlings resulted in RNA which was altered in composition from the RNA found in tissues not affected by water stress. This “altered” RNA was not used as effectively in protein synthesis, which could have contributed to the reduction in growth rates observed both by West and by Gates and Bonner. Ashton (1956) observed that alternating cycles of high and low soil moisture affected the photosynthetic rates of sugar cane as well as the plants’ ability to recover from the effects of water stress. Photosynthetic rates of sugar cane decreased markedly during periods of water stress. Recovery of photosynthetic rates, after the wilted plants were rewatered, increased with repeated exposures to cycles of low followed by high soil moisture. These findings support the empirical observation that stress at one period may decrease the severity of subsequent moisture stress. The effects of water stress depend not only on the duration and intensity of the stress, but on the timing in terms of plant development. Denmead and Shaw (1960) reported corn grain yield reductions of 25 per

45

FIELD PLANT PHYSIOLOGY

OPTICAL DENSITY ( 2 6 0 m u ) NP

0.02A

AOP

0 6 .

0 2 .

1 I

0

180 .

I40

,

,

40

,

80

,

,

,

,

,

,

,

,

200

160

120

,

,

240

280

AMP

14.7 A

st/(I

I00

-

40

GTP VTP

UDP AT P

UMP

06

CTP

0 2

I

0

.

1

40

,

.

80

. 120 . . 1.60 . 2 .0 0 Tuba Numbar

I

I

240

I

I

280

I

1

320

'

FIG.16. Comparative elution chromatograms of nucleotides from 6-day-old corn seedlings grown on nutrient solution of 0.02 and 14.7 atmospheres of osmotic pressure. West ( 1962).

D. E. M C CLOUD, R. J, BULA, AND R. H. SHAW

46

cent when moisture stress (depletion of soil moisture to the wilting point) occurred prior to silking (Table VIII). When the moisture stress occurred at silking the yield reduction was twice as severe. Moisture stress periods that occur early in the development of the plant indirectly affect yield by reducing the size of the plant. On the other hand, TABLE VIII Plant Height and Grain Yield of Corn Plants Subjected to Moisture Stresses at Various Stages of Development@ Stage of Development when moisture stress applied None Vegetative Silking Ear Vegetative and silking Vegetative and ear Sillcing and ear 0

Stalk height (em. 1 154 138 144 144 122 134 132

Grain yield ( g./plant 1 364 273 183 289 154 243 179

From data of Denmead and Shaw ( 1960).

moisture stress periods coinciding with a critical stage of development such as silking in corn have a direct and consequently more severe effect on yield since in addition to reducing the photosynthate available for grain production, other detrimental effects are induced. Asana and Saini ( 1958) reported that intermittent drought during the post-head emergence period of wheat depressed grain yield. However, if the moisture stress was not severe enough to affect the color of the head, no yield depression was noted. The inference can be drawn that photosynthesis in the head of wheat or other cereals plays a major role in development of grain and in accumulation of dry matter. Chinoy (1962) concludes that in wheat the adverse effects of periods of moisture stress manifest themselves essentially by delaying maturation. The observed reduction in yield could be attributed to the adverse effect of higher temperature normally encountered in the field during the later periods of the growng season. Early maturing wheat varieties were much less affected by drought or moisture stress conditions than were late maturing varieties. These observations lead Chinoy (1962) to the conclusion that among the wheat varieties tested no real genetic differences in drought resistance exist, but rather genetic differences in rates of plant development and maturity. These latter characters determine the susceptibility of any given genotype to moisture stress. These relationships between heat tolerance and transpiration rates in winter oats have also been discussed by Kinbacker (1963).

47

FIELD PLANT PHYSIOLOGY

Scott and Patterson ( 1962) reported additional evidence that tolerance to moisture stresses may change during plant development. Corn and sorghum plants 30 days old survived simulated drought conditions in the greenhouse better than 60-day-old plants. However, alfalfa appeared to survive equally well at both 30- and 60-day-old stages of development. TABLE IX Effects of Various Periods of Moisture Stress Imposed on Alfalfa, Corn, and Sorghum at Two Physiological Agesa Species and age at time moisture stress imposed

Meclicago satiua L.

(3 0 days) (60 days) Zea mays L. (30 days) (60 days) Sorghum uulgare Pers. ( 3 0 days) (60 days)

Per cent plants recovering after moisture stress duration of:

1 Week

2 Weeks

3 Weeks

4Weeks

100 100 100 100 100 100

100 100 100

100 100 100

50

0 100 0

100 100 0 0 100 0

100 0

From data of Scott and Patterson ( 1962).

When corn and sorghum were grown in association with alfalfa tolerances to the drought conditions were less (Table IX). IV. Controlled Environment Facilities as a Supplement to Field Research

A. SUPPLEMENTAL ROLE OF CONTROLLED ENVIRONMENT FACILITIES Agronomists are interested in controlled-environment facilities primarily from the standpoint of acquiring a more complete understanding of the relationships that exist between the plant and its natural or field environment. This differs from the approach taken by those whose interests are centered primarily on studying the effects of environment on plant behavior. In the latter studies, the environments employed may or may not be related to field environments. One of the premises for the use of controlled environment facilities is that a single factor can be varied and thereby its effects studied while other factors are held constant. To an agronomist, the interactions of the various factors may be as important as the effects of any single factor. Thus, to delineate the response of a plant to natural or field conditions where the various factors are constantly changing, a factorial approach seems to be more justifiable. An alternative approach which emphasizes the interrelationships between the environmental factors is the utilization of the new largecapacity electronic computers for multiple variant analysis of the effects

48

D. E. MC CLOUD, R. J. BULA, AND R. H. SHAW

of the many environmental factors affecting the plant in field studies. Sometimes as much knowledge about the plant-environment relationship can be gained from this approach as from the single factor method in a controlled environment chamber. However, the field approach is inadequate for many types of problems. For example, controlled environment facilities provide the only means of fully elucidating the cold-hardiness mechanism in plants. During the fall several climatic changes occur simultaneously: the photoperiod gradually shortens, the mean daily temperatures decrease, sporadic extreme temperatures occur, and a more pronounced diurnal temperature variation develops, Under field conditions it would be impossible to define the role, if any, of these interrelated environmental changes. Such a problem can be studied only under a controlled environment. Thus, it is imperative that the general nature of the research to be conducted in the controlled environment facility must be defined before effective use can be made of this tool. For some work, such as production of uniform plant material, a less elaborate facility is necessary than for a problem involving the effect of environmental factors on plant growth. Likewise, equipment for programming environmental factors would be essential to study the effects of the seasonal weather changes on plant behavior, whereas for constant temperature studies less complex equipment would be required.

B. CONTROL OF PLANT ENVIRONMENT FACTORS IN THE GROWTH ROOM 1. Radiation Radiation is a major element in the plant’s environment. The plant is affected by three aspects of radiation: the quantity or density of the radiant flux, the quality of that radiant flux, and the duration of that radiant flux. The latter is the photoperiod. Radiation is undoubtedly the most difficult environmental factor to reproduce in controlled environment facilities. Until recently, the light sources that were available usually provided intensities considerably below photosynthetic saturation levels of most crop species. Those light sources that did provide higher intensities had spectra qualities quite different from sunlight. Also, the spectral distribution from some sources was such that even under high intensities poor plant growth resulted. The carbon-arc source is the most intense source available for continuous operation, but it is not well suited for uniform illumination of growth rooms. The light source that has been used most widely in this country is the fluorescent lamp supplemented with a small percentage of illumina-

FIELD PLANT PHYSIOLOGY

49

tion from filament lamps. The main advantages of the fluorescent lamp are the high efficiency of conversion of electrical energy to visible radiant energy and the relatively low heat output, primarily because of little emission of radiation in the infrared. Also, fluorescent installation is relatively simple and inexpensive. Within the last several years considerable progress has been made by illumination engineers in developing high output fluorescent lamps which are capable of providing intensities in the visible spectrum approaching full sunlight. Intensities of one-third full sunlight are easily attained. The spectral distribution, in the visible, of these high output lamps closely approximates daylight. Additional developments which have improved fluorescent lighting are the changes in the ballast circuiting and components. Special components are used to replace conventional ballasts, and these components may be located away from the growth chambers. This system minimizes or eliminates the problem of heat from the ballasts near the growth chamber and also reduces the complexity of the wiring system encountered when the ballasts are mounted away from the chambers. An elementary though often overlooked aspect of growth chamber lighting is uniformity of radiation. The most uniform lighting is in the center of the chamber. Near the walls the intensity is reduced since the walls absorb some radiation and output of a fluorescent lamp falls off near the ends. High reflectance material on the walls, such as paint, aluminized plastic film, or special aluminum sheets, provides a good reflecting surface and reduces the drop of intensities near the walls. Likewise, several lamps perpendicular to the main bank tubes results in a marked improvement in uniformity of illumination. Although quantity or intensity aspects of field radiation probably are the most difficult to reproduce, the quality or spectral aspect also presents difficulties. The major consideration is that the artificial light source provide the range of wavelengths involved not only in photosynthesis, but in the other photobiological processes as well. For photosynthesis, it is generally accepted that any light quantum absorbed by the photosynthetic organic pigments produces the same chemical effect regardless of its energy content. Emerson and Lewis (1943) found that for Chlorellu sp. the efficiency of each quantum remained virtually constant throughout the chlorophyll absorption spectrum. However, a loss of efficiency was noted in the blue region, where the carotenoids absorb strongly, and in the green region, where chlorophyll absorbs weakly. A further consideration is that chlorophyll when irradiated retains only 41 kcal. per mole regardless of whether it has absorbed a red quantum with just this energy or a blue quantum with an energy of 70 kcal. per mole (Steward, 1960). Thus, it seems reason-

50

D. E. M C CLOUD, R. J. BULA, A N D R. H. SHAW

able to conclude that quality of the radiation would not be a major consideration in photosynthesis and consequently in the dry matter accumulating process. Where light quality does become a major consideration is in the various photobiological processes that affect plant development. Fortunately, these light-induced processes are activated at extremely low energy levels. The action spectra of the pigments involved does not correspond to that of chlorophyll, but the wavelengths known to be involved lie between 0.3 and Lop, For this reason, filament lamps that emit this long wave radiation are required, though only as a small percentage of the total light source. Normal plant growth and development are obtained from the combination of fluorescent and filament lamp light sources. The question remains, however, how closely such light sources reproduce the spectral characteristics of sunlight and what effect such a divergence may have in extrapolating controlled environment results to field observations? 2. Air Temperatures Temperature effects have been the second most frequent environmental factor studied under controlled conditions, From an engineering and design standpoint, much more information is available relative to temperature control than to light. In plant growth rooms the cooling requirements are not static, but vary with the lighting component. In most cases, standard cooling and heating equipment provide temperature regulation well within acceptable limits. Two principles of air conditioning that only recently have been applied to plant growth room conditioning merit consideration. The first is size of heat exchange unit. The heat exchange surface used to remove or add heat to the plant growth room should be as large as feasible. The larger heat exchange surface is advantageous because the temperature differential between the cooling or heating media and the ambient air of the plant growth room need not be wide to effect the same temperature regulation. Conditioning systems with large heat exchange surfaces have been developed where the cooling media may be only a degree or SO lower than the chamber air. It is obvious that such a cooling system would reduce temperature variations and also have little effect on the humidity or moisture content of the air being cooled. The second principle important in temperature regulation is proportional control. A proportional controlling system provides for variable amounts of the total cooling or heating capacity to be used to regulate the temperature. This is in contrast to an on-off system where the entire heating or cooling capacity is cycled as demanded for temperature con-

FIELD PLANT PHYSIOLOGY

51

trol. Again, it is obvious that the fluctuations in temperature would be reduced with a proportional controlling system than in the on-off system. Utilizing large heat exchange surfaces and proportional temperature controlling systems, temperature control within f 1°C. can easily and economically be attained.

3. Humidity The principles of humidity control, similar to those for temperature control, are reasonably well understood. The salient difference is that humidity control within close limits without disrupting temperature control is expensive. Reasonably adequate equipment is available for controlling humidity within f 5 per cent. The most common humidifying systems are either steam or water mists or a hot water reservoir over which the chamber air is passed and recirculated into the chamber. De Remer and Smith (1961) described a humidifying system which uses various concentrations of salt solutions that condition the moisture content of the chamber air as it passes the salt solution reservoir. Dehumidification is usually accomplished by passing the air over cold coils, thus lowering the temperature of the air below the dew point, and the moisture condenses on the cold coils. Chemical dehumidification, which absorbs the moisture from the air with desiccants has been used, but it is much more expensive and response is slow when large amounts of water are to be extracted from the air. Chemical dehumidification is most feasible when operating at temperatures below 50°F. and relative humidities below 50 per cent. Whatever systems of humidification and dehumidification are employed, it is important that humidity control have little effect on temperature. 4. Atmospheric GQses Other than water vapor control, little has been done in controlled environment facilities to control the content of the other atmospheric gases. The importance of atmospheric gas content, particularly carbon dioxide, is readily appreciated. The main problem appears to be one of measuring the gas content-a problem that is not impossible but rather expensive. Infrared gas analyzers are available for measuring carbon dioxide content of the atmosphere. The development of thermistor thermal-conductivity cells may offer economical means of measuring carbon dioxide if the sensitivity can be made adequate to measure carbon dioxide concentrations as low as that encountered in air. The other important atmospheric gas that has not received adequate consideration is oxygen. As with carbon dioxide, equipment for measuring oxygen concentration has not been generally available. Oxygen

52

D. E. MC CLOUD, R. J. BULA, AND R. H. SHAW

analyzers based on the paramagnetic properties of oxygen have been available, but they are instantaneous indicators not suitable for operation of a control mechanism, such as a valve for enriching the oxygen content of air, Recently, improved oxygen meters based on polarographic methods have been introduced. This equipment is reasonably priced and could be used to monitor and regulate the oxygen content of the atmosphere in plant growth rooms. Little is known about the effects that various levels or changes of atmospheric pressures may have on plant behavior. It has been shown that changes in atmospheric pressures do affect some organisms. There is a seasonal pattern in the frequency of these pressure changes, and a small but definite diurnal atmospheric pressure change occurs. Although the magnitude is very small, because of the definite periodicity one could speculate that this pressure change might affect plants. Other minor exogenous rhythms have known biological effects. Effects of these atmospheric pressure changes on plant behavior are virtually impossible to study under field conditions. Barometric pressure changes occur simultaneously with changes in other climatic factors, such as humidity, temperature, and light quantity. Unfortunately, few controlled environment facilities are suitable for providing environments of various pressure levels or changes. However, this appears to be an intriguing area of environmental research that warrants exploration.

5. Soil Temperatures Soil temperatures, including the relationships between air temperature and soil temperature as they exist in the field, have received considerabIe attention. In this connection, one of the more serious reservations regarding plant growth rooms may well be the effects imposed by the peculiar soil or root zone environment characteristic of growth chamber culture. The question of the validity of observations on soil temperature in plant growth chambers has deterred work along these lines. The simplest method of soil temperature control is to place the pot in a water bath. Such baths have been described recently by Mack and Barber (1960) and Willis et al. (1963). Precautions need to be taken when such baths are used so that the large water surfaces do not affect the humidity levels of the air. Also, if soil temperatures below freezing are required precautions should be taken to avoid plant toxicity from the antifreeze in the water bath, Electrical cables or cooling coils embedded in the soil have also been used to control soil temperatures. This method is better adapted to

FIELD PLANT PHYSIOLOGY

53

bench-type culture than to the pot culture system which is used more frequently in plant growth rooms.

C. PROGRAM-CONTROLLED ENVIRONMENTAL CONDITIONS Most of the large controlled environment facilities have a number of rooms that operate at a constant environment. If specific requirements are needed for diurnal environment changes, the plants are moved from room to room. In contrast to this mode of operation a number of facilities have been developed around the concept of programmed environmental control. In such a facility, a preselected sequence of environmental changes occurs within each chamber. Both the rate and degree of change are normally programmed. The program-controlled approach appears particularly suitable for agronomic work because programs could be developed to simulate natural climatic changes more closely than could be achieved in a facility where the plants are moved from one environment to another. Also, a program-controlled facility eliminates the large amount of time required to move the plants during the course of the experiment. Of the factors that can be programmed, temperature, both dry bulb and wet-bulb, can be most easily accomplished. A number of rather inexpensive controllers are available that operate on a cam principle. The cam can be cut to program a 24-hour or 7-day sequence. Voisey (1963) recently described a cam-type controller for electronic temperature controllers. This type of controller is adaptable to any other electronically controlled environmental factor. Radiation quantity and quality also can be programmed rather simply by the use of a series of time clocks. Separate time clocks for the filament and fluorescent lamps permits programming these two light sources separately. Thus, predominantly red or long wavelength radiation could be provided to correspond with “sunrise and sunset” if the filament lamps were timed to come on before and go off after the fluorescent lamps. Some fraction of the total fluorescent lamps can be operated by separate time clocks so that each added increment of lamps increases the intensity from zero to the full capacity of the lighting system. Dimming of filament lamps can be accomplished by varying the voltage to the filament lamps. Although dimming of fluorescent lamps can be accomplished, it is rather expensive. The use of program-controlled facilities will in all probability increase. Such facilities can be used for constant conditions, but they do have the added flexibility of being able to provide varying environments if the need arises. The design of growth rooms is largely an economic problem. In most cases, the amount of space and degree of control are

54

D. E. MC CLOUD, R. J. BULA, AND R. H. SHAW

primary considerations, Therefore, the additional cost of programmed controls have been of necessity eliminated. In agronomic studies programming may well be more important than degree of control. For such work the ability to simulate field climatic changes may be more desirable than to control any one factor within very narrow tolerance. As engineering developments continue to advance, their application to growth chamber technology may mitigate the present economic barriers to more effective controlled environmental facilities. These facilities can be expected in the future to supplement even more efficaciously research in field plant physiology. REFERENCES Abbe, C. 1905. U. S. Dept. Agr. Weather Bur. Bull. No. 36. Alberda, Th. 1962. Netherlunds I. Agr. Sci. 10, 325-333. Alciatore, H. F. 1915. Monthly Weather Rev. 43, 400-402. Angots, A. 1914. Monthly Weather Rev. 42, 625-629. Asana, R. D., and Saini, A. D. 1958. Physiol. Plantarum 11, 666-674. Ashton, F. M. 1956. Plant Physiol. 31, 266-274. Azzi, G . 1914. Riv. Meteorica-Agrar, 35 ( 1 4 ) . Baker, D. G., and Strub, J. H. 1963. Minn. Tech. Bull. Ballard, L. A. T., and Petri, A. H. K. 1936. Australian 1. Exptl. Biol. Med. Sci. 14, 135. Black, J. N. 1955. Australian J. Biol. Sci. 8, 330-343. Black, J. N. 1957. Herbage Abstr. 27, 89-98. Blackman, G. E. 1938. Ann. Botany (London) [N.S.] 2, 257-280. Blackman, V. H. 1920. New Phytologist 19, 97. Bonner, J. 1957. Eng. Sci. Mag. 20, 28-30. Boughner, C. C., and Kendall, G. R. 1959. Cir. 3203. Meteorological Branch Dept. of Transport, Toronto. Boysen-Jensen, P. 1918. Botan. Tidsskr. 36, 219. Boysen-Jensen, P. 1932. “Die Staffproduktion der Pflanze.” Fischer, Jena. Briggs, G. E., Kidd, F., and West, C. 1920. Ann. Appl. Biol. 7, 103. Brooks, F. A., and Kelly, C. F. 1951. Trans. Am. Ceophys. Union 32, 833-848. Brougham, R. W. 1958. Australian 1. Agr. Res. 9, 39-52. Brougham, R. W. 1960. Ann. Botany (London) [N.S.] 24, 463-474. Brown, E. M. 1939. Missouri Agr. Exptl. Sta. Res. Bull. 299. Chinoy, J. J. 1956. Physiol. Plantarum 9, 1-18. Chinoy, J. J. 1962. Phyton (Buenos Aires) 19, 5-10. Cooper, J. P. 1957. 1. Argic. Sci. 49, 361-383. Crowther, F. 1934. Ann. Botany (London) 48, 877. Daday, H. 1963. Personal communication. Davidson, J. L., and Philip, J. R. 1958. Proc. Canberra Symp. Climatol. Microclimatol. 1956 Arid Zone Research Vol. 11, p. 108. UNESCO publ., Paris. Deacon, E. L. 1950. Quart. 1. Roy. Meteorol. SOC.76, 479-483. Denmead, 0. T., and Shaw, R. H. 1960. Agron. 1. 52, 272-274. De Remer, E. D., and Smith, R. L. 1961. Agron. 1. 53, 382-384. de Wit, C . T. 1959. Neth. I. Agr. Sci. 7, 141-148.

FIELD PLANT PHYSIOLOGY

55

Donald, C. M. 1961. Symp. SOC. Erptl. Biol. 15, 282-313. Ekem, P. C. 1959. Agron. Abstr. pp. 4-5. Emecz, T. I. 1962. Ann. Botany (London) [N.S.] 26, 517-527. Emerson, R., and Lewis, C. M. 1943. Am. J . Botany 30, 165-178. Evans, L. T., ed. 1963. “Environmental Control of Plant Growth.” Academic Press, New York. Friend, D. J. C., Helson, V. A,, and Fisher, J. E. 1962a. Can. J. Botany 40, 939955. Friend, D. J. C., Helson, V. A., and Fisher, J. E. 1962b. Can. 1. Botany 40, 12991311. Fritschen, L. J. 1963. J. Appl. Meteorol. 2, 165-172. Fritschen, L. J., and Shaw, R. H. 1961. Agron. J. 54, 71-74. Gale, J. 1961. Physiol. Plantarum 14, 777-786. Gale, J. 1962. Phytopathology 52, 715-717. Gaastra, P. 1959. Mededel. Landbouwhogeschool Wageningen 59( 13), 1-68. Gates, C. T. 1955a. Australian J. Biol. Sci. 8, 196-214. Gates, C. T. 195513. Australian J. Biol. Sci. 8, 215-230. Gates, C. T., and Bonner, J. 1959. Plant Physiol. 34, 49-55. Gilmore, E. C., Jr., and Rogers, J. S. 1958. Agron. J. 50, 611-615. Gol’tsberg, I. A. 1963. Climatic Descriptions from the Point of View of Agricultural Production Requirements. Available office of Tech. Sew. U.S.D.C., Washington, D. C. Gordon, N. T. 1930. J . SOC.Motion Picture Engrs. 14, 332-343. Grafius, J. E. 1956. Agron. J . 48, 56-59. Gregory, F. G. 1917. Exptl. and Res. Sta. Turner’s Hill, Cheshunt, Herts, 3rd Ann. Rept. p. 19. Gurevich, T. V. 1958. “Types of Meteorological Stations for Collective and State Farms and for Scientific Agricultural Research Institutions,” Compendium of Reports to the Commission for Agricultural Meteorology, W.M.O. pp. 2-5. Hydrometeorological Publ. House, Moscow. Hagan, R. M., Peterson, M. L., Upchurch, R. P., and Jones, L. G. 1957. Soil Sci. SOC. Am. Proc. 21, 360-365. Heath, 0. V. S. 1937. Ann. Botany (London) “3.1 1, 565. Hirst, J. M. 1954. Quart. J . Roy. Meteorol. Soc. 80, 227-231. Holmes, R. M., and Robertson, G. W. 1959. Publ. 1042 Canada Dept. of Agr., Ottawa, Canada. Hopkins, A. D. 1938. U. S. Dept. Agr. Misc. Publ. No. 280. Howell, R. W., and Cartter, J. L. 1953. Agron. J. 45, 526-528. Jennings, E. G., and Monteith, J. L. 1954. Quart. J . Roy. Meteorol. SOC.80, 222226. Kasanaga, H., and Monsi, M. 1954. Japan. J . Botany 14, 304-324. Katz, Y. H. 1952. Agron. 1. 44, 74-78. Ketellapper, H. J. 1963. Plant Physiol. 38, 175-179. Ketellapper, H. J., and Bonner, J. 1961. Plant Physiol. 36, Suppl. 21. Kimball, M. H., and Brooks, F. A. 1959. Calif. Agr. 13(5), 7-12. Kinbacker, E. J. 1963. Crop Sci. 3, 466-468. Klotz, I. M. 1958. Science 128, 815-821. Kramer, P. J. 1939. Aduan. Agron. 11, 51-70. Lang, A. 1956. Naturwissenschaften 43, 284-285. Langridge, J. 1963. Ann. Reu. Plant Physiol. 14, 441-462.

56

D. E. MCCLOUD, R. J. BULA, AND R. H. SHAW

Langridge, J., and GrifEng, B. 1959. Australian J . Biol. Sci. 12, 117-135. Lemon, E. R. 1960. Agron. J. 52, 697-703. Levitt, J. 1962. J. Theoret. Biol. 3, 355-391. Livingston, B. E. 1908. Science 28, 319-320. Livingston, B. E. 1915. Monthly Weather Rev. 43, 126-131. Livingston, B. E. 1935. Ecology 16, 438-471. Lloyd, M. G. 1961. Bull. Am. Meteorol. SOC. 42, 572-580. Loomis, R. S., and Williams, W. A. 1963. Crop Sci. 3, 67-72. McCloud, D. E. 1963. Temperature responses of some sub-tropical forage grasses. Report of the 2nd meeting of FA0 Working Party on Pasture and Fodder Development in the Tropical Americas. December 3-8, 1962. p. 10. Sao Paulo, Brazil. McCloud, D. E. 1964. Crop Sci. In press. Mack, A. R., and Barber, S. A. 1960. Agron. J. 52, 299. McLean, F. T. 1917. Physiol. Res. 2, 129-208. Merriam, C. H. 1898. U. S. Dept. Agr., Dio. Biol. Sci. Bull. No. 10. Mitchell, K. J. 1953. Physiol. Plantarum 6, 21-46. Mitchell, K. J. 1956. New Zealand J . Sci. Technol. A38, 203-216. Mitchell, K. J., and Lucanus, R. 1960. New Zealand J . Agr. Res. 3, 647-655. Monsi, M., and Saeki, T. 1953. Japan. J . Botany 14, 22-52. Monteith, J. L. 1962. Neth. J. Agr. Sci. 10, 334-346. Monteith, J. L. 1963. In “Environmental Control of Plant Growth ( L . T. Evans, ed.), Chapt. 7. Academic Press, New York. Monteith, J. L., and Szeicz, G. 1960. Quart. J. Roy. Meteorol. SOC. 86, 205-214. Morley, F. H. W. 1958. Australian J. Agr. Res. 9, 745-753. Morley, F. H. W., Daday, H., and Peak,‘j. W. 1957. Austrialian J. Agr. Res. 8, 635-651. Newman, J. E. 1956. Purdue Univ. Agr. Expt. Sta. Agron. Memo. 133. Newman, J. E., and Beard, J. B. 1962. Agron. J. 54, 399-403. Newman, J. E., and Wang, J. Y. 1959. Agron. J. 51, 579-582. Newman, J. E., Shaw, R. H., and Suomi, V. E. 1959. Wisconsin Agr. Expt. Sta. Bull. 537. Nishida, K. 1963. Physiol. Plantarurn 16, 281-298. Nuttonson, M. Y. 1947. “International Agro-Climatological,” Ser. NO. 1. Am. Inst. Crop Ecol., Washington, D. C. Oertli, J. J. 1963. Agron. J. 55, 137-138. Olsen, S. R., Watanabe, F. S., Kemper, W. D., and Clark, F. E. 1962. Agron. 1. 54, 544-545. Oppenheimer, C. M., and Drost-Hansen, W. J. 1960. J. Bacteriol. 80, 21-24. Owen, P. C. 1958. New Phytologist 57, 318-325. Parker, M. W. 1946. Soil Sci. 62, 109-119. Pasquill, F. 1950. Quart. J. Roy. Meteorol. SOC. 86, 16. Peters, D. B. 1960. Agron. J . 53, 536-538. Peterson, M. L., and Bendixen, R. E. 1963. Crop Sci. 3, 79-82. Ponomarev, B. P. 1958. “Phenological Observations,” Compendium of Reports to the 2nd Session of the Commission for Agricultural Meteorology, W.M.O. pp. 14-18. Hydrometeorological Publ. House, Moscow. Purvis, 0. N., and Gregory, F. G. 1937. Ann. Botany (London) 1, 569-591. Raschke, K. 1958. Flora (Jena) 146, 546-578. Reaumur, R. A. F. 1735. Thermometric observations made at Paris during the

FIELD PLANT PHYSIOLOGY

57

year 1735, compared to those made below the equator on the Isle of Maurituis, at Algiers and on a few of our American islands. Paris Mem. Acad. Sci. Rider, N. E. 1958. Work.! Meteorol. Organ. Tech. Note 21, 3-17. Roberts, W. J. 1961. J. Geophys. Res. 66, 3309-3312. Robertson, G. W. 1955. World Meteorol. Organ. Publ. No. 42, TP No. 16, Tech. Note 11. Robertson, G. W., and Holmes, R. M. 1956. Field Husbandry, Soils, and Agr. Eng. Div., Sci. Paper, Expt. Farms Serv., Ottawa, Canada. Robertson, G. W., and Holmes, R. M. 1958. I.U.G.G., Intern. Assoc. Sci. Hydrol. 3, 399-406. Robertson, R. N., Highkin, H. R., Smydzuk, J., and Went, F. W. 1962. Australian J . Biol. Sci. 15, 1-15. Russell, M. B. 1959. Aduan. Agron. 11, 1-131. Russell, S. R., and Barber, D. A. 1960. Ann. Reu. Plant Physiol. 11, 127-140. Saeki, T. 1960. Botun. Mag. (Tokyo) 72, 404-408. Schaal, L. A., and Newman, J. E. 1958. Bull. Am. Meteorol. SOC.39, 121. Schnelle, F. 1955. “Plant Phenology : Problems in Bioclimatology,” Vol. 3. Geest and Portig, Leipzig. Schnelle, F., and Volkert, E. 1957. Meteorol. Rtlndschau 10, 130-133. Scott, W. O., and Patterson, F. L. 1962. Agron. J. 54, 242-244. Shaw, R. H. 1954. Iowa State College, Agron. Dept. Sci. Rept. No. 2, Contract AF19 (604), p. 589. Spector, W. S., ed. 1956. “Handbook of Biological Data,” pp. 447-449. Saunders, Philadelphia, Pennsylvania. Steward, F. C., ed. 1960. “Plant Physiology: A Treatise,” Vol. lB, p. 21. Academic Press, New York. Stone, E. C. 1957. Ecology 38, 407-413. Sullivan, J. T., and Sprague, V. G. 1949. Plant Physiol. 24, 706-719. Suomi, V. 1957. In “Exploring the Atmosphere’s First Mile” (H. H. Lettau and B. Davidson, eds.), pp. 24, 79. Macmillan (Pergamon Press), New York. Suomi, V. E., and Kuhn, P. M. 1958. Telhs 10, 160-163. Talling, J. F. 1961. Ann. Rev. Plant Physiol. 12, 133-154. Tanner, C. B. 1963. Agron. J. 55, 210-211. Tanner, C. B., and Suomi, V. E. 1956. Trans. Am. Geophys. Union 37, 413-430. Tanner, C. B., and Suomi, V. E. 1958. Trans. Am. Geophys. Union 39, 63-66. Taylor, C. F. 1956. Plant Disease Reptr. 40, 1025-1028. Thesis, T., and Calpouzos, L. 1957. Phytopathology 47, 746-747. Thornthwaite, C. W., and Holzman, B. 1942. U . S. Dept. Agr. Tech. Bull. NO. 817. U. S. Weather Bureau, U.S.D.C. 1955. Instructions for Climatological Observers. Circ. B, 10th ed. U. S. Gov’t. Printing Office, Washington, D. C. Vaadia, Y., Raney, F. C., and Hagan, R. M. 1961. Ann. Reo. Plant Physiol. 12, 265-292. Van der Veen, R., and Meijer, G. 1959. “Light and Plant Growth.” Macmillan, New York. Van Oorschot, J. L. P., and Belksma, M. 1961. Weed Res. 1, 245-257. Ventskevich, G. 2. 1958. Agrometeorology U. S. Dept. Comm. Office Tech. Sew., Washington, D. C. Verhagen, A. M. W., Wilson, J. H., and Britten, E. J. 1963. Ann. Botany (London) [N.S.] 27, 627-640.

58

D. E. M C CLOUD, R. J. BULA, AND R, H. SHAW

Vicenti-Chandler, J., Silva, S., and Figarella, J. 1959. Agron. 3. 51, 202-206. Voisey, P. W. 1963. Can. 3. Plant Sci. 43, 111-112. Wang, J. Y. 1960. Ecology 41, 785-790. Wang, J. Y., and Barger, G. L. 1962. “Bibliography of Agricultural Meteorology.” Univ. Wisconsin, Press, Madison, Wisconsin. Warren Wilson, J. 1960. New Phytologist 59, 1. Watson, D. J. 1946. Ann. Botany (London) [N.S.] 11, 43-76. Watson, D. J., and Witts, K. J. 1959. Ann. Botany (London) [N.S.] 23, 431. Went, F. W. 1953. Ann. Rev. Plant Physiol. 4, 347-362. Went, F. W. 1957. “Experimental Control of Plant Growth.” Chronica Botanica, Waltham, Massachusetts. West, S. H. 1962. Plant Physiol. 37, 565-571. Whyte, R.0. 1960. “Crop Production and Environment.” Faber & Faber. London. Wiggans, S. C. 1956. Agron. 3. 48, 21-25. Williams, R. F. 1936. Australian 3. Exptl. Biol. Med. Sci. 14, 165. Williams, R. F. 1939. Australian 3. Exptl. Biol. Med. Sci. 17, 123. Williams, R. F. 1946. Ann. Botany (London) [N.S.] 10, 41-72. Willis, W. O., Power, J. F., Beichman, G. A., and Grunes, D. D. 1963. Agron. 3. 55, 200. Winneberger, J. H. 1958. Physiol. Plantarum 11, 56-61. Withrow, R. B., and Withrow, A. P. 1956. In “Radiation Biology” (A. E. Hollaender, ed.), Vol. 111. McGraw-Hill, New York. World Meteorological Organization. 1958. Final Report of the 2nd Session of Commission for Agricultural Meteorology. WMO Publ. No. 83, RP 35. Warsaw, Poland. World Meteorological Organization. 1961. Guide to Meteorological Instruments and Observing Practices. WMO No. 8, Tech. Publ. 3. Geneva, Switzerland.

CROP RESPONSE TO FERTILIZERS I N RELATION TO CONTENT OF "AVAILABLE" PHOSPHORUS

G. L. Terman, W. M. Hoffman, and B. C. Wright Tennessee Valley Authority, Muscle Shoals, Alabama; United States Department of Agriculture, Beltsville, Maryland; and Mississippi State University, State College, Mississippi

I. 11. 111. IV. V. VI.

VII.

Page 59 Introduction ............................................... 60 Status of Chemial Methods in the United States and Other Countries Chemical and Physical Nature of Fertilizers Marketed in the United 66 States ..................................................... 73 Crop Response Results Prior to 1950 .......................... 77 Recent Crop Response Results ................................ A. Water Solubility of the Phosphorus and Granule Size Effects . . 78 B. Quality of the Water-Insoluble Phosphate Fractions of Fertilizers 83 Problems Concerned with Nonorthophosphates and Other Fertilizers 93 93 A. Liquid and Suspension Fertilizers ......................... B. Solid Ammonium Polyphosphates ......................... 94 94 C. Fused Potassium Phosphates ............................. D. Calcium Polyphosphates ................................. 96 E. Bulk Blends ............................................ 96 In Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 References ................................................. 98

I. Introduction

The description of the nutrient content of fertilizers is of prime importance to both producers and consumers of commercial fertilizers. For this reason, fertilizer control laws have been enacted in nearly all countries that pay particular attention to the promotion of agriculture. In the United States, regulations have been adopted by the individual States, but not by the federal government. They require guarantees of the minimum percentages of each of the three primary nutrients-total nitrogen ( N ) , available phosphorus (expressed as Pz05), and soluble potassium (expressed as KZO)-and other constituents if they are listed on the fertilizer label or tag. The development and publication of standardized methods of analysis, necessary for the successful operation of fertilizer control laws, is a function of the Association of m c i a l Agri59

60

G. L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

cultural Chemists (AOAC). Some States specify a choice of AOAC or State methods in their fertilizer laws. The AOAC method for the evaluation of phosphorus in fertilizers is based on the quantity of phosphorus dissolved by extraction with water followed by extraction with neutral ammonium citrate solution. The soluble phosphorus determined by this method is comnionly considered to be usable by plants and is termed “available” phosphorus (expressed as the pentoxide, P205) . Fertilizer technology has changed greatly in recent years, and many new fertilizers have been introduced. Meanwhile, the assemblage of salts in many of the fertilizers has also been altered greatly. Farmers also have recently made great changes in fertilizer practices. These various changes in fertilizers and in fertilizer practices raise the question how adequately the present AOAC methods for phosphorus characterize the fertilizers that are currently being used. It should be recognized, of course, that no one laboratory method may give results in agreement with crop response over a wide range of kinds of fertilizers and of growth environments. Physical characteristics, such as granule size and placement of the fertilizer, for example, may greatly affect crop response, but cannot be reflected in a laboratory analysis of a ground fertilizer sample. The chief purpose of this chapter is to discuss the effects of various fertilizer characteristics on crop response to phosphorus in applied fertilizers, with emphasis on chemical composition and dissolution in water and ammonium citrate solutions. II. Status of Chemical Methods in the United States and Other Countries

It was known as early as 1808 that sulfuric acid will decompose mineral phosphates. However, great interest in phosphorus solubility did not occur until after Liebig’s proposal (1840) that the insoluble phosphorus in bones be rendered soluble by means of sulfuric acid. The quality of the superphosphate produced was based mainly on the phosphorus in water-soluble form. In 1842, a patent was granted to Lawes, who began the commercial manufacture of superphosphate by treating phosphate rock with sulfuric acid. It was soon discovered that the water-soluble phosphorus in this superphosphate returned to less soluble forms quite rapidly. However, this reverted phosphorus had a more beneficial effect on vegetative growth when applied to the soil than unacidulated mineral phosphates. While searching for analytical methods to distinguish between the two forms of phosphorus of low solubility, European agricultural chemists found that the reverted type was largely soluble in am-

‘‘AVAILABLE’’ PHOSPHORUS

IN FERTILIZERS

61

monium citrate solution, whereas unacidulated phosphate rock was only slightly soluble. During this same period, vegetative tests had indicated the reverted, or ammonium citrate-soluble, phosphorus to be equal in value to water-soluble phosphorus. Thus, the value of superphosphate, as far as phosphorus was concerned, was based on the quantity of soluble phosphorus, i.e., water-soluble plus citrate-soluble phosphorus, that it contained. Procedures were developed for the evaluation of phosphate materials using neutral (Fresenius et al., 1871) and alkaline ( Joulie, 1873; Petermann, 1880) ammonium citrate solutions in addition to the method based on solubility in water. The method based on the use of 2 per cent citric acid (Wagner et al., 1903) was developed specifically for basic slag in order to prevent its adulteration with raw phosphate rock. In the United States, fertilizer control chemists directed their efforts almost exclusively to the procedure of Fresenius, Neubauer, and Luck. Their investigations culminated in the incorporation of neutral ammonium citrate into the official method for the determination of available phosphorus adopted at the organizational meeting of the Association of Official Agricultural Chemists (1884). The method called for digesting the water-insoluble residue from 2 g. of the phosphate fertilizer, prepared by washing the sample with water, in 100 ml. of neutral ammonium citrate solution at 65°C. for 30 minutes with frequent shaking; washing the residue with water at room temperature; and then analyzing it for phosphorus. Direct citrate digestion, without prior washing with water, for nonacidulated fertilizers was adopted at the second meeting of the Association of Official Agricultural Chemists ( 1885). These methods were used for regulatory purposes and quality control until 1931, except that in 1922 the sample size for nonacidulated phosphates was reduced to 1g. In 1931, this modification plus a change in citrate digestion time from 30 minutes to 1 hour was extended to all nonacidulated phosphates except basic slag, and to all acidulated fertilizers. The change was made as a result of evaluation studies on ammoniated superphosphates. In 1949, basic slag, which had been evaluated by the use of 2 per cent citric acid, was included under the neutral ammonium citrate procedure. In 1950 continuous agitation by mechanical means during the citrate digestion was adopted as an official procedure. Manual shaking at 5minute intervals was deleted in 1960. A procedure for the direct determination of available phosphorus in the combined water- and neutral ammonium citrate-soluble extracts rather than by an indirect determination (total phosphorus minus the

62

C. L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

citrate-insoluble phosphorus) was adopted by the Association of Official Agricultural Chemists in 1961. The methods of the Association of Official Agricultural Chemists (1960) are the official methods of the United States, Canada, and Mexico. The Association’s methods or adaptations thereof are also used wholly or in part in Colombia, Chile, India, Israel, Republic of the Philippines, and some of the states of Australia and Brazil. In the European countries, as in the United States, a great lack of uniformity in the methods of analysis existed prior to 1880. The actions taken in Germany to introduce standardization of fertilizer analysis are typical of those used in other countries. In the various Germanic kingdoms, fertilizer control by the agricultural experiment station over the fertilizer plants in their immediate vicinity was entirely voluntary, even though the fertilizer industry had asked several times to have the government establish compulsory control and uniform regulations applicable to the whole trade. Several conventions of agricultural chemists, directors of experiment stations, and fertilizer manufacturers-beginning with one at Magdeburg in 1872, followed by those in Danzig and Munich, and the last one at Halle in 1881-resulted in the adoption of standard methods of analysis of fertilizers. The method for the evaluation of phosphorus soluble in ammonium citrate was based on the use of neutral ammonium citrate. Some of the meetings of directors of agricultural experiment stations were international in scope. At Karlsruhe, Germany, Petermann ( 1880) director of the station at Gembloux, Belgium, reported on the similarity of the availability to crops of various water-soluble and water-insoluble phosphates to their chemical solubility as measured by alkaline ammonium citrate. Because of his findings, the work on evaluating soluble phosphates in the European countries shifted to the use of alkaline citrate solution and resulted in the adoption of the Petermann method, or modifications thereof, as an official procedure in most countries of Europe. Twelve of 14 countries whose methods appear in a publication of the Organization for European Economic Cooperation (1952) use alkaline ammonium citrate. Nearly every country in the world has regulations governing the marketing of fertilizers. These regulations include an analytical examination of the material for determining its nutrient content. Among the nutrients, phosphorus presents the most complex problem. Total phosphorus in a fertilizer is a definite fixed quantity and different methods for its determination should give comparable results. Inasmuch as watersoluble and citrate-soluble phosphorus are arbitrary quantities and are dependent on factors such as sample weight, time of digestion, fineness

“AVAILABLE” PHOSPHORUS IN FERTILIZERS

63

of sample, kind of solvent, agitation, which have been chosen to give values that harmonize with crop response tests, methods for their determination vary among materials and from country to country. The bases for guarantee and quality control of phosphorus in fertilizer materials and mixtures in 31 countries appear in Table I. This table contains updated information on 21 countries surveyed by Jacob and Hill (1953) plus similar data on 10 other countries, (Note: for the sake of continuity, the same form and abbreviations are used. ) All countries except Belgium, China (Taiwan), Denmark, and India require a guarantee for total phosphorus in at least one phosphate material. These are materials such as basic slag, bone meal, guano, or nonacidulated phosphate rock for direct application to the soil. In Japan and Switzerland, mixed fertilizers in which the phosphorus is solely in the forms of these materials require only a total phosphorus guarantee. All fertilizer materials and mixtures are analyzed for total phosphorus in Australia, Canada, New Zealand, and Sweden. There is no couniry in which water-soluble phosphorus is the sole basis of guarantee for all fertilizer materials and mixtures. However, many countries do require only this determination for superphosphates. Many countries, including Australia (except New South Wales), Brazil (Bahia and S5o Paulo), Canada, Chile, China (Taiwan), Colombia, India, Israel, Italy, Republic of the Philippines, and the United States, o5cially control phosphorus solubility with neutral ammonium citrate. Alkaline ammonium citrate is used in Japan as well as in most European countries. Either citrate solution may be used in the Netherlands. Neither procedure is o5cial for any fertilizer material or mixture in New South Wales, New Zealand, Republic of South Africa, or the United Kingdom. New South Wales and New Zealand are the only constituencies that specify the use of 2 per cent citric acid for all products. Most other countries apply this procedure to the control of phosphorus quality in basic slag. According to the AOAC neutral ammonium citrate method (Association of OfficialAgricultural Chemists, 1960), 1 g. of fertilizer is placed on a filter paper and leached with successive small portions of water until 250ml. of filtrate is collected in an hour’s time, The residue from the water extraction is then extracted with 100ml. of neutral ammonium citrate under prescribed conditions, The citrate-insoluble content of the residue is then determined. Total phosphorus content of the original fertilizer is determined in a separate sample. The “available” phosphorus content of the fertilizer is defined as the difference between the total and the citrate-insoluble phosphorus contents, or it may be determined

TABLE I Bases for Guarantee and Quality Control of Fertilizer Phosphorus in Several Countriesa, Phosphorus soluble in:

country United States Canada Mexico Brazil Bahia,d Sao Paulo Chile Colombiad Australia New South Wales Queensland, Tasmania,d Victoria, South Australia, Western Australia New Zealandf Republic of South Africa China (Taiwan)f India6 Israeld Japan Republic of Philippines6 Belgium Denmark France Germany, Federal Republic

Total phosphorus

Ammonium citrate solutionc Water

Neutral

Alkaline

-

F F F

-

2% Citric acid

-

Bm, R F Bm, G, R

-

R, s Bm, G, R G,R

-

S

Fm, S Fm, S, T p

-

Fe

-

F F

F F

-

-

F

-

-

F Bm, G, R

Fg S

-

-

-

Np, s R, pp, s F

-

F A, Bs, Fc, Fm, Sb

R Bm, Fm, R

-

R, pp, s S A, s

Bm, R

-

-

A, Fc, Fm, K,S S A, K Sa, Fm

Bm, Bs, R Bm, R, S

F

-

-

-

Pp, s

Fc, Fm, Mp, Pp. R, T p Bs, Fc Fc, Fm, Pp, S, T p Fm, Np, Pp, S, T p

P

r

Fm, R Bm, Bs, G , R

F

-

Bs, Fm, Pp, Ps, R, Sa

-

Bs, Fm

Bs Bs, Fm

m n

TABLE I (Continued) PhosDhorus soluble in: Country Hungary6 Irelandf Italy Netherlands

Total phosphorus R R Bs, R Bm, R

Ammonium citrate solutionc Water S S, Sb

-

Neutral

Fm, G, S A, Fc, Fm, Np, Sa, Pp, Tp

Alkaline

2% Citric acid

-

-

Fc, Fm

Bs Bs, Pp, R

-

A, Fc, Fm, A, Fc, Fm, Np, Np, Sa, S Sa, Pp, Tp Bs’ Np’ Fm, S Norway Bm, G, R Bs, Np, R NP7 RP R S Bs, Fm, Pp Polandd Bs,Pp Fm Bs,Fm Portugal Bin, Bs, R A, Fm, S R Fc, Fm, S Spain Al, Bm, Fc, Fm, Pp, S Bs Swedend F Fm, S Fm, Np, S Bs, R Fc, Fm, S Bm, Fm, G, R Switzerland PP, TP Bs United Kingdom Bm, Bs, G, Pp, Sb Bs Bd, Fc, Fm, S Bs, R S Fm, S Bs Y ugoslaviad a Updated and expanded Table XI1 from Jacob and Hill (1953). Sources of information: Association of Official Agricultural Chemists (1960), Organization for European Economic Cooperation (1952), official publications of the countries, and private cornmunications with officials in the governments and private industry. b The materials are identified as follows: A, ammonium phosphates; Al, aluminum phosphate; Bd, dissolved or vitriolized bones; Bm, bone meal; Bs, basic slag; F, all fertilizers; Fc, compound fertilizers; Fm, mixed fertilizers; G, guano; K, potassium phosphate; Mp, ammoniated magnesium phosphate; Np, nitric acid-phosphate rock products; Pp, precipitated calcium phosphates; Ps, phosphate rock-magnesium silicate glass; R, raw mineral phosphates for direct application to the soil; S, superphosphates; Sa, ammoniated superphosphates; Sb, basic superphosphate; and Tp, products of mixkures of mineral phosphates and alkali salts. c Includes water-soluble phosphorus. d Not included in survey of Jacob and Hill ( 1953). e Guarantee is not required for guano and rock. f Changed since survey of Jacob and Hill ( 1953). 9 Permitted but not required.

z‘

E

!: B

i

m

2 2

3

i!

6 B

66

G. L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

directly in the combined water-soluble and citrate-soluble extracts. In the official method as usually carried out, water-soluble phosphorus content is not determined separately, but is included with the citrate-soluble fraction as available phosphorus, The water-soluble phosphorus content is not a part of the legal description in any state. 111. Chemical and Physical Nature of Fertilizers Marketed in the United States

Prior to 1930 the principal solubilized phosphate fertilizers were ordinary superphosphate ( 16 to 20 per cent Pz05),triple superphosphate (43 to 50 per cent P205),and ammonium phosphates. The chief purpose of analyzing for the content of available phosphorus in fertilizers prior to 1930 was to determine how completely the raw phosphate rock had been acidulated. About 1928, manufacturers of mixed fertilizer initiated the process of ammoniating superphosphate, which has now become almost a universal practice in the United States. Keenan (1930) found that ammoniation produced a number of phase changes in the phosphorus components. Ross and Jacob (1931) noted that absorption of 2' per cent ammonia by ordinary superphosphate did not materially reduce the citrate solubility but that 6 per cent ammonia caused the content of citrate-insoluble PzO5 to approach or even exceed 6 per cent P205. [Note: Degree of ammoniation is expressed in this chapter as the pounds of free ammonia per unit (20 pounds) of AOAC available P205. In the case of superphosphate containing 20 per cent available PzOs, percentage of ammonia (NHs) or nitrogen ( N ) in the fertilizer (frequently termed per cent ammoniation) is equal to pounds per unit of available P205.] When not carried to excess, ammoniation is a very useful and desirable process in fertilizer formulation because it neutralizes some of the free acid usually associated with superphosphate, improves the physical condition of the fertilizer, enhances granulation, and incorporates nitrogen into the fertilizers from the cheapest possible source without excessive dilution of the phosphate. Because of the lower cost of nitrogen in solutions, superphosphates are commonly ammoniated to the highest practical degree, which is the point beyond which losses of ammonia become excessive. In the first stage of ammoniation, the monocalcium phosphate is converted to watersoluble ammonium phosphates and water-insoluble basic calcium phosphates. Anhydrous dicalcium phosphate can form if temperatures rise appreciably during ammoniation. With higher ammoniation rates, the

67

"AVAILABLE" PHOSPHORUS IN FERTILIZERS

change in phase composition is quite different for ordinary than for concentrated superphosphates. With ordinary superphosphate, water solubility of the phosphorus decreases progressively with increase in degree of ammoniation to the practical maximum of about 7 pounds of ammonia Der unit of Pz05. With concentrated superphosphate, water solubility A

'OOr-----l 90

80

w

ORDINARY SUPERPHOSPHATE

J

m

a

a3

70

Y

0

z

60

0 "

nN

y

so

m

3

A 0

u)

c;

40

CON C ENT R AT SUPERPHOSPH

w I-

s 30

20

10 0

I

2 3 4 5 6 LB. OF FREE NHs / 20 LB. AVAIL. P 2 0 ~

7

FIG. 1. Effect of degree of ammoniation on water solubility of phosphorus in ordinary (normal) and concentrated superphosphates. (Adapted from Hignett, 1956.)

decreases with increase in degree of ammoniation to a minimum of about 50 per cent of the total at 3 to 4 pounds of ammonia per unit of Pz05. With further ammoniation to the practical maximum of about 5 pounds of ammonia, water solubility of the phosphorus increases, owing to the formation of a greater proportion of ammonium phosphates. Effects of

68

G. L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

ammoniation on the water solubility of the phosphorus in ordinary (OSP) and concentrated (CSP) superphosphates are shown in Fig. 1. Recently, Hignett and Brabson (1961) studied the reversion of available Pz05 in ordinary superphosphates as affected by ammoniation. Their results indicated that dissolution of the water-insoluble phosphorus in alkaline ammonium citrate is similar to that of mixtures of dicalcium phosphate and hydroxyapatite. Terman et al. (1962) presented the results of petrographic studies of fertilizers made by TVA indicating that monoammonium phosphate and basic calcium phosphates similar to apatites are the principal phosphate compounds in heavily ammoniated superphosphates and that dicalcium phosphate is a rather minor constituent. Unreacted apatites are also present. Monoammonium and diammonium phosphates are the principal phosphate constituents of ammonium phosphate nitrate and ammonium phosphate sulfate fertilizers. Dicalcium phosphates and monoammonium phosphate constitute the major phases in nitric phosphate fertilizers. In later X-ray identification work by Ando et al. (1964) quantitative estimates were made of the various phosphate compounds present in a series of commercial fertilizers. The results are shown in Table 11. AOAC citrate-soluble and citrate-insoluble apatites comprised 25 to 58 per cent of the phosphorus in ammoniated ordinary superphosphate-base NPK fertilizers. Monoammonium phosphate comprised 7 to 44 per cent of the phosphorus. Monocalcium phosphate was a major component in only 3 of 9 of these fertilizers. Of 4 ammonium phosphate fertilizers examined, monoammonium phosphate was the major phosphate phase in 11-48-0 and 13-13-13, and diammonium phosphate was the major phase in 16-48-0 and 18-46-0. Dicalcium phosphate and apatite were the major phases in a 20-10-0 nitric phosphate. Monoammonium and dicalcium phosphate were the major phases in a 20-20-0 nitric phosphate. As Fig. 2 shows, the trend in the production of ordinary, or normal (NSP), and enriched superphosphates was upward from 1943 to 1952 but has declined since 1952. Production of CSP rose from 1945 until 1961, but fell about 8 per cent in 1962. Production of ammonium phosphates has risen rapidly since 1958. Production of other phosphates has remained at a rather low level. Several surveys have been made to determine the water-soluble phosphorus contents of commercial fertilizers sold in the United States. Archer and Thomas (1956) found that the average water-soluble phosphorus in 250 samples of commercial fertilizers was 48 per cent of the available phosphorus content. In about 23 per cent of the samples, less than 40 per cent of phosphorus was water soluble.

TABLE 11 Phosphate Compounds Present in Commercial Fertilizersa Per cent of total P,O,b Sample no.

Grade

Ammoniated superphosphate B-4 4-12-12 B1-4 4-12-12 D-3 3-12-12 1-4 4-12-12 1-6 6-8-8 W-4 4-12-12 5-10-15 w-5 W-8 8-16-0 w-10 10-10-10 Ammonium phosphate A-11 1148-0 A-16 1648-0 C-13 13-13-13 18-46-0 U-18 Nitric phosphate R-20-10 R-20-20 a b 0

d

MonoMonoamcalcium monium phosphate phosphate 0 8 20 0 0 0 16 0 0

25 13 22 14

0 0 0 0

85 45

Apatite

Other Citrate (by differinsolubled ence)

MonopoDiamtassium monium Dicalcium phosphate phosphate phosphate

Citrate solublec

5 0 0 8 0 0 0

0 0 0 0 0 0 0 0 0

11 11 20 8 12 13 15 16 15

33 41 35 45 55 26 44 21 36

16 17 6 8 3 2 7 4 1

0 0 4 0

0 50 0 84

0 0 0

0 0 2 0

0 0

7 30 12

44 30

77 3

4 0

0

1 0

11 10 -8

25 23 21 6 15 18 l5

5 16 13

F

Mr

2 2v

B

9 2 w

M

8

1 v)

20-10-0 20-20-0

0 0

0 24

Data from Ando et al. ( 1964). Calculated from X-ray analyses except as noted otherwise. Difference of total apatite and citrate-insoluble apatite. Calculated from the results of chemical analysis.

0 0

0 0

73 76

24 0

1 0

2 0

70

G. L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

Clark and Hoffman ( 1952) reported the solubilities (see tabulation) of 92 samples of superphosphates and 420 samples of multinutrient (mixed) fertilizers marketed in the United States in 1949 and 1950. SHORT TONS P,O,-

1,500

-

1,000

I

I‘,

OTHER

1945

1950

1960

1955

FIG.2. Production of normal and enriched superphosphates (NSP and ENSP), concentrated superphosphate (CSP), ammonium phosphate ( AP), and other phosphatic fertilizers, 1943 to 1962. (From U. S. Department Agriculture, 1963.)

Fifteen per cent of the samples contained less than 30 per cent of the total phosphorus in water-soluble forms. Percentage of total P,O, Available Fertilizer Superphosphates Mixed goods: NP grades PK grades NPK grades

Water-soluble

Range

Mean

Range

Mean

87-100 77-100 81- 99 51-100

97 95 95 93

5693 24-93 9-87 3-93

82 69 55 45

Clark et al. (1960) made a similar survey of fertilizers marketed in 1955 and 1956. The solubilities of 103 samples of superphosphates and 488 samples of mixed fertilizers are tabulated. Twenty-five per cent of the fertilizers contained less than 30 per cent of the total phosphorus in water-soluble forms.

71

“AVAILABLE” PHOSPHORUS IN FERTILIZERS

Percentage of total P205 Available Fertilizer Superphosphates Mixed goods: NP grades PK grades NPK grades

Range 71-100 88- 99 93- 99 49-100

Water-soluble Mean 97 95 97 94

Range 34-94 43-93 40-84 1-96

Mean 82 66 72 41

These results indicate that the solubilities of fertilizers sold in these two periods averaged about the same. This is not surprising since, as indicated in Fig. 2, ordinary superphosphate was the dominant phosphate material used for mixed goods in both periods. Production of CSP was increasing durirg this period, but marked expansion in ammonium phosphate production had not yet begun. Water solubility of the phosphorus was higher in both periods in the north central and western than in the southeastern sections of the United States, probably because of the greater use of ordinary superphosphate in the southeast and the lower solubilities resulting from its ammoniation. As Fig. 3 shows, there has been a slight upward trend in the percentage of available phosphorus and a trend downward in the percentage of water-soluble phosphorus in fertilizers sold in the United States from 1880 to 1956. Rogers and Ensminger (1961) reported that about 75 per cent of analyzed samples of 4-10-7 fertilizer sold in Alabama contained less than 40 per cent of the available phosphorus in water-soluble form. Less than 20 per cent was water soluble in a third of the samples. Gilliam (1963) reported the solubilities of 157 samples of fertilizer sold in Mississippi during 1959 and 1960. Water-soluble P20s as a percentage of the available ranged from 4 to 97 per cent. Among fertilizer grades containing 10 per cent or less of PZO5,largely formulated with ammoniated ordinary superphosphate, none had a mean value of more than 43 per cent. Another important change in fertilizers sold in the United States, as well as in European countries, is in the extent of granulation. As discussed by Hignett ( 1963), granular fertilizers were commonplace in Great Britain, Sweden, Germany, and the Netherlands by 1950. In the United States, however, granulation of mixed fertilizers has reached commercial importance only since that time. Annual consumption of granular fertilizers in 1954 and 1955 was only about 9 per cent of all mixed fertilizers. In 1957 it was estimated that 24 per cent of all mixed fertilizers sold were granular. In 1963, probably more than half of all solid mixed fertilizers were in granular form. Farmers now have a strong

WATER-SOLUBLE

0A V A I L A B L E

P O R T I O N OF T O T A L P205

PORTION

OF T O T A L P205

W A T E R - S O L U B L E P O R T I O N OF A V A I L A B L E

100

P205

100

80 I-

z W

0

a

60

w

a 40

20

0 1880

1890

1900

1910

1920

1925

1930

1935

li

1949-50

80

60

40

20

0

1955-56

FIG. 3. Solubility and availability of phosphorus contsnt of solid commercial fertilizers for 1800 to 1956. (From Clark et al., 1960.)

73

“AVAILABLE” PHOSPHORUS IN FERTILIZERS

preference for granular fertilizers, and the conversion is nearly complete in some sections of the United States and Europe. This upward trend in the use of granular fertilizers in the United States is shown in Fig. 4. The importance of granule size in relation to water solubility of the phosphorus for the fertilized crop is stressed in Sections IV and V. Laborat

I950

MORE FERTILIZERS ARE GRANULAR

I955

I960

I

1963

FIG.4. Trend in use of total and granular mixed fertilizers in the United States. (TVA chart.)

tory analyses, however, cannot reflect these relationships since, in order to obtain reproducible results, each fertilizer sample is ground finely before analysis. IV. Crop Response Results Prior to 1950

Superphosphate has been more widely used as a source of phosphorus in mixed fertilizers than any other phosphatic material. From the time of initiation of its manufacture by Lawes in 1842 until about 1928, superphosphate was used as a straight material and in mixed fertilizers without further treatment. About this time, however, ammoniation as a formulation process made its appearance and was rapidly accepted. Gerlach‘s pot experiments in Germany (1916) were cited as evidence that ammoniation does not reduce phosphate effectiveness. He found that ammoniated superphosphate applied on the basis of water-soluble Pz05 was superior for oats to a mixture of nonammoniated superphosphate and ammonium salts. A fairly large number of experiments were conducted during 1930

74

G. L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

to 1932 to determine what changes in availability of the phosphorus to crops might occur as a result of ammoniation of ordinary superphosphate. Results from these experiments have been summarized recently by Wright et al. (1963). Buie (1931) summarized results from 17 field experiments conducted in 1930 comparing ammoniated ordinary superphosphate with nonammoniated superphosphate for cotton. He interpreted the results to indicate that ammoniated superphosphate was as effective as that without ammoniation; however, no details of the experiments or experimental fertilizers were given. Parker (1931) stated that ammoniation to the extent of 3 per cent ammonia in the product was the highest then being used by the fertilizer industry and that 4 per cent ammonia was “about the maximum possible under fertilizer plant practice.” With present technological skill, however, ammoniation to the extent of 6.5 pounds of nitrogen per unit of available PzOa is commonly achieved. Parker pointed out that fine particle size of the reverted phosphate in ammoniated superphosphate should enhance the dissolution of the phosphorus and, thereby, its effectiveness. He further noted that the acid produced during the nitrification of the ammonium salts in a mixture with slightly soluble calcium phosphates should solubilize the phosphorus and render it more available. As a result of the interest of the AOAC, an extensive cooperative pot study was conducted at several experiment stations. Ross and Jacob (1931) and Ross et al. (1932), in reporting the results of this research, concluded that ammoniated ordinary superphosphate was at least 90 per cent as effective as monocalcium phosphate in soils below pH 6.0, but was much less effective in soils at higher pH levels. Fertilizers used in these pot experiments were thoroughly mixed with the soils prior to planting the test crop, which greatly enhances the effectiveness of slightly soluble basic calcium phosphates in acid soils, as compared to band application. The authors chose to exclude the results of tests conducted on soils above pH 6.0, reasoning that 75 per cent of the fertilizer used in 1932 was applied to soils below pH 6.0. They recommended a change in the AOAC official method for determining “available” phosphorus which would give a higher rating to fertilizers containing ammoniated ordinary superphosphate. Williamson (1935) summarized 185 experiments with cotton on different soil types and with different fertilizer treatments. Relative yield increases from the various phosphorus sources over no applied phosphorus were rated as follows: ordinary superphosphate, 100; superphosphate ammoniated to 2, per cent ammonia, 100; and superphosphate

“AVAILABLE” PHOSPHORUS IN FERTILIZERS

75

ammoniated to 4 per cent ammonia, 90. Inclusion of lime in the complete fertilizer greatly reduced the effectiveness of the ammoniated ordinary superphosphate. In these experiments, the average increase due to superphosphate was only 241 pounds of seed cotton per acre. The 60 pounds of PzO5 applied per acre may have been more than adequate to produce maximum yields with the 36 pounds nitrogen applied per acre. As pointed out by Terman (1960, 196l), there is little chance of detecting differences among sources of phosphorus under such circumstances. Salter and Barnes (1935) related soil pH to the effectiveness of various phosphorus sources for wheat grown in the field and Sudangrass in the greenhouse. Their results clearly demonstrated that ammoniation of ordinary superphosphate above 3 per cent ammonia markedly decreased its effectiveness, and this decrease in effectiveness became more pronounced as the soil pH increased from 5.5 to 7.0. At pH 7 the relative increase in yields over no applied phosphorus for superphosphates ammoniated to 0, 2.9, 5.4, and 7.1 per cent ammonia were 100, 72, 50, and 23, respectively, whereas on a soil at pH 6.0, the relative yield increases for these same fertilizers were 100, 87, 86, and 77, respectively. Gilbert and Pember (1936) concluded that for oats, barley, and millet grown in greenhouse pots, the phosphorus in superphosphate ammoniated to more than 4 per cent ammonia was always less effective than in those ammoniated to less than 4 per cent. The fertilizers in these tests were mixed thoroughly with acid soils, a procedure which, as pointed out above, tends to enhance the efficiency of basic calcium phosphates. Andrews (1942) reviewed the question of ammoniated superphosphates with particular reference to the AOAC official method for determining “available” phosphorus in fertilizers, He concluded that ( 1 ) superphosphate ammoniated to 2.4 per cent ammonia is a good source of phosphorus in acid-forming fertilizers on acid soils, and ( 2 ) superphosphate ammoniated to 3.0 per cent ammonia or higher is less valuable than nonammoniated superphosphate in acid-forming fertilizers, and much less valuable in neutral fertilizers. A further conclusion was that the AOAC official method for determining the availability of phosphate in regard to crop response gives too high a rating to basic calcium phosphates occurring in mixed fertilizers and to ammoniated ordinary superphosphate that contain as much as 3 per cent ammonia. Ross and associates (1947) reported results from pot experiments with fertilizers containing ammoniated superphosphates that had been subjected to varying storage conditions. The average relative yield increase over no applied phosphorus from fertilizers ammoniated to 0, 2, 3, 4, and 5 per cent ammonia were 100, 100, 88, 82, and 85, respectively,

76

G . L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

on acid soils and 100, 63, 69, 38, and 5, respectively, on a Houston clay (pH 8.2). The results of these experiments further showed that inclusion of dolomitic limestone in the fertilizer formulation greatly reduced the effectiveness of ammoniated superphosphates and that the AOAC method gives much too high a rating to superphosphate mixtures ammoniated higher than 2 per cent ammonia. They concluded that for superphosphates ammoniated more heavily than 2 per cent, “the official method gives higher availability values than those indicated by the pot tests for all crops, and that the spread between the two methods increases with increase in ammoniation of the mixture.” Rogers et al. (1953) concluded from their review of earlier results that ammoniation of ordinary superphosphate higher than 2 per cent ammonia results in a relatively small, but consistent, decrease in the effectiveness of the phosphorus. Several criticisms can be made regarding many of the field and greenhouse experiments just reviewed. Specifically, many of the experiments were conducted with finely divided, mixed-salt-type fertilizers that were mixed throughout the soil. On acid soils these conditions enhance the effectiveness of slightly soluble calcium phosphates and decrease the effectiveness of soluble phosphates such as superphosphate, so that the effectiveness of the fertilizer-soil reaction products resulting from slightly soluble compounds is maximal and that of the soluble compounds is minimal. Furthermore, many of the experiments were conducted at low yieId levels, and yield increases resulting from added phosphates were small. In some experiments the phosphorus sources were tested at rates which fell on the flat portion of the response curve. These conditions make it impossible to detect anything but gross differences among phosphate sources ( Terman, 1960, 1961 ). Finally, relative yields or relative increases in yield were used almost exclusively to evaluate the experimental fertilizers which, because yields may not be linearly related to the rate of fertilizer applied, is an unsuitable method to use in comparing similar fertilizers (Cooke, 1956; Cooke and Widdowson, 1959). Thus, when judged by present-day knowledge and standards, many of the older field experiments did not evaluate phosphorus sources very precisely. From this review, it may be concluded that ammoniation of ordinary superphosphate decreases the fertilizer efficiency of the phosphorus, and the loss of efficiency becomes greater as the degree of ammoniation increases. This effect is less pronounced in acid soils when the fertilizer is mixed throughout the soil but is outstanding in neutral or calcareous soils. Ammoniated ordinary superphosphate in mixed fertilizers containing lime is much less effective than in acid-forming fertilizers.

“AVAILABLE” PHOSPHORUS IN FERTILIZERS

77

V. Recent Crop Response Results

A source of confusion in the interpretation of results from water solubility-granule size and placement experiments with phosphates has been the difference often found between early growth response and final yields of grain or hay crops under field conditions. Most greenhouse pot experiments are conducted for two months or less, and the growth responses obtained are analogous to the early growth response in field experiments. As indicated below, marked increases in early growth response are usually found on phophorus-responsive soils with increase in water solubility of the phosphorus, with increase in granule size of highly water-soluble sources, with decrease in granule size of sources having a low water solubility, and with band placement as compared to mixing with the soil. Whether these early growth responses follow through to final yields of forage, fiber, or grain depends on numerous factors. These include the level of plant-available phosphorus in the soil, soil reaction, adequacy of supplies of other nutrients, moisture supply, kind of crop, length of season, and others. As the season progresses, the plants draw increasingly from soil phosphorus and less from the fertilizer applied for that particular crop. Final yield differences related to water solubility of the phosphorus occur more frequently with potatoes and other relatively short-season crops than with longer season crops such as corn, cotton, small grains, and forage species. Even though early growth response by a crop may not be reflected in final yields, such “starter effects” are still quite important in enabling a row crop to push ahead of weed growth and in providing a large area of leaf growth for rapid photosynthesis. Early, vigorous growth is particularly important for early market of vegetable crops and frequently results in higher profits. In addition to possible effects on yields and profits, early, vigorous growth has an intangible esthetic value to most farmers, an important point in selling fertilizers. Eight agronomists all indicated in a survey article (Webb et al., 1959) that water solubility of the phosphorus was important in various areas of the United States under certain specified conditions. In a second survey (Thomas, 1959), the majority of the agronomists contacted in 18 states indicated that fertilizers containing 40 to 60 per cent of their phosphorus in water-soluble forms were satisfactory for most soil and crop conditions. Seatz and Stanberry (1963) reviewed recent literature concerning the complex relationships among soil and fertilizer composition, granule

78

G. L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

size, and phosphorus water-solubility effects on crop response. Mattingly ( 1963) reviewed recent research on water- and citrate-soluble fertilizers at Rothamsted and elsewhere. Van Burg (1963) similarly reviewed results obtained in the Netherlands.

A. WATERSOLUBILITY OF THE PHOSPHORUS AND GRANULE SIZE EFFECTS

1. Greenhouse Pot Experiments Martin et al. (1963) reported that ammoniation of ordinary superphosphate to 4.5 per cent reduced the water solubility of the phophorus, but had no effect on yields of lettuce grown on four acid California soils. Yields were reduced markedly on two calcareous soils, however, by high ammoniation and the resulting reduction in water solubility. Lawton et al. (1956) found that crop response was closely related to the content of water-soluble phosphorus in granular 12-1212 fertilizers mixed with the soil or in banded pulverant fertilizers. If the pulverant fertilizer was mixed with the soil, however, crop response was not related to phosphorus water solubility. These fertilizers were prepared from slurries of ammonium nitrate, ammonium phosphate, dicalcium phosphate, and potassium chloride. Terman et al. (1956) similarly found that early growth response to phosphorus, as exhibited by oats and Sudangrass grown in greenhouse pots, increased with increase in granule size of NPK fertilizers high in water-soluble phosphorus (prepared with diammonium phosphate and ammoniated concentrated superphosphate). Early response also increased with decrease in size of granules low in water-soluble phosphorus (prepared with dicalcium phosphate or ammoniated ordinary superphosphate). In other greenhouse pot experiments, Terman et al. (1960) found that crop response to band-applied nonammoniated and ammoniated superphosphates was closely related to their content of water-soluble phosphorus. Other than with dicalcium phosphate, granule size was of little importance when the fertilizers were banded. With the phosphates mixed through the soil, both water solubility and granule size greatly influenced yields on most soils. Decrease in response with time of reaction with the soil prior to cropping was much less with granular than with fine superphosphates and dicalcium phosphate. Bouldin et al. (1960) studied the response of oats in greenhouse pot experiments to various granule sizes of monoammonium phosphate (water soluble) and dicalcium phosphate (low water solubility) and to mixtures of the two, which are commonly found in some ammoniated superphosphates and in nitric phosphates. The important fertilizer properties affecting response were found to be geometric surface area of

“AVAILABLE” PHOSPHORUS IN FERTILIZERS

79

the granules of dicalcium phosphate and water-soluble phosphorus content as monoammonium phosphate per granule. Granule size effects of mixtures of the two phosphates were intermediate between those of the single components, and there was no measurable interaction between them. With decrease in granule size, response to both phosphates approached a common level, indicating rather complete reaction of the fine phosphates with the soil.

2. Field Experiments Terman et al. (1956) reported that the early growth response of wheat forage and other crops grown in field experiments increased with increase in granule size of water-soluble phosphorus fertilizers and with decrease in granule size of those low in water solubility. These early responses did not persist in final yields of corn or wheat grain and of vegetables in most experiments. DeMent and Seatz (1956) similarly found that high-alumina nitric phosphates having higher phosphorus water solubility were more effective than phosphates lower in solubility for increasing yields of wheat or oat forage and as starter fertilizers for corn. The degree of water solubility was of minor importance in final yields of long-season crops grown in southeastern United States. Jordan (1964) also noted that the most consistent effects of water-soluble phosphorus content, granule size or placement of fertilizers occur in the early growth stages of crops grown in this region. In a series of experiments with hill-placed phosphates for corn in Iowa, Webb and Pesek (1958) reported that all 20 experiments showed consistent trends toward larger yield increases with increasing water solubility of the phosphorus. About 90 per cent of the yield increase attributable to water solubility was attained with fertilizers having 60 per cent of the phosphorus in water-soluble form. Early season growth response was correlated very closely with amount and water solubility of the applied phosphorus. With broadcast application of phosphorus for corn, however, Webb and Pesek (1959) found that yields increased with amount of phosphorus applied but were not affected by water solubility of the phosphates. Sources very low in water solubility tended to be less effective in a few experiments. For corn grown on calcareous soils, Webb et al. (1961b) concluded that highly water-soluble phosphates applied broadcast were more effective than most slightly soluble sources. Increasing the granule size of less soluble sources tended to reduce their effectiveness. Webb et al. (1961a) found that water solubility of the phosphorus was an important factor in effectiveness of phosphates applied for oats grown on calcareous soils. On acid soils, placement of the phosphates

80

G. L. TERMAN, W. M. HOFFMAN, AND B. C . WRIGHT

was more important than water solubility. Drilling the fertilizer with oat seed was significantly superior to broadcasting on acid soils but only slightly so on calcareous soils. A comprehensive investigation was carried out in Mississippi by Wright et al. (1963) to measure the effectiveness of ammoniated ordinary superphosphates as sources of phosphorus for corn, cotton, and wheat. Results were expressed in terms of superphosphate equivalents, i.e., the I20

I

I

I

e- WHEAT

1 0

I

I

I

I

I

I

I

I

I

I

FORAQE

I

I

I 2 3 4 5 6 7 AMMONIATION. LB. NH3 PER 2 0 LB. AVAILABLE PgOs

I

FIG.5. Percentage superphosphate equivalents for various crops, as affected by degree of ammoniation of ordinary superphosphate. (From Wright et al., 1963.)

amount of phosphorus in noammoniated 20 per cent superphosphate expressed as a percentage of the amounts of phosphorus in ammoniated superphosphates (or other phosphorus sources) required to produce the same yield of crop. A response curve obtained with several rates of superphosphate was included in all experiments. As an average for the 43 field experiments conducted, effectiveness of the ammoniated superphosphates decreased as follows: Degree of ammoniation: Pounds of NH, per 20 pounds of P,O, Per cent superphosphate equivalents

0 100

2.1 85

4.2 67

6.5 39

7.2 28

Results obtained with individual crops are shown in Fig. 5. Decreases in

81

“AVADLABLE”PHOSPHORUS IN FERTILIZERS

effectiveness found in the field experiments with increase in degree of ammoniation agree quite closely with results obtained in greenhouse pot experiments with a similar series of ammoniated ordinary superphosphates (Terman et ul., 1960). No evidence was found in the Mississippi study that corn, cotton, and wheat differed appreciably in their relative responses to ammoniated superphosphates. Granule size of the fertilizers was of minor importance in these experiments, in all of which the phosphates were band-applied. As shown in Fig. 6, AOAC-available

I

0

I

I 2 AMMONIATION,

I

I

I

I

I

3

4 I 6 7 LB. NHs PER 20 LBS. AVAILABLE PzOs

FIG.6. Effectiveness of ammoniated ordinary superphosphates, as related to dissolution in neutral ammonium citrate ( AOAC), alkaline ammonium citrate ( NAAC) , and water. (From Wright et al., 1963.)

phosphorus content of the test fertilizers decreased only slightly with increase in ammoniation, which resulted in a very low correlation with crop response, Decreases in solubility with increase in ammoniation, both in water and in alkaline citrate solution (NAAC method), however, were highly correlated with crop response. In a series of experiments comparing sources and rates of phosphorus for vegetable crops in western Washington during the 1951 to 1959 period (Mortensen et ul., 1964), the importance of water solubility for various crops decreased as follows: cucumbers > pole beans > potatoes > sweet corn. At a given rate of applied phosphate fertilizer, vegetable yields increased with content of water-soluble phosphorus. In some experiments limiting yields of the crops obtained with increasing rates of

a2

G. L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

applied phosphorus were higher with sources high in water solubility (Fig. 7 ) . Contents of phosphorus in the leaves during the early growth period were rather closely correlated with water solubility of the fertilizers and with final crop yields. Lewis (1962) grew snap beans in central Maine for two seasons with ammoniated superphosphates varying in content of water-soluble phosCUCUMBERS

8

6

4l

4 W

K

22

i,"i

0-DAP 0 - CSP A - OSP X DCPO 0 DCPA

-

2

-

0

!:F 2t w

20

4

POLE BEANS

-

0

POTATOES

I

0

50

0 50 100 I50 P 2 0 APPLIED ~ LB. PER ACRE

-

I

I

100

150

I

FIG.7. Yield response of vegetables to sources and rates of phosphorus in westem Washington, 1959. DAP (diammonium phosphate); CSP and OSP (concentrated and ordinary superphosphates) ; DCPD and DCPA ( dicalcium phosphate, dihydrate and anhydrate). (From Mortensen et al., 1964.)

phorus. Early growth response increased in all experiments with increase in phosphorus water solubility. Only in one experiment on a soil low in soluble phosphorus, however, did water solubility affect the yield of harvested product. Lingle ( 1960) obtained similar results with tomatoes. Yields of the first harvest were higher with fertilizers higher in phosphorus water solubility, but this advantage had disappeared by the time of the second harvest. Van Burg (1963) concluded that 50 per cent water solubility of the phosphorus was adequate in fertilizers for cereals and grassland in the Netherlands but that potatoes required a water solubility close to 100 per cent.

“AVAILABLE” PHOSPHORUS IN FERTILIZERS

83

3. Immediate us. Residual Efectiveness Several investigators ( Cooke, 1956; Ensminger and Pearson, 1957; Schmehl et al., 1955; Nelson and Stanford, 1958; Mattingly, 1963) have reported similar residual effects from acidulated phosphates differing widely in initial solubility and effect on the immediate crops. Responses by corn, millet, oats, and wheat (Terman et al., 1961b) increased in a similar manner at all seasons of the year with increase in the water-soluble phosphorus content of applied phosphate fertilizers. Banding resulted in greater response by corn than mixing of the phosphates with the soil just prior to planting. Effects of placement and of water solubility were less for a second corn crop, and there were no appreciable differences in response by a third crop to these variables. Residual phosphates have accumulated in large acreages of soils in Europe and the United States as a result of fertilizer applications continued for many years. On such soils, little or no yield response to phosphorus by the fertilized crop is obtained, and the chief purpose of continued applications of phosphorus fertilizers is to maintain the soluble soil phosphorus at a high level. For such maintenance applications any acidulated phosphate which will react with the soil is considered by Cooke (1963) to be a satisfactory fertilizer. Thus, as more soils reach rather high levels of residual phosphorus available to crops, the need for fertilizers having a high proportion of their phosphorus in water-soluble forms may decrease. B. QUALITY OF THE WATER-INSOLUBLE PHOSPHATE FRACTIONS OF

FERTILIZERS

Bouldin and Sample (1959) found a direct relationship between plant response and the geometric surface area of granules per unit of phosphorus in dicalcium phosphates. Size of crystals composing the granules was of much less importance than granule surface area. Terman et al. (1961a) found a similar relationship between crop response and granule surface area of a series of water-insoluble phosphates prepared by water leaching of ammoniated superphosphates and nitric phosphates. Fine granules and particles of these phosphates, however, tended to dissolve in and react with the soil, so that the plants obtained much of their phosphorus from the fertilizer-soil reaction products. In a later experiment, Bouldin and Sample (1963) found that plant response to phosphorus was correlated with geometric surface area of granular fertilizers prepared from mixtures of dicalcium phosphate with either glass beads or several nonphosphatic salts. Calcium phosphates more basic than

84

G . L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

dicalcium phosphate, which occur in heavily ammoniated ordinary superphosphates (AOSP) and nitric phosphates (NP), are less effective sources of phosphorus for crops than dicalcium phosphate (Beaton and Gough, 1962.;Rogers et al., 1953). Terman et al. (1961a) prepared a series of water-insoluble phosphate fractions varying in AOAC citrate solubility (neutral ammonium citrate) TABLE I11 Chemical Analyses of Water-Leached Phosphate Fractionsa AOAC fractions Water-leached residue fromb (1) DCP ( 2 ) NP (3) NP ( 4 ) ACSP ( 5 ) AOSP ( 6 ) AOSP ( 7 ) AOSP ( 8 ) P-4 ( 9 ) P-1 (10) P-2 (11) P-3 (12) AOSP (13) AOSP

TVA fertilizer number 255L 76L 151L 253L 248L 249L 259L 4L 1L 2L 3L 302L 304L

Total P,O, (%)

44.0 45.1 42.1 34.8 29.3 34.1 35.5 42.4 43.0 40.1 38.4 17.9 29.1

NAACc available Avail- ( % of total able P,O,)

( % of total P,O,) Water soluble

Citrate soluble

1 1 2 2 3 2 2 <1 <1

98 92 83 70 44 37 41 53 51 31 23 74 51

<1 <1 1 1

99 93 85 72 47 39 43 53

51 31 23 75 52

71 77 67 51 31 16 18 52 30 9 2 67 25

Data largely from Terman et al. (1961a). DCP, feed-grade anhydrous dicalcium phosphate; NP, nitric phosphate; ACSP, ammoniated concentrated superphosphate; AOSP, ammoniated ordinary superphosphate; P1-4, heavily ammoniated nitric phosphate residues; 1g. = sample size. c Netherlands alkaline ammonium citrate method. a b

from 23 to 99 per cent of the total phosphorus content by leaching AOSP and NP fertilizers with water. Response by two successive crops of corn was found to be rather closely related to the fractions of their total phosphorus content which dissolved in a neutral citrate solution. Samples of these water-insoluble fractions were also extracted by the Netherlands alkaline ammonium citrate (NAAC) method (Organization for European Economic Cooperation, 1952,). Results of chemical analyses are shown in Table 111. Plots of the relative effectiveness of phosphorus per unit of surface area of -16+20 mesh granules of six of these phosphates against percentage of phosphorus dissolved by the AOAC and NAAC

“AVAILABLE” PHOSPHORUS IN

85

FERTILIZERS

methods are shown in Fig. 8. Effectiveness of the phosphates increased with degree of dissolution in both extractants. Dissolution of all of the phosphates was lower in alkaline citrate. Results shown in Fig. 9 indicate that dissolution by the AOAC and NAAC methods (1.0-g, samples) of a number of water-insoluble phosphate fractions of AOSP and NP fertilizers and feed-grade dicalcium phosphate (DCP) are essentially linear and highly correlated for most of

l o ‘

d

? v) v)

W 2 W

-wc 0 W LL LL W

W

? c a J

W

K

01

I

I

I

1

I

I

the fractions, At zero dissolution in alkaline citrate, 20 per cent or more of the phosphorus is soluble in neutral citrate. Percentage dissolution of phosphorus in several of the original fertilizers (from which the waterinsoluble fractions were obtained) are also shown in Fig. 9. Amounts of P dissolved from all these ammoniated superphosphates were high by the AOAC method but varied widely by the NAAC method. Hignett and Brabson (1961) reported similar results with a series of 6-12-12 fertilizers formulated with ammoniated ordinary superphosphate. As shown in Fig. 10, they found that amounts of phosphorus dissolved by the NAAC method decreased with increase in degree of ammoniation of the super-

86

G . L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

phosphate, but that phosphorus in all the fertilizers was highly soluble by the AOAC method. Actually, after subtraction of the water-soluble phosphorus, a similar amount of phosphorus was dissolved by the alkaline citrate and a variable amount by the neutral citrate. The sum of the water-soluble and citrate-soluble phosphorus contents thus leads to decreasing amounts of “available PzO{ by the NAAC methods, but not with the AOAC method. The results presented in Fig. 9 suggest that the 100

I

I

I

I

J

a. I- 80

e

LL

0

s ;60 0

WATER- INSOLUBLE P FRACTIONS OF AOSP, NP AND D C P

C

w J m

2 40 a

3-12-12 AND 6 - 12-12 FERTILIZER

> a 0

9z

20

0 i

40 60 80 I00 AOAC AVAILABLE P 2 0 ~ % - OF TOTAL

FIG. 9. Relation between AOAC- and NAAC-available phosphorus in waterinsoluble phosphate fractions (Table 111) and in NPK fertilizers (Table IV) prepared with ammoniated ordinary superphosphate.

amount of phosphorus per sample had marked effects on the percentages of the phosphorus dissolved by the neutral and alkaline citrate solutions. Such effects of sample size were investigated further with 7 of the AOAC water-insoluble phosphate fractions ( TVA data obtained by D. R. Bouldin) (Fig. 11).With 1.0-g. samples, the relationship between amounts of phosphorus dissolved by the AOAC and NAAC methods is essentially the same as shown in Fig. 9. With decrease in sample size (0.5, 0.25, and 0.10 g.), percentages of the total phosphorus which dissolved in both extractants increased from all of the phosphates. The AOAC extraction dissolved 88 to 100 per cent and the NAAC extraction

87

“AVAILABLE” PHOSPHORUS IN FERTILIZERS

63 to 100 per cent of the phosphorus in 0.10-g. samples. The actual ranges of amounts of water-insoluble Pz06per sample were as follows: 1.0-g. samples: 0.5-g. samples: 0.25-g. samples: 0.10-g. samples:

0.18 to 0.45 g. 0.09 to 0.225 g. 0.045 to 0.113 g. 0.018 to 0.045 g.

In the case of the original 6-12-12 or 3-12-12, fertilizers prepared

a

t0

t-

IL

0

F 2

K W

n

0

AVAILABLE P 2 0 ~

A . 0 A.C

N A A C A V A I L A B L E P2Os

A WATER- SOLUBLE PzOs n

1

0

I

2 4 6 A M M O N I A T I O N OF ORDINARY SUPERPHOSPHATE IN 6-12-12 F E R T I L I Z E R , LB. N H s / U N I T AVAILABLE PzOs

8

FIG. 10. AOAC-available, NAAC-available, and water-soluble phosphorus content of ordinary superphosphate-base NPK fertilizers, as affected by degree of ammoniation. (From Hignett and Brabson, 1961.)

88

G. L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

0" &I00

5

-

I

I

I

80-

I-

60-

f

40-

% a

20-

I

..

-1

%

I

I .O g. SAMPLES

w

t 0 w

---

I

I

1

I

0 . 5 0 . SAMPLES

I

-

.*

-

--

-

--

-

--

-

'

I I

o

8o

I I

I I

1

I

0.25g. SAMPLES

100-

I

--

I

I

1

I

I

I

I I

t

t

1

.

I

0.10 g. SAMPLES

* *

1

20 '0

20 4 0 60 80 100 0 20 40 60 80 100 PzOs DISSOLVED BY NEUTRAL C I T R A T E - % OF T O T A L PpO,

FIG.11. P,O, in 7 water-insoluble phosphates dissolved by alkaline and neutral ammonium citrate, as affected by sample size. (TVA data obtained by D. R. Bouldin.)

TABLE 1V Chemical Analyses of Ammoniated Ordinary Superphosphate-Base 6-12-12 and 3-12-12 Fertilizersa Per cent of Total P,O, Fertilizer grade 6-12-12

3-12-12 @

Pounds NH, TVA per unit Total of avail- P,O, fertilizer (%) number able P,O,

AOAC method

302A 303A 199A 304A 305A 300A 307A

2.0 4.1 6.5 7.2 5.5 5.7 5.8

13.5 13.4 13.9 13.3 13.2 13.6 13.1

56 35 25 10 31 30 23

38 57 64 80

301A 306A

4.1 6.9

13.4 12.0

46 13

TVA data.

NAAC method

Water Citrate Avail- Citrate Availsoluble soluble able soluble able

66 73

94 92 89 90 96 96 96

31 30 24 28 27 26 29

87 65 49 38 58 56 52

50 82

96 95

32 27

78 40

65

89

“AVAILABLE” PHOSPHORUS IN FERTILIZERS

with ammoniated ordinary superphosphates (Table IV), 89 to 96 per cent of the total phosphorus was dissolved by the AOAC method (1.0-g. samples, water followed by neutral citrate) and 38 to 87 per cent by the NAAC method (2.0-g. samples, water followed by alkaline citrate). The ranges of amount of water-insoluble PZO5in these fertilizers per sample were 0.057 to 0.105 g. with the AOAC method and 0.114 to 0.210 g. with

01 0

I I I I I00 200 300 400 WATER- INSOLUBLE PzOe YO. PER SAMPLE

I

-

FIG. 12. Phosphorus dissolved from water-insoluble phosphate fractions by neutral and alkaline ammonium citrate, as affected by amount of water-insoluble P,05 per sample (0.10, 0.25, 0.5, and l . 0 g . of material). (TVA data obtained by D. R. Bouldin. )

the NAAC method. Thus the results shown in Figs. 9 and 11 are strongly influenced by the actual amount of water-insoluble P in the samples being extracted, as well as by the extractant. This is illustrated by Fig. 12, in which the percentage of the total Pz06 dissolved is plotted against the amount of water-insoluble PZO5per sample. Percentages of the phosphorus in both water-insoluble phosphate fractions (Table 111) decreased with increase in amount of water-insoluble phosphorus in the sample, the ammoniated ordinary superphosphate fraction ( AOSP, No. 6 ) much more so than the nitric phosphate fraction (NP, No. 3). Similar results were obtained with 5 other water-insoluble fractions.

90

G. L. TERMAN,

W.

M. HOFFMAN, AND B. C. WRIGHT

Because of the small amounts of water-insoluble PzO5 in the 1.0-g. samples of NPK fertilizers formulated with ammoniated ordinary superphosphate, 89 per cent or more was dissolved by the AOAC method. With 2.0-g. samples of these fertilizers extracted by the NAAC method, however, the percentage dissolutions were much less. As shown in Fig. 13, dissolution of the various fertilizers by the latter method de-

I

I

I

I

I

FIG. 13. Phosphorus dissolved from 3-12-12 and 6-12-12 fertilizers formulated with ammoniated ordinary superphosphates in relation to the amount of water-insoluble P,O, per sample. (TVA data.)

creased linearly with increase in amount of water-insoluble phosphorus in the sample. With constant amounts of Pz06 in the sample, percentage dissolution is somewhat different than with a constant 1.0-g. sample (AOAC) or 2.0-g. sample (NAAC). These results indicate that it would be desirable to standardize the amount of water-insoluble P206 per sample extracted by both the AOAC and NAAC methods for the values to be very meaningful. It is important to note, however, that with a constant sample weight of such fertilizers the amount of water-insoluble P205 per sample increases with degree of ammoniation. The relative abundance of basic phosphates of low solubility also increases with

“AVAILABLE” PHOSPHORUS IN FERTILIZERS

91

increasing ammoniation. Thus, it is difficult to separate cause and effect in terms of dissolution, as affected by size of sample and by differences in amounts of water-insoluble Pz05 in different samples. Haskins (1921) found that dissolution of a 1.0-g. sample of dicalcium phosphate in neutral ammonium citrate more nearly reflected its fertilizer value than dissolution of a 2.0-g. sample. Ross and Jacob (1931) noted a decrease in citrate-insoluble Pz05 in ammoniated superphosphates with decrease in sample size from 2.0 to 0.5 g. Howes and Jacobs (1931) found that citrate-soluble PzOB in ammoniated superphosphates (3.5 to 6.0 per cent NHB) increased greatly with decrease in size of sample from 2.0 to 1.0 g., with decrease in pH of the extractant, with increase in amount of solution, and with increase in time of digestion. Many other investigators have noted these effects since then. For example, Brosheer (1953) studied the dissolution in neutral ammonium citrate of precipitates formed during ammoniation of nitric acid extracts of phosphate rock. He found that the amount of phosphorus dissolved from 1.0-g. samples containing 0.3 to 0.5 g. of water-insoluble PzO, increased irregularly from 72 to 95 per cent of the total P20, content with decrease in estimated content of apatite from 90 to 5 per cent. However, from 97 to 100 per cent of the phosphorus in 0.25-g. samples was dissolved and dissolution was not reIated to content of apatite. In connection with the increase in dissolution with decrease in size of sample, it is of interest to note that a change in the AOAC procedure recommended by Ross et al. (1932) for use of a 2.0-g. sample when the fertilizer contained 10 per cent or less available Pz05 and a 1.0-g. sample when the fertilizer contained more than 10 per cent available Pz05 was not approved. Instead, the AOAC Subcommittee A on Recommendations of Referees recommended that a 1.0-g. sample be taken for all fertilizers regardless of the available P205 content. This latter recommendation was later incorporated into the official method. In another study carried out by TVA, NH4N03 and KCl were granulated with phosphate rock of two degrees of fineness. Results with the AOAC method, as shown in Fig. 14, indicated that the presence of the salts had increased the availability of the rock. However, analysis of the same amount of rock in samples without the salts indicated similar availabilities. Thus, decreasing the size of sample of phosphate increases its availability, as shown by the official method. More phosphorus was dissolved by the neutral ammonium citrate from the finer than from the coarser ground rock. Results from greenhouse pot tests with oats, how-

92

G. L. TERMAN, W. M. HOFFMAN, AND B. C.

WRIGHT

ever, failed to show any increase in plant utilization of phosphorus in the phosphate rock due to the presence of NH4N03or KC1. Increased dissolution of phosphorus by ammonium citrate with decrease in content of water-insoluble basic calcium phosphates per sample probably results largely from the corresponding decrease in content of calcium. Phosphorus dissolution is controlled largely by the

P205 ADDED, MO. PER 100 ML. OF CITRATE SOLUTION

FIG.14. Phosphorus dissolved by neutral ammonium citrate from phosphate rock in mixtures containing NH,N03 and KCl. (TVA data obtained by D. W. Rindt. )

extent to which calcium is complexed by the citrate. Rate of solution of the calcium phosphates also decreases with increasing size of the component crystals and content of apatite. Robertson (1914) described the citric acid test used in some countries (Table I ) as a test for calcium rather than for phosphorus. He showed that a second extraction of a mineral phosphate might dissolve more phosphorus than the first extraction. That is, the “citrate insoluble” residue was more “soluble” than the original material. Rosanow (1934) showed that the “‘insoluble” residues of certain mineral phosphates

“AVAILABLE”

PHOSPHORUS IN FERTILIZERS

93

extracted with citric acid were more active as fertilizers in pot experiments than the untreated phosphates. The removal of calcium as a result of complexing by citric acid thus increased the availability of the phosphate. These results indicate that estimates of “available” phosphorus, as determined by extraction with citric acid (or with ammonium citrate) do not necessarily separate more-active from less-active basic calcium phosphates. It has been a common practice in agronomic experiments comparing phosphorus sources and rates to apply each fertilizer on the basis of its content of AOAC available, rather than of total, PzO,. With fertilizers containing onIy smalI amounts of citrate-insoluble phosphorus, this practice is satisfactory, but as shown above, it can be misleading with some fertilizers containing high amounts of water-insoluble phosphorus. With such fertilizers, it would seem much more accurate to make applications on the basis of total PzOs content. VI. Problems Concerned with Nonorthophosphates and Other Fertilizers

Several new materials that in recent years have entered, or may soon enter, the commercial fertilizer market have created some new problems of analyses for fertilizer anaIysts and controI officials. These incIude liquid and suspension fertilizers, particularly those that contain polyphosphates; ammonium polyphosphates; calcium polyphosphates, including calcium metaphosphate; and bulk blends.

A. LIQUIDAND SUSPENSIONFERTILIZERS Expansion of the liquid mixed fertilizer industry and problems of production were discussed by Slack (1957) and more recently by Potts (1963). With use of orthophosphoric acid, an 8-244 is about the maximum that can be used without danger of salting out. With use of superphosphoric acid, in which about half of the phosphorus is present as various nonorthophosphates, liquid grades such as 11-33-0 ( Striplin et al., 1959) or 11-37-0 (Slack and Scott, 1962) can be made. A 13-43-0 base suspension is also produced by TVA by ammoniation of superphosphoric acid and adding attapulgite clay (Slack and Scott, 1962). These liquids and suspensions result in crop yields equal to those obtained with water-soluble solids applied so as to contact the same amount of soil and to supply equal amounts of nitrogen and phosphorus. The Association of Official Agricultural Chemists’ “Official Methods of Analysis” (1960) makes no provision in the method for water-soluble

94

G. L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

phosphorus to hydrolyze meta-, pyro-, and polyphosphates to orthophosphate, which is determined by the official method. MacIntire et al. (1937) recommended that solutions of metaphosphate in water or in dilute acid be boiled with HN03 to convert nonortho- to orthophosphates before determination of phosphorus. Because of numerous reports of incorrect analyses of liquid fertilizers containing polyphosphates, boiling of the filtrate containing the water-soluble phosphorus fraction of the fertilizer was made official in 1960 by Association of Official Agricultural Chemists (1961).

B. SOLDAMMONIUMPOLYPHOSPHATES Ammonium polyphosphates prepared in solid form (Getsinger et al., 1962) are produced by ammoniating superphosphoric acid under pressure. These fertilizers (grades ranging from about 18-56-0 to 15-61-0) can be used for direct application, for preparation of NPK grades, or for liquid fertilizers (Slack, 1962). Ammonium pyrophosphates and short-chain ammonium polyphosphates are water soluble. Consequently there is no special analytical problem except that of hydrolysis to the orthophosphate form. Stinson et al. (1956) found that some ammonium metaphosphates made by gasphase reaction of NH3 and P205were less soluble in neutral ammonium citrate than in water. Certain long-chain polyphosphates have recently been identified (unpublished TVA data) of which large fractions are not readily soluble in water. However, some of these experimental products have greater solubility in water than in ammonium citrate solution, apparently because of the NH4+ common ion effect. With these materials there is a problem of separating AOAC water-soluble and non-water-soluble fractions as discussed below. C. FUSEDPOTASSIUM PHOSPHATES Fusion products approximating the anaIyses of potassium metaphosphate ( KP03) and calcium potassium pyrophosphate ( CaK2P20T)have been produced and evaluated by TVA (DeMent et al., 1963). In powder form, these materials were essentially equivalent to CSP and KC1 as sources of phosphorus and potassium for crops. Effectiveness for the immediate crop decreased with increase in particle size. Results with pure calcium ammonium and with calcium potassium pyrophosphates as sources of nitrogen, phosphorus and potassium for corn have been reported by Lehr et al. (1964). Difficulties are encountered (Brabson, 1963) in determining the water solubility of potassium phosphates and other fertilizers containing

“AVAILABLE” PHOSPHORUS IN FERTILIZERS

95

partially water-soluble nonorthophosphates. The present official method (Association of Official Agricultural Chemists, 1960) of washing a 1.0-g. sample of fertilizer on a filter paper with successive small portions of water until 250 ml. of filtrate is obtained was adopted when normal superphosphate was the primary phosphatic constituent of fertilizers. The water-soluble monocalcium phosphate hydrolyzes in dilute solution and dicalcium phosphate precipitates. HNO, is added prior to diluting to volume to dissolve the precipitate (turbidity) and obtain a true measure of water solubility. When this method is applied to potassium metaphosphate or calcium potassium pyrophosphates and probably to partially water-soluble polyphosphates in general, the filtrate is usually turbid because of colloidal material passing through the filter paper, rather than because of formation of a precipitate. Acidification of this turbid filtrate leads to high contents of “water-soluble” phosphorus. An alternative procedure which has been used by TVA in connection with development research is that of placing a 1.0-g. sample in a flask, diluting to 250 ml., and shaking for 45 minutes. The suspension is then allowed to settle for 15 minutes, and the supernatant liquid is filtered prior to analysis for content of watersoluble phosphorus. Another problem is that amounts of phosphorus dissolved from meta-, pyro-, and polyphosphates by ammonium citrate solutions is more a function of their rate of solution and hydrolysis than of a true solubility. Rate of solution is strongly influenced by particle size, kind of nonorthophosphate, and other factors. Thus only total phosphorus content is particularly meaningful by official AOAC methods. This was recognized by DeMent et al. ( 1963). Determination of citrate-insoluble phosphorus in nonorthophosphate may, however, provide information on the content of apatite. Harris (1963) also found that dissolution of potassium metaphospliates in water and citric acid was not a good measure of content of available Pz06. The AOAC method was also unsatisfactory. It was concluded that the extraction of a 1.0-g. sample with 100 ml. of either 2 per cent sodium nitrate or sodium chloride solutions for 30 minutes did give a reliable estimate of available P205,provided that little potassium chloride was present. If chloride is present, the sample can be extracted with water and then with salt solution. Harris (1963) reported that the phosphorus content of the second and third cuttings of grass was higher with potassium metaphosphate than with equivalent amounts of phosphorus from superphosphate. Lehr et a,?.(1964) found that response by a first crop of corn was greater for 100-mesh than for -6+14 mesh calcium ammonium and calcium

96

G. L.

TERMAN, W.

M. HOFFMAN, AND B. C. WRIGHT

potassium pyrophosphates. Residual response by a second crop to the coarse granules, however, was much greater. These results indicate a slower rate of dissolution and hydrolysis of the nonorthophosphates, particularly from large particles. Potassium metaphosphate has been produced experimentally for many years, but is not produced commercially. As a result of commercial production of superphosphoric acid, various nonorthophosphate components of fertilizers may become increasingly abundant, and there appears to be a need for development of better methods to assess their chemical availability.

D. CALCIUM POLYPHOSPHATES One calcium polyphosphate ( calcium metaphosphate ), produced by TVA by reacting Pz06 gas with phosphate rock, has found considerable use for direct application, as an ingredient of mixed fertilizers without hydrolysis, and for mixing with other ingredients after partial hydrolysis to orthophosphates ( Nelson and Terman, 1963). Hoffman and Lundell (1937) reported that digestion for 30 minutes in the HCl-HN03 mixture specified by AOAC (2.018b, 1960) was adequate for the decomposition and hydrolysis of calcium metaphosphate. Brabson and Edwards (1951) found that 20-mesh calcium metaphosphate was satisfactory with one AOAC method (2.018b, 1960) and that 35-mesh material dissolves more rapidly, Grinding of calcium metaphosphate to 35 mesh was first specified in the official methods of AOAC in 1955. Continuous agitation of the sample for 1 hour is also specified, which results in a nearly complete dissolution of the phosphates present.

E. BULKBLENDS The problem with bulk-blended fertilizers is one of sampling rather than of analytical method. Analyses of samples, even from a single lot of fertilizer, may vary greatly because of segregation of various sizes of granules. Proper mixing, together with similarly sized ingredients, reduces segregation (Hoffmeister et al., 1964) and the variability in the analyses, not only of phosphorus but of other nutrients as well. VII. In Conclusion

As indicated in the preceding sections, there is a need for serious consideration of the type of information that can be obtained from a single chemical analysis of phosphate fertilizers. The AOAC method for determining available phosphorus measures only the sum of the phosphorus soluble in water and in neutral ammonium citrate. It is un-

“AVAILABLE” PHOSPHORUS IN FERTILIZERS

97

fortunate that the term “available” was used to describe these soluble phosphates, as this word has fostered the feeling that chemical availability and crop availability are identical. At present, the chief use of the AOAC method for phosphorus is to establish a minimum level of reactivity that is commonly termed “availability” in the chemical sense. It is used effectively by state control officials in their surveillance of fertilizers moving from the producer to the consumer and has ensured that unacidulated phosphate rock is not marketed as an acidulated phosphate. It is aIso highly desirable, of course, that the official method should indicate the commercial value of the fertilizer as a source of phosphorus for crops, since this is the reason for the manufacture and use of phosphorus fertilizers. The present method sometimes fails to do this. Obviously, no one chemical procedure will perform satisfactorily for all fertilizer materials and all crop, climate, and soil conditions; but if only one method is used, it should give a satisfactory indication of crop availability for most of the fertilizers being marketed. If this is not true, then a new chemical method should be adopted that will perform more satisfactorily, as confirmed by results obtained under soil and crop conditions now prevailing. For example, as shown by Wright et al. (1963) and Brabson and Burch (1964a, b ) , a modification of the NAAC method may be found more useful for fertilizers based on ammoniated ordinary superphosphates. Agronomists in many States are currently recommending that farmers use water-soluble phosphates under certain conditions and for certain crops. However, water-soluble phosphorus determinations are not routinely made by State fertilizer control laboratories, and the legal description of fertilizers offered for sale contains no information relative to that portion of the available phosphorus content which is water soluble. As a result, the user must rely on other sources of information concerning the water solubility, such as sales literature and his own previous experience. Making the water-soluble, as we11 as the “available,” phosphorus content of fertilizers a part of the legal description would remedy this situation, and in our opinion this should be done. This would be valuable to users of heavily ammoniated ordinary superphosphates, for which the AOAC “available P205))content is not a particularly good indicator of the agronomic value of the phosphorus. Because of the marked effect of amount of water-insoluble phosphorus per sample on the content of “available” phosphorus in fertilizers, we also recommend that the AOAC method be modified to standardize the water-insoluble content per sample within rather narrow limits. In regard to terminology, we recommend that the term “available

98

G. L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

PzOc be replaced by “water-soluble plus citrate-soluble phosphorus,” as determined under prescribed conditions. The AOAC method was not designed for a chemical evaluation of polyphosphate fertilizers, and there is some doubt whether it can be modified for their proper evaluation. There is therefore a need for research to find suitable chemical methods for evaluating polyphosphates. ACKNOWLEDGMENT This chapter was prepared under the auspices of the Fertilizer Evaluation Committee of the Soil Science Society of America. Other members of the committee are D. R. Bouldin, 0. P. Engelstad, W. R. Schmehl, and J. R. Webb, chairman. W. M. Hoffman is the AOAC associate referee on phosphorus in fertilizers. REFERENCES Ando, J., Siegel, M. R., and Jordan, J. E. 1964. Unpublished TVA data. Andrews, W. B. 1942. J . Assoc. Ofic. Agr. Chemists 25, 498-509. Archer, J. R., and Thomas, R. P. 1956. J . Agr. Food Chem. 4, 608-613. Association of Official Agricultural Chemists. 1884. Proc. 1st Ann. Meeting, A.O.A.C., Philadelphia, 1884. Association of Official Agricultural Chemists. 1885. U.S. Dept. Agr. Diu. Chern. Bull. 7. Association of Official Agricultural Chemists. 1960. “Official Methods of Analysis,” 9th ed. Washington, D. C. Association of Official Agricultural Chemists. 1961. J. Assoc. W c . Agr. Chemists 44, 133-134. Beaton, J. D., and Gough, N. A. 1962. Soil. Sci. SOC. Am. Proc. 26, 265-270. Bouldin, D. R.,and Sample, E. C. 1959. Soit Sci. SOC. Am. Proc. 23, 276-281. Bouldin, D. R., and Sample, E. C. 1963. J . Agr. Food Chern. 11, 212-214. Bouldin, D. R., DeMent, J. D., and Sample, E. C. 1960. J . Agr. Food Chem. 8, 470-474. Brabson, J. A. 1963. Personal communication, TVA. Brabson, J. A., and Burch, W. G. 1964a. J. Assoc. Ofic. Agr. Chemists. ( I n press). Brabson, J. A., and Burch, W. G. 1964b. J . Assoc. Ofic. Agr. Chemists. ( I n press). Brabson, J. A., and Edwards, 0. W. 1951. J . Assoc. Ofic. Agr. Chemists 34, 771777. Brosheer, J. C. 1953. Unpublished TVA data. Buie, T. S. 1931. Comm. Fert. 4 2 ( 3 ) , 27-28. Clark, K. G., and Hoffman, W. M. 1952. Farm. Chem. 115(5), 17-20, 21, 23. Clark, K. G., Hoffman, W. M., and Freeman, H. P. 1960. J . Agr. Food Chern. 8, 2-7. Cooke, G. W. 1956. J. Agr. Sci. 48, 74-103. Cooke, G. W. 1963. Personal communication to T. P. Hignett, TVA. Cooke, G. W., and Widdowson, F. V. 1959. J . Agr. Sci. 53, 46-63. DeMent, J. D., and Seatz, L. F. 1956. 1. Agr. Food Chem. 4, 432-435. DeMent, J. D., Terman, G. L., and Bradford, B. N. 1963. J. Agr. Food Chem. 11, 207-212.

‘‘AVAILABLE” PHOSPHORUS IN FERTILIZERS

99

Ensminger, L. E., and Pearson, R. W. 1957. Soil Sci. SOC. Am. Proc. 21, 80-84. Fresenius, R., Neubauer, C., and Luck, E. 1871. Z . Anal. Chem. 10, 133-158. Gerlach. 1916. Z. Angwew. Chem. 29( l ) , 13-14, 18-20. Getsinger, J. G., Siege], M. R., and Mann, H. C. 1962. J . Agr. Food Chem. 10, 341-344. Gilbert, B. E., and Pember, F. R. 1936. Rhode Island Uniu. Agr. Expt. Sta. Bull. 256. Gilliam, J. W. 1963. M. S. thesis. Mississippi State Univ., State College, Mississippi. Harris, F. J. 1963. Fertiliser SOC. (Engl.) Proc. 76. Haskins, H. D. 1921. J . Assoc. Opt. Agr. Chemists 4, 64-66. Hignett, T. P. 1956. Com. Fertilizer 9 2 ( 5 ) , 23-24, 26, 67. Hignett, T. P. 1963. Farm Chem. 126( l ) , 34-35. Hignett, T. P., and Brabson, J. A. 1961. J . Agr. Food Chem. 9, 272-276. Hoffman, J. I., and Lundell, G. E. F. 1937. J . Res. Natl. Bur. Std. 19, 59-64. Hoffmeister, G., Watkins, S. C., and Silverberg, J. 1964. J . Agr. Food Chem. 12, 64-69. Howes, C. C., and Jacobs, C. B. 1931. Ind. Eng. Chem. Anal. Ed. 3 , 70-72. Jacob, K. D., and Hill, W. L. 1953. In “Soil and Fertilizer Phosphorus in Crop Nutrition” (W. H. Pierre and A. G. Norman, eds.), Vol. 4, pp. 299-345. Academic Press, New York. Jordon, H. V. 1964. U . S. Dept. Agr. Tech. Bull. (In press), Joulie, H. 1873. Monit. Sci. 3, 563-584. Keenan, F. G. 1930. Ind. Eng. Chem. 22, 1378-1382. Lawton, K., Apostolakis, C., Cook, R. L., and Hill, W. L. 1956. Soil Sci. 82, 465-476. Lehr, J. R., Engelstad, 0. P., and Brown, E. H. 1964. Soil Sci. SOC. Am. Proc. (In press). Lewis, D. T. 1962. M. S. thesis. Univ. Maine, Orono, Maine. Liebig, J. 1840. “Organic Chemistry in Its Application to Agriculture and Physiology.” Lingle, J. C. 1960. Proc. Am. SOC. Hort. Sci. 76, 495-503. MacIntire, W. H., Hardin, L. J., and Oldham, F. D. 1937. Ind. Eng. Chem. 29, 224-234. Martin, W. E., Vlamis, J., and Quick, J. 1953. Soil Sci. 75, 41-49. Mattingly, G. E. G. 1963. Fertiliser SOC. (Engl.) Proc. 75, 55-97. Mortensen, W. P., Baker, A. S., and Tennan, G. L. 1964. Wash. State Uniu. Agr. Expt. Sta. Bull. p. 652. Nelson, W. L., and Stanford, G. 1958. Aduan. Agron. 10, 67-141. Nelson, W. L., and Terman, G . L. 1963. In “Fertilizer Technology and Usage.” ( M . H. McVickar, G. L. Bridger, and L. B. Nelson, eds.), pp, 379-427. Soil Sci. SOC.Am., Madison, Wisconsin. Organization for European Economic Cooperation. 1952. “Fertilizers: Methods of Analysis Used in OEEC Countries.” Paris. Parker, F. W. 1931. Corn. Fertilizer 4 2 ( 5 ) , 28-44. Petermann, A. 1880. Landwirtsch. Vers. Sta. 24, 310-350. Potts, J. M. 1963. Fertilizer Solutions Magazine 5 ( 2 ) , 18-21. Robertson, G. S. 1914. J. SOC. Chem. Ind. (London) 33, 9. Rogers, H. T., and Ensminger, L. E. 1961. Chem. Farming (Spring).

100

G . L. TERMAN, W. M. HOFFMAN, AND B. C. WRIGHT

Rogers, H. T., Pearson, R. W., and Ensminger, L. E. 1953. In “Soil and Fertilizer Phosphorus in Crop Nutrition” (W. H. Pierre and A. G. Norman, eds. ), Vol. 4, pp. 189-242. Academic Press, New York. Rosanow, S. N. 1934. Phophorsaeure 4, 641. Ross, W. H., and Jacob, K. D. 1931. J. Assoc. Ofic. Agr. Chemists 14, 182-196. Ross, W. H., Jacob, K. D., and Beeson, K. C. 1932. J . Assoc. Ofic. Agr. Chemists 15, 227-266. Ross, W. H., Adams, J. R., Hardesty, J. O., and Whittaker, C. W. 1947. J . Assoc. Ofic. Agr. Chemists 30, 624-640. Salter, R. M., and Barnes, E. E. 1935. Ohio. Agr. Expt. Sta. Bull. 553. Schmehl, W. R., Olsen, S. R., Gardner, R., Romsdal, S. D., and Kunkel, R. 1955. Colo. Agr. Expt. Sta. Tech. Bull. 58. Seatz, L. F., and Stanberry, C. 0. 1963. In “Fertilizer Technology and Usage” (M. H. McVickar, G. L. Bridger, and L. B. Nelson, eds.), pp. 155-187. Soil Sci. SOC.Am., Madison, Wisconsin. Slack, A. V. 1957. Com. Fertilizer 95, 28-29, 33, 35-37, 39-40. Slack, A. V. 1962. Farm Chem. 125( l l ) , 16, 18, 20. Slack, A. V., and Scott, W. C. 1962. Com. Fertilizer 105( 11), 24-26. Stinson, J. M., Striplin, M. M., Brown, N. A., and Seatz, L. F. 1956. I. Agr. Food Chem. 4, 248-254. Striplin, M. M., Stinson, J. M., and Wilbanks, J. A. 1959. J, Agr. Food Chem. 7, 623-628. Terman, G. L. 1960. Soil Sci. SOC. Am. Proc. 24, 356-360. Terman, G. L. 1961. Soil Sci. SOC. Am. Proc. 25, 49-52. Terman, G. L., Anthony, J. L., Mortensen, W. P., and Lutz, J. A. 1956. Soil. Sci. SOC.Am. Proc. 20, 551-556. Terman, G. L., DeMent, J. D., Clements, L. B., and Lutz, J. A. 1960. J. Agr. Food Chem. 8, 13-18. Terman, G. L., Bouldin, D. R., and Webb, J. R. 1961a. J. Agr. Food Chem. 9, 166-170. Terman, G. L., DeMent, J. D., and Engelstad, 0. P. 1961b. Agron. J. 53, 221224. Terman, G. L., Bouldin, D. R., and Webb, J. R. 1962. Aduan. Agron. 14, 265319. Thomas, R. P. 1959. Croplife 6 ( 2 6 ) . U. S. Department of Agriculture. 1963. “The Fertilizer Situation.” Govt. Printing Office, Washington, D. C. van Burg, P. F. J. 1963. Fertiliser SOC. (Engl.) Proc. 75, 5-54. Wagner, P., Dorsch, R., Aschoff, F., and Kunze, R. 1903. Mitt. Ver. Deut. Landw. Vers. Sta. 1. Webb, J. R., and Pesek, J. T. 1958. Soil Sci. SOC. Am. Proc. 22, 533-538. Webb, J. R., and Pesek, J. T. 1959. Soil Sci. SOC. Am. Proc. 23, 381-384. Webb, J. R., Lathwell, D. J., Caldwell, A. G., Terman, G. L., Schmehl, W. R., and Mortensen, W. P. 1959. Crops Soils 12( l ) , 12-15. Webb, J. R., Pesek, J. T., and Eik, K. 1961a. Soil Sci. SOC. Am. Proc. 25, 222-226. Webb, J. R., Eik, K., and Pesek, J. T. 1961b. Soil Sci. SOC. Am. Proc. 25, 232-236. Williamson, J. T. 1935. J. Am. SOC. Agron. 27, 724-728. Wright, B. C., Lancaster, J. D., and Anthony, J. L. 1963. Mississippi State Uniu. Agr. Expt. Sta. Tech. Bull. 52.

OBJECTIVES IN CORN IMPROVEMENT

G . H.

Stringfield

DeKalb Agricultural Association. fncorporoted. DeKalb. Illinois

Page I. Introduction ...................................... ...... 11. Hybrid Corn and Yield . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Inflated Estimates . . . . . . . . . . . . . . . . . .................... B. Less Blight versus the High Yielders ....................... C . Corn Borers versus the High Yielders ....................... D . Leaf Aphids versus the High Yielders ....................... E . Plant Population versus the High Yielders . . . . . . . . . . . . . . . . . . . F. A Need for Objectives ...................... I11. The Offense and the Defense ..................... A . The Moths Picked the Lush Plants ......................... B . Continuous Adjustments .................................. C . Adjustments to New Norms .................... ...... D. Domestication Adds Its Problems .......................... E . Two Categories of Goodness . . . . . . ..................... F. The Fallibility of Definition and Prediction of Yielding Capacity G. Judgments of “Where” and of “How” ...................... IV . Culture and Improvement .................................... A. Genotype and Mineral Accumulation ....................... B. The Genotype-Stand Problem ............................. C . Genotype-Stand Relations under Higher Soil Productivity . . . . . . D . Crowding Pressure and Stalk Quality ....................... E . An Experience with Stalk Troubles ........................ F . Crowding and Barrenness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G . Crowding Tolerance: A Major Objective .................... V. Breeding for Industrial Uses and Nutritive Value ................ A . Oil and Protein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Waxy and High Amylose Corn . . . . . . . . . . . . . . . . . . . . . . . . . . . . C . The Yellow Carotenoid Pigments .......................... 01. Parent Stocks ............................................... A. Growth Characters of Inbred Lines Reappear in Hybrid Progenies B. Smut Resistance of Inbred Lines Reappears in Hybrid Progenies C . Aphid Resistance of Inbred Lines Reappears in Hybrid Progenies D . Corn Borer Resistance of Inbred Lines Reappears in Hybrid Progenies ............................................... E . The Carotenoid Pigmentation of Inbred Lines Reappears in Hybrid Progenies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F. The Weak Combiner as a Common Tester Parent . . . . . . . . . . . . 101

102 103 104 104 106 106

108 108 110 111 111 113 113 114 114 114 116 116 118 119 119 119 119 121 122 122 122 124 125 126 127 127

102

VII. VIII. IX. X.

G.

H.STRINGFIELD

G . Poverty in Parent Lines Means Poverty in Hybrids . . . . . . . . . . . . H. Have We Honored Homozygous Lines Too Much? . . . . . . . . . . . I. Parent Stocks and Genetic Plasticity ....................... J, Seed Quality of Parent Stocks ............................ Exotic Germ Plasm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Cytoplasm .............................................. Tetraploid Corn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary and Conclusions .................................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

127 128 129 131 132 133 133 134 136

1. Introduction

“I think that when the final history of the agriculture of this period is written, they will refer to the chapter on hybrid seed corn as the golden kernel, and they will print the chapter in letters of gold.” SO spoke Raymond A. Iones of the Foreign Agricultural Service, U. S. Department of Agriculture. These words, from an authoritative speaker, made good listening to the several hundred breeders, researchers, producers, processors, and sellers of hybrid corn gathered at their Hybrid Corn IndustryResearch Conference in Chicago in 1962. It made us think that everything in corn improvement was under fine control. On second thought, if informed people are now of the opinion that this golden chapter might be written, it seems that the materials for it must already be in hand-and attributable in large measure to work already accomplished. At any rate, a bit of honest praise implies no assurance that our course is true for this decade. Indeed, the question being asked in seriousness and by responsible observers is whether the hybrid corn enterprise is not slowing toward a walk and possibly facing an era of stagnation until or unless new research shall uncover some markedly improved approach. After tramping the rows of corn Zea mays L. since the early 1920’s, the writer has accepted the task of putting down what now seems significant in terms of present and future opportunities and responsibilities in dent-corn improvement. References to past events and research will be made to support or refute current judgments. The writer has used his own observations and data and those of close colleagues more than any other writer would, should, or could. Everyone feels that he can best evaluate the work he knows best. Statements of fact and of belief about genetics are presented and citations of papers essentially genetic are made when the statements or citations seem to have a bearing on choice of objectives. This paper is not, however, a treatise on genetic theory or breeding methods; it is more a personal report of experience, observation, and viewpoint. The citations of literature are in no sense presumed to be either complete or limited

OBJECTIVES I N CORN IMPROVEMENT

103

to the most important papers. No papers appearing after September 1963 can be considered. Corn breeder’s jargon is used sparingly, but a few accepted terms save many words. For instance, the corn plants do their own breeding. But even so, the person who directs which plant shall mate with which plant is called a corn breeder. We may say a strain of plants is “resistant” to something, meaning that the strain has so far shown less susceptibility to that something than have other strains we know about. Again “resistance” as used here may indicate an active antibiosis, a low-grade or nonexistent attractiveness, or even a high degree of tolerance. It means a capacity to sustain an exposure with little or no injury. Usually we do not know why. The term “pressure” is used to indicate a depressing effect, especially some natural or artificial influence that favors one gene, character, or organism as opposed to another. If tall plants are selected from a population and short plants are refused, that is selection pressure for tallness. “Homozygous line” means a highly inbred strain. Its uncrossed progenies are not importantly different genetically. No strain remains completely homozygous. The meanings of most other words are briefly defined in the context, or they are intended to be understood as the dictionaries define them. It. Hybrid Corn and Yield

A. INFLATED ESTIMATES Overselling the yielding capacity of corn hybrids in the earlier phases of their use may be reacting now to dull some people’s appreciation of progress. In 1942, which was about the time the open-pollinated varieties were nearing their demise in Ohio (they accounted for 17% of the corn acreage), Stringfield et al. (1943) reported: “Open-pollinated corn in 27 tests averaged 83.3 bushels per acre. The average acre yield of hybrids of the same seasonal requirement as the open-pollinated corn and in the same tests, was 100.1 bushels. That yield represents a gain of 16.8 bushels, or 20.2%, for the use of good hybrids on good average soil.” It is important here that adjustment was made for seasonal requirement (silking date) and for stand. Too many of the earlier comparisons were of farm-stored variety seed, of mediocre germination and viability, tested against seed-house dried and processed seed of a later hybrid. In such comparisons, spurious gains of 50 per cent or more often were made by the hybrids. Then the problem of how much gain had been made would get complicated in comparisons wherein the varieties would blow over and

104

G . H. STRINGFIELD

the hybrids, or some of them, would stand and thus score a tremendous legitimate gain. If someone believes, however, that the hybrids which had become established in the early 1940’s were normally 50 per cent, or even 30 per cent, superior to the best local varieties, and with no advantage to the hybrids from more nitrogen and better care, his estimate of present yield expectations may well be too high. The facts, good as they are, surely will disappoint him. B. LEAFBLIGHTVERSUS

THE

HIGHYIELDERS

More complications in the yield comparisons were occasioned by a sharp upsurge of leaf blights (mostly Helrninthosporium turcicum Pass.) across southern Ohio in the late 1930’s. A startling number of hybrids that had been making good yields and rapid acreage gains in the area because they had been well fortified with “yield genes, would be taken with a leaf-spotting sickness in late July. By mid-August, the now rampant fungus would have obliterated most of the green leaf tissue. This sequence occurred on thousands of acres where there were favorable ecological circumstances for blight, including susceptible host hybrids. Early defoliation accentuated stalk rotting and breakage. Grain yields were low and of inferior quality. A number of “high-yielding” hybrids were given medical discharges. Ullstrup (1954) found losses of over 50 per cent from severe H . turcicum infection when the epidemic was established by the soft-dough stage. The local open-pollinated varieties too were hurt in the south Ohio epidemic, but much less so. Injury to the individual plants varied greatly, but the overall effect was in the moderate range. The varieties reversed the yield relation with the susceptible hybrids in these situations. Being old hands in the area, the open-pollinated varieties had coped with H. turcicum in many previous generations. Natural selection in variable populations had kept the varieties fortified with a complement of genes having a degree of protective function that resulted in only variable infection among plants. It is assumed that as farmers picked seed ears they did not always keep those from the most susceptible plants. Then, too, there was a large potential for segregation within each seed ear.

C. CORNBORERSVERSUS THE HIGHYIELDER^ Again in 1949 the corn borer Ostrinia nubilalis (Hiibner) had built up an overwintering population quite sufficient for a “population explosion,’’ provided the weather and availability of susceptible host tissue were right. This was so roughly in the western half of the northern Corn

OBJECTIVES IN CORN IMPROVEMENT

105

Belt. The weather and host tissue were cooperative (with the parasite). The loss cannot be known closely, but it was heavy. Brindley and Dicke (1963) cited studies showing losses of 3 per cent per borer per plant and higher for first-brood larvae, and of under 2 per cent for second brood. The Bureau of Entomology and Plant Quarantine reported corn borer losses in 1949 in the United States at 313,819,000 bushels, 99 per cent occurring in the North Central States. The report appeared in “North Central Regional Publication No. 22 (Revised),” published as Iowa Agricultural Extension Pamphlet 176. These statements were made: “Probably the major and most variable factor affecting the increase or decrease of corn borer numbers is weather.” “Progress is being made in the development of hybrid varieties resistant or tolerant to corn borer infestation . . . some have been developed and are recommended for use in certain areas.” The known methods of control were given. Sprays and dusts understandably had the major attention. However, a supplement to the pamphlet from the Ohio representative (C. R. Neiswander, a stout cooperator in the breeding work for borer resistance) stated, “. . . corn hybrids resistant to borer infestation are available to Ohio farmers.” The hybrids to grow (in 1952),and several not to grow, where corn borer threatened in Ohio were listed. Of insecticides the supplement stated in part, “. . . for early market garden sweet corn . . . some localities . . . some years . . . for second generation of borers in canning corn. It may be advisable in some years for the protection of high value per acre hybrid seed corn and in rare instances certain fields of commercial field corn.’’ Huber (1961) presented data showing an expected acre yield of 76 bushels for corn borer-susceptible hybrids compared with 100 bushels for only moderately resistant hybrids (one resistant line was substituted for one susceptible line in a 4-parent hybrid to convert susceptibility to resistance) at Van Wert, Ohio in 1939. However, yield loss due to susceptibility is highly variable. The important point here from the angle of improvement is that a relatively few borer-susceptible inbred lines, and their closely related derivatives, had been widely extended in hybrid combinations over the northern Corn Belt during the 1940’s. These inbred lines, of course, had been selected for the area in large part by standard yield-testing routines. They had good yield records. It would be presumptuous to charge the 1949 corn borer troubles entirely to germ plasm, and it would be unfair to castigate the able and sincere breeders who made the difficult choices of parent lines. This episode had much in common with the Ohio experience with leaf blight.

106

G. H. STRINGFIELD

Such “backlashings of nature” are “growing pains” in corn improvement. They present opportunities to learn by costly experience.

D. LEAFAPHIDSVERSUS THE HIGHYIELDERS In the summers of 1958 and 1959 entomologists and agronomists in Ohio were besieged with characteristic trouble calls centering in about a dozen lake-bed counties of the northwest quarter of the State-good corn territory. The little peace-loving aphid Rhopalosiphum m i d i s ( Fitch ) was sucking at the juices of most of the susceptible corn plants, messing up the leaves, throwing enormous numbers of plants into barrenness, and disturbing the expectations for many field-tested, “field-proved,’’ hybrids. Fortunately in this case there was a significant scattering of aphidresistant hybrids over the troubled area. One could usually identify a resistant or a susceptible field as soon as one got into it, and often before. Again, the amount of damage could not be closely estimated. From Pennsylvania, Huber ( 1961) reported four aphid-susceptible single crosses as averaging 114 bushels per acre at Chambersburg with zero aphid infestation compared to 87 bushels with 33 per cent aphid-infested plants at Hershey. These susceptible crosses were compared with four others each containing just one resistant line. The resistant crosses averaged close to the susceptibles (only 4 bushels more) in the absence of aphids at Chambersburg, but under aphid pressure at Hershey, the resistant crosses had but 3 per cent aphid infestation and yielded 112 bushels per acre, or 12,9 per cent of the susceptibles. Again, these differences were somewhat confounded with other variables, and the penalty for susceptibility under aphid infestation is far from a constant value.

E. PLANTPOPULATION VERSUS

THE

HIGHYIELDERS

It is not left for pests alone to disarrange the expectations of yield among corn hybrids. In 1957, while on the research staffs of the Agricultural Research Service, United States Department of Agriculture, and the Ohio Agricultural Experiment Station, the writer tested single crosses at Columbus in Central Ohio and at Hillsboro in the southwest quarter. Plot replication and field design were of an approved order. The plots were overplanted and thinned at both locations to the relatively thick stand for the area of 18,000 plants per acre. This created an unexpectedly heavy plant-population pressure on certain entries. In some entries, as high as 25 per cent of plants were virtually barren. Presumably, a mild dry period near silking accentuated this fruiting failure. Modal yield, however, was good. There was moderate incidence of corn borer (modal damage, light), of aphids (modal per cent infested, 5; maximum, 31), and of stalk rot

107

OBJECTIVES I N CORN IMPROVEMENT

(modal per cent soft, 9; maximum, 43). The dominant pressure, however, was crowding; an interaction of genotype with the complex of differentia1 pressures of limited light, water, minerals, ventilation, and doubtless other items. Two inbred lines, Oh29, nonprolific, with a good yield record, and Oh43, moderately prolific ( intermediate), were uniformly in single-cross combination with three nonprolific mates and with six intermediate mates. “Prolific” here means capacity to develop a second ear in growth situations favorable for plants as individuals, but not necessarily in heavy stands where interplant competition is strong. A summary of pertinent results is presented in Table I. TABLE I The Differential Contributions of Inbred Lines to Tolerance to Crowding Pressure in Single-Cross Combinations Grown at 18,000 Plants per Acre at Columbus and Hillsboro, Ohio, 19575 Nonprolific Oh29 Parameter Average, 3 nonprolific lines Average, 6 medium prolific lines Medium prolific minus nonDroIific

Bu./A.

Medium prolific Oh43

Ears/plantb Bu./A.

Medium prolific minus nonmolific

Ears/plant

Bu./A.

Ears/plant

25

0.16

91

0.80

116

0.96

120

0.98

121

0.96

1

29

0.18

5

0

-

-0.02

-

5 Courtesy, Ohio Agricultural Experiment Station and Agricultural Research Service, United States Department of Agriculture. b Continuous grain around the cob and extending across the palm of the hand was required for an “ear.”

Intermediate Oh43 greatly exceeded nonprolific Oh29 in ears per plant and in grain yield when both were crossed with the same nonprolific mates. But the two lines displayed no measurable difference when both were crossed with intermediate mates. Again, the intermediate lines were grossly superior to nonprolifics when crossed with a nonprolific mate but only moderately superior with an intermediate mate. Thus, a descending order of goodness was evident for the rather specific environmental circumstances of these two fields in 1957; i.e., prolific x prolific, best; nonprolific, slightly inferior; and nonprolific x nonprolific, prolific grossly inferior. Prolificacy as such, however, surely is a result, not a cause, We are not yet fully ready to accept it as a dependable indicator of crowding tolerance. Lang et al. (1956) studying the performance of hybrids over a plant

x

10s

C . H. STRINGFIELD

population range of 4 to 24 thousand per acre, found that the two hybrids showing the most multiple earing at the lightest stands had the lowest incidence of barrenness at the heaviest stands. Josephson (1957) studied both prolific and nonprolific hybrids at varying stands in Tennessee. His data showed that the important merit of prolificacy, or its concomitant character, was its capacity to contribute high yields at heavy stands ( 14,000).

F. A

NEED FOR OBJECTIVES

The experiences described above with leaf blight, corn borers, leaf aphids, and crowding pressure have important bearings on the formulation of objectives in corn improvement. Can objectives be set up to moderate such jolts or is pursuance of high yielding capacity only an infatuation with a love untrue? This discussion will be continued in the next section. 111. The Offense and the Defense

A. THEMOTHSPICKED THE LUSHPLANTS When the corn borer first settled down in earnest along the southern shores of Lake Erie, an Ohio agronomist decided in about 1926 to try a “quickie” survey hoping for a rough estimate of the yield loss per infested plant. Infested and proximate noninfested pairs of plants in a lightly infested field were harvested and dry grain weights were recorded. Grain weights showed that the infested plants had produced the more grain! Was a tunneling larva then boosting the grain production of its host? Not at all. The gravid moths had effectively chosen the larger, lusher plants for egg deposition. And these moth-selected plants by and large were above the norm in potential grain yield by a margin in excess of the damage by this light infestation. Is it chasing a phantom, then, to pursue the hope of combining in one genotype both effective borer resistance and a high potential of grain production? We have known since the early 1930’s that such genotypes are possible (Meyers et al., 1937; Dicke, 1955). If that combination of potentials was the only one needed, corn improvement might indeed have seen its heyday,

B. CONTINUOUS ADJUSTMENTS The combining of high yield potential with high resistance to any natural pest seems to get close to the nub of the corn improvement problem. Usually plants in the wild live along in the presence of persistent attack by assorted parasites. What often appears to be sporadic inci-

OBJECTIVES IN CORN IMPROVEMENT

109

dence of infection by a given parasite in a given host is a product of three constantly shifting variables: the parasite, the host, the remaining environment. (Parasite and host are environment to each other.) If the parasite gains the dominant place, selection pressure on the host is heavily in favor of those genotypes, possibly few, which contribute the complicated patterns of growth and chemistry that disturb the rapport between parasite and host tissue, At once the critical genes or gene combinations required for this favorable host response begin to take over in the host population. The species has for such exigencies not only variable genotypes as they appear in any current generation but also, we believe, a potential or latent variability “stored” as linked genes on separate chromosomes. This latent variability may subsequently be “freed by crossing over. The parasite, like the host, has the capacity to alter its gene frequencies and gene organization to contend with alterations in the environment, such as a new degree of resistance in its host. Thus, a countervailing antagonism between host and parasite may continue indefinitely, each constantly adjusting to meet temporarily successful adjustments in the other. In natural selection these adjustments may extend over long periods of time. But under the added and violent impact of artificial selection, even mass selection, the pace is accelerated so much that hazardous imbalances often are unwittingly established. Only a few of these imbalances were reported in the preceding section. Striking examples in corn of varietal modifications to meet new environmental pressures were the alteration of plant types after central Corn Belt varieties had been introduced into the more xerophytic Nebraska environment and required to grow there for a few decades. Here the morphology, especially the relative leaf area, was markedly changed (Kiesselbach, 1922). Leng ( 1962) reported an epochal mass-selection breeding experiment in Illinois. Beginning with single ears of open-pollinated corn, selection was applied in four different directions, namely for ( a ) high oil, ( b ) low oil, ( c ) high protein, and ( d ) low protein. After 38 generations of continuous selection the high-oil strain was at 13 per cent of oil compared with 1 per cent for the low-oil strain. The high-protein strain was at 19 per cent compared with 5 per cent for the low-protein strain. Reverse selection was then begun in each of the four lines and the reversed strains were carried along concurrently with the originals for 13 (now 16) additional generations. After 13 generations of reverse selection, each reversed strain had responded until it was cleanly separated from its original counterpart,

110

G. H. STRINGFIELD

In spite of drastic modification during the first 38 generations and in spite of close breeding, the genetic variance had either not been spent or new variance had appeared. This monumental experiment appears to typify the action of latent variance. There are other, and maybe better, explanations. Under any interpretation the capacity to segregate in a free-breeding evolutionary line that has survived the ordeals of ages will not be snuffed out by a few generations of selection. Objectives in breeding, one would think, may safely be erected on this assumption. Mutation, of course, plays an important part in supplying new variation in the evolutionary time scale. Mutation may be more significant than we had thought in the corn breeder’s time scale (Russell et al., 1963). However, gene segregation and recombination are of much more importance to the breeder,

C. ADJUSTMENTS TO NEW NORMS We may assume with confidence that mass selection, in the form of keeping for seed those ears that most closely approached the selector’s ideal, was the practice in corn improvement from its inception until varietal crossing and later the field testing of separate plant progenies (ear-row breeding) began. From the beginning of the hybrid-corn era until very recently, mass selection often was cited as an example of relative ineffectiveness in breeding for higher yields. Changes under mass selection came slowly, and a ceiling seemed soon to be reached (Lonnquist, 1961). But mass selection was known to effect changes in traits other than potential yield, and there was the example of a tremendous improvement over the centuries by the American Indians, unlearned in genetics. Undoubtedly, the Indians had patiently brought forth varieties far more productive than the wild plants which they first cultivated. But the final achievements of productive varieties were not necessarily pestridden nor incapable of withstanding normal weather hazards. The very mildness of the changes and the relative shortness of the successive northward migrations presumably provided time for the continual reorganization of genes, briefly described above, to effect adaptations, including resistances, to the slowly changing norms. The new norms were represented by populations geared to ( a ) heavier and heavier production of seed, and ( b ) new environmental pressures, including day length, temperature, moisture, and new pests. So our working assumptions include a reliance upon continuous segregation in an open-breeding corn population. Latent variance and

OBJECTIVES IN CORN IMPROVEMENT

111

associated ideas were well and convincingly discussed in a collection of papers by Darlington and Mather (1950). D. DOMESTICATION ADDS ITS PROBLEMS The assumption is made here that as a statistical concept every advance in grain yield was accompanied by an advance in susceptibility to trouble. The required fast growth, vegetative lushness, and soft tissue, are predisposing factors to attack from many sources. So, in spite of time and the process of natural adaptation, aided more or less by human selection, modern domesticated crops are heirs to more troubles than their wild counterparts. Corn has no known wild counterpart, but the principle holds, given native habitats for the wild counterparts. Yet resistance is normal and natural. Every plant has many resistances as a necessary heritage from the past, supposedly going back to the single-cell progenitor. Why does corn not get sick with a wheat rust, with oak galls, with the people’s common cold? It is not for lack of exposure. Corn resists these troubles, under the definition of resistance in the introduction. Corn sustains these exposures without harm. Moreover, important differentials between corn genotypes are so powerful that one genotype will be thrown out of production by a given pest or by a flush of hot sun following a chilly night whereas a competing genotype, even a good grain producer, will pass the same exposure with little or no injury, E. Two CATEGORIES OF GOODNESS So the most useful corn genotypes must be provided with two somewhat antagonistic categories of goodness. For the purposes of this discussion these categories may as well be dubbed the offense and the defense, which roughly they are. The offense is speed, vigor, bulk. The term heterosis largely connotes characters of the offense. The defense is protection. It involves climatological adaptation, resistances and tolerances to pressures, such as come from pests, sun, wind, and shade. Whatever the attainable maximum in either offense or defense may be, it seems reasonable that both maxima cannot be attained by the same organism. For a simple example, if wind resistance were not a requirement, breeding could be planned for shunting into grain production some of the energy now absorbed in building firm stalk and strong root anchorage. Obviously, this separation into two categories is neither absolute nor complete. Characters in different categories may interact strongly, and many nonadaptive characters would belong in neither group. But the

112

G.

H.STRINGFIELD

offense and the defense are not alike from a crop improvement standpoint. The point to be emphasized is that characters of the offense, so prominent in yield tests and in much of our studies of breeding methodology, contribute only a part, and maybe not the most important part, to corn improvement. Any gain in yield potential must be accepted as ( a ) specific for limited environmental variations, even local ones, and ( b ) more liable than not to follow the historical pattern of appearing with more or less loss of defense values. This latter loss could be through physiological incompatibilities or random genetic drift. In a breeding program selection with high intensity for character A fractionates the number of eligible candidates carrying any other desired character which is independent of A, This circumstance adds up to a practical certainty of some separation between top yielding capacity and top defense. What were the “yield genes” doing amid the local reversals already referred to and conditioned by pests and by an accelerated crowding pressure? On the other hand, what is the good of great defense in a hybrid that loses in all yield comparisons? No baseball manager ever built a winning team only from lumbering home-run hitters who could not catch pop flies, or only from sparkling fielders who could not hit out of the infield. So there is no real improvement if either offense or defense is neglected. The offense is not liable to be neglected. The defense is not the negative. Stalk quality is a combination of defense characters. And improved stalk quality was a major reason for the rapid acceptance of corn hybrids in the 1930’s. We have now put down some experiences and observations and have added some inferences about natural and artificial selection. Case histories are more dramatic than components of variance. And case histories can be put in the perspective that suits the prejudices of the writer. But see Sprague (1955a) for a discussion of hybrid x season interaction based upon a sound statistical analysis of 32 field performance experiments. The differential seasonal effect contributed materially more to the variance than did the hybrid effect. Sprague concluded: “If one may extrapolate from these tests . . . to the situation in an openpollinated variety of corn one might conclude that a particular genotype or series of genotypes would be favored in a given year and therefore contribute heavily to the progeny available for propagation in the ensuing year. In a subsequent year, climatic conditions might well favor a quite dissimilar genotype or series of genotypes. Since there may be no consistent and continuous selective advantage to any particular

OBJECTIVES IN CORN IMPROVEMENT

113

genotype, it would appear that the conditions necessary for a stable equilibrium are not met.” The inferences of Sprague seem reasonable and are cited to support the idea that genotype-environment equilibrium in corn is only an approximation. Surely all agree that the genotype which contributed to excellent production last season may meet troubles next season. And from these arguments the working assumption has developed that the defense (protective and adaptive traits) merits an importance coordinate with the offense (fast vigorous growth and potential for high production) in corn improvement plans. F. THEFALLIBILITY OF DEFINITION AND PREDICTION OF YBLDING CAPACITY

In this setting it becomes difficult to de€ine high yielding capacity. In passing, high yielding capacity is not conditioned by combining a genetic potential for number of fruits with another genetic potential for size of fruits, nor by combining genetic potentials for length and for girth. The product is not manufactured by its shape nor by its division into pieces. Place for place and season for season, it seems that high yielding capacity will be generated by the optimal balance of offensive and defensive characters. This idealized balance or combination will be as inconstant as winds and weather can make it. And it well may wane under the impact of an emergent or migrating pest biotype or by a changing cultural practice, such as a modified pattern of plant distribution. No enduring hybrid-environment equilibrium should be expected. The best of our current homozygous inbred parents are so restricted genetically, so nearly inflexible in contribution to progeny, that their practical values must always be accepted as tentative, and their use as a calculated risk. Now if this discussion appears to be in opposition to the statistical approach it is because the argument was poorly stated. The mathematical estimation of genotype x environment interaction is of proved value. The emphasis here is that the interaction has a potential of explosiveness -a potential that may not appear in a short-term study. The most carefully controlled field tests of genotypes often deceive us because they lack involvement with critical ecological variations which will have their innings in later seasons.

G.

JUDGMENTS OF

‘WHFXE” AND

OF

“How”

The improvement program based on the most sensitive predictions on the corn needs and hazards of the near future could win in a longterm yield race against the program based on the best breeding method-

114

G. H. STRINGFIELD

ology. But alertness in both prediction and method should produce the real winners. Thus, judgments of direction, of where to go with the improvement program, are fully as important and demanding as judgments of how to keep moving. What breeder has not seen hybrids of low defense value, in heavy trouble in farmers’ fields? And this in places where only minor injury was suffered by other hybrids that were better conceived but perhaps less glamorous with no troubles about. The writer’s vote is for moving in the direction of the high average yielder with the best chance of avoiding disaster status. Such a hybrid might not produce the most total bushels over a 10-year period. But a 50-bushel loss per acre all in one season would hurt so much worse than taking 5 bushels under the top in each of 10 seasons. The high average yielder with a good dosage of disaster “resistances” is likely to be out there growing again and again after the disaster-prone favorite is only an unhappy memory. But it is a matter of taking a better or a poorer calculated risk. No hybrid can be genetically protected against every possible environmental contingency. IV. Culture and Improvement

A. GENOTYPEAND MINERALACCUMULATION Corn genotypes exhibit marked individuality in their relative accumulations of mineral nutrients (Sayre, 1955). The knowledge of these variables has had only minor application in corn improvement. Some inbred lines will accumulate more than twice as much of a given essential element as other lines in a comparable environment. Sayre found that certain inbred lines would grow satisfactorily at the low level of magnesium availability typical of certain Ohio and Pennsylvania soils, whereas other lines in the same situations would develop damaging magnesium deficiencies. At present, farmers can af€ord to apply mineral nutrients so lavishly that refined efficiency in fertilizer ratios could not rate as a major breeding objective. But there may be something more fundamental in this genotype-mineral accumulation relation. B. THEGENOTYPE-STAND PROBLEM The agronomists at the Ohio Agricultural Experiment Station by 1924 had accumulated more than 20 years’ data on stands for open-pollinated corn. Their stands were thinned to 7112, 10,668, and 14,224 plants per acre. The ensuing recommendation for average Ohio conditions was three viable seeds per checkrowed hill-a little more than 11,000 seeds from which the farmer would do well to harvest 10,000 plants. Today we occasionally find final stands above 25,000 per acre.

OBFCTIVES I N CORN IMPROVEMENT

115

Were the Ohio agronomists wrong? They were not disputed by colleagues nor by successful farmers of the time. The agronomists of 1924 correctly interpreted their data. They were dealing with open-pollinated varieties far more susceptible to lodging than the hybrids of today. The average corn yields were not much above half of today's average. The corn of 1924 was harvested largely by hand, placing a premium on large ears. In the 1930's and 1940's we examined the tendency toward barrenness in hybrids compared with the old varieties. Stringfield and Thatcher (1947) noted ( a ) that adapted hybrids produced more ears per 100 plants than did the local open-pollinated varieties, and ( b ) that this disparity in ear number increased as stands were pushed upward. The disparity in yield followed the same pattern. The hybrids used in these studies were not strongly prolific, and it must be assumed that factors other than prolificacy were involved. Prolificacy was not recognized as the primary differentiating factor at the time; nor is it now. However, the strong single-ear habit which the admirers of over-sized ears had bred into the open-pollinated varieties, or some less obvious character associated with it, surely was contributing to a greater and greater disadvantage as crowding increased. In the 1950's (Ohio Agricultural Experiment Station, 1957, p. 63) average crowding tolerances were compared using early hybrids (southem Michigan maturity), mid-season hybrids (mid-Ohio maturity) , and late hybrids (northern Kentucky maturity). With a soil-season yield potential of just above 100 bushels per acre, the late hybrids exceeded the mid-season hybrids in average yield across a stand range of roughly 8000 to 19,000 plants per acre. But at 20,000 plants per acre the late hybrids had dropped below the mid-season competitors. The early hybrids were about 10 bushels per acre below the mid-season at 12,000 plants per acre, but they gradually approached the mid-season as stands were increased, until at 21,000 (the heaviest stand studied) the early hybrids were within 5 bushels of the mid-season hybrids and had caught up with the late competitors. Thus, it was learned that bald statements about seasonal requirement and yielding capacity had little relevance in the absence of information about growth conditions. In Section I1 it was shown how individual genotypes, apart from seasonal requirement, may respond differentially to crowding pressure. But why this emphasis on the seemingly simply matter of stand? Stand demands attention because of its relation to several other developments. Modern methods in the manufacture and distribution of commercial fertilizers, especially the developments in the fixation of atmospheric nitrogen, modern methods in soil and water conservation, research

116

G. H. STRINGFIELD

in fertilizer application, and other yield-boosting technologies cannot be fully exploited at 1924 or even 1960 stand levels. Increased stands were required better to exploit higher yield potentials. Then fertilizers were increased to support the higher stand, and so the spiral has been going. Reviews of the stand problem (including numbers of plants per unit of land area and also how the plants are distributed) covering numerous experiments have been published by Dungan et al. (1958), Hinkle and Garrett (1961), and Stringfield (1962).

C. GENOTYPE-STAND RELATIONS UNDER HIGHERSOIL PRODUCTIVITY The developments regarding stand bear on breeding objectives because the right stand for one corn hybrid may be dangerously wrong for another even in identical environmental situations. Here is the most critical current culture-improvement contact. Here is a reason for the flagging of interest in conventional yield tests, no critical pressures operating. Farmers continue to thicken their stands hoping to gain more and more return from more and more applied nutrients. Troubles have arisen from excesses in this leap-frog game, and more will follow. Breeders led the way first by providing hybrids with improved standing qualities and with further improvements that permitted nearly doubling the stands on the better corn farms since the 1920’s. But the attainable summit and the most difficult climb lie ahead. D. ‘CROWDING PRESSURE AND STALKQUALITY Early stalk disintegration with its accompanying ills and barren or near-barren plants are the most commonly observed troubles when crowding pressure is beyond the tolerance of a hybrid. The high importance of stalk quality to the modern problem warrants the following quotation from personal correspondence with Dr. A. J. Ullstrup, Plant Pathologist, ARS, U. S. Department of Agriculture, and Purdue University: “1. Root rot and stalk rot appear to be a single disease entity. Rotting of roots commences prior to rotting of stalks and generally the same fungi are involved. As the roots and stalks mature, lesions on the former become larger and more numerous and rotting proceeds into the crown and stalk tissue, Small, apparently static or quiescent lesions can be found on roots of both resistant and susceptible genotypes several weeks before silking. The onset of rapid colonization by Diplodia may& Gibberella zeae and one or two other weak or hemi-parasites, takes place after silking when the reproductive phase is complete. 2. Resistance to stalk rotting is directly related to the rate of senescence of stalk tissue. The reasons which I believe support this are: ( a ) Inoculum consisting of spores or mycelium of these hemi-parasites remains static and there is little or no spread of rot until after silking.

OBJECTIVE3 IN CORN IMPROVEMENT

117

( b ) Inoculation made by puncturing the stalk with a nail contaminated with soil presents the same picture as an inoculation made with a pure culture of one of the stalk-rotting fungi, i.e. no extension of rotted pith until after silking. Inoculations made from tissue inoculated by puncturing with a soil-contaminated nail yield fungi generally considered to be saprophytes or, at best, weak parasites-species of Trichoderma, Mucor, Alternuria, Fusarium (not G.z.) and a few others. These appear able to colonize and disorganize only senescent tissuetissue which is becoming progressively less and less active physiologically. ( c ) Resistance to colonization or rotting can be modsed by removing leaf tissue. This appears to hasten maturity and hence increases susceptibility. The opposite effect is attained by keeping plants barren. Thus, by preventing reproduction the plant is kept in a juvenescent condition and resistant to colonization by these weak parasites. ( d ) Resistance to colonization can be enhanced to some degree by the application of nutrients. It is well known that KC1 added to the soil tends to decrease stalk rot. This seems proportional to the amount of K already present-if low, the benefits are marked; if adequate, little effect is shown. Deficient K seems to hasten maturity. ( e ) A mutant that fails to attain the reproductive stage and remains juvenescent does not succumb to colonization by weak parasites. Lesions are present on the roots of these mutant plants, but there is no increase in lesion size or numbers. “These are some of the reasons why I believe stalk rotting and the accompanying stalk lodging is directly related to maturation. “The roots and stalks of resistant genotypes mature at a slower rate than those that are susceptible. Maturation of roots and stalks is not necessarily directly correlated with maturation of the ear-some late inbreds are susceptible and some early inbreds relatively resistant. “The pathogenesis of root rot and stalk rot is something different than of certain other diseases of corn-smut, leaf blight, or pythium stalk rot. The latter, incited by P . butleri, may take place well before silking. “In addition to stalk rot resistance, as an expression of maturation, there is the matter of morphological resistance (rind thickness). This seems important from a practical standpoint in that a thick rind will tend to resist breakage for a longer period than a genotype with a thin rind. This probably has little if anything to do with actual invasion by colonizing fungi.”

Thus stalk quality is made up in part of resistance to rotting organisms and in part of rind strength, The resistance to rotting is importantly conditioned by delayed vitality relative to the grain-filling period. But it is not necessarily correlated with maturation of the ear. In Ohio experiments, corn ears in the shock did not dry measurably faster than those on standing (slower drying) stalks. Furthermore, killing corn leaves chemically failed to hasten ear drying. The grain dried through the husks without significant relation to the moisture content of the supporting

118

G. € STRINGFIELD I.

stalks (Ohio Agricultural Experiment Station, 1957, p. 68). Again Crane (1958) reported no material effect on grain moisture following either late leaf spraying or hand defoliation. The contribution of rind strength to lodging resistance and the evaluation and inheritance of rind strength have been established by Zuber and Grogan (1961) and by Loesch et al. ( 1963). The phenotypic expression of differentials in potential lodging resistance is highly sensitive to environment-even between one end of a 10-hill row and the other end. Since it requires repeated observations to separate the genetics from the environment, the character or combination of characters has low heritability. But like other characters, when resistance to lodging is established as a strain mark through generations, it is heritable and can be managed in breeding.

E. AN EXPERIENCE WITH STALKTROUBLES The 1955 growing season was one of the worst for stalk rot known in Ohio agricultural history. Thirty-two hybrids were given a second year of testing in a river-bottom field near Columbus, following eight years of continuous corn. This was a deliberately established favorable situation for stalk rot and other troubles. The test included three population levels, with appropriate plot size, replication, and other testing detail. Six of the 32 hybrids were selected as “resistant” to stalk rot. The basis for selection of the six was the soft-stalk records at a moderate lodging level in the previous year. Under the heavy lodging pressure of 1955 the average stalk breakage for the entire 32 entries in percentages of plants broken was 15, 38, and 52 at acre stands of 13, 17, and 21 thousands, respectively. Comparable values for the six entries previously designated as resistant were 8, 23, and 36. The six resistant entries also exceeded the mean of all 32 in harvestable yield by 13, 15, and 18 bushels per acre, respectively, at the three population levels, (Ohio Agricultural Experiment Station, 1957, p. 69). The total Ohio loss from stalk-rot trouble in 1955, being widespread, may well have exceeded the total loss combined from aphids, corn borer, and leaf blight in any one season. Stalk rot occurs every autumn, otherwise the stalks would have to be burned. When it comes early and destructively the agricultural public is less perturbed than when comparable destruction appears as visible lesions or crawling things. But the Ohio losses from stalk rots in 1955 represent another of those “backlashings of nature” and one that could have been greatly moderated by attainable defense qualities in area hybrids as a whole.

OBFCTIVES IN CORN IMPROVEMENT

119

F. CROWDING AND BARRENNESS Scatter-grain ears and complete barrenness often appear in overcrowded fields. Applicable knowledge of the underlying physiology in this situation is too scant. Building up that knowledge, however, is an important objective which is under active research by alert plant physiologists (Hesketh and Musgrave, 1962, and papers cited therein). Stinson and Moss (1960) applied artificial shading to one group of hybrids known by field testing to be tolerant of high plant populations, and to another group known to be relatively intolerant. With adequate soil moisture and fertility the two groups of hybrids yielded approximately alike in the absence of shade. In shade, however, the tolerant hybrids yielded significantly the more. The yield differential was expressed more as ear barrenness than as ear weight. G. CROWDING TOLERANCE: A MAJOROBJECTIVE Surely the stage is now set for acceptance as working assumptions that (1) genotypes differ greatly in contributions to both standing and grain yield under comparable crowding pressures; ( 2 ) high tolerance to crowding is a requirement for efficient exploitation of the increasingly high levels of soil productivity; ( 3 ) breeding has already done much, but its most significant contributions to this complex are yet to come; (4)breeding objectives that apply to the problem of crowding tolerance are major in terms of responsibility and opportunity; and (5) these contributions can be made the breeders’ greatest service to corn in this decade. V. Breeding for Industrial Uses a n d Nutritive Value

A. OIL AND PROTEIN Sprague ( 195513) and Schneider (1955) have summarized the modern situation in industrial utilization and in the nutritive values of corn. Irving (1962), in presenting a perspective of research in industrial utilization, said, “By its nature, utilization research is unlikely to make a sudden impact upon our surplus problem, but the cumulative markets for farm products it will create will ultimately permit us to use profitably all the American farmer can produce.” In important respects breeding and industrial utilization have common interests. Economically needed are genotypes contributing good agronomic qualities and also carrying a genetic potential to double or triple the normal fraction of oil or protein in the grain. Increasing the percentage of total oil or of total protein can be

120

G. H. STRINGFIELD

managed (Jugenheimer, 1958, Chapt. 6 ) . But neither oil nor protein is of one piece. There are choices to be made between the fatty acids of oil and between the amino acids of protein. Corn protein has the disturbing habit of dropping in quality as quantity is increased. At this point breeding and chemistry must operate together. The research-team approach has demonstrated that remarkable modifications can be made. In terms of percentage composition great progress has been made. The successful breeding of corn for modified chemical or physical properties involves three categories of uncertainty: ( 1) establishing the genotypes required of the modification, ( 2 ) preserving satisfactory agronomic qualities, and ( 3 ) establishing market or farm demand once reasonably dependable genotypes are available. On the agronomic side the desired genotypes would have to be classed as extreme variants. After a long process of genetic adjustment normal corn has a potential to ration the seed-stored photosynthate in fairly well-standardized proportions. If this rationing was conditioned in its evolution by physiological necessity or advantage to the plant, selecting gross variations in the rationing is asking for trouble. If the rationing was conditioned by seed needs, gross alterations can readily be managed. Extreme variants, whether in depth of leaf color, in brace root development, leaf width, husk size, plant height, ear size, or tassel size, usually collect disqualifying troubles of one sort or another. It is no surprise, therefore, that doubling or tripling the oil or protein percentage should follow the same pattern. The recessive gene su when homozygous depresses the normal conversion of sugar to starch in the endosperm. Most of the delectable sweetcorns are su su. This incomplete sugar-to-starch conversion may be looked upon as an intermediate point between the normal situation in dent corn and complete inhibition of fruiting which is easily effected by covering ear shoots and silks. Inhibition of fruiting or subnormal fruiting depresses total photosynthate production, Brunson and Latshaw ( 1934). Thus, the internal physiology of the fruit may exert a seemingly backward effect on the total plant production. Apparently, the question is not answered whether a gross departure from the established norms in oil or protein percentages in the grain necessarily imposes a restriction on total nutritive value stored in the grain. The considerable and able efforts in sweet-corn breeding have so far not resulted in yielding capacities comparable with the dent corns. One view would be that the species cannot adjust to any of these new norms. However, recalling species history of adjusting to new norms, it seems a reasonable view that by appropriate breeding procedures an

0BF.CTrVEs IN CORN IMPROVEMENT

121

acceptable degree of rapport between modifying growth genes and “oil genes” or “protein genes” is attainable. For instance, the brachytic-2 character, in which the lower internodes are drastically shortened, was typically associated with disease and low grain production when first transferred to otherwise standard hybrids. Even under the depressing infiuence of self fertilization, breeders have assembled minor genes which modify the dwarfed brachytics until they are beginning to look and perform more like respectable corn. Hopefully, we infer that the same can be done with high oil or high protein. Breeding for high oil and for high protein are not new objectives. They are nevertheless objectives still worth effort. High-oil corn would go to the processors. High-protein corn would be fed on farms where its modified nutritive content would be of significance especially to ruminants, and it would be sold to processors. Until these proposed new corns can be made to compete agronomically with standard hybrids, farmers would require a premium for producing them-one that processors or feeders could profitably pay. The “ifs” arising at this point are beyond present evaluation. Sweet-corn culture has been made a commercia1 success because of its high acre value. Surely high oil and high protein have as much likelihood for successful production as would sweet corn were it newly discovered.

B. WAXYAND HIGHAMYLOSECORN Waxy-maize starch, being a satisfactory replacement for tapioca starch, provides, among other things, the filler ingredients for fruit pies and puddings the country over. Its initial development is a completed objective. Improvements will come. As in sweet corn the conversion of sugar to the waxy type of starch has so far been accompanied by lower efficiencyin grain production. Again high product value has offset lower yields permitting profitable production on a very limited contracted acreage. A worthy objective, now being seriously investigated, is the development of high-amylose corn (Bear et al., 1958). Amylose is a component of all cereal starches. Because of the physical characteristics of the amylose molecule, cereal chemists believe it may have tremendous possibilities for industrial uses once it can be produced economically. It is commercially separated from amylopectin, the more plentiful endosperm starch, with a difficulty which warrants direct breeding for an endosperm content of three-fourths amylose or higher. Normal corn endosperm is about one-fourth amylose. The attainment of the objective here is a difficult one, and there is the possibility that processing tech-

122

G . H. STRINGFIELD

nology may find a practicable method of amylose separation. But again, the one who waits for all the answers would not be a breeder. The objective is approaching realization.

C. THE YELLOW CAROTENOID PIGMENTS A thin covering of protein about the starch granule carries (in yellow corn) a variable content of distinct but related compounds known to chemists as carotenoids. Within the group there are members important as sources of provitamin A. Others are noted for contributing color to egg yolks and poultry skin. Consumers pay cash for colors they like in animal products. Of course, coloring matter may cost little, but it cannot cost less than when produced by gene action in a product already in demand. Genetic production of high carotenoid content in the endosperm apparently creates no new agronomic or breeding problems beyond the addition of selection criteria. Here is another worthy objective that already has had some effective attention and merits more (Watson, 1962, and papers cited therein). VI. Parent Stocks

A. GROWTHCHARACTERS OF INBREDLINESREAPPEAR IN HYBRIDPROGENIES The first “crop” of inbred lines, isolated by the writer, came from open-pollinated varieties, They were tested as Sis X Ss tester strains in 1928 at the Ohio Agricultural Experiment Station. The S3)s had been placed in more or less distinct phenotypic groups which included: Sturdy-Short, thick culm; sensibly resistant to horizontal pull; no lodging StifF-Faultless standing record; normal height or tall High leaf area-By observation, and confirmed by measurements Vigorous-Unusual vigor as noted by observation Single crosses involving these S3 cultures were field tested in 10-fold replication of single-row 10-hill plots thinned to 14,000 plants per acre. The yield results are shown in Table 11. Within the test crosses, leaf area was positively correlated with silking date: r = 0.85 k 0.02; but in this test, silking date was not closely correlated with grain yield: T = 0.22 -C 0.08. Thus, the visually identitiable and heritable character, high leaf area, seemed to be advantageous with this material in this environment (including stand). However, leaf area and grain per unit of leaf area were negatively correlated (T = -0.40 0.06), a result suggesting that as leaf area was added its

123

OBJECTIVE3 IN CORN IMPROVEMENT

efficiency decreased, even at the moderate stand of 14,000 plants per acre. But 14,000 was a full stand in 1928, and the Ohio open-pollinated source material had had no vigorous selection for tolerance of self shading and other concomitants of population pressure. These data were interpreted in 1928 as supporting the simple Mendelism-based assumption that where most “growth genes” had been assembled in inbred lines, most “growth genes” and growth performance were reaped in hybrid progenies. It seemed to follow that inbred lines could visually be assorted into categories that would display differential and measurably predictable field performance qualities in their progenies. In our naivety we continued breeding work on these assumptions. The assumptions have not yet been changed. TABLE I1 Yield Contributions of S, Inbred Parents Classified by Type5 Number of single crosses

Average grain yield per acre (bu.)

Sturdy x Sturdy Stiff x Stiff High Leaf x High Leaf Vigorous x Vigorous

14 20 8 7

91.7 91.9 101.0 110.9

All crosses All crosses All crosses All crosses

20 30 18 22

90.3 92.8 99.5 104.5

Parents (type)

a

with with with with

Sturdy Stiff High Leaf Vigorous

Courtesy Ohio Agricultural Experiment Station.

Kiesselbach (1922) after comparing inbred lines and their F1hybrids concluded, “There appears to be some general correlation between productivity of the pure line parents and that of their hybrid offspring. Exceptions to this general rela tion occur.” Jugenheimer ( 1958) under the chapter heading, “Evaluating Inbred Lines,” discusses investigations by the U. S. Department of Agriculture and in Louisiana, Minnesota, Iowa, and Illinois, all showing measurable phenotypic correlation between inbred lines and their hybrid progenies. Among the characters studied were plant vigor, size, seasonable requirement, ear size and shape, leaf area, brace roots, pulling resistance, standability, several morphological characters, and grain yield. Hayes and Johnson (1939) stated, “If the Mendelian explanation of hybrid vigor proves the correct one, it furnishes a genetical basis for the belief that improved inbred lines can be developed by the same breeding methods that have been so successfully applied to self-pollinated plants.” They got a partial multiple correlation of 0.53 between test-cross yields

124

G . H. STRINGFIELD

and l k other characters of the inbreds when silking date was held constant. True, the relations between field performances of lines and their test crosses were not very close. But, similarly, the correlation between two sets of test crosses (same lines but different tester and testing circumstances) may be very low or even negative. It seems that these data from Ohio and Minnesota which indicate that part of the variation in contribution to yield is observable in the parent lines, are quite in harmony with the results from a recent neatly designed and important experiment in Iowa (Penny et al., 1962). Briefly a homozygous line was test crossed with two segregating populations, A and B. The test crosses were repeated after one cycle and also after two cycles of recurrent selection. The selection was in both A and B and both for high yield and for low yield in cross combination with the homozygous line. The selection was effective for the stated objectives. Now the succeeding phases of both A and B became lower yielding as such (without outcrossing) as their combining values with the tester decreased. And both A and B became higher in yield as their combining values with the tester increased. Furthermore, crosses between the comparable phases of A and B followed the same yield patterns as did the uncrossed populations. The phases of A and B when crossed together gave highest yields when the respective phases also had given best yields in crossed combination with the tester. In other words, where genes that made for higher grain yield in these tests were concentrated, higher yields were found. And in the cases of the uncrossed populations the assembly of more “yield genes gave higher yields in spite of a slight inbreeding estimated at 3.6 to 5 per cent in cycles two. Genter and Alexander (1962) found that, “Sl progeny performance was as closely related to test-cross performance as performance of one test-cross was to the other . , , .’’ Yield and several other characters were studied. Of course, an S1 is not homozygous, but it is halfway down the scale. Note that this was a preliminary evaluation of previously untested lines. Advanced lines long selected for contributing to desired agronomic characters would not be expected to act similarly for characters of the heterotic type (Leng, 1963).

B. SMUT RESISTANCEOF INBRED LINESREAPPEARSIN HYBRID PROGENIES The essential interest here is that inbred lines, if well evaluated directly, will provide not complete but useful predictive information, on potential hybrid progenies. Further support for this view is in work with smut, Ustilago maydis (D. C.) Cda. Stringfield and Bowman (1942) identified 12, inbred lines as being relatively smut resistant as lines. This

125

OBJECTIVES IN CORN IMPROVEMENT

identification was based upon repeated year-location nursery observations. Smutted plants were then counted in 48 statewide field-performance tests over a 3-year period. A total of 1052 hybrid entries (one hybrid in one test) were included and a minimum of 4 open-pollinated varieties were entered as controls in each test. Of the 1052 entries, 624 were related by half or more to lines of the “resistant” twelve. All entries were rated relative to the average of the comparable controls. Distributions of the ratings shown in Table I11 provide clear evidence from a rough experiment, but one exposed to a wider environmental TABLE I11 Frequency Distributions of Smut Susceptibility Ratings Comparing Hybrids of Half or No “Resistant” Parentage with Miscellaneous Hybrids and with Open-Pollinated Varieties as Contro1sa.b

Class of entries Controls All hybrids “Resistant” parentage

Number of entries 192+ 1052 624

Class centers Ratings 17

53

89

125

Higher

5

18 28 27

38 19 12

26 9 4

14 13

31 50

7

(I After Stringfield and Bowman (1942). Ohio Agricultural Experiment Station, and ARS, U. S. Department of Agriculture. * Low ratings indicate low infection, high ratings, high infection. The controls are distributed about a mean rating of 100.

scope than most, that lines selected over a few seasons for a given protective trait had contributed powerfully to that trait in hybrid progenies grown in other seasons and largely in other places.

C. APHIDRESISTANCE OF INBRED LINESREAPPEARS IN HYBRID PROGENIES Huber and Stringfield (1942) studied natural infestations of the leaf aphid Rhopalosiphum maidis ( Fitch) . They established susceptibility ratings on lines taken at such times and places as natural infestation permitted. The ratings were adjusted through a common susceptible control (because of the time-place effect on infestation intensity). Predicted aphid-susceptibility ratings were made for hybrids by averaging the separate ratings of the parent lines in proportion to genetic contribution, i.e., 1:l:l:lfor a double cross, and 1:1:2 for a 3-way cross. Observed and predicted data on 28 hybrids were assembled. The data on the hybrids were gleaned from about 50 over-state field tests where the required combination of control strain, one or more of the needed hybrids, and an aphid infestation, permitted. Not every hybrid always performed as predicted. But for data averaged over several experiments

126

G . H. STRINGFIELD

and thrown into quartiles of seven, the agreement is amazingly close (Table IV). The regression of the observed on the predicted rating is TABLE IV Aphid Ratings on 28 Double Crosses, Condensed to Quartile Averagesa

Quartile

Number of hybrids

1 2 3 4

7 7 7 7

Average number of tests

Observed

Predicted

4.0

35

3.9

52

5.9 6.7

75

48 86 80 98

Adjusted ratingsb

88

After Huber and Stringfield (1942). Ohio Agricultural Experiment Station and ARS, U. S. Department of Agriculture, b “Observed” is from actual counts of infested or aphid-caused barren plants in miscellaneous experiments on natural infestation, adjusted to susceptible hybrid Ohio K35 = 100. “Predicted” is from counts similarly taken but on the respective inbred parents and adjusted to susceptible Indiana Wf9 = 100. The prediction for a given hybrid is the average of the adjusted values for the 4 parent lines. a

essentially unity, Again, inbred lines contributed a protective trait to hybrid progenies under fairly diverse growth conditions.

D. CORNBORERRESISTANCE OF INBRED LINESREAPPEARS IN HYBRID PROGENIES Guthrie and Stringfield (1961) studied the relative contributions to corn borer resistance of 40 Sz lines. Damage ratings were taken following artificial infestation on ( a ) the lines, ( b ) test crosses of the lines x a susceptible tester, and ( c ) test crosses of the lines x a resistant tester. Estimates of correlation coefficients ( r ) and of proportions of coincident variation ( r “ ) were as follows: ( a >x ( b ) (a) x (c) (b) x (c)

= 0.60 = 0.46 T = 0.40 r r

T2 = 0.36 r2 = 0.21 r2 = 0.16

Since the closest agreement was between ( a ) and ( b ) it seems a reasonable inference that either ( a ) or ( b ) gave the best estimate of the relative heritable contributions of the lines. That would point to ( c ) , test crosses involving the resistant tester, as poorest and it would promote ( a ) , the lines as such, to no worse than an intermediate position. Fleming et al. (1958) reported dominance of resistance to corn borer leaf feeding in some crosses, while other crosses tended to be intermediate. Dominance would seem to favor the direct testing of lines for preliminary evaluation. Test-crossing with a susceptible tester conceiv-

OBJECTIVES IN CORN IMPROVEMENT

127

ably would be better, but the efficient screening of test crosses, an expensive operation, demands that there be simultaneous evaluation for as many pertinent characters as possible. It is not practical to develop a common tester that is recessive or susceptible for all important characters.

E. THECAROTENOJD PIGMENTATION OF INBRED LINESREAPPEARS IN

HYBRID PROGENIES

In a project on breeding corn for higher content of carotenoid pigments in the grain, Brunson and Quackenbush (1962), found for six lowcarotene lines, mean values of calculated provitamin A of 1.3 and 1.7 micrograms per gram, respectively, for the lines and for the mean of their 15 possible single crosses, Comparable values for six high-carotene lines were 6.5 and 7.5. Similar agreement was found for total carotenoid pigments, the values being 18.2 and 18.4, respectively, for the low-carotenoid lines and their crosses, and the comparable values were 40.2 and 36.1 for the high carotenoid lines and their crosses. These comparisons show rather close agreement in pigment content between the lines as such and their hybrid progenies,

F. THEWEAKCOMBINER AS A COMMON TESTERPARENT Most corn breeders argue that the best tester for preliminary and general line evaluation is a weak tester, a common source of gametes that contribute little to the total goodness of the test cross (Rawlings and Thompson, 1962). If the argument is sound, why not carry it to its logical conclusion? The gametes of a homozygous line to be evahated bring in no masking dominants, initiate no confusing epistasis. If the line carries a residue of heterozygosis, then the masking and epistatic situations characteristic of a more heterozygous test cross are at a minimum when the line is tested as such. G. POVERTY IN PARENT LINESMEANSPOVERTY IN HYBFUDS It would be easy to marshal more experimental evidence that a wealth of unmined information useful in predicting hybrids is available from more critical direct evaluation of the inbred lines. This is a pragmatic conclusion and a generalized one ignoring many individual exceptions. The gist of the argument is that the choicest inbred lines, based upon direct evaluation, do not necessarily make the best hybrids; but the poorer performing lines, especially in terms of defense characters, have shown no likelihood of waking to glorious action in hybrid combinations. Giving new inbred lines a high grade of direct evaluation, especially for protective traits, remains largely an unrealized objective.

128

G. H. STRINGFIELD

H. HAVEWE HONORED HOMOZYGOUS LINESToo MUCH? There is something about favorite inbred lines that just will not be completely separated from sentiment. But sometimes the little inbred idols tumble down, and the question arises whether they may have been honored too much. A homozygous line is highly predictable genetically; it is useful as a storehouse for genes and characters; it may have a special beauty. The refined genetic predicability, however, is not necessarily a practicable predictability amid the unpredictability of environment. The storage and esthetic values would not warrant the heavy current investment in inbred-line production, On the strictly debit side the homozygous lines are genetically inflexible. If one of them begins falling prey to a new pest, or if any need arises which it fails to meet, the usefulness of the line is crippled or dead. How many advanced homozygous lines fail as parents to accommodate the heavier-stand movement, or for the areas where leaf blight, or the southwestern stalk borer, Diatrueu grundioselk Dyar, are invading, or fail to help restore normal production where root worms, Diabroticu spp., have been catching up with conventional insecticides? ( I t seems here that breeding has no responsibility to support so ill a practice as long-time continuous corn.) But when a change is needed, time is too precious to resort to backcrossing or to begin anew. When a destructive invader or any new demand appears, the unprepared homozygous line is outdated by its inflexibility, and genetic purity, the very quality we had looked upon as merit. Must we go back again to the evolutionists and inquire what are the fortunes of a natural species when it gets hemmed in with an incapacity to vary? It is on its way to extinction, maybe slowly, but surely. The situation is not entirely comparable, to be sure. If one established inbred line fails, another can be tried, or related derivatives can be isolated. The substitute, however, has little likelihood of both repeating the performance of the replaced line and meeting the new demand. And again, the time is too short. Why not build up a tremendous backlog of standby lines? No one can say this system has not had a reasonably good trial. On the debit side too is the narrow genetic base of homozygous lines. They are the products of continuous self-pollination. The violence of self-fertilization, fixing as homozygous (nonsegregating) half the store of heterozygous (segregating) gene pairs in each generation, creates a torrent of gene loss. Genetic drift seems too mild a term here. Sprague (1955a) has stated this general situation well: “The necessity

OBJECTIVES IN CORN IMPROVEMENT

129

for handling large numbers of lines is a direct result of the deficiencies of the methods used. These deficiencies result, in part at least, from the large number of genes affecting most traits of interest and the rapid fixation of alleles during continuous self-fertilization. Both of these factors tend to minimize the opportunites for selection. The extensive testing in hybrid combinations of the lines developed is required to evaluate additive and non-additive effects over a series of environments.” In spite of the violence of self-fertilization a breeder often does come up with a homozygous line carrying a preponderance of the genes and gene combinations needed for the characters he wants. The degree of this preponderance depends upon the source material and the intensity, judgment, skill, and luck in selection. But during the self-breeding process a significant fraction of the fixing is of genes having inferior, indifferent, or negative values. Thus the scope of preferred gene action in a homozygous line is relatively limited, and we say a homozygous line has a narrow genetic base. The line must be outcrossed with a distantly related genotype to broaden the genetic base before corn plants of normal growth potential are realized. Whatever the final explanation of heterosis may be, the statement just made serves very well as a working assumption in practical breeding. Since the genetic base of a homozygous line cannot be widened, short of gambling on outcrossing and beginning anew, its genetic restriction, along with its genetic inflexibility, must be placed on the debit side. However, the genetic restriction being more or less compensated in first generation hybrids, is a lesser evil by far than is genetic inflexibility. The high cost of maintenance, seed increase, and crossing in volume of homozygous lines is a debit item, But that cost is a concomitant of genetic restriction.

I. PARENT STOCKS AND GENETIC PLASTICITY Of course, we are not presently prepared to go on without homozygous lines. But the time seems ripe to admit as a serious objective the development of genetically pliable (nonhomozygous) parent stocks for corn. Recurrent selection with its remarkable power to accumulate genetic potential for specified characters lies close to the area of the proposed objective. Recurrent selection involves successive cycles of two alternating phases: (1) a short period of gene fixing by inbreeding supported by appropriate evaluation, and ( 2 ) mixing by intercrossing related, selected lines or by alternative mixing procedures. There are numerous variations ( see Lonnquist, 1961, and papers cited therein).

130

G. H. STRINGFIELD

If the ultimate predictability of homozygous lines is not worth its several costs, then there seems no reason why recurrent selection cannot be used to develop gene pools with high ratings in suitability as parent stocks for hybrids. This is not a new idea. Comstock et al. (1949) suggested that recurrent selection could advantageously be applied reciprocally to the two parents of a double cross, A new phase of the hybrid could be made at any time by crossing the respective gene pools. Dr. J. H. Lonnquist has reported to the writer that the Nebraska thirdand fourth-cycle populations of recurrent-selection stocks are phenotypically more uniform than open-pollinated varieties and that crosses of these materials are comparable to the better hybrids. The probability that a given gene pool will contribute to disaster status in a hybrid should be lower than the corresponding probability of a homozygous line. A gene pool will have a degree of genetic plasticity depending upon its heterozygosity. Plasticity would permit continuing adaptive adjustment to emerging cultural requirements, to pest biotypes, or to other selection pressures. It seems likely that the pinnacle of potential yielding capacity might sooner be reached in single crosses selected for high specific combining ability in a given environment, but at the price of going back to genetic fixity. The other alternative would look interesting. The writer has a group of crosses that appear to embody the essentials of good hybrids. Other people agree that they look like good hybrids. Appearance does have its importance. We are all esthetes of sorts. Each of these hybrids is the first generation progeny of two parent gene pools that are being called “broad lines.” The “broad lines” were developed by a system closely akin to phenotypic assortative mating. The system involves no self-pollination. Obviously, the scheme can have variation in detail, and it has had variation. The objective is to challenge, in practical breeding, the half-century reign of the homozygous line. A natural sexually reproducing species usually appears nearly uniform to the eye. Yet beneath the apparent uniformity is a great capacity to vary-a latent variance. The assumption is that the gross morphology is conditioned by relatively few fixed major genes. In “broad line” development there is selection for an approximate uniformity in plant height, tassel type, and ear color. So much for the esthetic requirement. Most of the remaining morphology could vary with only little dependence on these characters (Thompson, 1957). The “broad lines” are, of course, subjected continuously to multiple selection pressures. Can “broad lines” be maintained with enough genetic variance for

OBTECTIVES IN CORN JMPROVEMENT

131

effective selection? Related stocks are always available for planned introgression. What about random genetic drift? It seems that those who are able to live with the genetic drift in a selfing program should be able to manage the much milder drift in a gene pool. If a practical breeder waits for all the answers before setting up and pursuing an objective, he will still be in the answer-seeking stage when his legs finally give out. Worry about genetic drift can begin when or if a reason for concern appears. How could this plan compete with the top first-generation hybrids where the best of specific combining ability is utilized? Maybe, crosses of “broad lines” or of other selected gene pools cannot compete. The hope for the soundness of the objective here lies in the possibility of achieving a greater gain from wider adaptation and continuous selection than may be lost through not having the top in more specific combining ability and more specific environmental adaptation. Stated in another way, can broad lines,” or any kind of gene pools wherein continuous massive gene adjustment supplants homozygous lines en m s e , provide the improved combinations of predictability and plasticity better to meet the variety of needs and adjustments required by fluctuating environment and evolving agriculture?

J. SEEDQUALITY

OF

PARENT STOCKS

Better phyiscal qualities of seed corn constitute improvement if they make for more efficient production. At the Ohio Agricultural Experiment Station in cooperation with ARS, U. S. Department of Agriculture, five seed sizes of one double hybrid were studied by the writer on four levels of soil productivity in 1940 and 1941. Sizes from small to large as measured by the number of seed pound were 4500, 3000, 1500, 1100, and 900. Planting rate was four seeds per checkrowed hill spaced 42 inches each way. In 1940, the average yields in bushels of grain per acre over all productivity levels were 90, 87, 93, 95, and 92 from the smallest seed to the largest, respectively. The comparable periods from planting to silking in days were 78, 77, 76, 75, and 73, again from the smallest to largest seed, respectively. Comparable values in percentage of grain moisture at harvest were 28, 27, 27, 26, and 26. The differences associated with seed size were highly significant (odds better than 99 to 1 against randomness as cause) in all three categories. However, the differences were somewhat confounded with

132

G . H. STRINGFIELD

final stands. A similar study in 1941 confirmed the 1940 trends, but only the data on silking date and grain moisture were significant. Funk et al. (1982), after a study of seed quality, concluded that differences in seed quality may have a genetic base and that the quality of corn seed should be improved by the selection of better seed-parent single crosses. The concensus of several investigators seems to be that seed quality being equal, the larger seed gives earlier emergence, more vigorous seedlings, a little earlier silking and maturity, and probably a small average advantage in grain yield. Seed size being equal, the more dense seeds are the more dependable for quick, vigorous seedling growth. Thus, we must favor the larger, denser seeds because they will produce as much as other seeds and will do the job in a shorter time. The plants they produce are therefore more efficient. Where parent stocks are produced randomly (an inexcusable waste), good seed parents are rare. Where the objective to get good seed parents or good pollen parents is early recognized and unvaryingly pursued in selection, good parent stocks are not too difficult to obtain. It must be recognized from the first pollination until a parent stock is produced that top seed parentage and top pollen parentage do not appear in the same genotype. Vil. Exotic Germ Plasm

The planned introgression of exotic germ plasm into local corn is perhaps more an item of method than of objective. However, the incorporation of new genes and new characters of exotic source seems certain to expedite the realization of more strictly objective goals (Kramer and Ullstrup, 1959). The many indigenous races of corn in Latin America have been largely collected and classified. They are being preserved, and samples are available for use by breeders, Descriptions and classifications appear in nine publications. The latest publication (Grobman et al., 1961) lists the entire series. There is a temptation to mix exotic strains to suit a temporary whim or expediency. Confusion later on in regard to relationships of breeding stocks could be much reduced if breeding groups of Latin American and other exotic races were established now. The breeding-group plan, as applied to inbred lines, is in operation in the North Central States. The plan seems to have acceptance at a majority of the breeding centers. It is not the responsibility here to suggest the mechanics or the personnel for organizing available exotic races into breeding groups. It is a worthy organizational objective even if applied only within a breeding project.

OBJECTIVES IN CORN IMPROVEMENT

133

VIII. The Cytoplasm

Fleming et al. (1960) found significant differential cytoplasm effects with genotype constant. The differential held for silking date, ear height, plant height, percentage of erect plants, grain yield, and budworm damage. The authors suggested that it would seem highly desirable to bring together the most efficient cytoplasm-genotype combination for a commercial hybrid. Duvick ( 1958) found that certain cytoplasm substitutions affected performance patterns with respect to grain yield, barrenness, and tillering. The cytoplasm effect was complicated, showing interaction with genotype and with components of the environment during the growing season. As it works out, the choice of cytoplasm usually is fixed by the more criticai choice of a genotype believed to be superior as a seed parent. The choice of a pollinator genotype compatible with the cytoplasmgenotype combination chosen as seed parent then becomes important. The real objective, however, is building up a greater body of information about corn cytoplasm. To what extent is it environment, heredity, or both? More quantitative information is needed on items such as the sensitivity of cytoplasm ( exclusive of transient materials ) to environment, and of genotype to cytoplasm. Some work is under way in establishing a wide range of cytoplasms with constant genotype and vice versa. That much is a good beginning. The whole study of corn cytoplasms in relation to corn improvement is not currently as acutely needed as are several other studies. But in the long run, the cytoplasm must be given a place of importance. The cytoplasms of exotic corns would seem to offer a most interesting area for further investigation. The use of male sterility resulting from cytoplasm-genotype incompatibilities to facilitate commercial control of pollination is a concern of seed production more than of corn improvement. IX. Tetraploid Corn

Since serious students are giving attention to tetraploidy (four sets of chromosomes instead of the normal two sets) in corn, we must include the development and possible use of tetraploids as a worthy objective in corn improvement (Alexander, 1960). Polyploidy in wheat, Triticum spp., and in a number of other field and horticultural plants has been much to man’s benefit. So drastic a change, however, must be expected to require many generations of gene reorganization and phenotypic

134

G. H. STRINGFIELD

selection before commercially satisfactory adjustments to the new norms are derived. Since it seems unlikely that tetraploidy is an immediately promising objective, it would fail to look like a good gamble for every breeding program. X. Summary and Conclusions

When hybrid corn first appeared, many appraisals of it carried inflation factors. These early high estimates seem to be reacting now to dull appreciation of continued progress ( Section I). Dependence upon high potential grain yielding capacity as a merit in itself is shown to rest upon frail assumptions. Expectations of superior yields, based upon the results of good field tests, have been reversed by leaf blights (11, B), by corn borers (11, C), by leaf aphids (11, D ) , and by crowding pressure (11, E ) , Selecting for higher and higher potential yield requires that the species adjust to new norms, and every gain in potential yield adds to the liability of new susceptibility to troubles. But the species has acquired workable adjustments to the demands of new norms and new susceptibilities by drawing from reserves of latent variation. These adjustments are importantly in the form of protective or defense values required in times of stress. Thus, the alert corn breeder’s objectives will continuously be directed to building and maintaining defense qualities (largely resistances and adaptive traits ) commensurate with the somewhat opposite and antagonistic qualities of the offense (lush growth, high yielding potential). No enduring genotype-environment equilibrium is to be expected (111). Modern advances in soil productivity requiring denser stands for their profitable exploitation are adding heavily to the breeder’s responsibility and opportunity. He must provide the genotypes that contribute adaptability to crowding pressures. The objectives required of this need, if well pursued, may lead to the greatest contribution of breeding to corn improvement in this decade (IV). Breeding for chemical modifications in the grain to effect useful improved nutritive value, or improved or new suitability for industrial uses, typically encompasses three uncertainties: ( 1) assembly of the genetic potential for the hoped-for modifications; (2) new organization of numerous modifying genes to reestablish, in part at least, “normal” plant growth, and ( 3 ) a market demand that will support upward price adjustments and workable contractual plans with farmers. Included under the framework of (1) is farmer willingness to grow a modified, and probably lower-yielding, corn because of its higher unit feed value in his own operations. The odds seem good that the three uncertainties

OBjECTIVES IN CORN IMPROVEMENT

135

will become workable realities with respect to several chemical modifications. The objectives seem worthy of continued vigorous effort at stations where the considerable technical involvements can be accepted ( V ) . Parent stocks, the foundation materials of hybrid corn, have been overproduced numerically and underevaluated as strains in their own right. The inbred lines that yield the most grain do not necessarily make the best hybrids, but the poorer performing inbred lines, especially in defense characters, make inferior hybrids. Adequate direct evaluation of the inbred candidates for status as parent stocks remains largely an unrealized and important objective (VI, A to G ) . The genetic fixity, and to some extent the narrow genetic base, of homozygous parent stocks are liabilities (VI, H). The state of corn improvement seems to be ripe for setting up as an objective the development of superior parent stocks having sufficient genetic plasticity for continuous, effective inter se selection. This objective, like most of the others discussed, has already had its beginning (VI, I ) . Establishing objectives in the early selection phase of a breeding project aimed at isolating parent stocks is essential. Even the early decision whether the selection is to be for pollen parentage or for seed parentage is an essential one (VI, J ) . Genes and characters from exotic races recently have been made available in profusion. Surely much improvement will come from this material, especially if it is blended in with local germ plasm in planned fashion (VII). Less immediately demanding than the previously noted objectives, but of very possible value over the longer haul, are (1) objectives aimed at utilization or improved utilization of the cytoplasm (VIII); (2) modification in genome number (IX); and ( 3 ) genotype differentials in contribution to mineral accumulation (IV, A). We must conclude that the future second chapter ( I ) on what happened in hybrid corn during the latter half of Century 20 need not fail to have its letters of gold because there was no challenge, no worthy objectives. A dominating “break through” in research is not required, and from this viewpoint none is in the offing. Corn improvement now is a field of opportunity, but it is more difficult than ever before. That new chapter is up to us (all the corn men-not breeders alone), especially to the younger of us, and in them this writer has every confidence. ACKNOWLEDGMENTS

I must express my profound gratitude to the Management of the Company which employs me, and whose only directive was that I write freely on the corn improvement situation as I see it. I am very grateful also to Advances in Agronomy for

136

G. H. STRINGFIELD

this opportunity, to Dr. Ullstrup for the valued inclusion which he gave me leave to use, to the Ohio Agricultural Experiment Station and the ARS, U. S. Department of Agriculture for approval to use several items of information previously unpublished, and to Robert Shrode of the DeKalb Agricultural Association, Inc., for a critical reading of the manuscript. REFERENCES Alexander, D. E. 1960. Proc. Ann. Hybrid Corn 1nd.-Res. Conf. 1960 Vol. 15, pp. 68-74. Am. Seed Trade Assoc., Washington, D. C. Bear, R. P., Vineyard, M. L., MacMasters, M. M., and Deatherage, W. L. 1958. Agron. J. 50, 598-602. Brindley, T. A., and Dicke, F. F. 1963. Ann. Rev. Entomol. 8, 155-176. Brunson, A. M., and Latshaw, W. L. 1934. J. Agr. Res. 49, 45-53. Brunson, A. M., and Quackenbush, F. W. 1962. Crop Sci. 2, 344-347. Comstock, R. E., Robinson, H. F., and Harvey, P. H. 1949. Agron. J. 41, 360-367. Crane, P. L. 1958. Agton. J. 50, 35-36. Darlington, C. D., and Mather, K. 1950. “Genes, Plants, and People.” McGrawHill (Blakiston), New York. Dicke, F. F. 1955. “Corn and Corn Improvement” (G. F. Sprague, ed.), pp. 537-612. Academic Press, New York. Dungan, G. H., Lang, A. L., and Pendleton, J. W. 1958. Aduan. Agron. 10, 435473. Duvick, D. N. 1958. Agron. J. 50, 121-125. Fleming, A. A., Rameshwar, S., Hayes, H. K., and Pinnell, E. L. 1958. Mlnn. Agr. Expt. Sta. Tech. Bull. 266. Fleming, A. A., Kozelnicky, G. M., and Browne, E. B. 1960. Agron. J . 52, 112115. Funk, C. R., Anderson, J. C., Johnson, M. W., and Atkinson, R. W. 1962. Crop Sci. 2, 318-320. Genter, C. F., and Alexander, M. W. 1962. Crop Sci. 2, 516-519. Grobman, A., Salhuana, W., Sevilla, R., and Mangelsdorf, P. C. 1961. Natl. Acad. Sci.-Natl. Res. Council, Publ. 915. Guthrie, W. D., and Stringfield, G. H. 1981. J. Econ. Entomol. 54, 784-787. Hayes, H. K., and Johnson, I. J. 1939. J . Am. SOC.Agron. 31, 710-724. Hesketh, J. D., and Musgrave, R. B. 1962. Crop Sci. 4, 311-315. Hinkle, D. A,, and Garrett, J. D. 1961. Arkansas Agr. Expt. Sta. BUZZ. 635. Huber, L. L. 1961. Pennsylvania Agr. Expt. Sta. Bull. 679. Huber, L. L., and Stringfield, G. H. 1942. J. Agr. Res. 64, 283-291. Irving, G. W., Jr. 1962. Proc. Ann. Hybrid Corn Ind.-Res. Conf. 1962 Vol. 17, pp. 101-110. Am. Seed Trade Assoc., Washington, D. C . Josephson, L. M. 1957. Proc. Ann. Hybrid Corn 1nd.-Res. Conf. 1957 Vol. 12, pp. 71-79. Am. Seed Trade Assoc., Washington, D. C. Jugenheimer, R. W. 1958. Food and Agr. Organ. U . N . FA0 Agr. Develop. Paper 62. Kiesselbach, T. A. 1922. Nebraska Agt. Expt. Sta. Res. Bull. 20. Kramer, H. H., and Ullstrup, A. J. 1959. Agron. J. 51, 687-689. Lang, A. L., Pendleton, J. W., and Dungan, G. H. 1956. Agron. J. 48, 284-289. Leng, E. R. 1962. Crop Sci. 2, 167-170. Leng, E. R. 1963. Crop. Sci. 3, 187-190. Loesch, P. J., Jr., Zuber, M. S., and Grogan, C. 0. 1963. Crop Sci. 3, 173-174.

OBJECTIVES IN CORN IMPROVEMENT

137

Lonnquist, J. H. 1961. Nebraska Agr. Expt. Sta. Res. Bull. 197. Meyers, M. T., Huber, L. L., Neiswander, C. R., Richey, F. D., and Stringfield, G. H. 1937. U. S . Dept. Agr. Tech. Bull. 583. Ohio Agricultural Experiment Station. 1957. “Ohio Experiments in Agronomy,” Book Ser. B-3.Ohio Agricultural Experiment Station, Wooster, Ohio. Penny, L. H., Russell, W. A., and Sprague, G. F. 1962. Crop Sci. 2, 341-344. Rawlings, J. O., and Thompson, D. L. 1962. Crop Sci. 2, 217-220. Russell, W. A,, Sprague, G. F., and Penny, L. H. 1963. Crop Sci. 3, 175-178. Sayre, J. D. 1955. In “Corn and Corn Improvement” (G. F. Sprague, ed.), pp. 293-314. Academic Press, New York. Schneider, B. H. 1955. In “Corn and Corn Improvement” (G. F. Sprague, ed.), pp. 637-678. Academic Press, New York. Sprague, G. F. 1955a. Cold Spring Harbor Symp. Quant. Biol. 20, 87-92. Sprague, G. F. 195513. In “Corn and Corn Improvement” (G. F. Sprague, ed.), pp. 613-636. Academic Press, New York. Stinson, H. T., Jr., and Moss, D. N. 1960. Agron. J . 52, 482-484. Stringfield, G. H. 1962. Proc. Ann. Hybrid. Corn I d . - R e s . Conf. 1962 Vol. 17, pp. 61-68. Am. Seed Trade Assoc., Washington, D. C. Stringfield, G. H., and Bowman, D. H. 1942. J . Am. SOC. Agron. 34, 486-494. Stringfield, G. H., and Thatcher, L. E. 1947. 1. Am. SOC. Agron. 39, 995-1010. Stringfield, G . H., Lewis, R. D., and Pfaff, H. L. 1943. Ohio. Agr. Expt. Sta. Spec. Circ. 66. Thompson, J. C. 1957. M. S. Thesis, Ohio State Univ., Columbus, Ohio. Ullstrup, A. J. 1954. Proc. Ann. Hybrid Corn 1nd.-Res. Conf. 1954 Vol. 9, pp. 35-37. Am. Seed Trade Assoc., Washington, D. C. Watson, S. A. 1962. Proc. Ann. Hybrid Corn 1nd.-Res. Conf. 1962 Vol. 17, pp. 92-100. Am. Seed Trade Assoc., Washington, D. C. Zuber, M. S., and Grogan, C. 0. 1961. Crop Sci. 1, 378-380.

This Page Intentionally Left Blank

SALINITY IN RELATION TO IRRIGATION Lowell E. Allison United States Salinity laboratory. Riverside. California

I. Introduction ................................................ I1. Salinity of Irrigation Waters .................................. A. Classification of Waters ................................... B. Interpretation of Irrigation Water Analyses .................. I11. Effect of Salts on Soils ........................................ A . Characteristics of Salt-Affected Soils ........................ B. Diagnosis of Saline and Sodic Soils ........................ IV. Effect of Salts on Crops ...................................... A. General and Specific Effects .............................. B. Salt. Sodium. and Boron Tolerance of Crops ................. V . Reclamation of Salt-Affected Lands ............................ A. Leaching to Remove Soluble Salts ......................... B. Leaching to Remove Boron ............................... C. Reclamation Procedures .................................. VI . Management Practices for Salt-Affected Land .................... A Leaching Requirement for Salinity Control .................. B. Drainage Requirement for Salinity Control . . . . . . . . . . . . . . . . . . C . Special Planting Procedures ............................... D Improving Irrigation Water Quality ......................... E . Method and Frequency of Irrigation ......................... VII . Conclusions ................................................. References ..................................................

.

.

Page 139 140 140 143 146 146 152 156 156 158 164 164 165 166 171 171 172 174 175 176 177 178

.

1 Introduction

History reveals that civilization began in an environment of irrigation agriculture. The Nile Valley in Egypt and much of the land in China have been irrigated for more than 4000 years. and still they produce good yields. These are but two examples of successful long-time irrigation developments. Despite successes in some areas. failures occurred in others. notably in Mesopotamia. where an early great civilization developed in the valley formed by the Tigris and Euphrates Rivers. Although the downfall of this civilization has been attributed to many and varied causes. most authorities agree that finality was determined by waterlogging and salinity. 139

140

LOWELL E. ALLISON

Salinity is the major and ever-present threat to the permanence of irrigation agriculture. In 1958, about 31 million acres of land were under irrigation in the 17 western States and Hawaii and, according to Hayward (1958), approximately 27 per cent of this land is salt affected to some degree. Unless salinity is controlled, productivity decreases, land values drop, and, in severe cases, the land is completely abandoned. In fact, during the dacade 1929 to 1939, over 1 million acres of irrigated land in the 17 western States were abandoned because of the accumulation of salt and sodium. However, most of this abandoned land has been restored to production. This report is concerned with the technical aspects of the problems of irrigation agriculture on salt-affected land, with emphasis on factors or practices that are important for the development of a permanent irrigation agriculture. 11. Salinity of Irrigation Waters

The salt content of most irrigation waters ranges from 0.1 to 5 tons of salt per acre-foot (70 to 3500 p.p.m.). Therefore, a knowledge of water quality is exceedingly important because it greatly influences irrigation and drainage practices, the choice of crops grown, and to some extent other management practices.

A. CLASSIFICATION OF WATERS Significant contributions have been made to our knowledge of water quality by Hilgard (1906), Kelley et al. (1939; Kelley and Brown, 1928), Scofield (1936; Scofield and Headley, 1921; Scofield and Wilcox, 1931), Eaton (1935, 1936, 1950), Doneen (1954), Thorne and Thorne (1951), Wilcox (1948, 1955), and U. S. Salinity Laboratory Staff (1954). Although the several proposed methods of classifying irrigation waters differ somewhat, they agree reasonably well with respect to criteria and limits. Christiansen and Lyerly (1952) believe that probably too much emphasis has been placed on an attempt to answer the question, How good is the water? rather than, What can be done with this water? Somewhat similar views have been expressed by Eaton (1958) and Kelley (1962) to the effect that too rigid a dependence on any classscation is questionable. Even so, classification data provide a basis for anticipating with reasonable confidence, the general effect of an irrigation water on the soil and on the plant. The classification recommended by the U. S . Salinity Laboratory Staff (1954) and Wilcox (1955) incorporates many of the desirable features of the early classifications, together with more recent developments based

SALINITY IN RELATION TO IRRIGATION

141

both on research and on field observations. However, this classification is tentative and should be used for genera1 guidance only. The four gen-

erally recognized criteria of irrigation water quality are ( 1 ) total salinity, ( 2 ) sodium, ( 3 ) boron, and ( 4 ) bicarbonate. 1 . Salinity Hazard Total salt concentration is probably the most important single criterion of irrigation water quality. On the basis of electrical conductivity (EC) measurements, waters are divided into four classes: low salinity, medium salinity, high salinity, and very high salinity, the dividing points between classes being 250, 750, and 2250 pmho./cm. This range includes waters that can be used for irrigation of most crops on most soils, to waters that are not suitable for irrigation under ordinary conditions. More than half of the irrigation waters in the western United States have conductivity values of less than 750 pmho./cm. (500 pap.m. dissolved solids) and less than 10 per cent of the waters have conductivities in excess of 2250 pmho./cm. ( 1500 p.p.m. dissolved solids).

2. Sodium Hazard The sodium adsorption ratio (SAR), described in Section 111, B, 3, for soil extracts, is used to evaluate the sodium, or alkali hazard of irrigation waters. This ratio expresses the relative activity of sodium ions in cation-exchange reactions with the soil. The SAR is more significant than the soluble-sodium percentage (SSP) for use as an index of the sodium, or alkali hazard, of the water because it relates more directly to the adsorption of sodium by the soil. However, because irrigation waters become concentrated in the root zone, the SAR will indicate a minimum, but not necessarily the ultimate effect of a particular water on the sodium status of the soil. Waters are divided into four classes with respect to the sodium hazard: low, medium, high, and very high, depending upon the values for SAR and EC. At EC values of 100 pmho./cm., the dividing points are at SAR values of 10, 18, and 26, but with increasing salinity, these SAR values decrease progressively until at 2250 pmho./cm., where the corresponding dividing points are at SAR values of approximately 4, 9, and 14. With respect to sodium hazard, waters range from those that can be used for irrigation on almost all soils to those that are generally unsatisfactory for irrigation. 3. Boron Hazard Boron is very toxic to plants at low concentrations in the soil solution. Because boron tends to accumulate in the soil from low concentrations in the irrigation waters, it is necessary to consider this constituent

142

LOWELL

E. ALLISON

in assessing the quality of irrigation waters. The classification in Table I uses the limits originally proposed by Scofield (1936). TABLE I Permissible Limits of Boron for Several Classes of Irrigation Watersa ~

For irrigation of Boron class

Sensitive crops

< 0.33 0.33 to 0.67 0.67 to 1.00 1.00 to 1.25 > 1.25 Values as parts per million.

1 2 3 4 5 0

Semitolerant crops

Tolerant crops

< 0.67 0.67 to 1.33 1.33 to 2.00 2.00 to 2.50 > 2.50

< 1.00 1.00 to 2.00 2.00 to 3.00 3.OO-to 3.75 > 3.75

The specific toxicity of boron to crops is discussed in Section IV, B, 3. Table VI indicates the relative tolerance of some common crops in relation to the boron present in irrigation waters.

4. Bicarbonate Hazard The bicarbonate ion is primarily important because of its tendency to precipitate calcium, and to some extent magnesium, in the soil solution as their normal carbonates. Carbonate ions are seldom present in waters, but bicarbonate ions may represent an appreciable proportion of the total anions present in irrigation waters. In many tropical waters, bicarbonate is often the main anion present. The effect of the bicarbonate ion concentration on water quality is expressed in terms of the residual sodium carbonate (RSC) concept of Eaton (1950),defined in the foIIowing equation: RSC = (CO,--

+ HCO,-)

- (Ca+

+

+ Mg++)

in which the concentration is expressed in milliequivalents per liter. It is obvious from this equation that as calcium and magnesium are lost from the soil solution by precipitation, the relative proportion of sodium remaining in the water is increased. Thus, the sodium hazard as defined by the SAR (cf. Section 111, B, 3 ) is increased. Laboratory and field studies (Wilcox et al., 1954) have resulted in the tentative conclusion that waters containing less than 1.25meq./l. RSC are probabIy safe for most purposes. Waters containing 1.25to 2.5 meq.fl. are marginal, and those containing more than 2.5 meq./l. of RSC are not suitable for irrigation purposes. However, with good management practices, it should be possible to use some of these marginal waters successfully for irrigation. These practices would be (1)

SALINITY IN RELATION TO IRRIGATION

143

adequate leaching, which would tend to maintain a low level of bicarbonate in the soil solution, and ( 2 ) the application of gypsum or any other source of soluble calcium, to maintain a favorable Ca:Na ratio in the soil solution. Either of these practices would tend to retard sodium accumulation in the exchange complex.

5. Other Hazards a. Lithium. Lithium toxicity to citrus has been observed in California (Bradford, 1953). Where the problem occurs, the lithium content of the irrigation water is 0.05 p.p.m. or more. A survey of 400 waters from throughout the State showed that about one-fourth of these waters contained toxic levels of lithium (0.05 to 0.50 p.p.m.) and that this constituent was usually associated with low magnesium, or high sodium, or both. It is significant that the order of tolerance of lithium-sensitive plants closely parallels their sodium tolerance (Bingham et al., 1964). b. Pollutants. Many substances that are discharged as industrial wastes into surface streams may have phytotoxic properties. Wilcox (1959) reported information on many substances that are known to be toxic to plants. Great caution should be exercised in the use of an irrigation water that is suspected of containing phytotoxic pollutants.

B. INTERPRETATION OF IRRIGATION WATERANALYSES The analysis of a water should provide information on its suitability for irrigation and, in addition, suggest the management practices to be followed. The successful use of a particular water may depend not alone on quality, but also on the drainage characteristics of the soil and on management practices. In appraising waters for irrigation, first consideration should be given to (1) the salinity hazard and ( 2 ) the sodium hazard, followed by the independent characteristics ( 3 ) boron, or other toxic elements, and ( 4 ) bicarbonate, any of which may change the quality rating. Table I1 gives the analysis of several waters that vary widely in their characteristics. The following description of these waters illustrates the management practices required for their successful use for irrigation. Columbia River water (1) represents a typical mountain water, low in total salts and containing chiefly calcium and bicarbonate ions. It presents no hazard for irrigation. Colorado River water ( 2 ) is a high calcium (gypsiferous) water with a moderately high salinity hazard, but a low sodium hazard (EC = 1130 pmho./cm.; SAR = 2.8). It is used successfully under good management for all but the most salt-sensitive crops on about 600,000 acres of land

TABLE I1 Analyses of Some Typical Irrigation Waterso.

Water

Location sampled

E C X Dis106 solved at solids 25°C. (p.p.m.) Ca 85 1.06 140

E b

Milliequivalents per liter Mg 0.35

Na K 0.08 0.03

CO, HCO, SO, - 1.21 0.26

B C1 (p.p.m.) SAR TAF RSC

0.02 0.04 0.1 0.11 0 Grand Coulee, Washington - 2.77 6.25 2.71 0.18 2.8 1.0 0 Yuma, Arizona 1130 753 4.59 2.14 5.05 0.13 (below Imperial Dam) (3) Pecos River a. Alamogordo 1000 741 8.48 1.56 1.52 - - 2.08 8.43 0.99 - 0.7 1.0 0 D m b. Artesia, New 2500 1860 15.60 4.19 9.66 - - 2.44 17.99 8.97 - 3.1 2.5 0 Mexico c. Red Bluff 8620 5900 24.50 12.09 60.46 - - 1.97 34.14 60.10 - 14 8.0 0 Dam ( 4 ) Gila River Gillespie Dam, 8160 5620 14.87 10.94 60.90 0.31 - 3.18 27.90 56.96 2.6 17 7.6 0 Arizona ( 5 ) Well Water Van Horn, Texas 500 476 0.10 0.10 5.90 0.10 - 3.80 1.70 0.50 - 19 0.4 3.60 No. 85 1950 1210 1.00 1.00 16.00 0.50 - 3.50 3.70 11.20 - 16 1.6 1.50 ( 6 ) Well Water El Paso, Texas No. 55 6000 3775 22.20 4.80 30.90 0.30 - 7.30 18.30 34.50 - 8.4 5.4 0 ( 7 ) Well Water El Paso, Texas No. 49 ( 8 ) RuziziRiver Belgian Congo - 918 0.45 5.78 4.72 1.87 3.48 7.89 1.00 0.71 - 2.7 1.2 5.14 Analyses were obtained as follows: Waters (1) to ( 4 ) from U. S. G. S . Water Supply Paper 1575 (1938); Waters ( 5 ) to ( 7 ) from Texas Agr. Ezpt. Sta. Circ. 132 (1952); and Water (8) by private communication. b EC, electrical conductivity; SAR, sodium adsorption ratio; TAF, tons per acre-foot; RSC, residual sodium carbonate. (1) Columbia River ( 2 ) Colorado River

3 m

8 Z

SALINITY IN RELATION TO IRRIGATION

145

in Southern California. Adequate drainage is a first requirement for success with this water. Pecos River water (3) is moderately saline at its source (EC = 1000 pmho./cm.) in northern New Mexico, but it rapidly gathers calcium, sodium, sulfates, and chlorides and becomes extremely saline as it flows southward toward the Rio Grande (EC = 8620 pmho./cm. at Orla, Texas). This water is gypsiferous, which is a point in its favor for irrigation. Water (3b) is used with moderate success in the Pecos Valley in New Mexico where the soils are permeable and well drained, but it requires high leaching for salinity control. Gila River water (4)has three major hazards for irrigation-a very high salt content ( EC = 8160 pmho./cm. ), high sodium ( SAR = 17), and high boron (2.6 p.p.m.). The use of this water is obviously limited to the growth of crops having both high-salt, and high-boron tolerance under a high-leaching regime. This water can be used only on very permeable, well-drained land. Well waters ( 5 ) , (S), and ( 7 ) represent the extreme variations in concentration and composition often observed in pumped waters. Water ( 5 ) , although low in salt content, is very high in percentage of sodium (SAR = 20) and bicarbonate (RSC = 3.60 meq./l.). The quality of this water for irrigation is questionable in view of its extremely low calcium and magnesium content. Since the chief anion is bicarbonate, gypsum should be added either to the water or to the soil to prevent high-sodium accumulation in the exchange complex with resulting loss of permeability. This water should be applied only to coarse-textured soils with at least moderate leaching for removal of released sodium salts from the root zone. Water ( 6 ) is a high-salt, high-sodium water in which chloride is the chief anion. Owing to its high-sodium hazard (SAR = 16), gypsum should be applied occasionally, followed by adequate leaching to prevent loss of permeability. Water (7) resembles Gila River water ( 4 ) except that it presents no boron hazard. Ruzizi River water, ( 7 ) of the Belgian Congo is included to illustrate a problem often encountered with tropical waters. This water is high in both bicarbonate and carbonate and is being considered for limited, dry-season irrigation of sugar cane in an area receiving 30 inches of rainfall annually. The soils to be irrigated are coarse-textured latosols. This water has a moderate salinity (EC not given), and a rather lowsodium hazard (SAR = 2.7), but, because of its high content of both carbonate and bicarbonate (RSC = 5.14 meq./l.), the calcium content is extremely low. If used extensively for irrigation, considerable difficulty may arise, especially on medium- or fine-textured soil, owing to rapid

146

LOWELL E. ALLISON

sodium accumulation. Moreover, nutritional disturbances may arise due to the extremely low ca1cium:magnesium ratio of 0.08. In this case, gypsum should be applied in generous amounts to the soil. However, if this water is used only for Iimited (supplemental) irrigation on coarsetextured soils, under conditions of a moderate to high rainfall, little or no difficulty should be encountered. 111. Effect of Salts on Soils

A. CHARACTERISTICS OF SALT-AFFECTEDSOILS Salt-affected soils are characterized by the fact that they contain sufficient soluble salts or exchangeable sodium, or both, to restrict plant growth. Agriculturally, they are regarded as a class of problem soils that require special remedial measures and management practices. A knowledge of the chemical and physical characteristics of the several kinds of salt-affected soils is essential to serve as a basis for their diagnosis, treatment, and management ( Richards and Hayward, 1957). 1 . Source and Composition of Salts The ultimate source of all the salt constituents found in soils is the primary minerals in the exposed rocks of the earth's crust. The processes of chemical weathering, which include hydrolysis, hydration, solution, oxidation, and carbonation, gradually release these salt constituents to the surrounding water. The ocean may be the source of salts, as where marine deposits have been uplifted and drainage therefrom affects sources of irrigation water. The Mancos shales in Colorado, Wyoming, and Utah are examples of saline marine deposits. The ocean may be a direct source of so-called cyclic salt along the seashore through windblown sprays (Teakle, 1937). However, the main source of salt affecting irrigation agriculture is from surface and ground waters. The soluble salts that effectively contribute to soil salinity consist mostly of various proportions of the cations calcium, magnesium, and sodium and of the anions chloride, sulfate, bicarbonate, and sometimes carbonate. Potassium occurs to a lesser extent than any of the other three cations. Among the anions, bicarbonate and carbonate are usually present in minor amounts as compared to chloride and sulfate. Bicarbonate ions form as a result of the solution of carbon dioxide in water, which may be of atmospheric or biological origin. Bicarbonate and carbonate ions are interrelated, the relative amounts of each being a function of the pH value of the soil solution. Appreciable amounts of carbonate ions occur in soils only at pH values of 9 or higher.

SALINITY IN RELATION TO IRRIGATION

147

Boron, owing to its marked toxicity to plants at concentrations of only a few parts per million, deserves mention, even though its salts make no important contribution to total soil salinity. The reclamation of boron-rich soils is discussed in Section V, B, and the toxic effect of boron on plant growth is discussed in Section IV, B, 3.

2. Salination and Alkalization a. Salination. Salination is the process whereby soluble salts accumulate in the root zone of the soil. Irrigation waters contain solubIe salts varying from 0.1 to 5 tons/ acre-foot and the annual application may range from 3 to 5 feet. In the absence of leaching, the salt contained in the applied irrigation water is deposited in the root zone of the soil due to evapotranspiration, or consumptive use. Restricted drainage is a factor that contributes to the salination of irrigated soils. This may involve the presence of a high ground water table, or low permeability of the soil, or both. Where a high water table exists within 4 or 5 feet of the soil surface, upward movement of saline ground water, combined with the evaporation of applied irrigation water, may result in the formation of a saline soil. In severe cases, salts may accumulate at the soil surface, as shown in Fig. 1, with total loss of production. Low permeability of the soil, causing waterlogging, may be due to an extremely fine-textured condition of the soil to well below the root zone, or to the presence of an impermeable barrier below the root zone. The latter may consist of a clay lens (clay pan), caliche layer, or a silica hardpan. De Sigmond (1924) considered the presence of an impermeable layer essential for the formation of the saline soils in Hungary. Many thousands of acres of irrigated lands in the United States are affected by high water table conditions and require an effective drainage system for economic crop production. b. Alkalixation. This is the process whereby the exchangeable-sodium content of a soil is increased, leading to the formation of a sodic soil. It involves both salination and change in composition of the accumulated salts. Calcium and magnesium are the dominant cations found in normal soils in arid regions. However, as soluble salts accumulate from irrigation waters and become more concentrated in the soil, owing to consumptive use and to the lack of leaching, certain composition changes occur. The solubility limits of calcium sulfate, calcium carbonate, and magnesium carbonate are exceeded, causing calcium and magnesium to

M

FIG.1. Border-irrigated alfalfa field that became highly salinized as a result of poor management, i.e., lack of adequate drainage for salinity control.

SALINITY IN RELATION TO IRRIGATION

149

precipitate. This causes a corresponding increase in the relative proportion of sodium in the soil solution, i.e., the soluble-sodium percentage (SSP) increases. Due to the dynamic equilibrium between soluble and adsorbed ions, sodium replaces some of the calcium and magnesium originally present on the exchange complex of the soil (Kelley, 1951). In general, half or more of the soluble cations must be sodium ( SSP > 50) before appreciable amounts of this ion are absorbed by the cationexchange complex. In some saline soils, practically all the soluble cations are sodium and, therefore, sodium is the predominant adsorbed cation. As long as soluble salts are present in the soil solution in appreciable quantities, the soil ( saline-sodic) remains flocculated and permeable, and the pH is less than 8.5. If the soluble salts are removed by leaching, the soil (sodic) may become very impermeable because of the dispersing effect of the adsorbed sodium ion on the exchange complex. Some of the sodium hydrolyzes off the exchange complex, forming traces of sodium hydroxide, and the pH increases, often as high as 10. Arany (1956) pointed out the importance of the anions associated with sodium in the formation of sodic soils. The alkaline sodium salts ( bicarbonate, carbonate, silicate ) favor nearly complete displacement of calcium by sodium because of the low solubility of the corresponding calcium salts. Conversely, the neutral sulfate and chloride salts of sodium produce only partial replacement of calcium by sodium. Therefore, the rate of alkalization is more rapid for the basic than for the neutral salts. 3. Saline Soils

Saline soils contain soluble salts in such quantity that they interfere with the growth of most crop plants. By definition, the electrical conductivity of a saturation extract (EC,) of a saline soil is greater than 4 mmho./cm. and the exchangeable-sodium percentage (ESP) is less than 15. The pH of the saturated soil is usually less than 8.5, but if the soil is gypsiferous the pH seldom exceeds 8.2. Saline soils correspond to Hilgards “white alkali” and to “solonchak” as defined by Russian scientists. The salts present in saline soils consist mainly of neutral salts, such as the chlorides and sulfates of sodium, calcium, and magnesium. Sodium seldom comprises more than half of the soluble cations and, therefore, it is not adsorbed to any significant extent in the soil exchange complex. Because saline soils are generally flocculated, their tillage properties and permeability to water are often equal to or better than those of similar nonsaline soils. Restricted plant growth is almost directly related to the total salt concentration of the soil solution and is largely independ-

150

LOWELL E. ALLISON

ent of the kind of salts present. If drainage is adequate and the excess salts are removed by leaching, saline soils become normal soils. Saline soils may be recognized by the presence of a white inflorescence on the surface (Fig. 1) or by an oily looking surface devoid of vegetation. Plants indicate the presence of salinity by stunted growth and sometimes by considerable variability in size within the field. The foliage is often a deep green color with occasional tipburn or marginal burn on the leaves. 4. Saline-Sodic Soils

Saline-sodic soils contain sufficient quantities of both soluble salts and adsorbed sodium to reduce the yield of most plants. For the purpose of definition, the exchangeable-sodium percentage is greater than 15, and the electrical conductivity of the saturation extract is greater than 4 mmho./cm. The pH reading of the saturated soil is usually less than 8.5, but if gypsum is present in appreciable quantities, the pH may be as low as 8.2. These soils form as the result of the combined processes of salination and alkalization ( accumulation of exchangeable sodium). As long as excess salts are present, the appearance (Fig. 1) and properties of these soils are generally similar to those of saline soils. If gypsum is present, leaching converts these soils to the nonsaline condition due to the replacement of exchangeable sodium by calcium resulting from the solution of gypsum during the leaching process. However, if gypsum is neither present nor supplied as an amendment, leaching causes the soil to become strongly alkaline ( p H above 8 . 5 ) , the colloids disperse, and the soil becomes unfavorable for the entry and movement of water and for tillage. Although the return of soluble salts may restore the soil to a flocculated condition, the management of saline-sodic soils continues to be a problem until the excess salts and exchangeable sodium are removed from the root zone and a favorable physical condition of the soil is reestablished.

5. Sodic Soils Sodic soils contain sufficient exchangeable sodium to interfere with the growth of most crop plants, but do not contain appreciable quantities of soluble salts. By definition, the ESP is greater than 15 and the EC, is less than 4 mmho./cm. The pH reading of the saturated soil is usually greater than 8.5 and sometimes as high as 10. Sodic soils correspond to Hilgard’s “black alkali” and in some cases to “solonetz,” as the latter term is used by Russian scientists. The term sodic us used here instead of alkali because the latter has

SALINITY IN RELATION TO IRRIGATION

151

had various and somewhat indefinite meanings. Some scientists have used the term alkali to include soils affected both by salts and by exchangeable sodium (Hilgard, 1906), and in this respect its meaning is similar to the term salt-affected. Others have used “alkali” to indicate soils affected mainly by exchangeable sodium (U. S. Salinity Laboratory Staff, 1954). In Iraq, farmers use the term alkali to describe any soil having the appearance of ashes in the surface, regardless of the chemical condition (Buringh, 1960). Thus, the term sodic describes the soil condition in terms of its cause, that of exchangeable sodium. The soil solution of sodic soiIs is relatively low in soluble salts and the ionic composition differs considerably from that of saline soils. The predominant cation is sodium because, at high pH and in the presence of the carbonate ion, calcium and magnesium are largely precipitated as calcium and magnesium carbonate. The anions present consist mostly of chloride, sulfate, and bicarbonate, with small to moderate amounts of carbonate, depending on the pH of the soil. If carbonates are present in detectable amounts in the saturation extract, then the pH must be above 9. The exchangeable sodium present has a marked influence on the chemical and physical properties of sodic soils. As the proportion of exchangeable sodium increases, the soil tends to become dispersed, less permeable to water, and exhibits poor tilth. Sodic soils are usually plastic and sticky when wet and form large clods or crusts on drying. Their crusting tendency is a serious hazard to seedling emergence, and it often accounts for a poor stand of crops, causing reduced yield. A special method of reclaiming high-sodium soils is discussed in Section V, C, 5.

6. Salination Due to Euaporation Considerable salt may accumulate at the soil surface by evaporation from a shallow, saline water table during the fallow period between crops. This is especially significant if the intercrop period is long and the cIimate is arid. Donnan et al. ( 1954) and Bradshaw and Donnan (1953) recommended that newly reclaimed, fine-textured land having a water table near the root zone should be kept under production (irrigated) continuously to prevent salt from returning to the surface by capillary rise from the water table. Recent studies of evaporation losses in the Imperial Valley by Doering (1963) provide data to illustrate salination of land during a fallow period. For this purpose, the following revised formula (U. S. Salinity Laboratory, 1954, p. 36) is used in which groundwater (gw) is substituted for irrigation water (iw).

152

LOWELL E. ALLISON

EC, = D m D,

EC,, d, 100 SP d,

D, = depth of groundwater evaporated D, = depth of soil in which salts accumulate EC, d, d8 SP

= EC of groundwater evaporated

= density of water = bulk density of soil = saturation percentage of soil

The soil studied was a silty clay to a depth of 60 cm. over silt loam, with a water table at 150 cm. The bulk density of this silty clay soil was 1.4 gJcm.3 and the SP was 63. Water loss determined by the chloride accumulation method, which agreed favorably with evaporation from a special evaporimeter, was about 0.09 cm./day, or 33 cm./year. For a summer-fallow period of 4 months between regular crops in the rotation, the corresponding water loss was 11cm. Based on a groundwater salinity of EC = 10 mmho./cm., which is not uncommon, the increase in EC, of the surface 30 cm. of silty clay soil was calculated to be 4.1 mmho./ cm. for the 4-month period. Hence, the salinity of the surface soil might increase twofold as the result of capillary rise of salt from a shallow water table during a 4-month fallow period. The assumption has been made that all the salts remain in solution. If gypsum and/or carbonates precipitate from solution, the change in EC, is reduced accordingly. The application of a few irrigations during a prolonged intercrop period might prove profitable in terms of maintaining a more favorable salt balance for salt-sensitive crops to follow. B. DIAGNOSIS OF SALINEAND SODICSOILS

An effective diagnostic system for appraising the salinity and sodium status of soils must take into consideration field moisture conditions because plants are responsive to the salt concentration of the soil solution, which reflects osmotic pressure conditions. Ideally, salinity and sodium measurements would be most reliable if made on extracts of the soil solutions within field moisture range. With pressure-membrane equipment, this is possible but the difficulty of obtaining such extracts makes this system prohibitive for routine use (U. S . Salinity Laboratory Staff, 1954). A test kit for making salinity and sodium tests on soils and waters has been described (Richards et al., 1956) and is commercially available. Several systems, presently in use, classify soils on the basis of total salt content. In Russia, soils are considered slightly saline at 0.3 per cent salt, moderately saline at 0.7 per cent, and strongly saline at 1.0 per cent, etc. These systems, while providing a measure of total salts in the soil,

SALINITY IN RELATION TO IRRIGATION

153

do not evaluate salinity in terms of the force to which plants are responsive-that of the osmotic pressure of the soil solution. 1. Soil Sampling Generally, the major root activity occurs in the less saline parts of the soil, and this fact should be borne in mind in determining the salt and sodium status of the soil with reference to plant response. For instance, samples collected from the surface soil around the base (ridge) of row-planted crops at later stages of growth may contain 5 per cent salt or more, representing an EC, of 50 mmho./cm. or higher. This condition represents an accumulation of salt due to moisture movement into the ridge where it evaporates, not the salt concentration in the active root zone under the irrigation furrow. Therefore, in correlating crop growth with salinity, soil samples should be taken from the active root zone (furrow), which is uncontaminated by surface encrustations of salt. The larger the number of surface samples and the more carefully they are selected, the more accurate will be the appraisal; but with experience a satisfactory appraisal may be made on the analyses of relatively few samples. Because often the salt content varies with depth, it is advisable to take a few profile samples by genetic horizons or by depositional layers, if present. However, it is recognized that salt does not necessarily accumulate according to genetic horizons. In the absence of layering in the profile, a sample should be taken to plow depth, usually to 6 or 7 inches, and succeeding samples may be taken at intervals of 6 to 18, 18 to 36, and 36 to 72 inches, or at other convenient depths, depending on depth of the root zone, the nature of the problem, and the detail required. The size of sample will depend on the number of measurements to be made, but usually 1000 g. of soil per sample is adequate.

2. Determination of Salinity Hazard a. Saturation-extract method. The electrical conductivity of the saturation extract ( EC,) expressed in millimhos per centimeter at 25°C. is recommended for appraising the salinity hazard in relation to plant growth (U. S. Salinity Laboratory Staff, 1954). The unique feature of this method is that the salt concentration in the saturation extract is about one-half the concentration of the soil solution at the upper end (field capacity), and about one-fourth the concentration at the lower end (permanent-wilting percentage) of the field moisture range for plant growth. This relationship makes it possible to interpret salinity measurements ( a t saturation) directly in terms of

154

LOWELL E. ALLISON

field moisture conditions, which is not possible when dilution extracts are used. The procedure involves making a saturated-soil paste by stirring the soil, during the addition of distilled water, until a characteristic end point is reached. The end point is reasonably definite, and with a little training good agreement can be obtained among various operators. A suction filter is used to obtain a sufficient quantity of extract for making the conductivity measurement and for the determination of soluble cations and anions, if desired. EC, values may be related to plant growth by reference to Table 111. b. Dilution-extract method. The estimation of salinity hazard from EC measurements made on dilution extracts at 1:1,15, or 1:lOsoi1:water ratios may have certain advantages, but the limitations of the method should be clearly understood. For rapid salinity determinations, these higher dilutions are often convenient where repeated samplings are to be made on the same textural soil (as in plot experiments) to determine the change in salinity with time or treatment. However, salinity measurements made on dilutions higher than saturation do not relate satisfactorily with salt-tolerance data given in Table IV. For determination of ESP, dilution extracts are definitely unsatisfactory because, with increase in dilution, calcium replaces sodium in the exchange complex in accordance with the valence-dilution principle. Therefore, the soluble-sodium percentage ( SSP ) increases with dilution, resulting in lower values for exchangeable sodium. This gives ESP values that are much farther out of line with the actual sodium status of the soil under growing plants than is provided when determinations are made at the SP.

3. Determination of Sodium Hazard The exchangeable-sodium percentage (ESP) has long been used to indicate the sodium status or hazard of salt-affected soils. In fact, the definition of sodic soils is based on the ESP determination. The determination of ESP by conventional chemical methods requires the following separate determinations: (1) soluble sodium on a saturation extract of the soil, ( 2 ) total sodium on an ammonium acetate extract of the soil, and ( 3 ) the cation-exchange capacity of the soil. When so determined, much time is required and appreciable errors may occur, owing to the indirect nature of its determination as indicated by the formula: ESP = (total sodium - soluble sodium) x 100/cation-exchange capacity. Recent research (U. S . Salinity Laboratory Staff, 1954) demonstrated that the most reliable index of sodium status is the sodium adsorption ratio (SAR). It is calculated value and is defined as

155

SALINITY IN RELATION TO IRRIGATION

Na+

SAR =

+

Ca++

+ hfg+i/2

where concentrations are expressed in milliequivalents per liter. This ratio is based on cation-exchange equations of the mass-action type ( Gapon, 1933). N a* M EO./ L

Cia'++

.

Mf'

MEO./L.

/

4.

'2

7

EXCHANGEABLE SODIUM PERCENTAGE

A0 +

i 30

31

40

45

-80

A

B

FIG.2. Nomogram for determining the sodium adsorption ratio (SAR) value of a saturation extract and for estimating the corresponding exchangeable-sodium percentage (ESP) value of soil in equilibrium with the extract.

The unique feature of the SAR is that it takes into consideration changes in both concentration and composition of the salts present in the soil solution that are in equilibrium with the soil. Previously, no formula or equation made this possible. Moreover, the results of a recent

156

LOWELL E. ALLISON

study involving 218 soil samples from 24 countries (Bower, 1962), indicated that the SAR can be used with confidence in all arid countries of the world to determine the sodium status of salt-affected soils. To determine the SAR, it is necessary merely to prepare a saturation extract of the soil, determine the concentration in milliequivalents per liter of the caIcium, magnesium (EDTA titration), and sodium (flame photometer) ions in the extract, and substitute these concentrations in the nomogram shown in Fig. 2. The ESP is then interpreted from the SAR line of this nomogram. Besides flame photometry, sodium may be determined by means of a special glass electrode (Bower, 1960). IV. Effect of Salts an Crops

A. GENERALAND SPECIFICEFFECTS Salts affect plant growth directly (1) by increasing the osmotic pressure of the soil solution, ( 2 ) by accumulating certain ions in toxic concentrations in plant tissue, and ( 3 ) by altering the plant’s mineral nutrition. Plant growth responses to salinity have been discussed by Hayward and Wadleigh ( 1949), Grillot ( 1956), Bernstein and Hayward (1958), and Bernstein ( 1962). 1. Kind and Function of Ions The principal ions in the soil solution of salt-affected soils are calcium, magnesium, sodium, potassium, chloride, sulfate, and bicarbonate. Some ions (calcium, magnesium, potassium, and sulfate) provide major essential elements for growth, and still others (boron and lithium) may be toxic at very low concentrations. Soviet scientists attach considerable importance to the composition of the salts in soils in relation to plant growth, but the emphasis is on the anionic composition (Bower et al., 1962). Thus, chloride salinity and sulfate salinity are frequently used terms in their soil science literature. American scientists, in contrast, emphasize the cationic composition of the salts, because the cations are known to undergo exchange reactions with the soil and thereby control its chemical and physical properties.

2. Osmotic Pressure and Plant Growth Gauch and Wadleigh ( 1944) demonstrated progressive growth reduction of beans in solution culture with increasing salinity, and the equivalent effects of sodium and calcium salts regardless of the anion (C1 or SO,) present. This indicates that the total concentration of solute particles in the solution, rather than their chemical nature, is mainly

SALINITY IN RELATION TO IRRIGATION

157

responsible for the inhibitory effects of saline solutions on the growth of crop plants. The cause of growth reduction, associated with increasing osmotic pressure (OP) of the rooting medium, has been attributed to decreased water entry or availability, based on observations of Hayward and Spurr (1943, 1944). However, recent evidence by Bernstein (1961) indicates that water-absorption capacity is relatively unaffected by salinity. The reduced growth associated with osmotic stress is attributed to the process of building up the OP of developing cells (which is contingent upon accumulation of solutes) to meet the increasing OP of the rooting medium and still maintain turgor. This theory suggests that salt tolerance may be defined as the degree to which osmotic adjustment can be made without sacrifice in growth. 3. Specific Ion and Other Efects

Certain ions exert specific effects that depress growth and yields independent of osmotic effects. These specific ion effects may be toxic or nutritional in nature. A toxic effect is considered to be due to the presence of an ion in the solution which causes direct damage to the plant. Injury is usually associated with the accumulation of harmful concentrations of the toxic ion in the plant tissue, which may show no other significant change in mineral composition. a. Sodium and chloride. The toxicity of sodium and chloride ions may be the major factor in salt damage to specifically sensitive fruit crops (Bernstein and Hayward, 1958). Direct toxicities by these two ions have been demonstrated in stone fruits for such crops as peaches, plums, apricots, and also for other crops, for example, citrus, avocados, grapes, strawberries, and blackberries. These toxic effects may occur at osmotic concentrations in the substrate which are below the level that normally restrict yield for these crops. Chloride may accumulate in the leaves to about 1 or 2 per cent of the dry weight when this anion occurs in the root medium in only moderate concentrations (700 p.p.m. to 1500 p.p.m. in the soil moisture, or nutrient solution). At these concentrations in the leaves, marginal burn develops, leading ultimately to leaf drop, twig dieback, and even death of the plant. The selection of rootstocks for characteristics of low chloride, and also low boron accumulation, offers considerable promise for fruit-crop improvement. Sodium accumulation in leaves less than 0.05 per cent of the dry leaf weight produces similar leaf-burn symptoms and extensive injury. Lilleland et al. (1945) observed sodium injury in almond trees growing in

158

LOWELL E. ALLISON

essentially nonsaline soil containing less than 5 per cent of exchangeable sodium. b. Boron. The specific effects of boron on plant growth are discussed in Section IV, B, 3. c. Bicarbonate. The bicarbonate ion in excess may be toxic to plants, although sensitivity varies with different species. As an example, beans and Dallisgrass are very sensitive to the presence of bicarbonate ions in the substrate, whereas Rhodesgrass and beets are relatively tolerant (Wadleigh and Brown, 1952; Brown and Wadleigh, 1955; Gauch and Wadleigh, 1951). Bicarbonate excess may give rise to ion chlorosis, a problem that has been reviewed by Brown (1956). The bicarbonate ion does not frequently occur in sufficient concentration in irrigation waters to produce direct toxic effects, although bicarbonate-induced chlorosis in apple orchards has been observed in Washington (Harley and Lindner, 1945). d. Nutritional effects. Salinity may inhibit growth in plants because of effects on plant nutrition. Because plants vary widely in their nutrient requirements and in their ability to absorb specific nutrients, the effects of salinity on nutrition differ markedly from species to species. High concentrations of sulfate generally decrease the uptake of calcium while promoting the uptake of sodium (Hayward and Wadleigh, 1949). Some lettuce varieties are known to develop calcium-deficiency symptoms in the presence of high sulfate concentrations (Doneen and Grogan, 1954). By promoting the uptake of the sodium ion, sulfate may induce sodium toxicity in susceptible species (Brown et al., 1953). Salinity also tends to increase the incidence of blossom-end rot of tomatoes (Geraldson, 1960), a disease likewise attributed to calcium deficiency. On the other hand, high concentrations of calcium may restrict the uptake of essential potassium by beans and some carrot varieties ( Bernstein and Hayward, 1958). AND BORON TOLERANCE OF CROPS B. SALT,SODIUM,

1. Relative Tolerance of Crops to Salts Salt tolerance refers to the ability of plants to tolerate concentrations of soluble salts in the root medium. Information on the salt tolerance of crops is important in diagnosing suspected salt injury in the field, in selecting tolerant crops for saline lands, and in determining irrigation and drainage requirements and management practices for salt-affected land. Table I11 indicates the response of crops to increasing salinity, expressed as EC, values. The salt tolerance of many crops has been appraised, and the data

SAJJNITY

159

IN RELATION TO IRRIGATION

were initially reported in terms of the salinity level (EC,) that would be expected to give a 50 per cent decrease in yield (U. S. Salinity Laboratory Staff, 1954, p. 67). More recently, these data representing field, vegetable, and forage crops have been revised and published in popular bulletin form (Bernstein, 1958, 1959, 1960).

Salinity effects Yields of very Yields of many Only tolerant Only a few very mostly neglisensitive crops crops yield tolerant crops gible crops may be restricted satisfactorily yield satisfacrestricted torily

Electrical conductivity of saturation extract in millimhos per centimeter (EC, at 25°C.

x

103)

Table IV gives data from these publications for the salinity levels that would permit yields equivalent to 85 or 90 per cent of the yields obtained on comparable nonsaline soil. The position of each crop in this table reflects its relative salt tolerance under management practices that are customarily employed when this crop is grown under irrigation agriculture. Crops are listed within each group in the order of decreasing salt tolerance, but a difference of two or three places in the column may not be significant in some cases. Significant varietal differences were found for Bermudagrass, barley, and smooth brome, whereas for truck crops, such as green beans, lettuce, onions, and carrots, varietal differences were of little or no practical significance. In applying the information in Table IV, it is important to remember that climatic conditions may influence the sensitivity of plants to salinity. Onions, for example, are more sensitive to a given level of salinity in hot, dry areas than in cooler, more humid areas (Magistad et al., 1943). Consequently, information on salt-tolerant crops should be evaluated with reference to the condition under which they are to be grown. a. Stage of development. The listing of crops according to salt tolerance (Table IV) fails to reveal certain specific problems because some plants are especially sensitive to salinity during certain stages of development and tolerant at other stages (Bernstein and Hayward, 1958; Bernstein, 1961). For example, rice is quite tolerant during germination but becomes very sensitive during the seedling stage, and again somewhat so during the fertilization of the florets (Pearson and Bernstein, 1959). Corn appears to be appreciably more tolerant during germination than at later stages of growth. Sugar beets, on the other hand, can

160

LOWELL E. ALLISON

tolerate salinity levels of only about 4 mmho./cm. in the saturation extract during germination but can easily tolerate three times this salt level once the young seedlings are well established. TABLE IV Relative Tolerance of Crops to Salinity Arranged According to Decreasing Tolerance within Groups CroD Field

Tolerant Barley Sugar beet Rape Cotton

Truck

8-5 mmho./cm. Garden beets Kale Asparagus Spinach

Forage

Moderatelv tolerant

Sensitive

8-4 mmho./cm.

3-2 mmho./cm.

12-8 mmho./cm.

12-6 mmho./cm.

Rye Wheat Oats Sorghum Sorgo Soybeans Sesbania

Broadbean Corn Rice Flax Sunflower Castorbean

5-3 mmho./cm. Tomato Broccoli Cabbage Cauliflower Lettuce Sweet corn Potatoes Sweet potato

Fruit

8 mmho./cm. Date palm

3-2 mmho./cm.

Yam Bell pepper Carrot Onion Peas Cantaloupe Squash Cucumber

6-3 mmho./cm.

Sweetclover Saltgrass Perennial ryegrass Bennudagrass Tall wheatgrass Mountain brome Rhodesgrass Harding grass Canada wildrye Beardless wildrye Western wheatgrass Strawberry clover Tall fescue Dallisgrass Barley (hay) Sudangrass Birdsfoot trefoil Hubam clover Alfalfa Rye (hay) Wheat (hay)

6-3 mmho./cm. Pomegranate Fig Olive Grape

Field beans

Radish Celery Green beans

3-2 mmho./cm.

Oats (hay) Orchardgrass Blue grama Meadow fescue Reed canary Big trefoil Smooth brome Tall meadow oatgrass Milkvetch Sourclover

White dutch clover Meadow foxtail Alsike clover Red clover Ladino clover Burnet

3-1.5 mmho./cm. Orange Grapefruit Lemon Apple Pear Plum Prune Almond

Peach Apricot Boysenberries Blackberries Raspberries Avocado Strawberry

SALINITY IN RELATION TO IRRIGATION

161

Many crops are more sensitive during the seedling stage than at later stages of growth. Rice can germinate at salinities up to 10 or 15 mmho./cm., but the plants usually die if the salinity is in excess of 5 or 6 mmho./cm. during the seedling stage (Pearson and Ayers, 1958). Barley is like rice in being more sensitive to salinity during the seedling stage than at earlier or later growth stages. Occasionally, special practices may be required to permit a crop to survive during phases of minimum salt tolerance. For example, the paddy field is sometimes drained and refilled with fresh water to lower the salinity during the critical, sensitive flowering stage of rice. Special bedding practices have been developed to minimize salt accumulation around the germinating seeds, the condition responsible for poor stands of furrow-irrigated row crops. These practices are discussed in Section VI, c. 2. Relative Tolerance of Crops to Sodium ~

The chemical and physical characteristics of sodic soils were described in Section 111, A, 2-5, and the literature on the factors affecting plant growth on such soils has been reviewed (Bernstein and Hayward, 1958). The sodium-tolerance data recently reported by Pearson ( 1960) for several important agricultural crops are shown in Table V. The response of plants to exchangeable sodium is complicated by a number of factors, such as direct toxic effects in the case of sodiumsensitive species, indirect effects due to structural deterioration in sodic soils, and nutritional effects. Ratner (1935, 1944) pointed out that decreased absorption of calcium by plants is often due to the presence of increasing levels of exchangeable sodium in the soil. Sodium-sensitive plants, such as avocado, almond, citrus and stone fruits, exhibit characteristic leaf-burn symptoms when sodium accumulation in leaves becomes excessive (Martin and Bingham, 1954; Jones et al., 1957). They may become injured at ESP levels too low to give unfavorable soil physical conditions. Sodium appears to be directly toxic for these species, since no evidence was detected of the calcium deficiency or any other nutrient unbalance. Plants normally tolerant to sodium may be inhibited in their growth primarily by the adverse physical conditions in sodic soils, which restricts moisture transmission and aeration, and may physically impede root elongation and seedling emergence. Those sodium-tolerant crops that may be primarily affected by poor soil structure include beets, Rhodesgrass, cotton, tomatoes, and some of the other grain crops. Moderately tolerant crops, alfalfa, clover, Dallisgrass and certain others, may exhibit a nutritional component in the overall growth reduction, as well as effects due to poor soil structure.

162

LOWELL E. ALLISON

From their studies in the Kanpur District of India, Agarwal and Yadav (1956) suggested an additive effect of exchangeable sodium and salinity on crop growth. More recently, Bernstein (1962) found that when a good physical condition of the root zone of the soil is maintained by synthetic soil conditioners, the ESP exerts a pronounced effect on growth at low salt concentrations but not at higher salt concentrations in the root medium. Lagerwerff and Holland (1960) found TABLE V Tolerance of Various Crops to Exchangeable-Sodium Percentage ( ESP) Range of ESP values affecting growth Extremely sensitive

2-10

Sensitive

10-20

Moderately tolerant

20-40

Tolerant

40-60

Most tolerant

> 60

Crop Deciduous fruits Nuts Citrus Avocado Beans

Clover Oats Tall fescue Rice Dallisgrass Wheat Cotton Alfalfa Barley Tomatoes Beets Crested wheatgrass Fairway wheatgrass Tall wheatgrass Rhodesgrass

Growth response under field conditions Sodium toxicity symptoms at low ESP

Stunted growth at low ESP, despite favorable soil structure Stunted growth due to nutritional factors and poor soil structure

Stunted growth, usually due to poor soil structure

Stunted growth, usually due to poor soil structure

similar effects in sand cultures containing exchange resins to simulate the adsorption properties of the soil. Thus, an additive effect does not occur and, under saline conditions where the physical condition of the soil is not too bad, the ESP within limits becomes less critical than under nonsaline conditions. However, yields under such saline-sodic conditions are restricted eventually by the salinity level.

3. Relative Tolerance of Crops to Boron Boron is essential for plant growth at very low concentrations, but it becomes toxic at only a few parts per million in the soil solution. Its presence in both waters and soils has been extensively studied (Eaton,

163

SALINITY IN RELATION TO IRRIGATION

1935; Wilcox, 1955, 1960; Kelley and Brown, 1928). Table VI gives the relative boron tolerance of several crops, as determined by Eaton (1935) and modified by Wilcox (1958, 1960) based on more recent field observations. The chief source of boron in agriculture is from irrigation water since all natural waters contain some boron. Only a few surface streams are contaminated, but a large number of well waters are high in boron. When used for irrigation, these waters produce toxic concentrations of TABLE VI Relative Tolerance of Crops to Boron Tolerant

a

Semitolerant

Sensitive

4 p.p.m.a

2 p.p.m,

1 p.p.m.

Asparagus Date palm Sugar beet Mange1 Garden beet Alfalfa Broadbean Onion Turnip Cabbage Lettuce Carrot

Potato Cotton Tomato Radish Field pea Olive Barley Wheat Corn Milo Oat Pumpkin Bell pepper Sweet potato Lima bean

Pecan Walnut Navy bean Plum Pear Apple Grape Kadota fig Cherry Peach Apricot Blackberry Orange Avocado Grapefruit Lemon

2 p.p.m.a

1 p.p.m.

0.3 p.p.m.

Indicates limits of tolerance for boron in irrigation waters.

boron in the soil, especially under conditions of poor drainage. Boron has been found in toxic concentrations in the soils of many arid regions of the world. In the United States, it is confined almost exclusively to the irrigated areas of the arid West. The total area in which boron toxicity is a problem is not large, but the injury sometimes is very severe. The classification of waters on the basis of the boron hazard is given in Table I. Boron is translocated to the leaves where it accumulates in the tip and in the margin. Injury can be recognized by the very characteristic patterns developed on leaves of many crops (Wilcox, 1960). Tip burn, which starts with yellowing followed by browning and death of the leaf tip, affects such plants as lemon, orange, grapefruit, and black walnut. Both tip and marginal bum are characteristic symptoms on cereals and

164

LOWELL E. ALLISON

grasses, such as oats, milo, corn, wheat, and barley. Marginal burn occurs on the leaves of many broadleaf plants, including bush berries, alfalfa, and cotton. However, many crops show no characteristic pattern of boron injury. This includes grapes, figs, beans, bell peppers, tomatoes, potatoes, avocados, pumpkin, peas, radish, sunflower, turnips, and beets, among others. Foliar analysis of leaf tissue is preferred over leaf symptoms as a basis for diagnosing boron injury, and often it provides a more reliable basis for diagnosis than the analysis of soil or water. The boron content of normal mature leaves is about 50 to 100 p.p.m. Boron contents of 20 p.p.m., or less, indicate deficiency, while values above 250 p.p.m. are usually associated with boron toxicity. If a boron toxicity is indicated, its source should be determined whether from irrigation water, from soil or from fertilizer. If the irrigation water is contaminated, it may be possible to mix it with enough low-boron irrigation water so that the mixed water is safe for use. If this is not possible, then the only aternative is to plant boron-tolerant crops, as listed in the left-hand column of Table VI. If the soil is found to be contaminated with boron, reclamation will be necessary, entailing leaching with a large quantity of water, as described in Section V, B. V. Reclamation of Salt-Affected Lands

Many studies have contributed to our understanding of the methods and principles of the reclamation of saline and sodic soils (Hilgard, 1906; Harris, 1920; Kelley and Thomas, 1928; Burgess, 1928; Wursten and Powers, 1934; Snyder et al., 1940; Reeve et al., 1948, 1955; Kelley, 1937, 1951; Bower et al., 1951; Overstreet et al., 1951, 1955).

A. LEACHING TO REMOVE SOLUBLESALTS The reclamation by prolonged leaching of salt-affected land in new project areas, and also of previously irrigated but abandoned land in older projects, is an essential part of the overall program of irrigated agriculture. For instance, about one-third of a possible 100,000 acres of irrigable land in the Coachella Valley, California, requires extensive leaching before good production is possible, and most of the nearly one-half million acres now under irrigation in the Imperial Valley required intensive leaching at one time or another. Approximately 1,000,OOO acres went out of production in the United States during the period 1930 to 1940, but much of this land has since been reclaimed. Large acreages in the Delta area, Utah, previously irrigated for many years, also have been reclaimed.

SALINITY IN RELATION TO IRRIGATION

165

The amount of water required to reclaim salty land has been determined on the basis of leaching experiments conducted in Utah (Reeve et al., 1948) and in California (Reeve et al., 1955). The basic findings from these two experiments are expressed in a salt-leaching curve given in Fig. 3. This curve indicates that to reduce the salt content of the soil to about 20 per cent of the initial high value, 1 foot of water is required for each foot of soil considered. Thus, to effectively reclaim a highly

$

DEPTH OF WATER APPLIED PER UNIT DEPTH OF SOIL

FIG.3. Percentage of initial salt and boron remaining in the soil in relation to depth of leaching water applied per unit depth of soil. (After Reeve et al., 1955.)

saline soil to a depth of 4 feet, at least 4 feet of water should be applied as a continuous application. The results of these two leaching experiments are in good general agreement with a recent theoretical analysis of the leaching problem (Gardner and Brooks, 1957); hence, the data should find useful application in other areas.

B. LEACHING TO REMOVEBORON Boron salts are found in toxic concentrations in many arid soils of the world, including the Coachella Valley, California (Kelley and Brown, 1928; Reeve et al., 1955).Where the boron content is onIy slightly above crop tolerance, it and the accompanying soluble salts may be removed

166

LOWELL E. ALLISON

by leaching with about 4 feet of water, the usual reclamation procedure in this valley. However, in a leaching experiment (Reeve et al., 1955) where the soil contained 54 p.p.m. of boron in the saturation extract, at least 3 times as much water was required to reduce the boron to a safe TABLE VII Rate of Leaching of Soluble Salts and Boron@ In saturation extract

Leaching treatment None (initial) 4 feet of water 8 feet of water 12 feet of water

0

Boron EC, x 10s 64.0 4.2

3.4 3.3 After U. S. Salinity Laboratory Staff (1954).

(p.p.m. 1 54.0 6.9 2.4 1.8

level (1.8 p.p.m.) for moderately boron-tolerant crops as was required to remove the soluble salts. These data, shown in Table VII, are in agreement with the results expressed in the boron-leaching curve in Fig. 3. C. RECLAMATION PROCEDURES

1. Flushing Rapid flushing of water over the soil surface is practiced in some parts of the world to remove surface accumulations of salt, often referred to as salt crusts. Reeve et al. (1955) conducted experiments to determine the reclaiming value of a series of surface flushings with Colorado River water on a highly saline (EC, = 75 mmho./cm.), silty-clay loam soil in the Coachella Valley. Since the quantity of salt removed was equal to only 1 per cent of the salt present in the surface 2 feet of soil, it was concluded that surface flushing appears to have no essential reclamation value for soils that have sufficient permeability to permit leaching in the usual manner. 2. Basin Method This method, which resembles the border method of irrigation, is extensively used for leaching highly saline land of low permeability. However, it requires heavy machinery for construction of the large borders needed for safely ponding water for long periods of time. Some irrigation districts require official inspection of the borders for proper construction before water is allocated for leaching purposes. This practice resulted from past experience with improperly constructed borders, which resulted in washouts and severe damage to District drainage facilities and often to farmers’ crops in adjacent fields.

SALINITY IN RELATION TO IRRIGATION

167

If the land is properly leveled, parallel borders may be used; otherwise, contour borders are required. Ponded waters on highly saline soils of very low permeability, especially in regions of high aridity and high temperatures, may increase two- or threefold in salinity on prolonged standing in the basins for up to 90 days or longer in some cases. For that reason, a certain amount of the impounded waters may be allowed to waste away at the lower end of the basins to maintain a favorable salt balance in the leaching water. When prolonged leaching is required, it is common practice to dry the basins after 90 to 120 days of ponding and test samples of the leached soil for completeness of salt removal, as a basis for determining the need for additional leaching. The basin method of leaching has at least two disadvantages. The construction of large borders, often 3 feet high and from 6 to 8 feet wide at the base, requires large costly machinery and much labor. A serious di5culty sometimes results from the accumulation of large quantities of salts within the borders. When the borders are broken down in the finish-leveling process, this salt is scattered over the reclaimed surface, which often results in spotted stands of the initial crop. Moreover, the soil below the borders is not always effectively leached of salts and this may cause a series of sterile strips in the first few crops grown following reclamation. Because of the high cost of border construction, leaching operations seldom are interrupted to change the position of the borders to prevent the occurrence of sterile strips. Obviously, only the more salttolerant crops (Table IV) should be grown on land di5cult to reclaim. 3. Furrow-Basin Method This method combines the furrow and the basin, or border, methods used for irrigation. It has been used effectively in the Coachella Valley, California, where the soils on the nearly level valley floor vary in texture from loamy sand to loam, and are stratified with discontinuous clay lens in the subsoil, which retard downward movement of water. As for the basin method, the land should be properly leveled (level in one direction with not more than 0.1 per cent slope in the other direction) and provided with adequate underdrainage. The land is first plowed 18 inches deep to turn under any salt crust present, then it is smoothed with an appropriate implement and furrowed in the level direction. Small narrow borders, approximately twice the height of the furrow ridges, are constructed downslope about 40 feet apart. The borders are alternately cross-diked at about every sixth furrow so that the water applied at the high end of the field meanders slowly back and forth between the main borders as it moves slowly downslope. The water should at no time submerge the ridges of the furrows, and

168

LOWELL

E. ALLISON

movement should be slow, up to one week being required for the water to reach the far end of a 40-acre block. After a 4-foot depth of water has been applied in this manner, with no runoff, the basins are allowed to dry. After drying, the field is harrowed to level the borders and ridges, smoothed, and plowed again 18 inches deep. This second plowing turns under any salt that accumulated in the ridges and borders from the first leaching. After a smoothing operation, the furrows and borders are again reestablished, as before, but with the borders in offset position. The land is again leached with from 2 to 4 feet of water, as may be required, to effectively remove salts from the root zone into the drains below. The unique feature of the furrow-basin method of improving land difficult to reclaim is that it provides more effective reclamation, at lower costs and in much less time, than does the standard basin method of leaching using large permanent borders. Moreover, no sterile strips occur due to incomplete leaching under the borders. The small borders can be constructed with a small farm tractor, normally found on most farms. The furrow-basin method of leaching appears to be highly successful in the Coachella Valley where the soils are coarse to medium textured, but it is doubtful that this method would be equally successful on the deep, fine-textured and less permeable soils of the Imperial Valley. 4. Trenching

A trenching procedure, shown in Fig. 4, is used successfully in the Coachella Valley for “spot” reclamation of small areas in otherwise reclaimed fields, which resist improvement after leaching with 4 feet of water in the usual manner. These unreclaimed areas are caused by the presence of clay lens in the subsoil, usually too deep to be broken up by chiseling or subsoiling. They are irregularly shaped and generally vary in size from less than 1 to more than 3 acres. These resistant areas are trenched about 5 feet deep at 8-foot intervals and in a direction parallel to the tile drains. The tile are usually at about 7 feet below the surface. These 8-inch wide trenches are constructed with a chain-bucket trenching device attached to the rear of a small farm tractor. The trenches are allowed to remain open for several days, or long enough to facilitate drying and cracking of the walls, after which they are backfilled by subsoiling the affected area in a direction perpendicular to the trenches. The loose fill is settled by running water down the trench. After a drying period to permit drainage and settling of the fll, the trenched area is finish-leveled, diked, and heavily leached, as before. Where the clay lens is less than 5 feet below the surface, i.e., above the bottom of the trench, and the space between bottom of trench

SALINITY IN RELATION TO IRRIGATION

169

8 .B c! a

170

LOWELL E. ALLISON

and tile is free draining, this method has proved very successful in completing the job of reclamation.

5. Use of High-Salt Waters Sodic soils are difficult, often impossible, to reclaim because of their extremely low permeability. Complete reclamation depends upon the movement of water through the soil (1) to bring about the exchange of calcium for sodium on the exchange complex and ( 2 ) to remove the released sodium salts from the root zone. Recent studies by Reeve and Bower (1960),including both laboratory and field experiments, have demonstrated that high-sodium soils can be HYDRAULIC CONDUCTIVITY ICM./HR.l

EXCHANGEABLE SODIUM

(21

h 3 0:l

SUCCESSIVE DILUTIONS (SEAWATER COLORADO RIVER WAlER)

50 40

30 20

10 0

COLO'RADO RIVER W A l t R ONLY

FIG.5. Hydraulic conductivity and exchangeable-sodium percentage of soil as related to successive leachings with various dilutions of sea water. (After Reeve and Bower, 1960.)

reclaimed in a relatively short period of time by first saturating the soil with a high-salt water, such as seawater or from other sources, to rapidly flocculate the soil and render it permeable. However, it is important that the sodium adsorption ratio (SAR) of the initial (saline) leaching water be appreciably lower than the SAR of the saturation extract of the soil being reclaimed; otherwise, sodium adsorption, rather than release will occur. The SAR principle as applied to salination and reclamation is discussed in Section 111, B, 3. This initial leaching is followed by successive leachings with dilutions of the high-salt water and the irrigation water. Each successive leaching, beginning with the first dilution of the salt water with the irrigation water, must be continued until the soil (SAR) comes into equilibrium with the diluted water (SAR), after which it is possible to continue leaching at the next lower dilution of the salty

SALINITY IN RELATION TO IRRIGATION

171

water. If the dilutions are too wide, then loss of permeability may result and the reclamation operation ceases. This indicates that very careful technical control over this type of reclamation is essential for success. Figure 5 shows the trend in permeability resulting from successive leachings with diluted seawater and also the sodium removal from the exchange complex by a series of dilutions. It should be noted that Colorado River water used alone did not increase the permeability of this highly sodic soil sufficiently to make reclamation possible. VI. Management Practices for Salt-Affected Land

A. LEACHING REQUIREMENT FOR SALINITYCONTROL The relationship between the quantity of salt brought into an area ( a farm or an irrigation project) with the irrigation water and the quantity of salt removed in the drainage water has been referred to by Scofield (1940) as the “salt balance” of the area. If a favorable salt balance occurs, the output of salt must equal or exceed the input. Studies of salt balance in large irrigated areas, initially begun by Scofield, have been extended by Wilcox (1963). Other contributions to the subject of “salt balance” and leaching requirement have been made by Hill ( 1961 ) , Klintworth ( 1952), and Eaton ( 1954). The fraction of the irrigation water that must be leached through the root zone to control salinity at any specified level has been defined by the U. S. Salinity Laboratory Staff (1954) as the leaching requirement, which may be calculated by Eq. ( 3 ) . L R = - -Ddw Diw

ECiw

(3) ECdW

The LR is expressed in the second part of this formula as the ratio of the conductivity of the applied irrigation water ( ECiw),to the conductivity of the drainage water ( ECdw) required to maintain salinity at a specified level (favorable salt balance) at the bottom of the root zone. It may be expressed as a fraction or as a percentage. This concept has greatest usefulness when applied to steady state water-flow rates. Values for ECdwrepresent the maximum salinity toIerated by the crop species grown. The leaching requirement for any particular crop may be illustrated by use of the salt-tolerance data in Table IV. For cotton, where a value of ECdw = 8 mmho./cm. can be tolerated, LR = ECiw/8. Thus, for irrigation waters with conductivities of 1, 2, and 3 mmho./cm., respectively, the leaching requirement will be 12, 25, and 38 per cent. These are maximal values, since removal of salt by rainfall, by the crop, and by precipitation of salts such as calcium carbonate or gypsum in the soil,

172

LOWELL E. ALLISON

are seldom zero. If properly taken into account, these factors would tend to reduce the predicted value of the leaching requirement. Kelley et al. (1949) stated that as the precipitation decreases below 15 inches, increasing amounts of irrigation water must be applied to provide adequate leaching. In many arid regions, the rainfall effective from the standpoint of leaching (total rainfall - runoff and evaporation) is generally less than 10 inches, and often less than 5 inches. This amount of rainfall is insufficient to appreciably affect the leaching requirement (Eaton, 1954). Failure of the LR formula to consider removal of salts by rainfall, by the crop, or by chemical precipitation, provides a justifiable margin of safety, which is necessary to ensure a productive irrigation agriculture. If possible, LR values of more than 25 per cent should be avoided because they are wasteful of water, giving rise to low irrigation efficiency and increasing the drainage requirement. Low subsoil permeability may, and often does, determine whether high LR values for salt-sensitive crops can be met. Other factors that may be decisive in this matter are the cost of water and its availability for irrigation. The alternative to high LR requirements is to grow salt-tolerant crops that have lower leaching requirements, and thus conserve irrigation water.

B. DRAINAGE REQUIREMENTS FOR SALINITY CONTROL Drainage problems often arise in irrigated areas due to low efficiencies in the conveyance and application of irrigation water, or they may arise from subsurface flows out of overirrigated higher areas (Israelson and Hansen, 1962). Many saline and sodic soils have developed as the result of restricted drainage due to low permeability of the subsoil, or as the result of a high water table in an otherwise permeable subsoil (Thorne and Peterson, 1954; Kelley, 1951; Magistad and Christianson, 1944; Fireman et al., 1950; Fireman and Hayward, 1955; U. S. Salinity Laboratory Staff, 1954).

1. Drainage Requirements In arid regions, drainage is primarily for salinity control. The requirements for drainage include both the adequacy of drainage and the quantity of water to be drained (Reeve, 1957; Edminister and Reeve, 1957 ) . The adequacy of drainage for irrigated land is primarily related to the control of the water table (at 5 feet or deeper) to maintain a favorable salt balance within the rooting zone of growing crops. If the water table is allowed to rise to within 3 or 4 feet of the soil surface, lands irrigated with saline water often go out of production because of rapid

SALINITY IN RELATION TO IRRIGATION

173

accumulation of salts in the root zone. The depth to the water table must be such that upward flow of saline ground water, by capillary movement into the root zone, is prevented or greatly reduced. In irrigated land, both high salts and high fluctuating water-table conditions often occur together. Roots that extend and flourish during a period of receding water table may be badly damaged, or often are killed, if the water table rises and inundates the roots. Under saline conditions, the detrimental effects of a high water table are even more severe because of salts in the root zone and the resultant moisture stress to which the plant is subjected. A fluctuating, saline water table at some depth below the root zone is of little consequence except as it may affect the upward movements of salts into the vicinity of actively growing roots. Pearson and Goss (1953) demonstrated the deleterious effect of both high salt and a high water table on grapefruit trees that were grown in lysimeters equipped for water-table adjustment (Allison and Reeve, 1955). At high salinity, a water table 2 feet below the soil surface caused distinct mottling of the leaves, followed by yellowing and progressive deterioration. High salt (EC, = 9 to 11 mmho./cm.) caused bronzing and burning of the leaves, followed by defoliation. However, the highsalt treatment combined with a high, fluctuating water table caused dieback as well as defoliation. Where clay lens or hardpan formations have created perched water tables, often it becomes necessary to break up these impervious layers by subsoiling, deep plowing, or other means. Deep plowing has proved helpful in some, but not all, cases in the Imperial Valley of California, where the soils are fine textured, deep, and stratified, with sand lens. A trenching procedure for establishing drainage through clay lens that has proved successful in medium-textured soils of the Coachella Valley, is described in Section V, C, 4.

2. Quantity of Water To Be Drained A drainage system must be adequate to remove from the soil the equivalent depth of water that must be passed through the root zone in order to maintain a favorable salt balance. In this regard, the leachingrequirement equation (3) given in Section VI, A serves to establish a lower limit for drainage. However, the total quantity of water that actually must be drained will be increased by the inefficiencies in water conveyance and application, and by other sources of excess water, which tend to maintain a high water table. The minimum depth of water required to be drained from the root

174

LOWELL E. ALLISON

zone, when expressed in terms of the consumptive use and the leaching requirement, is as follows: D,

ECIW

(rnin.) =

DCW

ECdw

- ECW

(4)

This gives a lower limit for the quantity of water to be drained, Ddw(min.),which is expressed in terms of the salt content of the irrigation water, ECI,, and other conditions determined by the crop and climate, namely, consumptive use, D,,, and the salt tolerance of the crop. The salt tolerance of the crop (Table IV) is taken into account in the selection of permissible salinity values of drainage water, ECd,. Using this formula, the minimum drainage requirement for alfalfa grown in an arid region, such as the Imperial Valley of California for example, may be calculated. Colorado River irrigation water has an electrical conductivity of 1.1mmho./cm. and alfalfa has a consumptive use requirement of about 56 inches, with a salt tolerance of 5 mmho./cm. (see Table IV). By substituting these values in Eq. (4),the minimum quantity of water to be moved beyond the bottom of the root zone (drainage requirement) is approximately 16 inches per year. Hence, the applied irrigation water should be 72 inches per year, of which 56 inches is for consumptive use and 16 inches is for drainage or leaching requirement. C. SPECIALPLANTINGPROCEDURES Furrow-irrigated row crops often fail to produce satisfactory stands owing to the accumulation of salt in the seed row, which prevents germination. However, properly shaped planting beds minimize salt accumulation around the seed, as shown in Fig. 6 (Bernstein and Fireman, 1957). In flat-topped beds, the salt initially present in the soil is transported in the wetting front and accumulates in a thin layer along the top of the bed and under the bed center where opposing wetting fronts meet. Salinities 5 to 10 times greater than the initial salinity of the soil are developed in the center of such beds, which precludes the possibility of a good stand for center plantings. However, good emergence is obtained for plantings along the edges of the bed where salinity is at a minimum. Rounded beds are much less suitable for double-row plantings than the flat-topped beds. For success, it is essential that uniform water infiltration occurs from alternate furrows into double-row beds; otherwise, salt moves near to one edge or the other, depending on the differential in infiltration rate, and prevents germination.

SALINITY IN RELATION TO IRRIGATION

175

With sloping beds, the wetting front effectively sweeps salt along, leaving nonsaline soil at the planting position, despite initial salinities in excess of 25 mmho./cm. in the saturation extract. The sloping bed is, therefore, very effective in limiting salt damage to the seed, and in

FIG. 6. Wetting and salt accumulation (stippled area) patterns for flat-topped and sloping seed beds. (Diagram modified after Bernstein and Fireman, 1957.)

obviating the need for a thorough leaching prior to planting on soils of limiting salinity.

D. IMPROVING IRRIGATION WATERQUALITY The quality of certain waters that have a high alkalizing effect (high SAR) but are low in total salts (such as water No. 5 in Table 11) can be improved by increasing the content of calcium. This may be accomplished by the use of a simple machine (Doneen, 1947) that drops powdered gypsum into running water in a standpipe or head ditch. Such machines are commercially available. Turbulence of the moving water keeps the gypsum in suspension until dissolved. This results in a highcalcium water (lowers the SAR, or sodium hazard). The use of such treated waters has improved the permeability of soils previously injured by irrigation with untreated, high-sodium waters. Machines for metering dry fertilizers and soil amendments into irrigation waters have been described ( Fullmer, 1950). Mixing of high-salt waters with low-salt waters is another means of improving water quality for irrigation. This is especially feasible in areas where pumping is practiced primarily for drainage. Occasionally, a pumped water may contain nitrate in excess of crop requirements for fertilization, with resulting injury to yields. Blending of such waters with low-nitrate waters is a corrective measure, Research on desalting of ocean water is in progress by both government and private agencies. Several different processes ( electrolysis, dis-

176

LOWELL E. ALLISON

tillation, etc.) for desalinating waters are under study but, to date, no economically feasible method has been discovered.

E. METHODAND FREQUENCY OF IRRIGATION The method of applying irrigation water and the quantity of water applied are important from the standpoint of salinity control. At very low salinity levels, furrow irrigation can be used effectively. With increasing salinity, however, the shape of the ridge or planting bed, and the position of the seed with respect to the water line in the furrow, require greater attention, as mentioned in Section VI, C. For still higher salinity conditions, irrigation is best accomplished by the border method of flooding, as for alfalfa. This method gives uniform areal application of water and affords the most effective method for minimizing salt accumulation in the root zone. In any case, the application of irrigation water in excess of consumptive-use requirement, and in accordance with plant tolerance, is essential if salinity is to be controlled at a satisfactory level for maximum crop production, Sprinkler irrigation provides uniform application and penetration and, therefore, offers greater efficiency in water use and salinity control through elimination of much of the waste associated with surface methods of irrigation. Uniform application of sprinkler-applied water also prevents the surface accumulation of salts in raised planting beds, ridges, or borders. However, the wetting of foliage poses a serious problem for those plants that absorb sodium and chloride in harmful quantities, as mentioned in Section IV, A, 3. In this connection, some crops, such as stone fruits and citrus, are very sensitive to foliar-absorption injury, whereas others, such as strawberries and avocados, are resistant to injury by sprinkler irrigation. The frequency of irrigation profoundly affects response of plants under saline conditions (Ayers et aE., 1943), primarily because it controls the osmotic pressure of the soil solution. The relation of osmotic pressure to the salt concentration of the soil solution and its effect on plant growth was mentioned under Section IV, A, 2. After an irrigation, the soilmoisture content is at a maximum and the salt concentration (OP) is at a minimum, favorable for growth. As the soil dries out because of evapotranspiration, the salt concentration ( OP ) increases progressively, the dryer the soil is allowed to become before reirrigation. Thus, infrequent irrigation aggravates salinity effects on growth. Conversely, more frequent irrigations, by keeping the soil at a higher moisture content, prevents high salt concentrations in the soil solution and tends to minimize the harmful effects of a given level of salinity. For most plants, therefore,

SALINITY IN RELATION TO IRRIGATION

177

saline soils should be irrigated when the moisture content is considerably above the permanent-wilting percentage. VII. Conclusions

The permanence of irrigation agriculture has often been questioned, especially in view of historic failures in many parts of the world. The survival of irrigation in Egypt and China for over 4000 years, however, provides convincing evidence that irrigation, under proper conditions, can be permanent. Based on lessons of the past, it is obvious that the development of a sound irrigated agriculture depends upon a catena, or chain of related factors, involving soils, waters, crops, and man. Failure of any one of these links can bring hardship, or even disaster to an irrigation enterprise. In the past, man was primarily responsible for many of the historic irrigation (civilization) failures. Regarding soil conditions, relatively few sizable land areas of the world are known where, owing to physical limitations alone, irrigation cannot be practiced successfully. This presupposes that an adequate supply of reasonably good quality water is available for irrigation and that good management practices are followed. Such management practices may include the establishment of an effective drainage system for deep, fine-textured soils of low permeability. Sufficient data are available to provide a sound basis for reclaiming highly saline, and also highly sodic, soils. Success in either case depends, of course, upon adequate subsoil drainage for removing salt from the root zone. Irrigation waters vary greatly in quality with respect to salt content (concentration factor), sodium percentage (composition factor) , and boron content ( phytotoxic factor). Despite quality limitations in some cases, nearly all surface waters of the western United States and many ground waters of adequate supply are being used successfully for irrigation. The increase in salinity of some surface water supplies, owing to fuller utilization and reuse, is a matter of increasing concern to downstream irrigated areas, where ultimately changes in management with respect to leaching, drainage, and kind of crops grown may be required. Most economic crops have been classified with regard to their response to salinity on the basis of measurements of the EC of a saturation extract of the soil. These data provide an excellent basis for determining the leaching requirement for maintaining a favorable salt balance in the root zone and maximal crop yields. The leaching requirement relates the capacity of a particular crop to tolerate salinity in the root zone to

178

LOWELL E. ALLISON

the salinity of the applied irrigation water and, thus, serves to determine irrigation practice. Within the past few decades, research and education have provided a better understanding of the physical and chemical problems involved in irrigation agriculture in the United States and other developed countries. This knowledge is being extended to newly developed countries throughout the world and should contribute significantly in meeting the need for food and fiber for expanding populations. Based on our present scientific knowledge of soil and water problems and management practices, there appears to be no valid reason why irrigation agriculture, once established, should not remain permanently successful. REFERENCES

Agarwal, R. R., and Yadav, J. S. P. 1956. I. Indian SOC. Soil Sci. 4, 141-145. Allison, L. E., and Reeve, R. C. 1955. Soil Sci. 79, 81-91. Arany, S. 1956. Congr. Intern. Sci. Sol, 1956 Rappt. 6B, pp. 655-659. Ayers, A. D., Wadleigh, C. H., and Gauch, H. G. 1943. Plant Physiol. 18, 151. Banerjee, S. 1959. Soil Sci. 88, 45-50. Bemstein, L. 1958. U. S. Dept. Agr. Agr. Inform. Bull. 194. Bemstein, L. 1959. U. S. Dept. Agr. Agr. Inform. Bull. 205. Bemstein, L. 1960. U. S . Dept. Agr. Agr. Inform. Bull. 217. Bemstein, L. 1961. Am. J . Botany 48, 909-918. Bemstein, L. 1962. PTOC.Park Symp. UNESCO Arid Zone Res. 1960 XVIII, pp. 139-174. UNESCO, Paris, France. Bemstein, L., and Fireman, M. 1957. Soil Sci. 83, 249-263. Bemstein, L., and Hayward, H. E. 1958. Ann. Rev. Plant Physiol. 9, 25-46. Bingham, F. T., Page, A. L., and Bradford, G. R. 1964. Soil Scl. 98 (July issue). Bower, C. A. 1960. Trans. 7th Intern. Congr. Soil Sci. Madison, Wisconsin, 1960 VOI. 2, pp. 16-21. Bower, C. A. 1962. Unpublished data. Bower, C. A., Swamer, L. R., Marsh, A. W., and Tileston, F. M. 1951. Oregon State Coll. Agr. Expt. Sta. Tech, Bull. 22. Bower, C. A., Haise, H. R., Legg, J., Reeve, R. C., Carlson, R., Dregne, H. E., and Whitney, R. S. 1962. U. S. Dept. Agr., Agr. Res. Serv. Rept. Tech. Study Group 41 pp. Bradford, G. R. 1963. Soil Sci. 96, 77-81. Bradshaw, G. B., and Donnan, W. W. 1953. Unpublished data. Brown, J. C. 1956. Ann. Rev. Plant Physiol. 7, 171-190. Brown, J. W., and Wadleigh, C. H. 1955. Botun Gaz. 116, 201-209. Brown, J. W., Wadleigh, C. H., and Hayward, H. E. 1953. Proc. Am. SOC. H o e . Sci. 61, 49-55. Burgess, P. S. 1928. Arizona Univ. Agr. Expt. Sta. Bull. 123. Buringh, P. 1960. “Soils and Soil Conditions in Iraq.” Iraq Ministry of Agriculture, Iraq. Christiansen, P. D., and Lyerly, P. J. 1952. Texas Agr. Expt. Sta. Circ. 132. de Sigmond, A. A. J. 1924. Soil. Sci. 18, 379-381. Doering, E. J. 1963. Soil Sci. 96, 191-195.

SALINITY IN RELATION TO IRRIGATION

179

Doneen, L. D. 1954. Trans. Am. Geophys. Union 35, 943-950. Doneen, L. D. 1947. Unpublished data. Doneen, L. D., and Grogan, R. G. 1954. Unpublished data. Donnan, W. W., Bradshaw, G. B., and Blaney, H. F. 1954. U . S. Dept. Agr., Soil Conserv. Sew. Tech. Publ. 120. Eaton, F. M. 1935. U . S . Dept. Agr. Tech. Bull. 448. Eaton, F. M. 1936. Trans. Am. Geophys. Union Pt. 11, 512-516. Eaton, F. M. 1950. Soil Sci. 69, 123-133. Eaton, F. M. 1954. Texas Agr. Erpt. Sta. Misc. Publ. 111. Eaton, F. M. 1958. Calif. Citrograph 44, 20-22. Edminister, T. W., and Reeve, R. C. 1957. Yearbook Agr. ( U . S. Dept. Agr.) pp. 378-385. Fireman, M. F., and Hayward, H. E. 1955. Yearbook Agr. ( U . S. Dept. Agr.) pp. 321-327. Fireman, M. F., Mogen, C. A., and Baker, G. 0. 1950. ldaho Univ. Agr. Expt. Sta. Res. Bull. 17. Fullmer, F. S. 1950. Better Crops Plant Food 34, 8-14. Gapon, E. N. 1933. J . Gen. Chem. USSR (English Transl.) 3, 114-152. Gardner, W. R., and Brooks, R. H. 1957. Soil Sci. 83, 295-304. Gauch, H. G., and Wadleigh, C. H. 1944. Botan. Gaz. 105, 379-387. Gauch, H. G., and Wadleigh, C. H. 1951. Botan. Gaz. 112, 259-271. Geraldson, C. M. 1960. Sunshine State Agr. Res. Rept. 1, 10-11. Grillot, G. 1956. Reviews of Research. UNESCO Arid Zone Research, IV, pp. 9-35. UNESCO, Paris. Harley, C. P., and Lindner, R. C. 1945. Proc. Am. SOC. Hort. Sci. 46, 35-44. Harris, F. S. 1920. “Soil Alkali, Its Origin, Nature and Treatment.” Wiley, New York. Hayward, H. E. 1958. Unpublished data. Hayward, H. E., and Spurr, W. B. 1943. Botan. Gaz. 105, 152-164. Hayward, H. E., and Spurr, W. B. 1944. Botan. Gaz. 106, 131-139. Hayward, H. E., and Wadleigh, C. H. 1949. Advan. Agron. 1, 1-38. Hilgard, E. W. 1906. “Soils, Their Formation, Properties, Composition and Relation to Climate and Plant Growth.” Macmillan, New York. Hill, R. A. 1961. J. Irrigution Drainage Div., Am. SOC. Civil Engrs. 87, 1-5. Israelson, 0. W., and Hansen, V. E. 1962. “Irrigation Principles and Practices,” 3rd ed. Wiley, New York. Jones, W. W., Martin, J. P., and Bitters, W. P. 1957. Proc. Am. SOC. Hort. Sci. 69, 189-196. Kelley, W. P. 1937. Calif. Unio. Agr. Erpt. Sta. Bull. 617. Kelley, W. P. 1951. “Alkali Soils, Their Formation, Properties and Reclamation.” Reinhold, New York. Kelley, W. P. 1962. Soil Sci. 95, 385-391. Kelley, W. P., and Brown, S. M. 1928. Hilgardia 3, 445-458. Kelley, W. P., Brown, S. M., and Liebig, G. F., Jr. 1939. Soil Sci. 49, 95-107. Kelley, W. P., Laurence, B. M., and Chapman, H. P. 1949. Hilgardia 18, 635-665. Kelley, W. P., and Thomas, E. E. 1928. Calif. Agr. Expt. Sta. Bull. 455. Klintworth, H. 1952. Farming S. Africa 27, 45-51. Lagerwe&, J. V., and Holland, J. P. 1960. Agron. J . 52, 603-608. Lilleland, O., Brown, J. C . , and Swanson, C. 1954. Almond Facts 9, 1, 5. Magistad, 0. C., and Christianson, J. E. 1944. U . S. Dept. Agr. Circ. 707.

180

LOWELL E. ALLISON

Magistad, 0. C., Ayers, A. D., Wadleigh, C. H., and Gauch, H. G. 1943. Plant Physiol. 18, 151-166. Martin, J. P., and Bingham, F. T. 1954. Soil Sci. 78, 349-360. Overstreet, R., Martin, J. C., and King, H. M. 1951. Hilgardia 21, 113-127. Overstreet, R., Martin, J. C., Shultz, R. K., and McCutcheon, 0. P. 1955. Hilgardia 24,53-68. Pearson, G. A. 1960. U.S . Dept. Agr. Agr. Inform. BuU. 216. Pearson, G. A., and Ayers, A. D. 1958. Unpublished data. Pearson, G. A,, and Bemstein, L. 1958. Soil Sci. 86, 254-261. Pearson, G. A,, and Bemstein, L. 1959. Agron. 1. 51, 654-657. Pearson, G. A., and Goss, J. A. 1953. Proc. Rio Grande Valley H w t . Inst. 7, 1-6. Ratner, E. I. 1935. Soil Sci. 4Q, 459-471. Ratner, E. I. 1944. Pochuouedenie 4-5, 205-207. Reeve, R. C. 1957. Proc. 3rd Congr. Intern. Comm. Irrigation Drainage, San Francisco, Calif., 1957 pp. 10.175-10.187 ( preprint). Reeve, R. C., and Bower, C. A. 1960. Soil Sci. 90, 139-144. Reeve, R. C., Allison, L. E., and Peterson, D. F., Jr. 1948. Utah State Agr. Coll. Expt. Sta. Bull. 335. Reeve, R. C., Pillsbury, A. F., and Wilcox, L. V. 1955. Hilgardia 24, 89-91. Richards, L. A., and Hayward, H. E. 1957. Proc. 1st Intersoc. Conf. Irrigation Drainage, Sun Francisco, Calif. 1957 pp. 93-96. Richards, L. A., Bower, C. A., and Fireman, M. F. 1956. U.S. Dept. Agr. C ~ T982. . Scofield, C. S. 1936. Smithsonian Inst. Ann. Rept. 1934-1935, 275-287. Scofield, C. S. 1940. 1. Agr. Res. 61, 17-39. Scofield, C. S., and Headley, F. B. 1921. J. Agr. Res. 21, 265-278. Scofield, C. S., and Wilcox, L. V. 1931. U.S. Dept. Agr. Tech. Bull. 254. Snyder, R. S., Kulp, M. R., Baker, G. O., and Man; J. C. 1940. Idaho Unlu. Agr. Expt. Sta. Bull. 233. Teakle, L. J. H. 1937. Australia ( W e s t . ) Dept. Agr. [2] 14, 115-123. Thome, J. P., and Thome, D. W. 1951. Utah State Agr. Coll. Expt. Sta. Bull. 346. Thome, D. W., and Peterson, H. B. 1954. “Irrigated Soils,” 2nd ed. McGraw-Hill (Blakiston), New York. U. S. Salinity Laboratory Staff. 1954. U.S. Dept. Agr. Agr. Handbook 80, 160 pp. Wadleigh, C. H., and Brown, J. W. 1952. Botan. Gaz. 113, 373-392. Wilcox, L. V. 1948. U.S. Dept. Agr. Tech. Bull. 962. Wilcox, L. V. 1955. U.S. Dept. Agr. Circ. 969. Wilcox, L. V. 1958. U.S. Dept. Agr. Agr. Inform. Bull. 197. Wilcox, L. V. 1959. Am. SOC. Testing Muter. Spec. Tech. Publ. 273, 58-64. Wilcox, L. V. 1960. U.S. Dept. Agr. Agr. Inform. Bull. 211. Wilcox, L. V. 1963. U.S. Dept. Agr. Tech. Bull. 1290. Wilcox, L. V., Blair, G. Y., and Bower, C. A. 1954. Soil Sci. 77, 259-266. Wursten, J. L., and Powers, W. L. 1934. 1. Am. SOC. Agron. 26, 752-762.

RESPONSE OF PLANTS TO THE PHYSICAL EFFECTS OF SOIL COMPACTION’ Norman J. Rosenberg University of Nebraska, Lincoln, Nebraska

Page I. Causes of Soil Compaction .................................. 11. Compaction Effects on Soil Productivity ....................... A. Mechanisms ............................................ 111. Plant Response to Soil Compaction ............................ A. Response of Selected Crops .............................. B. Parabolic Relationship between Bulk Density and Plant Response IV. Experimental Difficulties ..................................... V. A Mechanistic Study of Compaction Effects on Plant Growth ..... VI. Outlook .................................................... References .................................................

181 182 182 185 185 191 191 192 194 195

1. Causes of Soil Compaction

Excessive compaction is believed to cause or, at least, to be related to decreases in the productivity of many soils. Undesirable compaction levels may be due to genetically derived soil conditions (Winters and Simonson, 1951) or to rapid oxidation of the organic fraction and consequent loss of water stability in virgin soils when these are brought into cultivation. The annotated bibliography of soil compaction (Gill et al., 1959) makes clear that the manipulation of primary soil particles and aggregates by traction and tillage implements, notably the compression of soils by vehicular tra5c, is considered to be a major cause of compaction. Barley (1954) showed that even the growing root may temporarily compact the soil by reducing the “root-free pore space.” This discussion will be confined to review of plant response on what Raney et al. (1955) have called “induced pans.” This term is defined as applying to “those soils where the restrictive layer is the result of a recently applied compacting force, such as implement tra5c or trampling, upon a soil that had, under virgin conditions, physical properties favorable to the penetration of roots and water.” Further, no attempt will 1 Published with the approval of the Director as Journal Paper No. 1448, Journal Series, Nebraska Agricultural Experiment Station.

181

182

NORMAN J. ROSENBERG

be made to treat in any depth the literature of soil fertility-compaction interactions on plant growth. II. Compaction Effects on Soil Productivity

Fontaine (1958) has pointed out that productivity of compacted soils is affected by the increased mechanical impedance, reduced aeration, altered moisture availability and heat flux which follow from increased soil density and reduced pore space. At any one time one or more of these factors may become critical for the growth of plants. Which of the factors actually does become critical will depend upon the soil type, the climatic conditions, the plant species, and possibly upon the stage of development of the plant when its roots encounter compact soil conditions. Whether a given density increment will hamper or improve plant growth depends then upon whether the soil is looser than, at, or more compact than, the optimal density for the season and stage of growth of the crop growing in the soil. Because of the great complexity of the problem, as the forthcoming summary of literature will verify, reliable quantitative expressions relating soil compaction to plant growth response are yet to be achieved. Plant response has, however, been related to specific soil physical phenomena that arise as the result of soil compaction. These studies provide a framework for an understanding of the mechanisms which act on the plant in compacted soil. A. MECHANISMS 1. Mechanical Impedance The effects of mechanical impedance upon root penetration have been discussed in detail by Lutz (1952). The work of Veihmeyer and Hendrickson (1948) was adequately covered in that review but is cited here to illustrate the response of plants to a wide range of soil densities. Veihmeyer and Hendrickson (1948) found no common pIant whose roots penetrated soils with a bulk density of 1.9 g./cc. For clays a bulk density of 1.6 or 1.7 g./cc. was occasionally the critical limit for root penetration, although 1.5 g./cc. was sufficient to prevent rooting on an Aiken clay. Bertrand and Kohnke (1957) reported that corn roots would not penetrate Fincastle silty clay loam compacted to a bulk density of 1.5 g./cc. Forristall and Gessel (1955) reported the ability of trees to branch roots through compact layers to vary with the species. Western red cedar penetrated layers with a bulk density of 1.8 g./cc., whereas hemlock and Douglas fir could not penetrate where the bulk density was 1.3 g./cc. even at shallower depths on the same sites,

PLANT RESWNSE TO SOIL COMPACTION

183

Trouse and Humbert (1961) report that rooting of sugar cane was restricted in a Hydro1 Humic Latosol at a bulk density of 1.1 g./cc. and in a Grey Hydromorphic Clay at a bulk density of 1.75 g./cc. Roots became flattened at much lower bulk densities in the two soils: 0.70 g./cc. and about 1.30 g./cc., respectively. These results were based on cores placed in Mitscherlich pots and amply watered and fertilized. Cores at bulk densities determined as critical for root penetration were buried in field soils and were not penetrated by roots during a two-year period. Zimmerman and Kardos (1961) studied root penetration into cores compacted with an apparatus adapted from a Uhland sampler. It was found that a bulk density of 1.8 g./cc. excluded roots from penetrating Hagerstown silty clay and Shirley silty clay soils, whereas a density of 1.9 to 2.0 g./cc. excluded root penetration into Hublersburg sandy loam and sandy clay loam. Bunt (1961) mixed various proportions of three soil types with peat and granitic grit and compacted these mixtures to varying levels of bulk density. One such mixture, composed of 60 per cent brick earth, 25 per cent peat, and 15 per cent granitic grit was compressed to levels of bulk density ranging from 0.807 to 1.261 g./cc. Bunt found tomato root penetration greatly restricted in the most dense treatment. These results he attributed to the high mechanical impedance and longer periods during which moisture was held at high tensions in the compacted soil, rather than to impeded aeration. Flocker and Menary (1960) also noted root branching by tomatoes to have been restricted to the top inch in pots of soil compacted to bulk density 1.7 g./cc. and to the 4 to 6 inch layer at bulk density 1.4 g./cc. Mechanical impedance has often been observed qualitatively through root behavior. For quantitative measures of mechanical impedance, researchers have depended largely on penetrometer or bulk density data. Gill and Miller (1956) in an effort to measure true mechanical impedance devised an apparatus to determine the “wall pressure” required to restrict root expansion. Another direct approach is due to Wiersum (1957), who grew the seedlings of numerous plant species in tubes of varying diameters containing sands packed to uniform bulk density. Penetration decreased with decreasing tube diameter, and this was interpreted as response to increasing mechanical impedance. Phillips and Kirkham (1962a) found in laboratory experiments that rate of corn seedling root elongation in cores of Colo clay compacted to varying levels of density decreased linearly with the increased bulk density and with decreased depth of needle penetrometer penetration. The relationship held independently of free pore space, which was regulated by maintaining the

184

NORMAN J. ROSENBERG

soil a t 10 and 100 cm. of water suction. Since no significant differences in root penetration occurred as a result of differing moisture levels, the authors consider that aeration was not limiting and that differences in mechanical impedance explain the differing root penetration responses. Also noted was the fact that depth of corn seedling root penetration into sand confined in glass tubes of differing diameters increased linearly with depth of penetrometer penetration.

2. Aeration A comprehensive review by Peterson (1950) dealt with the physiological need of plant roots for adequate oxygen supply, and Russell (1952) in a review article, discussed the agronomic implication of soil physical condition, most particularly aeration, on the growth of plants. Wiersma (1959) has recently reviewed the interaction of soil structural conditions and aeration on root growth and development. Experiments designed to establish critical oxygen diffusion rates for various crops have been illuminating. Bertrand and Kohnke (1957) regulated oxygen diffusion rate by altering soil bulk density and found a g. cm.-2 min.-l to be critical for the growth of corn rate of 25 x roots for the first 5 weeks from seed. Cline and Erickson (1959) maintained different water table levels in a noncompacted soil in order to regulate oxygen diffusion rates and found 70 x g. cm.-2 min.-l to be the critical rate for peas throughout the entire growing season. Recently an oxygen diffusion rate in the order of 20 x g. cm.-2 min.-l has been proposed as a critical index of aeration. Below this rate normal root growth will not occur (Stolzy et a,?.,1961). Knowledge of critical oxygen diffusion rates is becoming increasingly available. Lemon and Kristensen (1961) have shown that the oxygen diffusion rate is not measurable when volume fraction of moisture is less than 0.3. When the oxygen diffusion rate is too great to be measured, aeration is apparently not limiting. It is becoming increasingly evident, particularly on soiIs of medium and coarse texture, that compaction effects on plant growth need not necessarily involve impeded aeration (Bunt, 1961; Phillips and Kirkham, 1962a; Rosenberg and Willits, 1962,). 3. Soil Moisture A comprehensive review of the interrelationships of soil moisture and plant growth has been given by L. A. Richards and Wadleigh (1952). The relations between soil moisture and soil structure were discussed by Marshall ( 1959). Baver (1938) demonstrated that compaction of a soil causes an increase in the percentage of moisture at any suction greater than approxi-

PLANT RESPONSE TO SOIL COMPACTION

185

mately 60 millibars. Substantiation of these results is to be found in the work of Rao and Ramacharlu (1959) with Indian alluvium. Heinonen ( 1954) has remarked that bulk density (over a relatively small range in natural soils) is positively correlated with the available water of sands and silts and negatively correlated with other textures. On the other hand, at extreme bulk densities a diminution of the capillary pore space may occur to such extent as to decrease the available water content of a soil. Such a condition was described by Veihmeyer and Hendrickson (1946). S. J. Richards et al. (1961) report an experiment in which young lemon trees were grown in drums containing a sandy loam top soil at two levels of compaction, 1.43 g./cc. and 1.59 g./cc. Increasing the compaction raised the water retention capacity of the soil, resulting in a lower rate of change of soil suction measured at the 24-inch depth with tensiometers. Therefore, the adverse effects of high density were minimized by the beneficial soil moisture conditions, and no growth differences due to compaction treatments were observed.

4. Soil Temperature The general subject of soil temperature has been covered from the physical and agronomic viewpoints by S . J. Richards et al. (1952). Heat flux is clearly related to compaction in soils since the thermal conductivity of any porous material depends on the proportion of the matrix occupied by the solid, liquid, and gaseous phases and upon the conductivity of each of the phases. Geiger (1956) described conditions in Europe where the incidence of low temperature at ground level was related to the tillage and loosening of the soil. Stickler (1962) showed the use of press wheels to have raised wheat and barley yields in Kansas through encouraging a higher level of winter hardiness in the plants. A suggested explanation is that the compaction resulting from use of the press wheels increases thermal conductivity of the soil allowing a longer period during which the seedling may develop winter hardiness. The higher moisture content of the compacted soil would also tend to increase its heat content. Thus a longer exposure to freezing temperatures would be required in order for ice to form. 111.

Plant Response to Soil Compaction

A. RESPONSE OF SELECTED CROPS The behavior of selected crop species grown in compacted soil cannot be characterized simply. A brief review of experience with a few major economic species grown in compacted soil is presented below and will illustrate the uncertainty that exists in this area of soil research.

186

NORMAN J. ROSENBERG

1. Field Corn In a field of Brookston clay in Canada excessive tillage induced by means of tractor wheel compaction altered total and aeration pore space on soil which had previously been in bluegrass sod, alfalfa sod, and row crops. Aeration pore space ranged from 7.4 to 13.2 per cent according to treatment and crop history. Corn yields were not reduced by the excessive tillage during the wet year 1954. Little or no relation between soil porosity and corn yield was found ( Bolton and Aylesworth, 1959). Phillips and Kirkham (1962b) reported a field compaction experiment in which Colo clay soil was compacted to various bulk density levels by vehicular traffic, Soil compaction reduced stands of corn and, in addition, reduced yields on equalized stands. Fertility was maintained at optinial levels of N, P, and K during the three-year duration of the study. Yields decreased with increasing bulk density regardless of fertility level, indicating a lack of interaction between compaction and fertility in this case. It was observed that plant height, nitrogen, phosphorus, and potassium content of the leaves and the quantity of roots were reduced by compaction. Tasseling and silking dates were delayed by compaction. Bulk density and needle penetrometer data were equally well correlated with yields. Corn yields in Minnesota were reduced by 7.5 per cent when Beardon silty clay loam and Waukegan silt loam were surface compacted by vehicular traffic and by 14.5 per cent when both subsoil and surface were compacted in this manner. Surface soil density rose from 1.07 to 1.19 g./cc. as a result of the compaction treatments; permeability and penetrability were reduced. Corn germination was slowed and population decreased. Corn on compacted soil matured more slowly than on untreated soil ( Adams et al., 1960). Swanson and Jacobson (1956) were able to correlate corn growth to penetrometer measurements and to the number of hammer blows on a core sampler. The regression indicated an 80 per cent drop in corn yield with an increase from 12 to 40 blows required to drive a core sampler into the soil. In a greenhouse study in New Jersey, Nixon sandy loam was vibrated for 5 seconds in a pot to establish a “compaction” treatment. Bulk density was increased slightly but not significantly over the control and over a third treatment in which aeration was improved in the normal greenhouse soil receptacle. Soil oxygen diffusion rate was greatly reduced by the compaction (perhaps more properly called “consolidation”) treatment, but root distribution was unaffected. Yield of plant material was

PLANT RESPONSE TO SOIL COMPACTION

187

little affected and variability between replicates was increased by the consolidation treatment (Van Diest, 1962). 2. Tomatoes Tomatoes have been used extensively by Flocker and co-workers as a test crop in greenhouse (Flocker et al., 195913; Flocker and Menary, 1960) and field (Flocker et al., 1960) compaction studies. A general increase in yield with increasing bulk density to an optimal value was noted in the greenhouse using three soils ranging in texture from a fine sandy loam to a clay loam (Flocker et al., 1959a). This was followed by marked yield decreases with further compaction. For the sandy soils, yield rose with bulk density increases from 1.0 to 1.3 g./cc. Further compaction reduced yields. However, at a bulk density of 1.6 g./cc., yields were not necessarily as low as at bulk density 1.0 g./cc. For the finer textured soil, yield reduction also began at a bulk density of 1.3 g./cc. Although slight compaction had improved germination, budding was decreased by compaction beyond certain critical levels of air space. As a result of a later study, Flocker and Menary (1960) were able to note certain physiological responses of the tomato plant to the compaction of Yolo fine sandy loam. Emergence was delayed by one day at bulk density 1.7 g./cc. as compared with bulk density 1.1 g./cc. Low density favored higher bud numbers. High protein and low sugar content were found in tomatoes grown on the compacted soils. Plants on the compacted soils showed high anthocyanin accumulations in the veins and stems. Yield response was of the parabolic type with production at bulk density 1.4 g./cc. increased but not significantly higher than that at bulk density 1.1 g./cc. Yield at 1.7 g./cc. was significantly lower than at any other density level. In a field compaction experiment with tomatoes as the test crop, Flocker et al. (1960) compacted a wetted Yolo fine sandy loam by the traffic of a loaded truck. Bulk density of the 3 to 6 inch soil depth was increased from 1.29 in the check to 1.54 g./cc. in the most severe treatment. The fresh weight of plants, measured after 4 weeks of growth, decreased with increasing compaction, but the differences in fresh weight had vanished by the time the total yields had been accumulated. 3. Potatoes

Bushnell (1953) surveyed soil physical conditions on Wooster silt loam, a soil series used extensively for potato growing in Ohio. He determined that poor yields of potatoes were obtained when the total pore space was less than 49 per cent of the soil volume. A 1 or 2 per cent

188

NORMAN J. ROSENBERG

shift in total pore space in either direction was accompanied by large yield differences. Dunn and Lyford (1946) also related potato growth to pore space. Newmarket fine sandy loam was tamped into pails of 14-quart capacity to varying degrees of compactness, They found that the most dense treatment (air space of 11.6 per cent) produced the lowest yields, whereas the next most dense treatment (12.8 per cent) yielded higher than any of the other less compact treatments. Potatoes were grown in the traffic-induced compaction plots of Adams et al. (1961). Yields were decreased by as much as 54 per cent, tubers grew on the average 2.2cm. nearer the ground, and specific gravity of the tubers was lowered in the compacted plots. Bulk density of the soil surface layer had been increased from 1.07 to 1.19 g./cc. by packing. Potatoes were also used as a test crop in the work of Flocker et al. (1960). Moderate compaction resulted in an increased yield of the Kennebec variety, but the same treatment had no effect on White Rose potatoes. Under severe compaction both varieties yielded lower than the control. The yield of Kennebec, for example, was reduced by 50 per cent. Tillage of the severely compacted soil resulted in extreme cloddiness, and potato shoot emergence was adversely affected by the large air voids. Root development on the naturally shallow-rooted potato plants was further restricted by the increased mechanical impedance due to compaction. Observations made in New Jersey (Alderfer, 1961) indicate that potatoes do not respond to radical changes in the physical condition of Matapeake silt loam. Vinyl acetate-maleic acid ( VAMA ) conditioner applied at rates up to one and one-half tons per acre resulted in a twentyfold increase in hydraulic conductivity, a change which unquestionably reflects sharp increases in total porosity. Potato yield response to improved physical condition was nevertheless negligible. 4. Sugar Beets and Other Root Crops

Eden and Maskell (1928) attempted to correlate the growth of rutabagas with the draw bar pull of tillage implements on artificially compacted soil. High draw bar pull was positively correlated with volume of roots, simply because the cultivator rode out of the soil and cut less on the harder, more compacted soil. Baver and Farnsworth (1940) noted a 50 per cent decrease in yield of sugar beets in soils where the noncapillary porosity was reduced to 2 per cent of soil volume. Yield increase was found to be linear with increasing noncapillary porosity over the range of 2 to 8 per cent of

PLANT RESPONSE TO SOIL COMPACTION

189

soil volume. Beets growing in the most poorly aerated soil were short, stubby, and had many auxiliary roots. Blake et al. (1960) indicated that yield of sugar beets dropped by 13 per cent while sugar content was not significantly reduced when beets were grown in traffic-compacted Beardon silty clay loam. It was noted, however, that the percentage of sprangled beets was doubled by soil compaction. In a greenhouse experiment, Smith and Cook (1946) grew sugar beets on Brookston clay loam. Treatments were factorials of forced aeration, compaction, and excess moisture. Compaction was shown to decrease sugar beet yield. Excess moisture reduced yields but not to the same extent as did compaction, while excess moisture on the compaction treatment resulted in the lowest yields. Forced aeration alleviated the results of compaction. The authors suggested that an effective air capacity (volume of total pores minus volume of water at moisture equivalent) of 30 per cent is critical for sugar beet growth. Stout et al. (1960) found that uptake of moisture by sugar beet seedballs is more rapid over the first few hours in compacted clay loam than in loose soil. Later, seedballs in loose soil have a higher moisture content. This is attributed to the more difficult resupply of moisture through the dried-out volume adjacent to the seedball in the compact soil. The effects noted were consistent over a bulk density range of 0.66 to 1.37 g./cc. 5. Cotton Results of a long-term experiment studying effects of compaction on the growth of cotton in South Africa have been reported by Heath ( 1937). Nonirrigated cotton was grown under loosened, normal tillage and compacted treatments of a “medium red loam” soil. Differences in probe penetrability were correlated with physiological measurements in the early stages of growth. Stands were better on the normal and compressed soils than on the loosened soil. Total dry weight, dry weight of buds and flowers, height, number of green leaves, number of green bolls and open bolls were all superior on the compressed treatment. The normal treatment followed and the loosened treatment was inferior in these measurements. Differences in the above-mentioned plant characteristics were not, however, statistically significant. This experiment was conducted for four years (with annual renewal of treatments) during which the compressed and normal treatments gave the highest yields of lint cotton. These yields were never significantly different from one another although each treatment produced the highest yield for two of the four years. Also, during two of the four years these treatments were significantly superior to the loosened treatment. Heath attributed his

190

NORMAN J, ROSENBERG

results to the increased volume of available water held by the more dense soil treatments and further stated that excessive aeration may be detrimental in tropical climates. That compaction may be clearly beneficial to the growth of cotton plants was shown by Hubbell and Staten (1951). Moderate and severe compaction treatments were prepared by tractor traffic 12 and 4 days after flood irrigation, respectively. The treatments were renewed three times a season for three years. The check treatment was tilled by hand hoeing or by horse cultivation. Moderate compaction produced higher yields than did the check treatment. The difference was significant at the 5 per cent level during the first year. During the second year, both compaction treatments outyielded the check, both being significantly superior at the 5 per cent level. The severe compaction treatment gave the highest yields during the third year. Penetrometer data indicated that the treatments caused real differences in soil hardness, but unfortunately no other physical measurements were made. One should not conclude, however, that cotton is always favored by poor soil physical condition. Taubenhaus et al. (1931) described a situation in the Texas Blacklands, where cotton roots were “strangled by rain-packed and sun-hardened fine-textured soil. Enlarged clavate bases of the stem indicated that translocation had been impeded by the pressure of the soil on the plant. Doneen and Henderson (1953) noted a fourfold increase in lint cotton yield on Hesperia sandy loam when the plow pan on this soil was disrupted by chiseling. The authors felt that hydraulic conductivity was a more sensitive measure of compaction than bulk density to which to relate yields, since any slight alteration in bulk density at low density levels resulted in marked differences in hydraulic conductivity, whereas at high density levels alteration caused virtually no change. Yield of cotton in Louisiana was increased by two-thirds bale per acre where dense Commerce silt loam was broken during the dry years 1954 and 1956. During 1955 when moisture was adequate no effect on yield was noted ( Saveson et al., 1958). Similar results were obtained by Raney et al. (1954). Jamison and Domby (1956) grew cotton in excavated cylinders of Commerce silt loam in which a traffic pan had existed. Drainage was restricted by means of a water table maintained at 24 inches. Oxygen diffusion was restricted, and retarded growth of cotton resulted. The authors consider that although root extension of cotton was affected by the compact layer, except at very high densities, the influence of the pan arose more from air-water relationships than from physical resistance to root penetration.

PLANT RESPONSE TO SOIL COMPACTION

191

B. PARABOLIC RELATIONSHIPBETWEEN BULKDENSITY AND PLANT REsPoNsE As seen, few if any relationships of universal value describing the dependence of plant growth on soil physical condition have appeared. One fact is apparent however: yield response will have a parabolic relationship with soil bulk density if a great enough range in bulk density is encountered. Vomocil (1955) described this parabolic nature of the regression of plant yieIds on bulk density and expressed the relationship by the equation Yo - Y = C ( D b *- D b ) ? ;Y is plant yield, and Yo is the maximum relative yield of a given crop from a given soil under given weather conditions (including irrigation). is the optimal 4 to 16 inch average profile bulk density, and C is a sensitivity constant (probably dependent on the crop and weather). This relationship was derived from observations of the yields of field corn, sweet corn, and potatoes in New Jersey. These yields in general declined abruptly at a profile bulk density of about 1.55 g./cc. Finding that the regressions differed for each crop during a given season, Vomocil concluded that each crop varies in its sensitivity to aeration, and that below critical bulk densities yield variation is due to factors other than soil physical conditions. He also noted that the regression of yields on average bulk density varied not only with crop, but with season and with the use of irrigation. Vomocil’s ( 1955) suggestions illustrate how difficult the achievement of universally acceptable quantitative expressions relating plant growth to soil physical condition may be. Such expressions appear almost as remote today.

zb

IV. Experimental Difficulties

Some, though certainly not all, of the confusion and nonreproducibility of results of compaction studies must stem from the lack of uniformity in methods whereby soil physical conditions are altered for experimental purposes. Soil masses of uniform density are essential when it is required that soil density be the independent variable against which plant response is studied. Generally soils have been pressed, tamped, or otherwise compacted to average densities. Such procedures, particularly when a soil is plastic, may create laminar structures. Smearing between, and even within, the layers may be considerable and microvariability of density may be great. Such conditions can alter the entire course of an experiment. Veihmeyer and Hendrickson ( 1948 ) were entireIy cognizant of this limitation when they reported critical densities for root penetration in soil compacted by applying surface pressure to one or more layers. Bunt’s results (1961) wherein tomato roots were restricted to the

192

NORMAN J. ROSENBERG

top 1 inch of soil which had been compressed into pots may possibly have been due to such occurrences. A different approach to preparation of uniformly compacted soils has been through use of dynamic compaction (Rosenberg, 1959, 1960). Soils were compacted by means of vibrators and exhibited a relatively high degree of uniformity with no evidence of layering or laminations. Vibrocompaction actually increased the uniformity of a treated soil over that of a soil undisturbed for twenty years. V. A Mechanistic Study of Compaction Effects on Plant Growth

In an attempt to study quantitative yield and physiological responses of selected grain and vegetable crops and to find a mechanistic explanation for these responses, Rosenberg and Willits (1962) carried out a series of studies in New Jersey. Two Atlantic Coastal Plain soils, Galestown sand and Freehold loamy sand and a Piedmont soil, Penn silt loam, were compacted with a vibrating probe (Rosenberg, 1959). Density of the soils ranged from approximately 1.30 to 1.65 g./cc. for the sandy soils and 1.0 to 1.35 g./cc. for the Penn soil. Six density levels were established for the sandy soils and five levels for the silt loam. All treatments were replicated 4 times. Barley was sown in the fall of 1958. No supplemental irrigations were applied to the barley. Compaction reduced the yields of barley forage (early cutting), grain, and straw on the Freehold loamy sand. Yields were depressed on the Penn silt loam but differences were not significant. Yields on the Galestown soil, however, increased markedly with compaction. Wheat was sown in the fall of 1959 in the drums, and although no yield data were obtained in the spring of 1980, root distribution was observed. Particularly on the Freehold soil, root penetration and branching was sharply reduced by compaction. In the other soils, depth of penetration was not affected although root ramification was decreased by compaction. Green beans were grown in the drums during the late summer of 1959. Supplemental irrigation was applied to this crop. Only on the Galestown soil were yields significantly reduced by compaction. The yield response to compaction was parabolic. In harvest of the green beans, the sum of the first two pickings, expressed as a percentage of total yield, served as an index of concentration of maturity. Concentration of maturity was increased by compaction of the Perm silt loam, followed a parabolic pattern on the Freehold sand with maximum concentration at an intermediate density level, and was unaffected on the Galestown sand. The grain yields and green bean yield and maturity data served as

PLANT RESPONSE TO SOIL COMPACTION

193

dependent variables with which physical measurements were compared in a multiple correlation study. These physical measurements included noncapillary pore space (pores drained by 60 millibar suction), hydraulic conductivity, oxygen diffusion rate at 60 mb. suction for the sands and 60 and 330 mb. suction for the silt loam, available water content (volumetric basis), and bulk density. Table I presents correlations that occurred between the barley yield response and the soil physical parameters. The interpretation given these TABLE I r Values for Linear Correlation of Soil Physical Measurements with Growth Parameters of Barley Grown on Three Soi1sa.b

soil Galestown Freehold Penn

Bulk

Plant remonse

densitv

Forage Grain Forage Grain Forage Grain

+0.28 +0.55* -0.69'" -0.60"' -0.54'* -0.26

Oxygen diffusion rate at 60 millibars suction

-

+0.54" +0.19

After Rosenberg and Willits (1962). f, ', "", indicate r values significant at the bility, respectively.

Hydraulic conductivity

Available water

-0.18 -0.49' +0.58"' +0.25 +0.55"" +0.41"

+0.21 +0.51" -0.42f -0.13 -0.36t -0.47'

Q

J

lo%, 5%, and 1% level

of proba-

correlations is discussed below. Barley yield on Galestown soil was positively correlated with bulk density, but equally well with available water content of the soil. Spring, 1959, was very dry in New Jersey, particularly during the period when barley was heading out. Grain and forage yields on the Freehold soil were negatively correlated with bulk density and forage yield had a highly significant correlation with hydraulic conductivity. The later grain yield was not well related to hydraulic conductivity. On the Penn soil forage yield was negatively correlated with bulk density and with oxygen diffusion rate at 60 mb. suction and hydraulic conductivity. Grain yields were positively correlated with hydraulic conductivity and negatively with available water. In a greenhouse study (Rosenberg, 1961) the same three soils were compacted and wheat plants were grown under a moisture regime whereby water was applied when the soils had dried to a suction of 1 atmosphere. Yields were significantly reduced by compaction of the Galestown sand and the Freehold loamy sand. Response of the Penn soil was parabolic, with yield on the most dense treatment (bulk density

194

NORMAN J. ROSENBERG

1.50 g./cc. ) significantly higher than with a bulk density approximately 1.0 g./cc. The evidence indicated that the response of barley grown on Galestown sand in the drum study resulted from the improved moisture relationships induced by compaction of that soil. Response on the Freehold soil was apparently due to sharply increased mechanical impedance as evidenced by the impeded growth of wheat roots in the same drums and by the very high and positive correlation with hydraulic conductivity. Barley growth on Penn soil was limited, at least during part of the growing season, by impeded aeration which resulted from compaction. This is evidenced by the significant positive correlation with oxygen diffusion rate. VI. Outlook

Results of the study discussed in Section V and the literature of soil compaction research indicate that plant response to compaction will vary with soil type, plant species, and climate. It is clear that the plant in compacted soil may respond to alterations in mechanical impedance, aeration, moisture availability, and heat flux of the soil. Plant response can be attributed to any of these phenomena within critical density ranges for a given soil under a given set of climatic conditions. Further, it is clear that plant response to compaction, if expressed over a wide enough range, is parabolic. Since a parabolic relationship implies that interacting factors are affecting plant growth, statistical methods for the study of the interactions are applicable. Multiple correlation and regression are suitable methods for study of the effect of the separate independent variables and their interacting effects upon a dependent variable. Kirkham (1W1) has emphasized the need for consideration of weather in correlations of physically measured properties of the soil with the growth of plants. In much of the literature available to us today, however, indirect measures of soil physical phenomena are employed as independent variables in regression equations. Hammer blows on a core sampler are not invariably related to the mechanical impedance encountered by a growing root, for example. Regressions based on such indirect measurements can be used to predict only the effects to be expected on the soil tested, for the crop grown, and under identical weather conditions as obtained during the experimental period. The correlative approach will not lead to useful quantitative prediction tools until the mechanisms by which plants are affected by compaction are more thoroughly understood and measurable with accuracy. As further progress is made, particularly in mensuration of the mechan-

PLANT RESPONSE TO SOIL COMPACTION

195

ical impedance and aeration aspects of soil compaction, the correlative approach will provide relationships of greater usefulness. The status of our understanding of the problem today does, however, permit the agricultural scientist with access to information on soil texture, pore size distribution, hydraulic conductivity, available water content, and oxygen diffusion rate to anticipate intelligently where compaction problems are likely to arise and to diagnose the existence of soil conditions unfavorable to plant growth because of compaction. ACKNOWLEDGMENTS The author is indebted to Professor John F. Stone of Oklahoma State University and Professors Leon Chesnin and A. P. Mazurak of the University of Nebraska for their suggestions and criticism of this review, and to Professor N. A. Willits, formerly of Rutgers University, who guided the author’s thesis research in soil compaction. REFERENCES Adams, E. P., Blake, G . R., Martin, W. P., and Boelter, D. H. 1961. Trans. 7th Intern. Congr. Soil Sci. Madison, Wisconsin, 1960 Vol. 1, pp. 607-615. Elsevier, Amsterdam. Alderfer, R. B. 1961. Personal communication. Barley, K. P. 1954. Soil Sci. 78, 205-210. Baver, L. D. 1938. Soil Sci. Soc. Am. Proc. 3, 52-56. Baver, L. D., and Farnsworth, R. B. 1940. Soil Sci. SOC. Am. Proc. 5, 45-48. Bertrand, A. R., and Kohnke, H. 1957. Soil Sci. Soc. Am. Proc. 21, 135-140. Blake, G. R., Ogden, D. B., and Adams, E. P. 1960. 1. Am. SOC. Sugar Beet Technologists 11, 236-242. Bolton, E. F., and Aylesworth, J. W. 1959. Can. J. Soil Sci. 39, 98-102. Bunt, A. C. 1961. Plant Soil 13, 322-332. Bushnell, J. 1953. Ohio Agr. Expt. Sta. Tech. Res. Bull. 726. Cline, R. A., and Erickson, A, E. 1959. Soil Sci. SOC. Am. Proc. 23, 333-335. Doneen, L. D., and Henderson, D. W. 1953. Agr. Eng. 34, 94-95, 102. Dunn, S . , and Lyford, W. H., Jr. 1946. New Hampshire Agr. Expt. Sta. Tech. Bull. 90. Eden, T., and Maskell, E. J. 1928. J. Agr. Sci. 18, 163-185. Flocker, W. J., and Menary, R. C. 1960. Hilgardia 30, 101-121. Flocker, W. J., Lingle, J. C., and Vomicil, J. A. 1959a. Soil Sci. 88, 247-250. Flocker, W. J., Vomicil, J. A., and Howard, F. D. 195913. Soil Sci. SOC. Am. Proc. 23, 188-191. Flocker, W. J., Timm, H., and Vomicil, J. A. 1960. Agron. 1. 52, 345-348. Fontaine, E. R. 1959. Mededel. Landbouwhogeschool Opzoekings-Sta. Staat Ghent 24, 30-35. Forristall, F. F., and Gessel, S. P. 1955. Soil Sci. Soc. Am. Proc. 19, 384-389. Geiger, R. 1956. “The Climate Near the Ground,” 2nd ed. rev., pp. 144-148. Harvard Univ. Press, Cambridge, Massachusetts. Gill, W. R., and Miller, R. D. 1956. Soil Sci. SOC. Am. Proc. 20, 154-157. Gill, W. R., Haise, H. R., and Hagan, R. M. 1959. “Annotated Bibliography on Soil Compaction,” Am Soc. Agr. Eng., St. Joseph, Michigan. Heath, 0. V. S. 1937. J. Agr. Sci. 27,511-540. Heinonen, R. 1954. Agrogeol. Julkaisuia 62, 75-82. English summary.

196

NORMAN J. ROSENBERG

Hubbell, D. S., and Staten, G. 1951. New Mexico Agr. Expt. Sta. Tech. Bull. 363. Jamison, V. C., and Domby, C. W. 1956. Soil Sci. SOC. Am. PTOC.20, 447-453. Kirkham, D. 1961. Soil Sci. SOC. Am. Proc. 25, 423-427. Lemon, E., and Kristensen, J. 1961. Trans. 7th Intern. Congr. Soil Sci. Madison, Wisconsin, 1960 Vol. 1, pp. 232-240. Elsevier, Amsterdam. Lutz, J. F. 1952. In “Soil Physical Conditions and Plant Growth” (B. T. Shaw, ed.), pp. 43-71. Academic Press, New York. Marshall, T. J. 1959. “Relations Between Water and Soil,” Commonwealth Agricultural Bureaux Tech, Commun. 50. Famham Royal, Bucks, England. Peterson, J. B. 1950. Soil Scl. 70, 175-185. Phillips, R. E., and Kirkham, D. 1962a. Soil Sci. SOC. Am. Proc. 26, 319-322. Phillips, R. E., and Kirkham, D. 1962b. Agron. J. 54, 29-34. Raney, W. A., Grissom, P. H., Wooten, 0. B., Jones, T. N., and Williamson, B. F. 1954. Mississippi Farm Res. 17( l ) , 1 and 3. Raney, W. A., Edminster, T. W., and Allaway, W. H. 1955. Soil Sci. SOC. Am. Proc. 19, 423-428. Rao, K . S., and Ramacharlu, P. T. 1959. Soil Sci. 87, 174-178. Richards, L. A., and Wadleigh, C. H. 1952. In “Soil Physical Conditions and Plant Growth” (B. T. Shaw, ed.), pp. 73-251. Academic Press, New York. Richards, S. J., Hagan, R. M., and McCalla, T. M. 1952. In “Soil Physical Conditions and Plant Growth” (B. T. Shaw, ed.), pp. 303-480. Academic Press, New York. Richards, S. J., Weeks, L. V., and Erickson, L. C. 1961. Soil Sci. 91, 347-350. Rosenberg, N. J. 1959. Soil Sci. 88, 288-290. Rosenberg, N. J. 1960. Soil Sci. 90, 365-368. Rosenberg, N. J. 1961. Ph.D. thesis, Rutgers, State Univ. New Jersey, New Brunswick, New Jersey. Rosenberg, N. J., and Willits, N. A. 1962. Soil Sci. SOC.Am. Proc. 26, 78-82. Russell, M. B. 1952. In “Soil Physical Conditions and Plant Growth” (B. T. Shaw, ed. ), pp. 253-301. Academic Press, New York. Saveson, I. L., Lund, Z. F., and Sloane, L. W. 1958. Louisiana Agr. Expt. Sta. Circ. 53. Smith, F. W., and Cook, R. L. 1946, Soil Sci. SOC. Am. Proc. 11, 402-406. Stickler, F. C. 1962. Agron. J. 54, 492-494. Stolzy, L. H., Letey, J,, Szuszkiewicz, T. E., and Lunt, 0. R. 1961. Soil Sci. SOC. Am. PTOC.25, 463-467. Stout, B. A., Snyder, F. W., and Buchele, W. F. 1960. Mich. Agr. Expt. Sta. Quart. Bull. 42, 548-557. Swanson, C. L. W., and Jacobson, H. G. M. 1956. Soil Sci. SOC.Am. Proc. 20,161-170. Taubenhaus, J. J., Ezekial, W. N., and Rea, H. E. 1931. Plant Physiol. 6, 161-166. Trouse, A. C., Jr., and Humbert, R. P. 1961. Soil Sci. 91, 208-217. Veihmeyer, F. J., and Hendrickson, A. H. 1946. Soil Sci. 62, 451-456. Veihmeyer, F. J., and Hendrickson, A. H. 1948. Soil Sci. 65, 487-493. Van Diest, A. 1982. Agron. J. 54, 515-518. Vomocil, J. A. 1955. Ph.D. thesis, Rutgers, State Univ. New Jersey, New Brunswick, New Jersey. Wiersma, D. 1959. Aduan. Agron. 11, 43-51. Wiersum, L. K. 1957. PZant Soil 9, 75-85. Winters, E., and Simonson, R. W. 1951. Adoan. Agron. 3, 1-92. Zimmennan, R. P., and Kardos, L. T. 1961. Soil Sci. 91, 280-288.

NITRATE ACCUMULATION IN CROPS AND NITRATE POISONING IN ANIMALS Madison J. Wright and Kenneth L. Davison Cornell University, Ithoca, New York

Page I. Introduction ................................................. 11. Recognition of Nitrate as a Toxic Agent ......................... 111. Accumulation of Nitrate by Plants ............................. A. General Metabolic Role of Nitrate ......................... B. Internal Factors Governing Accumulation .................... C. External Factors Governing Accumulation . . . . . . . . . . . . . . . . . . . IV. Postharvest Losses ........................................... V. Toxicity of Nitrate to Animals ................................. A. Ruminants vs. Nonruminants .............................. B. Involvement with Vascular System ......................... C. Types of Toxicity ........................................ D. Potential Hazards to Man ................................. VI. Conclusions ................................................. References ..................................................

1.

197 198 201 201 202 210 220 221 223 226 226 236 240 241

Introduction

The abundant and mobile nitrate ion occupies a position of primary importance in the normal metabolism of higher plants. It is the major nutrient form of nitrogen in most soils. It is often the first factor limiting plant growth. In the great majority of cases it is assimilated so rapidly following absorption that its concentration within plant tissues never rises above a few hundred parts per million. Over the past three-quarters of a century, however, nitrate has at times been observed to accumulate within plants to abnormally high concentrations, with results disastrous to animals fed these plants, or to animals and humans exposed to the gaseous decomposition products. Considering the ubiquitousness of the nitrate ion, it is not surprising to find that these cases have occurred in many parts of the world. But they tend to be concentrated in certain geographic areas, suggesting that local conditions contribute to the abnormal accumulation. In agriculture, feeds high in nitrate have been found to cause acute poisoning in cattle, sheep, and other livestock (death, collapse, abor197

198

MADISON J. WRIGHT AND KENNETH L. DAVISON

tion). Many cases of chronic poisoning also have been reported. Gases issuing from silos recently filled with high-nitrate forage have caused injury or death to livestock. Water supplies contaminated with nitrate have contributed to, or been entirely responsible for, poisoning of animals. Human health also has been jeopardized, and in some cases lives have been lost, because of silo gases or contaminated water supplies. Ingestion by humans of lethal amounts of nitrate accumulated in plant materials has not been reported and seems unlikely to occur. Most of the research so far conducted on nitrate accumulation and toxicity has been stimulated by sporadic but heavy economic losses. The literature consequently consists in large part of surveys of the nitrate content of vegetation and reports of responses of livestock to deliberate administration of nitrate or nitrite salts. From these findings various research centers have developed recommendations for limiting risks. More recently research has been directed along two main lines: ( 1 ) attempts to describe, or to substantiate the occurrence of, chronic nitrate poisoning, and ( 2 ) attempts to elucidate the processes leading to an accumulation of nitrate in plants. In general the research is becoming increasingly physiological and biochemical. This review is an attempt to summarize the present state of knowledge regarding the problem of nitrate accumulation in plants and the consequences of feeding nitrate or nitrite to animals. It is not intended as a review of the physiology of nitrate in plants in the manner of Burstrom (1946) or McKee (1962). The latter work provides an especially lucid and comprehensive review. Brief consideration will be given here to the related hazards to humans. II. Recognition of Nitrate as a Toxic Agent

Mayo’s (1895) report of several livestock poisonings in Kansas following feeding of constalks is usually cited as the earliest authentic and detailed account of nitrate toxicity. His descriptions of symptoms exhibited by the animals and conditions under which the crop was grown have been faithfully reproduced in accounts from other areas. Mayo’s examination of the corn plants, prompted in part by a reminder from the station chemist that corn analyzed in a previous case proved to be high in KN03, revealed the presence of abundant crystals of potassium nitrate in the leaf axils, at the cut surfaces, and inside the stalks. His statement that the stalks burned like firecracker fuses when lighted with a match has been substantiated by subsequent experimenters ( Ackerson, 1963); Theron ( 1957) has reported that turnip midrib tissue containing nearly 4 per cent NO,-N exploded during ashing.

NITRATE ACCUMULATION AND POISONING

199

Two of Mayo’s lethal samples proved to contain 2.6 and 3.5 per cent NO,-N.l After observing the crystals, Mayo administered KNO, to three cattle, and all three died. He reached the conclusion that their “symptoms were those of potash poisoning,” in so doing perhaps inadvertently placing the stigma on the cation component. The cation was shown to be of no consequence in experiments conducted several decades later ( Crawford, 1960). An indictment of nitrate, as distinguished from its salts, was made in work at the Wyoming station (Bradley et al., 193913, 1940b) subsequent to an outbreak of “oat hay poisoning cases in the High Plains region during the 1920’s and 1930s.‘ Numerous vain attempts had been made to demonstrate the presence of cyanide in the hays (Newsom et al., 1937), following the discovery that sorghum poisoning was due to a cyanogenetic glycoside. Obtaining oat hay from stacks fed immediately preceding large-scale losses, Bradley and his associates were able to show that either the hays or water extracts thereof would produce intoxication typical of oat hay poisoning. More importantly, they detected methemoglobin in the blood of poisoned animals while making a spectroscopic examination. Large amounts of nitrite were also found on autopsy, especially in the bile and urine. A typical analysis of these oat hays was: heads, 0.10%; stalks, 0.35%; leaves, 0.58% NO,-N. By administering to three calves a water extract of oats, a solution of KNO, of comparable strength, and a modified oat extract from which 1 The discovery of crystals led early investigators to report analyses as per cent KNO,. More recently the prevailing practice has been to report per cent NO,-, or per cent NO,-N. A majority of the 36 research workers and industrial representatives attending a Conference on Nitrate Accumulation and Toxicity held in New York City on April 15-16, 1963, favored standardization of reporting as follows: per cent NO,-N, for feeds and tissues; millimoles per milliliter, for physiological solutions. This usage is followed in this review. Conversion: NO,-N to NO,, multiply by 4.43; NO,-N to KNO,, multiply by 7.22. Further standardization also appears desirable in toxicological work with nitrate. The most common usage in recent reports from English-speaking areas is the hybrid, grams per 100 pounds body weight. Toxicological tables express dosages as milligrams per kilogram body weight, and this usage has been followed here. Conversion: g./lOO lb. to mg./kg., multiply by 22.05. 2 Most United States workers have credited Bradley and associates with first recognizing feed-induced methemoglobinemia, but it seems clear that priority more properly belongs to Rimington and Quin (1933) for their studies of TribuZus poisoning in South Africa. This penetrating piece of research may be found in somewhat condensed form in S. African J . Sci. 30, 472-482 (1933). Steyn (1934) makes reference to even earlier, but apparently unpublished, research by Green in 1926, noting that he detected methemoglobinemia in drenching sheep with TribuZus sp.

200

MADISON J. WRIGHT AND KENNETH L. DAVISON

70 per cent of the nitrate was removed by crystallization, the Wyoming workers demonstrated that the high concentration of nitrate was responsible for the toxicity. Only the calf given the de-nitrated extract survived; the others died with typical symptoms. The Wyoming experiments, and those conducted in Saskatchewan (Davidson et al., 1941) and elsewhere shortly thereafter, made it clear that nitrate accumulated in plants was reduced to nitrite after ingestion, and that the nitrite reacted with hemoglobin in the blood to form methemoglobin. The administration of sodium nitrite as an antidote in cases misdiagnosed as cyanide poisoning was thus shown to aggravate the poisoning rather than relieve it. Although the animals could tolerate low levels of methemoglobin for long periods, they suffered from anoxia when a high percentage of hemoglobin was converted. The Canadian experimenters also detected the odor of oxides of nitrogen during autopsy of poisoned animals, reinforcing the idea that reduction of nitrate was involved in the toxicity. In a few cases, reduction of nitrate to nitrite has occurred before feeding. Bacterial activity in wet tissue or aqueous solution was apparently responsible. Considerable difficulties were encountered in setting limits of safety for the nitrate content of feedstuffs in the ensuing period. (These difEculties continue to the present day, and are discussed in Section V, C of this review.) Administration of nitrate at presumably dangerous dosages did not invariably produce characteristic symptoms of nitrate poisoning, nor did it always duplicate the symptoms of “oat hay,” “cornstalk,” or other feed-induced poisoning. It became increasingly clear that differences between species of livestock, among individual animals within species, and among rations were responsible. Since 1955 there has been active discussion and experimentation concerned with “chronic,” or long-term sublethal, nitrate poisoning. Reduced rates of gain or milk production, depressed appetite, abortion, and other adverse effects have been attributed to diets containing substantial concentrations of nitrate. In general, a lack of clear-cut symptoms has restricted progress in defining or substantiating the role of nitrate. Impetus has thus been given to more basic studies of rumen function, oxygen transport, and reproductive physiology in the animal. Meanwhile agronomists were collecting and analyzing large numbers of plant samples from range land, farm fields, and experimental plots in an attempt to make an inventory of potential sources of high-nitrate feeds and to learn what factors govern accumulation. They found that while certain species regularly accumulated more nitrate than others, they were unable to produce plants containing predetermined levels of

NlTRATE ACCUMULATION AND POISONING

201

nitrate in quantities sufficient for feeding trials; this has hampered research in cooperation with animal scientists. Of necessity they, too, began to turn their attention to more basic studies. In retrospect it can be seen that agronomists, veterinarians, and animal husbandmen combined their efforts from the outset, but the parallel advances in toxicology, animal and plant physiology, and biochemistry have not always been exploited in research on nitrate accumulation and nitrate poisoning. High-nitrate well waters were identified as a cause of infant methemoglobinemia in 1945 and have received increasing attention from health authorities since that time. Silo gases (NO, NO2, NzO+), reported by several investigators over the years, were especially abundant in the western Corn Belt in the mid-1950’s and were shown to be derived from nitrate in the ensiled plant tissues. Recent spectacular losses due to nitrate poisoning or silo gases have increased public awareness of the fact that nitrate and its derivatives are potentially dangerous, Unfortunately there has been a tendency to label many ill-defined livestock problems as “nitrate poisoning.” This is understandable in view of the uncertainties surrounding the entire subject of chronic nitrate toxicity (discussed in Section V, C, 2), but many such “diagnoses” have been made by individuals who are not well acquainted with the problem or its manifestations. 111.

Accumulation of Nitrate

by

Plants

A. GENERAL METABOLICROLEOF NITRATE Most of the chemically combined nitrogen absorbed by plants is in the form of nitrate. The accumulation of nitrate thus implies that the rate of assimilation has not kept pace with the rate of uptake. Often the accumulation is only temporary, diminishing as the plant ages until at maturity little or no nitrate can be detected. Concentrations of nitrate sufficient to endanger livestock might sustain an actively growing plant for only a few days if further uptake were to cease. The presence of nitrate within the tissues of certain species of plants is normal. In some crops the nitrate content has been shown to be positively associated with ultimate yield, and tissue testing of these crops has been advocated as a guide to optimum fertilization. An accumulation of nitrate is not injurious to the plant, so far as is known. Although some of the conditions that promote nitrate accumulation, such as drought and mineral deficiencies, do produce characteristic symptoms suggesting that chemical analysis is in order, individual plants high in nitrate are often indistinguishable from those low in nitrate.

202

MADISON J. WRIGHT AND KENNETH L. DAVISON

There are several reports of high-nitrate plants that exhibited symptoms of nitrogen deficiency (Barker, 1962; Mulder, 1948; Nightingale et al., 1930). Each atom of nitrogen that is absorbed as nitrate and is assimilated as protein undergoes an 8-electron change of valence, from 5 to -3. In higher plants, the reduction is assumed to proceed along the general lines recently reviewed by Nason (1962), although much of the information on pathways has been derived from studies of microorganisms. Uncertainty persists as to the number and nature of intermediates that lie between nitrite (+3 ) , the first reduction product, and ammonia: amino nitrogen (- 3 ) . Intermediates other than nitrite have rarely been detected in more than trace amounts, and they appear to be toxic at higher concentrations. Even nitrite is not found in many analyses of highnitrate tissue. It appears likely that the interference with assimilation that brings about nitrate accumulation usually occurs in the conversion of nitrate to nitrite. Agronomists are familiar with the denitrification process in soils, by which gaseous nitrogen is liberated. They are perhaps less familiar with losses of nitrogenous gases from moist, high-nitrate plant tissues in limited-oxygen environments, e.g., masses of freshly ensiled corn. The “silo gases” evolved include not only molecular nitrogen (valence 0 ) but also nitrous oxide ( + l ) ,nitric oxide ( +2 ) , and nitrogen dioxide ( 4)-the latter two being dangerously corrosive-all lying between nitrate and ammonia in oxidation state. These gaseous products are not normally found in intact plants, but they are members of the oxidation series that must be negotiated or circumvented by the plant in assimilating nitrate. The formation of one or another of these gases, or of ammonia, has been shown to depend upon pH, availability of oxygen, and the presence of enzyme systems.

+

+

B. INTERNAL FACTORS GOVERNING ACCUMULATION 1 . Specific Differences Taxonomic units of plants are known to differ in tendency to accumulate nitrate. Both surveys of vegetation and planned comparisons in experiments have provided examples of varietal, specific, generic, and familial rankings. There are serious objections, however, to strict reliance on these rankings, especially those derived from surveys rather than experiments. Both stage of development and a number of environmental factors are known to influence nitrate content, and in most survey reports these variables are not tabulated. Plant-to-plant variation may be enormous (Kendrick et al., 1955). There is ample evidence that plants nor-

NITRATE ACCUMULATION AND POISONING

203

mally low in nitrate may under special conditions acumulate it to dangerous levels. As an example, the perennial forage grasses have been found to be relatively low in nitrate in many tests, and discounted as a source of trouble in areas where nitrate poisoning is an important problem (Gilbert et al., 1946; Whitehead and Moxon, 1952), but they have been observed to accumulate high levels of nitrate in other areas (Kretschmer, 1958; Olofsson, 1962; Stillings et al., 1961). Among the more comprehensive surveys ( Campbell, 1924; Gilbert et al., 1946; Olson and Whitehead, 1940; Sund and Wright, 1959; Webb, 1952,; Wilson, 1943), the recurrence of high nitrate content in certain families is notable. Members of the Amaranthaceae, Chenopodiaceae, Cruciferae, Compositae, Gramineae, and Solanaceae are listed especially often as accumulators. But these are large and economically important families in the regions where nitrate poisoning has occurred, and presumably they have been sampled more thoroughly than many others. So far as can be determined from the published reports, no single family of higher plants, nor even a single genus of a dozen or more species, has been systematically sampled for nitrate content. Taxonomic distinctions based on floral morphology may or may not prove to be associated with physiologic traits when research eventually provides this information. One of the few published comparisons of species grown in association and harvested at comparable stages of growth was conducted by Wilson (1943), who measured the nitrate content of sap expressed from each species: NO:<-N ( p.p.m. ) Plants growing in association Glycine m,Amaranthus retroflexus, 226, 323, 1329 Portulaca oleracea Panicum miliaceum, Helianthus tuberoms 87, 346 122, 207, 316 Panicum capillare, Fagopyrum esculentum, Amaranthus retroflexus Amaranthus retroflexus, Portulaca oleracea 1255, 1329 117, 226, 411 Panicum capillare, Glijcine m a , Amaranthus retroftexus Unfortunately the values cited in several other comparisons included by Wilson do not correspond to data presented in another part of the same paper. Generic or specific differences in tendency to store nitrate are, of course, of great importance in planning grazing or feeding programs. Varietal differences in nitrate content have been studied in considerable detail in oats ( G d and Kolp, 1960; Crawford et al., 1961), corn (Hoener and DeTurk, 1938; Zieserl et al., 1963); sugar beets (Soren-

204

MADISON J. WRIGHT AND KENNETH L. DAVISON

sen, 1962), perennial ryegrass and timothy (ap Griffith and Johnston, 1960), and cotton ( MacKenzie et al., 1963). In all cases varietal differences were found to be consistent though small. In testing 12 oat varieties at two locations, Gul and Kolp found that the rankings remained nearly the same although the levels at one station averaged nearly twice as high as at the other (Table I), The correlation between yield and nitrate TABLE I Nitrate-Nitrogen Content, as Per Cent of Dry Matter, of Oat Varieties Grown for Hay under Irrigation and Fertilization (Laramie) and under Dryland Conditions (Archer) in Wyominga Stageb: Variety

Overall rank

Laramie 1

2

3

Archer

4

1

2

3

4

Overall mean

Lowest 1.22 0.87 0.78 0.85 0.40 0.41 0.20 0.18 0.59 Highest 1.36 1.06 0.84 0.87 1.00 0.64 0.42 0.19 0.80 12-Variety average 1.32 1.00 0.81 0.66 0.58 0.48 0.33 0.20 0.67

IMPROVED GARRY SWEDISH SELECT

Data of Gul and Kolp (1960). Stages: 1-25% of heads flowering; 2-50% of heads in milk; 3-50% in soft dough; 4 4 0 % of heads in hard dough. a b

of heads

content was nonsignificant. It has been suggested that differences of this kind offer opportunities for breeding low-nitrate varieties, especially since such a variety is likely to be high in available carbohydrate (Jones et al., 1961) or resistant to effects of mutual shading (Knipmeyer et al., 1962). 2. Localization The fact that nitrate is not uniformly distributed throughout the various plant tissues was recognized very early (Berthelot, 1884). Much of the later research on nitrate accumulation has involved measurements of levels in leaf, stem, and flowering structures. In terms of gross structure, stems usually contain more nitrate than leaves, and leaves in turn more nitrate than floral parts. Published measurements of the nitrate content of roots are few, but suggest a level lower than that in stems yet above that in leaves. The lower portions of stems tend to be higher in nitrate content than the upper portions. Data of Whitehead and Moxon (1952) are presented in Table I1 as typical of many findings, including those of Hanway (1962). The older, outer leaves of fodder sugar beets were found to be lowest in total nitrogen but highest in nitrate nitrogen (N03-N) in analyses made by Sorensen (1962). As is usual, the petioles accumulated several times as much nitrate as the blades (Table 111). Sorensen attributed the distribution of nitrate to a relatively slow metabolism in the older parts.

TABLE I1 Distribution of Nitrate-Nitrogen in Dent Corn Plants Sampled September 4, 1945, Expressed as Per Cent of Air-Dry Weight'

2

Nodal position from root to tassel Sample Leaf lamina Leaf midrib Leaf sheath Internode

Shank Ear (fertile) Tassel

1

2

0.049 0.081 0.126 0.293

0.034 0.063 0.060 0.269 0.085

-

-

3

4

0.024 0.014 0.052 0.048 0.060 0.043 0.202 '0.139 0.046 0.049

-

Data of Whitehead and Moxon (1952).

-

5

6

7

8

9

10

0.007 0.031 0.032 0.120 0.027 0.008

0.007 0.031 0.024 0.097 0.020 0.011

0.006

0.007

0.006

0.008

0.004

0.004

-

-

-

0.021 0.092 0.015 0.010

0.012 0.092

0.012 0.080

0.010 0.081

0.024 0.063

-

-

-

-

-

-

-

-

-

11 0.006

-

12

13

0.007

0.008

-

-

0.020 0.066

0.006 0.095

-

-

-

0.003 0.008

-

-- 3 0.020

8 2:

*

21

206

MADISON J. WRIGHT AND KENNETH L. DAVISON

In a more detailed study conducted by Harker and Kamau (1961) on a shoot of Napiergrass, Pennisetum purpureum, nitrate was found to be distributed in a distinct basipetal pattern within leaves. A portion of these data are presented in Table IV. Localization of nitrate in particular tissues or cells has been studied very little since the early work of Schimper ( 1888) and Zacharias ( 1884), although this appears to be a logical line of investigation for those concerned with dangerous accumulations. TABLE I11 Content of Total Nitrogen and Nitrate Nitrogen in Leaves of Fodder Sugar Beets Harvested August 27. Values are Percentages on a Dry Matter Basisa Parameter Petioles Total N NO,-N NO,-N, Blades Total N NO,-N NO,-N, a

Center

Leaf position Inner Middle

Outer

Weighted mean

% of total

3.80 0.23 6.0

2.03 0.45 22.1

1.73 0.74 42.5

1.89 1.28 67.5

1.87 0.83 44.2

% of total

5.70 0.07 1.2

4.80 0.06 1.4

4.15 0.10 2.4

3.60 0.18 5.0

4.10 0.12 3.0

Data of Sorensen (1962).

A major consideration in the localization of nitrate accumulation is the site of reduction. In a review of transport in the xylem, Bollard (1960) states that in woody plants as a group, nitrate is reduced in the root system. He found no nitrate at all in the xylem sap of 68 of the 110 dicotyledonous species he examined (Bollard, 1957). Some nitrogen was transported in organic (reduced) form in all species tested. The accumulation of nitrate in the aerial portions of many herbaceous plants has been taken to indicate that reduction does not occur in their roots, although Bollard is of the opinion that substantial amounts of reduced nitrogen compounds would be found in the xylem sap if methods of extraction could be developed. In support of his opinion is Hageman and Flesher’s (1960) observation that the roots of corn seedlings had 80 per cent as much nitrate reductase activity as the shoots. It may be noted in this connection that in tissue testing procedures for trees and sugar cane, the leaf is assayed for total N whereas in corn, cotton, and sugar beet, the usual test is for nitrate. The more general opinion is that reductive activity in herbaceous species is concentrated in the photosynthetic part of the leaves. Localization of radioactive molybdenum in tomato has supported this opinion

TABLE IV

Fj

Distribution of Nitrate Nitrogen in 2-Inch Segments of Leaves on a Shoot of Pennisetum purpureuma. b Leaf Outside ( lowest ) leaf Fourth leaf Eighth leaf

Sheath segments 2-4” p6“

&p

0 3

0 3 ( concealed )

Blade segments 0-y

M”

4-6“

G-8”

8-10”

0 1 1

-

0 3

-

-

0 2

0 2

0 2

Data of Harker and Kamau (1961). Rough quantitative analysis with diphenylamine: 0 NO,-N or more.

0 3

10-12” 0 0

12-14

1P16”

-

-

0

0

a b

= none;

1

= 1 p.p.m. NO,-N; 2 = 2 p.p.m. NO,-N; 3 =

6 p.p.m.

Bs 4

cl

$ 4 3 2

Bz 8

208

MADISON J. WRIGHT AND KENNETH L. DAVISON

(Stout and Meagher, 1948). There is no doubt that the xylem sap of many plants does contain nitrate; in some species, concentrations are well above 250 p.p.m. N03-N (Lowry et al., 1936; Holley et al., 1931). No explanation has been offered as to why woody plants should have a different site of reduction than herbaceous plants, and McKee (1962) has judged the generalization to be premature, since leaves of certain woody species do contain nitrate. Little discussion of the reasons for localization of nitrate appears in the numerous studies of its occurrence. The explanation that appears to have been tacitly accepted is that most of the nitrate is in the vascular system en route to a site of reduction, and that its passage has been somehow retarded. Since the source is the soil and assimilatory demands are mainly in meristems, a progressive diminution of nitrate concentration with height above the roots has been assumed. This conception of localization is entirely inadequate in cases of extreme accumulation, for the volume of the xylem is too small to account for observed concentrations even if the system were packed with crystals. The very high levels sometimes found in stems, midribs, and petioles probably represent active accumulation by tissues adjacent to the vessels. Vascular parenchyma and other highly protoplasmic cells might thus be the main sites of heavy accumulation. Possibly the lag in nitrate accumulation that has been observed in very young plants is in some way related to the course of development of cells capable of accumulation. An extension of studies on nitrate reductase to the nitrate-storing tissues might lead eventually to a sounder conception of the process of accumulation.

3. Age Experiments involving periodic sampling of plants through a cycle of growth have shown that nitrate content first rises and then, after reaching a peak about the prebloom stage, declines as the plant matures. The importance of age itself is unknown, and alternative explanations are usually available. One of these is that changing proportions of stem, leaf, and fruit accompany maturation. Fruits or seeds usually contain very little nitrate, and as they increase in dry matter the effect of high nitrate content in other parts tends to be diluted. Secondly, the formation of fruit or seed makes a heavy demand on available nitrogen (Whitehead et d.,1948). In instances such as those reported by Davidson et al., (1941) with oats, and Flynn et al. (1957) with corn, drought prevented or greatly reduced grain formation and the nitrate content of the vegetative parts was maintained at a level much higher than normal.

NITFtATE ACCUMULATION AND POISONING

209

A third explanation for the apparent association between age and nitrate content is that the nitrogen supplying power of the soil usually diminishes as maturation approaches, permitting the plant to assimilate most of the nitrate that accumulated when more was available. Several experimenters have found that repeated (Stahl and Shive, 1933; Crawford et a!., 1961) or delayed (Nowakowski, 1961) applications of nitrogen through the season will partly overcome the usual downward trend in nitrate content. In other cases, however, heavy mid-season applications of nitrogen to corn have not raised the nitrate content. In some, but not all, of these cases the soil may have been too dry to move the nitrogen to active roots. 4. Other Internal Factors Influencing Accumulation In recent years much attention has been given to the enzyme nitrate reductase as it relates to nitrogen accumulation. Hageman and associates at Illinois have demonstrated that level of nitrate reductase activity in corn is genetically regulated (Zieserl and Hageman, 1962) and heritable (Hageman et al., 1963); that the enzyme is light activated (Hageman and Flesher, 1960) and exhibits diurnal (Hageman et al., 1961) and seasonal (Zieserl et al., 1963) fluctuations in activity; that the activity is influenced by self-shading and mutual shading at high populations (Zieserl et al., 1963), and by degree of hydration (Hageman, 1963); and that it is substrate dependent, requiring nitrate (Hageman and Flesher, 1960). Research on cauliflower at the Long Ashton station in England has developed similar information, especially with reference to molybdenum (Candela et al., 1957). The nitrate reductase activity of four hybrids, as averaged over the season, was shown to be negatively correlated with the nitrate content of these hybrids in one experiment (Zieserl et al., 1963). However, not all tissues high in nitrate are necessarily low in reductase activity, since nitrate accumulation indicates only that there is a difference between rates of input and output or conversion, nor can it be assumed that all tissues low in nitrate have a high rate of activity. There is good reason to believe that the initial (i.e., nitrate to nitrite) step limits the rate of the process by which nitrate is reduced to amino nitrogen (Hageman et al., 1962), and Hageman thinks it likely that in corn this rate also limits protein synthesis and ultimately growth and yield. Hence the nitrate reductase activity of a breeding line might suggest its yield potential, especially if it is known how greatly the activity is influenced by pertinent agronomic variables. In connection with the nitrate accumulation problem it might be desirable to select for high reductase activity in roots.

210

MADISON J. WRIGHT AND KENNETH L. DAVISON

A source of energy is needed, along with the enzyme and an electron donor, to reduce nitrate. Although nitrate reduction in some plants has been shown to be closely linked to a photoreaction that provides energy and electrons directly, the large number of species that reduce nitrate in the roots furnish proof that a respiratory mechanism is also important. Several studies have dealt with the relationship between carbohydrate content and nitrate accumulation. In most of these a negative relationship between the two has been clearly shown. Evidently the presence of a large supply of nutrient nitrogen stimulates the plant to draw upon its supply of available carbohydrate for reductive energy and carbon skeletons, and eventually the carbohydrate reserves may be insufficient to keep pace with nitrate uptake. Thus in daily analyses of corn seedlings grown in nutrient solutions Burt (1963) found that as the total available carbohydrate content fell from 22 per cent of the dry matter to 14 per cent, the nitrate-nitrogen content rose from 0.02 to 0.80 per cent. In discussing tissue tests for sugar beets, Ulrich (1961) recommends maintaining a nitrate content of not less than 1000 p.p.m. N03-N in petioles until the last few weeks preceding harvest, but warns that a nitrogen deficiency is then required to promote the accumulation of sugar. Applications of nitrogen fertilizer on several dates to two forage grasses increased the nitrate content and decreased the water-soluble carbohydrate content in each case (Jones et al., 1961). In oats a decline in the nitrogen-free extract fraction at head emergence coincided with a peak of nitrate content (Smith, 1960). A practical consideration in the carbohydrate-nitrate relationship is the desirability of feeding an energy-rich diet to animals that are ingesting nitrate (Crawford and Kennedy, 1960). The findings reviewed here indicate that a high-nitrate plant is likely to be low in readily available carbohydrates. C. EXTERNAL FACTORS GOVERNING ACCUMULATION 1. Nutrient SupZt~ The relationship between nutrient supply and nitrate content has received more attention than any other aspect of research on nitrate in plants. a. Nitrogen. It appears obvious that nitrate will not accumulate within a plant unless the external medium can furnish it at a rate faster than the rate of assimilation. This statement would be refuted if it could be shown that oxidation of some other nitrogenous nutrient, possibly ammonium, to nitrate, can occur in quantity after absorption; this is without general experimental support and seems inconsistent with what is known of the disposal of excess reduced nitrogen.

NITRATE ACCUMULATION AND POISONING

211

Numerous experiments might be cited in which an increase in external nutrient nitrogen has effected an increase in nitrate accumulation, but only a few will be mentioned here. Essentially a straight-line relationship was reported by ap GrifEth (1960) from experiments in which rates of nitrogen from 0 to 252 pounds per acre were applied to three forage mixtures. Oats growing in nutrient solutions were found to contain nitrate in amounts that "varied with, but not in proportion to" the nitrate in the solutions, reaching a maximum of 2,.3 per cent N03-N on a dry

.).:

9.000

8,000

JUNE

JULY

AUGUST

FIG. 1. Effect of rate and timing of fertilization on concentration of NO,-N in sugar beet petioles at Davis, California (Ulrich et al., 1959).

matter basis (Sessions and Shive, 1933). Nitrate contents of sugar beets have been especially responsive to fertilizer treatments in experiments conducted in Denmark (Sorensen, 1960) and California (Ulrich et al., 1959). An example of the latter data can be seen in Fig. 1. Three successive fertilizer-induced increases in nitrate content in a single season were recorded by Hvidsten et al. (1959). On the other hand, Ackerson (1963), Becker and Oslage (1955), Wilson ( 1943), and Wolfer (1941) reported no appreciable increase in nitrate content following heavy applications of nitrogenous fertilizers, and recent workers at the Illinois (Smith, 1963), New York (Davison and Wright, 1963), and Wisconsin (Sund, 1959) stations have failed at times to achieve high accumulation despite previous successes. No more

212

MADISON J. WRIGHT AND KENNETH L. DAVISON

than a tentative explanation is available for most of these seeming anomalies. A plateau of nitrate accumulation has been observed in solution culture work by Burt (1963), Hoener and DeTurk (1938), and Barker (1962), and in the field by Crawford et al. (1961) and Flynn et al. (1957). In some other field trials, however, rates of application at or near loo0 pounds nitrogen per acre have not produced a clear-cut maximum of nitrate content (Crawford et al., 1961; Wright et al., 1960). Such discrepancies are doubtless traceable to the rates at which the nitrate was absorbed from the substrate and reduced within the plant, rates that are influenced by the factors under review. In Burt’s experiments a very low concentration of nitrate in the nutrient solution was sufficient to produce maximal internal accumulation if the concentration was maintained by constant replenishment. It should be emphasized that commercial fertilizer nitrogen is not essential for the production of high-nitrate plants. Nitrate poisoning studies began in areas where no commercial fertilizers of any kind were applied to crops or ranges. In some cases abnormal accumulation has been associated with heavy dressings of animal manure Mayo’s (1895) two highest-nitrate corn samples came from a former cattle corral and hog yard. The Florida Everglades (Kretschmer, 1958) and the Central Wisconsin peat marshes (Sund et al., 1957) have produced high-nitrate crops and weeds without fertilization. Clean fallowing has been shown to increase the supply of available nitrate (Finnell, 1932; Ogata and Caldwell, 1960) and thereby increase the nitrate content of plants (Doughty and Warder, 1942; Kretschmer, 1958). Considerable difficulty has dogged attempts to describe the relationship between nitrate content of soils, as determined by soil tests, and nitrate accumulation. For every experiment in which a close relationship was shown (Whitehead and Moxon, 1952) there have been several in which only a weak or irregular association could be recognized, or there was none at all. Perhaps, as Welch and Bartholomew (1963) have suggested, the expectation that a soil test will permit prediction of accumulation is unreasonable in view of the other factors involved. A recent statement by Bould (cf. Bould and Hewitt, 1963, p. 55) seems pertinent: “Because nitrate is entirely contained in the soil solution, and is rapidly absorbed by plants, the rate at which it is replenished is more important than the total amount of nitrate nitrogen present in the soil.” Timing of nitrogen fertilization has been shown to have a marked influence on the nitrate content of pasture grasses, applications soon before harvest tending to increase accumulation. The timing effect also shows clearly in Fig. 1.

NlTlUm ACCUh4ULATION AND POISONING

213

Most comparisons of nitrogen carriers added to soils have led to the conclusion that differences in effects on nitrate accumulation are small, or lacking altogether (ap Griffith and Johnston, 1960; Crawford et al., 1961; Gilbert et al., 1946; Sund, 1959). Nowakowski (1961)) however, while minimizing the practical importance of the differences, found substantially higher nitrate contents associated with applications of Ca( N03)2 and NH4N03as opposed to urea and ( NH4)$04 at equivalent rates. He also found that swards treated with solid forms of these four carriers accumulated more nitrate than those given solutions. No explanation was offered. In nutrient solutions, where various carriers may retain their identity until absorption occurs, differences associated with these carriers may be more readiIy observed. There has been a great amount of experimentation involving comparisons of ammonium and nitrate, but uncertainty persists. Environmental influences on uptake ( Lycklama, 1983), microbial conversions, and other complications cast doubt on the validity of certain conclusions that have been published ( McKee, 1962). b. Other nutrients. From the first identification of potassium nitrate crystals there has been evidence of a strong positive relationship between nitrate content and potassium content. Fairly high yields of potassium nitrate have been obtained by extraction (Bradley et al., 1939b; Rimington and Quin, 1933). It has been assumed that potassium is the cation most readily available to preserve electrical neutrality when nitrate accumulates. These two ions constitute the best examples of luxury consumption in higher plants. Solution culture experiments have shown that nitrate is taken up more readily from solutions prepared with potassium nitrate than from those prepared with calcium or sodium nitrate (Tottingham et al., 1934), and that increases in the level of potassium in the solution promote an accumulation of nitrate in oats (Sund and Mao, 1958)) and in corn when nitrate is also present at a high level (Barker, 1962). But it has also been shown (Nightingale et al., 1930) that tomato plants growing in solutions devoid of potassium may accumulate high levels of nitrate while showing typical nitrogen-deficiency symptoms. In field experiments applications of potassium have had much less influence on nitrate accumulation. On low-nitrogen, unlimed soils Lowry et al. (1936) found that adding potassium decreased the nitrate content of xylem sap exuded by corn plants. Nitrogen raised the nitrate content; nitrogen and potassium together produced a nitrate level equal to that in sap from the untreated check. Crawford et al. (1961) detected only a slight increase in nitrate content of oats in response to an application of 83 pounds of potassium per acre, and this effect was lost as the oats

214

MADISON J. WRIGHT AND KENNETH L. DAVISON

passed from the vegetative to the dough stage. No significant effect on nitrate content of blue panicgrass was associated with three levels of potassium fertilization in a factorial trial conducted by Wright et al. (1960), nor was the N X K interaction significant. Puzzling results were obtained by Sorensen (1960) when he analyzed roots, petioles, and blades of fodder sugar beets given 0, 300, or 600 kg. of nitrogen per hectare. The nitrate contents ranked as expected, but in the petioles, where nitrate contents were highest, the maximum level of potassium (4.59 per cent of the dry matter) was found in plants given no introgen and the minimum (2.72 per cent) in those given 600 kg. per hectare. Phosphorus fertilization of soils has raised the nitrate content of plants in some experiments (Doughty and Warder, 1942; Hanway and Moldenhauer, 1964), whereas in others it has lowered nitrate content (Gilbert et al., 1946), had no effect (Wright et al., 1960) or variable effects. Presumably it has an indirect effect on nitrate content through its multiple functions in plant metabolism. In solution culture high concentrations of phosphate have tended to depress the absorption of nitrate. Deficiencies of two micronutrients, molybdenum and manganese, have been shown to produce accumulations of nitrate in higher plants as well as microorganisms. From recent research it has been shown that molybdenum is the metallic component of the enzyme nitrate reductase, while manganese performs a similar function in hydroxylamine reductase (Nason, 1962). Although a correction of manganese deficiency brought about a disappearance of nitrate from tissues of oats and Phahris minor grown on two Australian soils (Leeper, 1941), and restoration of molybdenum to nutrient solutions lowered the nitrate content of leaves and stems of tomato plants (Mulder, 1948; Stout and Meagher, 1948), limited attempts to reduce nitrate accumulations by applying these two elements to nondeficient soils have not been completely successful. Reports on the influence of other nutrients on nitrate accumulation are not sufficiently numerous or complete to support firm conclusions. It has been proposed that accumulation of nitrate might be limited by “balancing a high level of available nitrogen in the soil with heavy applications of other plant nutrients. Since both nitrogen and potassium can accumulate within the plant to levels much above those needed for maximum growth, and since increases in available potassium and phosphorus may encourage accumulation of nitrate, the attempt at balancing appears unlikely to provide dependable control of nitrate levels within the plant unless total growth rate is being limited by a correctable nutrient deficiency.

NITRATE ACCUMULATION AND POISONING

215

2. Moisture Nitrate poisoning cases have most often been concentrated in semiarid and subhumid regions, and elsewhere they have tended to occur when marked departures from the normal amount or distribution of rainfall were experienced. There has been, however, only a limited amount of experimental inquiry into the relationship between moisture supply and nitrate accumulation. Most of the conclusions that have been reached are products of field observations during severe outbreaks. The corn held responsible for two of the Kansas poisonings reported by Mayo was in one case drought stricken, and in the other harvested at an immature stage, doubtless for the same reason. Similarly the oats investigated by Davidson et nl. (1941) had been affected by drought, as had the cornstalks grazed by cattle in Missouri in 1954 (Garner, 1958). A long-sustained drought, on the other hand, is not as likely to bring about nitrate accumulation as a brief one (Davidson et al., 1941; Gilbert et al., 1946). Several moisture-dependent processes evidently contribute to the accumulation of nitrate, In the soil, nutrient nitrogen is released from more complex organic forms by microbial activity that requires moisture. The nitrogen thus released, as well as any added as fertilizer, must then move through water to the absorbing roots, and then must travel in solution in the xylem to reach the parts normally used as feed. The events so far described require at least a moderate degree of wetness in the soil. And from a practical point of view, moisture must be available long enough to permit nitrate-accumulating seedling plants to develop a sizable volume of tissue capable of storing nitrate. Established plants that have been in a state of dormancy due to drought may require only a very few days to accumulate a dangerous amount of nitrate once microbial release and processes of transport are stimulated by rains (Williams and Hines, 1940; Diven, 1963). When a plant experiences a moisture shortage there is a general disturbance of assimilatory processes. The rate of reduction of nitrate is slowed, in part by a drop in nitrate reductase activity. The soil may still be wet enough, however, to sustain mineralization and delivery of nitrogen, especially if the plant is losing turgor only during the hours of maximum evaporative load and regaining it at night. Under these conditions accumulation of nitrate may proceed rapidly. If dry weather continues, the normal decline associated with rapid translocation and assimilation of nitrogen during fruiting may never occur. References are made to the barrenness of toxic corn plants or the shriveled heads in poisonous oat hays (Davidson et aZ., 1941; Flynn et al., 1957; Hanway and Engel-

216

MADISON J. WRIGHT AND KENNETH L. DAVISON

horn, 1958; Mayo, 1895); conversely, nitrate was rarely found in corn stalks above ears that filled (Brady et al., 1955). Experimentation with moisture stresses has been limited, perhaps because of the difficulty of establishing and maintaining known moisture stresses. Wright and Trautman (1962) in Arizona were unsuccessful in attempts to influence nitrate accumulation by deferring irrigation of blue panicgrass, Panicum antidotale, until soil moisture approached the wilting point at 6, 12, 18, or 24 inches of depth. McKenzie et al. (1963) achieved slight increases in nitrate Ievel of cotton petioles by irrigating only at higher soil moisture tensions, but observed that such increases were unimportant compared to those associated with fertilization. Doughty and Warder (1942) did succeed in raising the nitrate content of oat straw in most of their pot cultures by subjecting them to “a partial drought” in a manner unspecified, and in another Saskatchewan greenhouse experiment cited by Whitehead and Moxon (1952) oat plants also responded rapidly to a cessation of moisture stress (see tabulation). NO,-N, per cent of D.M. Soil conditions First 4 weeks

Fifth week

After first 4 weeks

After week

Moist Near wilting

Near wilting Moist

0.92 1.57

1.30 1.04

fifth

There seem to be good opportunities for further research in this area.

3. Light The influence of the intensity, the duration, and the quality of light on nitrate accumulation has been investigated by agronomists, who were stimulated largely by the shading experiments of Gilbert and associates (1946) that proceeded from an observation that much of the toxic oat hay was produced in narrow valleys; and by plant physiologists, who were seeking to elucidate the mechanism for assimilation of nitrogen. Light intensity has been altered in a large number of experiments. This has usually been done by attenuation of sunlight through layers of gauze, glass, or plastic film, although more recently environmental control chambers have come into use for investigations of nitrate accumulation. The results of all these trials agree closely; a reduction in light intensity has been associated with an increase in nitrate content. As examples, white clover plants contained 4630, 6920, and 9270 p.p.m. N03-N, respectively, when given 100, 50, and 28 per cent of incident light (Bathurst and Mitchell, 1958); corn had twice as much nitrate under 35 per cent shade as when unshaded, and nitrate content increased still

NITRATE ACCUMULATION AM) POISONING

217

further as the shading was intensified to 70,80, or 90 per cent (Knipmeyer et al., 1962); and both sugar beet and mustard plants contained progressively less nitrate as light intensity was raised from 630 to 940 to 1240 foot-candles, regardless of herbicidal treatment (Whitehead et al., 1956). Day-to-day and diurnal variations in nitrate concentration have been observed (Crawford et al., 1961; Breniman et al., 1961b; Frank and Grigsby, 1957; Hageman et al., 1961; Sorensen, 1962). Nitrate concentration tends to be high on cloudy days and at the end of the dark period, and has been shown to be inversely related to nitrate reductase activity by Hageman and associates and Candela et al. (1957). Sorensen has suggested that the diurnal cycle of nitrate content is explainable in terms of relative rates of photosynthesis and nitrogen assimilation. Certainly it has been clearly indicated that hour of sampling should be specified in future reports, along with notations of weather conditions. Duration of illumination was shown to influence nitrate content in greenhouse-grown oats (Whitehead and Olson, 1941 ). Plants given continuous illumination contained only about two-thirds as much nitrate as those receiving daylight only. In a greenhouse or growth chamber (Monson, 1963) such an effect may suggest insufficient light intensity, however, rather than a true photoperiodic response. Effects of light quality on both uptake and assimilation of nitrogen were investigated vigorously during the 1920's and 1930's, and differences in nitrate content were reported to be due to spectral distribution. Because of the difficulties that were encountered in providing radiant energy of high and equal intensity in different portions of the visible and near-visible ranges, little current interest attaches to those attempts to establish an action spectrum distinct from that for photosynthesis. More recent research by Stoy (1956) has renewed the suggestion, however, that nitrate assimilation may be assisted by energy from a part of the spectrum that is unimportant photosynthetically. 4. Herbicides

Shortly after the large-scale adoption of selective herbicides, Stahler and Whitehead (1950) reported that sugar beet leaves inadvertently sprayed with 2,4-D contained 20 times as much nitrate as untreated leaves from nearby farms. The maximum value found, 1.21 per cent N03-N, was almost 40 times the mean for untreated leaves. At about the same time, other reports suggested that herbicidal treatment might enhance the palatability of plants normally refused, including some known to be high in nitrate (Willard, 1950).

218

MADISON J. W I G H T AND KENNETH L. DAVISON

The stage was thus set for numerous “diagnoses” of nitrate poisoning due to accumulations induced by herbicides ( Fertig, 1952), and experiments on the relationship between herbicidal treatment and nitrate content were instituted. The most extensive of these appear to be the works of Berg and McElroy (1953) and Frank and Grisby (1957). The former analyzed a large number of crops and weeds before and after treatment with 2,4-D at three rates and found no accumulation save in four instances, the most clear-cut increase being found in Russian pigweed (Axyris amaranthoides). Frank and Grisby confined their study to 14 species growing on muck soil, but treated each with six herbicides at sublethal concentrations. They observed a variety of short-term and long-term effects in the array of possible species-chemical combinations. Some species were seemingly unresponsive, whereas others increased or decreased in nitrate content following certain chemical treatments. An indication of short-term and long-term effects was also provided in experiments by Whitehead and associates (1956) and Fertig (1952). The attractiveness of weeds to livestock has been reported by some workers to be increased by spraying (Whitehead and Moxon, 1952; Willard, 1950), possibly because of increased succulence where hormonetype materials are used. Although Grisby and Farwell (1950) observed a general rejection of sprayed alfalfa-bromegrass by several kinds of animals, the high rates they used and the short time elapsed suggest that gross surface contamination rather than internal change was responsible. When the large numbers of species, chemicals, rates of application, ecological influences, and opportunities for selection by animals are taken into account, it is understandable that satisfactory generalizations regarding the influence of herbicides on nitrate poisoning are not available. An illustration of this uncertainty is provided by the Wisconsin marshland situation described by Sund and associates (1957; Sund and Wright, 1959), in which herbicides were used successfully to eliminate dozens of nongrassy species from pastures, thereby greatly reducing the average nitrate content of available vegetation and concurrently preventing noninfectious abortions of cattle grazing there, A complete analysis of the chemical composition of each species before and after spraying, and its degree of preference and use, was clearly out of the question in such circumstances. The Wisconsin workers were therefore limited to placing the responsibility on “plants high in nitrate,” since other constituents might have been involved. At present the adoption of herbicides is far more rapid than the elucidation of their mode of action. When such materials are applied to large numbers of crops and weeds, occasional side effects such as nitrate accumulation are to be expected.

NITRATE ACCUMULATION AND POISONING

219

5. Other External Factors The mechanism by which temperature influences nitrate content is not altogether clear, but undoubtedly a relationship does exist. A striking example was provided by Nightingale et al. (1930), who showed two soybean seedlings much alike in size and development. One of these contained three times as much nitrogen as the other, and 56 per cent of it was in the form of nitrate; in the low-nitrogen plant no nitrate was present. The high-nitrate plant was grown at low temperature. Studies of winter oats under field conditions in Florida, however, have demonstrated a strong association between high temperature and high nitrate content (Kretschmer, 1958). Nightingale et al. (1930) attributed the nitrate accumulation at low temperature to a slow rate of assimilation relative to uptake, whereas in Kretschmer's experiments an accelerated rate of nitrification in the virgin peat-muck soil may have been responsible. Evidence that the temperature effect varies with species was provided by the comprehensive experiments carried out by Bathurst and Mitchell (1958) in growth cabinets. Three species were grown at each of several temperatures, as follows: Dallisgrass (Paspalurndilatatum), 45" to 95°F. by 10°F. intervals; short-rotation ryegrass (Lolium sp.), 45" to 85"; subterranean clover ( Trifolium subterraneum), 45" to 75". Nitrate content in P q u l u m was lowest at 65", in ryegrass at 45", and in subterranean clover at 45". The chemical fractionation of these plants revealed a negative relationship between fructosan and nitrate contents in the grasses, and between total sugars and nitrate in subterranean clover. The general impression that temperature effects on a constituent such as nitrate are indirect is supported by these experiments and may account in part for the dearth of research along these lines. Nitrate has been observed to accumulate at different rates in plants grown on different soil types (Gilbert et al., 1946; Doughty and Warder, 1942), but the characterization of the soil was insufficient to establish whether these differences were due to factors considered above or to others, and Cook (1930) concluded that the influence of soil type was subordinate to that of fertility. Emphasis in the Wyoming reports is on parent material. Hanway and associates (1963) have drawn attention to the importance of degree of aeration of soils. I t appears that nothing has been reported regarding the effect of parasitization by diseases or insects on the nitrate content of the host, unless the symptoms reported by Wilson (1943) on muskmelon and snapdragon were due to pathogens rather than to physiological disturbances. It would be expected that parasites, by depleting the food reserves of plants, would accentuate nitrate accumulation.

220

MADISON J. WRIGHT AND KENNETH L. DAVISON

IV. Postharvest Losses

Methods of harvesting and handling influence the degree to which nitrate within plants is retained. On the one hand, shattering of leaves may enhance the toxicity of a feed by increasing the proportion of stems, or some of the nitrate may be converted to nitrite by microbial activity. On the other hand, nitrate and its decomposition products may be partially, or even completely, eliminated from the feed. The postharvest conversion of nitrate to nitrite has been reported in oat hay (Gilbert et aE., 1946; Olson and Moxon, 1942; Riggs, 1945), Amaranthus hay (Olson and Moxon, 1942), mangels and turnips prepared by heating (Robinson, 1942; McIntosh et al., 1943), Tribulus slurries (Rimington and Quin, 1933), and corn (Whitehead et al., 1948). It may also have been responsible for the deaths of cattle feeding on sugar beet tops that in previous weeks had been harmless (Savage, 1949). Two minor experiments have shown that the loss of nitrate on slow drying of high-moisture, high-nitrate plants was rather small: Iess than 10 per cent in 64 hours from sugar beets (Sorensen, 1962), and less than 16 per cent in 3 days from mixed forage (Kretschmer, 1958). The extent of loss of nitrate from feeds by leaching has not been reported. Losses of nitrate during ensiling may be rapid, large, and attended by extreme hazards to life. By an anaerobic reductive process diverging from the normal assimilatory path at the nitrite stage, various oxides of nitrogen and molecular nitrogen itself are produced. This denitrification has generally been assumed to be the work of microorganisms, although Wang and Burris (1960) detected these same gases when aseptically grown soybean seedlings were packed in “silo tubes.” A succinct review of the literature on denitrification of plant materials may be found in the paper by Peterson et al. ( 1958). The evolution of gaseous oxides of nitrogen from farm silos and laboratory vessels was followed in a series of experiments employing mass spectrometry (Peterson et al., 1958; Wang and Burris, 1960). It was found that nitric oxide was the first to appear, and usually reached a peak concentration in 20 to 30 hours. Nitrous oxide reached a peak concentration at about 50 hours. The time of most rapid production of Nz appeared to be delayed, but it was not precisely determined. The nitrogen dioxide responsible for most of the hazard to health is formed when NO subsequently reacts with oxygen, and the NO2 in turn may dimerize to form Nz04. Experiments by the same research group with nitrate and amino acids labeled with NIB indicated that some of the gaseous nitrogen had been

NITRATE ACCUMULATION AND POISONING

221

formed from amino nitrogen, probably by reaction with nitrite, but about 10 times as much was recovered from labeled nitrate, marking it as the primary source of the oxides, The safety of the high-nitrate silages that have undergone a gaseous denitrification has been an important economic consideration. Research has demonstrated that while the nitrate content falls, it does not always reach safe levels (Whitehead and Moxon, 1952; Brady et al., 1955; Whitehead, 1961). As research on the subject increased, it appeared that the loss was influenced by several factors, some of which could be regulated. In experiments with forages of several moisture contents, Jacobson and Wiseman (1963) observed that only about 20 per cent of the nitrate disappeared from material ensiled at 55 per cent moisture, whereas 61 to 98 per cent disappeared from material stored at 80 per cent moisture. The addition of sugar limited the loss of nitrate from highmoisture silages, as did laceration prior to storage. Chloroform and toluene stopped bacterial action and with it the evolution of oxides of nitrogen, thus maintaining the nitrate content (Peterson et al., 1958). Sodium metabisulfite greatly inhibited the normal progression of gases both in the laboratory and in farm silos (Breniman et al., 1961a; Wang and Burris, 1960), while several other additives had minor effects (Breniman et al., 1961a). For rapid denitrification to occur in plant materials it appears that a rapid reduction of nitrate to nitrite must be followed by autoxidation of nitrite: 3 HNOz + 2 NO H N 0 3 H20. Although the initial reduction is biological, the reaction shown is chemical. It requires an acid medium (below pH 5.5), so the microbial system reducing nitrate to nitrite must be acid tolerant if the evolution of gases is to be maintained (Alexander, 1961). This may account for the inhibition of losses observed in lacerated silages or in those prepared with sugar, since either modification promotes acidification. Conversely, in silages that never achieved a low pH, or lost it after a period of several weeks, there was a rapid loss of nitrate (Jacobson and Wiseman, 1963). These results suggest that rapid loss of nitrate from silos by denitrification is likely to occur under conditions associated with undesirable changes in feeding value.

+

+

V. Toxicity of Nitrate to Animals

The nitrate ion per se is relatively nontoxic to animals. Dogs have received daily doses as high as 280 mg. NO3-N per kilogram as sodium nitrate orally, or 570 mg. N03-N per kilogram intravenously over several weeks without death or alteration of blood pressure, although water consumption and urine excretion were increased (Greene and Hiatt,

222

MADISON J. WRIGHT AND KENNETH L. DAVISON

1954). The nitrate was distributed throughout the body and as much as 90 per cent was recovered in the urine. The urinary excretion of sodium and chloride was greatly increased with as much as 40 per cent more sodium being excreted than injected. Dogs injected repeatedly with ammonium nitrate over a period of 13 weeks have lost as much as 70 per cent of their total body chloride, which was replaced by nitrate (Hiatt, 1940). By this time they had lost their appetite, could not pant, and were dry around the eyes, nose, and mouth. The chloride was rapidly replaced by normal processes upon cessation of the nitrate treatment. A similarly high tolerance to oral administration of nitrate was observed in rats and rabbits (Kilgore et al., 1959). Although the treatments were not so extensive, the data of Pfander et al. (1957) indicate that the same thing can happen when nitrate is given intravenously to sheep. These findings can therefore probably be extended to include all animal species. The term “nitrate toxicity,” as commonly used, is actually “nitrite toxicity” and is produced following the reduction of nitrate to nitrite within the gastrointestinal tract. Both nitrate and nitrite, being highly water soluble, freely traverse the gastrointestinal wall into the bloodstream. Nitrite, but not nitrate, oxidizes the ferrous iron of the red blood pigment called hemoglobin to ferric iron, producing a modified browncolored pigment called methemoglobin which is incapable of transporting or releasing oxygen to the body tissues; animals so affected are said to be suffering from methemoglobinemia. This oxidation-reduction reaction should not be confused with normal oxygenation of hemoglobin which produces oxyhemoglobin, a temporary association of oxygen with hemoglobin. Nitrite is found only in small concentrations in living plant tissue. Although nitrite may be formed by bacterial action prior to consumption (see Section IV) , toxicity developing in this manner is probably uncommon. The toxin normally consumed, nitrate, is reduced to nitrite somewhere within the animal body. Once consumed, some of the nitrate quickly finds its way into the circulatory system. Nitrate is not reduced in blood (Winter, 1962), or in muscle tissue, with the possible exception of rat and guinea pig muscle ( Bernheim and Dixon, 1928). Although the latter workers found that enzymes in liver tissue from many species, including cattle, sheep, and swine, could reduce nitrate to nitrite, the rate must be slow, since nitrate has not produced methemoglobinemia when given intravenously to sheep (Pfander et al., 1957), or intragastrically to rats (Wanntorp and Swahn, 1953), and produced only mild methemoglobinemia when given parenterally to cattle (Winter, 1962). It has been known for some time that nitrate is a potent diuretic and

NITRATE ACCUMULATION AND POISONING

223

that the kidney is an important organ for removal of nitrate from the body. Some nitrate can, however, be recycled from the blood into the gastrointestinal tract by salivary and gastrointestinal secretions ( Bloomfield et al., 1 9 6 2 ~Kearley ; et nl., 1962). In fact, the stomach may concentrate nitrate when the level in the plasma is below 4.5 to 5.0 mM per liter. It is highly unlikely that an animal would voluntarily consume enough nitrate to be poisoned by the nitrate ion per se. The anatomy and miroflora of the digestive system of ruminants and horses, however, are especially favorable for reduction of nitrate to nitrite prior to absorption into the circulatory system. A. RUMINANTSvs. NONRUMINANTS The anatomy of the digestive tracts of ruminants and nonruminants will be briefly compared because this is essential to an understanding of why ruminants are more susceptible to nitrate toxicity than other animals. In higher animals the basic digestive tract includes the mouth, esophagus, stomach, small intestine, large intestine or colon, rectum, and anus. A blind pouch called the caecum (appendix in man) is located at the union of the small and large intestines. It is relatively small in all animal species except members of the horse family, in which it and the colon are both greatly enlarged. The stomach of ruminant animals (e.g., cattle and sheep) is called the abomasum, or “true” stomach. A very large organ, the rumen, and two smaller organs, the reticulum and omasum, are interposed in this order between the esophagus and the abomasum. Passage of materials through the mouth, esophagus, rectum, and anus is very rapid and absorption generally does not occur. Passive absorption may occur in all other organs. Active absorption occurs chiefly in the small intestine, although a few compounds, butyric acid for example, may be actively absorbed from the rumen. Nitrate is considered to be passively absorbed in all animal species. Digestive enzymes are secreted only in the mouth, stomach and small intestine. Microorganisms may be found in the stomach and small intestine, but because of abrupt differences in p H they are probably relatively inactive except under conditions of diarrhea and other digestive disturbances. The enzyme systems present in the rumen are of plant and microbial origin, those of microbial origin probably being of greater significance. Digestion in the caecum and colon is likewise due almost entirely to microorganisms. The reduction of nitrate within the gastrointestinal tract is attributed solely to microorganisms. Although the rat, a monogastric animal, is almost immune to nitrate toxicity, owing both to the nature of its diet and to the anatomy of its

224

MADISON J. WRIGHT AND KENNETH L. DAVISON

digestive tract, it has been used most extensively in studying the physiological actions of nitrate and nitrite. The chief reason for using the rat is the lower cost of the experiments compared to the cost of those with large animals, particularly ruminants. Experiments with rats may also possess a greater degree of flexibility. The physiological action of nitrate and nitrite has been studied in other monogastric animals also, including mice, dogs, pigs, and human infants. Monogastric animals, although not easily poisoned by nitrate, are highly susceptible to nitrite, and toxicities have occurred when this compound was ingested (Robinson, 1942; McIntoch et al., 1943; Wanntorp and Swahn, 1953; Riggs, 1945). On the other hand, nitrate consumed by monogastric animals undergoes little reduction to nitrite until it has reached the colon near the posterior of the gut. During the time required for nitrate to reach the colon, much of it has been absorbed and excreted by the kidneys (Whelan, 1935). Some has been reduced by the liver, and some finds its way back into the gut. In the colon much of the residual nitrate is probably reduced to nitrite and other reduction products by bacterial action and either absorbed or excreted in the feces. In short, a heavy feeding of nitrate to a monogastric animal would produce diuresis, and possibly death, but the animal would not suffer more than mild methemoglobinemia. Nitrate toxicity has been reported in horses (Bradley et aZ., 1940b; Davidson et al., 1941; Schwarte et al., 1939) and a mule (Steyn, 1951). The enlarged caecum and colon of the horse provides an opportunity for microbial reduction of nitrate to nitrite. For this reason, the horse, or mule, is more susceptible to toxicity from ingested nitrate than other monogastric animals. Unfortunately the data available are too scanty to be of much value for establishing toxic levels in the horse. Ruminants are especially susceptible to nitrate toxicity because much metabolism and absorption of nitrate and its reduction products occurs in the rumen, anterior to the stomach. Bradley et al. (1939a,b) in the United States, and Rimington and Quin (1933) in South Africa, suggested that the toxicity was due to the reduction of ingested nitrate to nitrite which produced methemoglobin. It was later shown that nitrite was an intermediate in the normal reduction of nitrate by rumen microorganisms (Sapiro et al., 1949; Lewis, 1951a). Lewis presented indirect evidence that nitrite could be further reduced to ammonia in the rumen, and this was later verified by Wang et a2. (1961) in experiments employing N15labeled nitrate. It has been speculated that the reduction of nitrate to nitrite proceeds at a faster rate than the reduction of nitrite to ammonia, and that the excess nitrite is absorbed. When a massive dose of nitrate is placed directly into the rumen of

NITRATE ACCUMULATION AND POISONING

225

cattle, it quickly disappears and large quantities of nitrate appear in the blood, as shown in Fig. 2. But when cattle are fed equally large quantities of nitrate on their hay, the levels of nitrate that appear in the blood are substantially lower (Davison et al., 1962). These lower levels are explained by the slower rate of administration that permits the animal to reduce more of the nitrate in the rumen, or otherwise rid itself of the nitrate as it is consumed. Once nitrate reaches the blood, it is apparently metabolized and excreted in a manner similar to that observed in non-

FIG.2. Blood and ruminal nitrate levels in cattle following a single intraruminal dose of 120 g. NaNO,.

ruminants (Kearley et al., 1962). Nitrite is not normally present in significant quantities in the blood of ruminant or nonruminant animals even when nitrate has been fed or placed directly in the rumen (Rath and Krantz, 1942a; Winter, 1962,;Wang et al., 1961)) and may be present for only a few minutes after nitrite is injected directly into the blood (Rath and Krantz, 1942b) or the rumen (Pfander et al., 1957). Lewis ( 1951b) suggested hydroxylamine as another reduction product of nitrate in the rumen and found that rumen organisms could reduce this to ammonia, but Winter (1962) demonstrated that hydroxylamine was present in insignificant quantities, if at all, in the blood of cattle fed nitrate. He also showed that hydroxylamine disappeared rapidly when added to blood in vitro, and that methemoglobin was produced. Measure-

226

MADISON J. WRIGHT A N D KENNETH L. DAVISON

ments of hydroxylamine in the rumen have been reported (Jamieson, 1958), but it is doubtful that this substance is stable enough in the rumen to make the transfer into the blood. Other reduction products have not been studied as possible toxins in animals consuming nitrate, and it would probably be only of academic interest to do so since death from nitrite probably occurs before toxic quantities of these other intermediates, all of which are unstable, can collect.

B. INVOLVEMENT WITH VASCULAR SYSTEM It has been known for some time that nitrite and organic nitrates are potent vasodilators when injected in or fed to monogastric animals (Hueper and Landsberg, 1940; Rath and Krantz, 1942b). Holtenius (1957) speculated that this may be one aspect of nitrate toxicity in sheep; he found that he could give temporary relief to stricken animals by injecting vasoconstrictory drugs, which is not too surprising since these drugs could be a temporary stimulant to animals suffering from many illnesses. Holtenius did not measure blood pressure. The blood pressure of cattle fed 100 to 150mg. N03-N per kilogram daily did not differ from control animals (Davison and Wright, 1963). I t appears that some animals have a very generous reserve capacity in their oxygen transport mechanism that helps them through emergencies. Cattle and sheep have been observed to survive a conversion of nearly 90 per cent of their hemoglobin to methemoglobin (Dodd and Coup, 1957; Helwig and Setchell, 1960; Davison et al., 1962), although symptoms of distress were evident. Holtenius (1957) was unable to demonstrate anoxia in the tissues of sheep experiencing as high as 65 per cent conversion. Furthermore, cattle and sheep can adapt to some extent to prolonged high levels of nitrate feeding. This is accomplished by an increase in their total red cell volume, with a resultant increase of both hemoglobin concentration and hemoglobin content ( Rimington and Quinn, 1933; Jainudeen et al., 1963). The increase may reach nearly 50 per cent. The oxygen-carrying capacity of blood from the cattle in Jainudeen’s experiments had enough residual hemoglobin even at 75 per cent conversion to methemoglobin to carry 8 per cent oxygen by volume, or about one-half of the accepted normal oxygen-carrying capacity of cow blood. OF TOXICITY C. TYPES

Toxicity is generally considered to be either acute or chronic. In acute nitrate toxicity the animal may die within a few hours after consumption of the poison, or the animal may collapse or go into a coma and then recover spontaneously, or with treatment. In contrast, in chronic

NITRATE ACCUMULATION AND POISONING

227

toxicity one or more lesions are formed and poor production is observed over a long period of time, after which the animal may or may not die. Prior to 1954, nitrate toxicity was considered to be acute only, with complete recovery if the animal survived. Since then there has been much discussion of sublethal effects, or chronic toxicity. Toxicologists express the toxicity of a substance as the dose in milligrams per kilogram body weight required to produce no deaths (LD,), 50 per cent deaths ( LDBo)or 100 per cent deaths ( LDloo).The LDo and LD,,,, may be determined by extrapolation from data obtained with doses intermediate to these points. A large number of animals are required for this sort of experimentation and it therefore becomes costly, and usually impractical, to establish with precision an LD,,) through tests with farm animals. Estimates of the nitrate LDb0in cattle range from 74 mg. N03-N per kilogram when given as a single dose by drenching (Bradley et al., 1940b) to 224 mg. N03-N per kilogram when fed over a 24-hour period as a salt solution sprinkled on chopped hay (Crawford, 1960). In contrast to these levels, the LDBofor nitrate in rats is about 800 mg. NO3N per kilogram when injected directly into the stomach in a single dose. Crawford may have overestimated the LDSo, for he experienced considerable difficulty with feed consumption. Cattle and sheep will at first refuse experimental diets to which nitrate salts have been added, but will gradually become accustomed to them and after 2 to 4 weeks will eat as much as control animals. I t now appears that the LDSo for ruminants lies somewhere between 160 and 224 mg. NO3-N per kilogram, when nitrate is fed with the roughage portion of the diet. 1. Acute Toxicity a. Symptom. The first sign to appear is a grayish to brownish discoloration of white areas on the skin and the nonpigmented mucous membranes of the mouth, nose, eyes, and vulva. These lesions result from the chocolate-brown discoloration of the blood. As the syndrome progresses, a staggering gait, rapid pulse, frequent urination, and labored breathing develop, followed by collapse and coma, with or without convulsions, and death. These symptoms may occur very rapidly, but some animals may collapse and then recover spontaneously and completely. Pregnant animals so affected may abort a few days later. Level of methemoglobin in the blood and, because other compounds may produce methemoglobin, the level of nitrate in the feed or water are useful in diagnosing nitrate toxicity. Because time will not permit these analyses when animals are sick, the treatment described in Section V, C, 1, c. should be given without delay as soon as brown blood is observed.

228

MADISON J. WRIGHT AND KENNETH L. DAVISON

The brownish discoloration of the tissues and blood caused by the accumulation of methemoglobin has frequently been confused with cyanosis. Cyanosis occurs in carbon dioxide poisoning, when the blood turns blue because of a lack of oxyhemoglobin and an abundance of nonoxygenated hemoglobin. The term cyanosis does not refer to the effects of cyanide poisoning (hydrocyanic acid or prussic acid poisoning), in which the blood turns cherry red because of an abundance of oxyhemoglobin, and the skin appears flushed. Cyanide blocks the utilization of oxygen by the cytochrome enzyme systems during oxidation of foodstuffs within the body cells whereas nitrite blocks oxygen transport to the tissues by the formation of methemoglobin. Strangely, methemoglobin has a strong affinity for cyanide, and in cases of cyanide poisoning, sodium nitrite or sodium thiosulfate are injected to produce low levels of methemoglobin. It picks up the cyanide, thereby unblocking the cytochrome oxidation systems. In the late 1930’s “oat hay poisoning” was in some instances misdiagnosed as cyanide poisoning and some affected animals were treated by giving them sodium nitrite; of course death resulted immediately. Some investigators feel that in some areas nitrate poisoning is still misdiagnosed as cyanide poisoning and the wrong treatment therefore given. Examination of the blood reveals the brown color characteristic of nitrate poisoning and helps prevent mistreatment. Animals dying for any reason should be autopsied immediately because postmortem decomposition of many tissues occurs rapidly and may mask lesions that might otherwise be observed ( McEntee, 1963). The kidneys, liver, and spleen may soften in less than 12 hours after death, especially in large or fat animals which cool slowly. The digestive tract becomes stretched and distended with gas and the internal lining sloughs in 1 to 3 hours. All these lesions are indicative of normal decay and point out the need for prompt autopsy. The most characteristic lesion observed in animals that have died of nitrate poisoning is the chocolate brown discoloration of the blood and tissues. Other lesions, which may or may not be present, are evidence of a massive hemorrhage in heart muscle; edema of the lungs; and foamy material in the trachea. All are indicative of respiratory distress and exertion prior to death. The odor of gaseous nitrogen oxides has been detected in freshly killed animals as the rumen and intestines were opened (Davidson et al., 1941). Nitrate is moderately stable in blood, and may be detected if examination is delayed too long to permit observation of the symptoms previously described. Other lesions have been described in the literature (Simon et al., 1958, 1959b), but these apparently are not specific for nitrate toxicity because many of them have

NITRATE ACCUMULATION AND POISONING

229

been observed in other situations not caused by nitrate, or have been absent from cases of known nitrate toxicity (Setchell and Williams, 1962; McEntee, 1963). b. Factors aflecting severity. The rate and quantity of consumption, type of forage, energy level or adequacy of the diet; adaptation, health and species of the animal; and pregnancy all affect the susceptibility of an animal to nitrate poisoning. It has already been pointed out that smaller doses of nitrate are required to cause death when given suddenly, as by drenching, than when given more slowly by feeding in or on hay. Hungry animals eat more rapidly and may be poisoned when grazing forages that have been grazed safely by other animals (Kretschmer, 1958). Sheep and cattle fed inadequate diets have been more susceptible to nitrate than those fed adequate diets (Holtenius, 1957; Crawford and Kennedy, 1960). Studies with rumen microorganisms in oitro have shown that those obtained from animals fed alfalfa hay could reduce both nitrate and nitrite at faster rates than those from sheep fed grass hays that presumably were inadequate in both protein and carbohydrate (Sapiro et al., 1949; Pfander et al., 1957). The addition of carbohydrate in the form of glucose, lactate, and similar compounds, or of dried grass which would supply nitrogen and minerals as well as carbohydrate, has also been shown to increase the rate of reduction of both nitrate and nitrite by rumen organisms in vitro (Lewis, 1951b; Barnett and Bowman, 1957), while Sapiro et al. (1949) observed that the feeding of glucose to sheep would enable them to withstand higher levels of nitrate, presumably because of the increased rate of nitrite reduction in the rumen. The preceding observations can probably be attributed to combinations of the following factors in the rumen: a favorable environment for certain species of microorganisms; an increase in total number of organisms; better nutrition of the organisms; and provision of energy needed for nitrate and nitrite reduction. Better nutrition of the host animal with a consequent increase in resistance to the stresses imposed by methemoglobinemia may also be important. Cattle and sheep may adapt, within limits, to high levels of nitrate intake (Rimington and Quin, 1933; Davison et al., 1962; Sokolowski et al., 1960). This probably involves an increased ability of the rumen organisms to reduce both nitrate and nitrite more rapidly, as well as an increase in the number of red cells in the blood. Some nitrite may be produced in wet forages (Olson and Moxon, 1942; Davidson et al., 1941), and since most analytical procedures for nitrate include the sum of nitrate and nitrite, these forages would be more toxic than their nitrate content would indicate. The LDao for nitrite

230

MADISON J. WRIGHT AND KENNETH L. DAVISON

in cattle is about 20 mg. N02-N per kilogram when given as a drench, or 45 to 50 mg. N02-N per kilogram when given more slowly to simulate rate of eating (Stormorken, 1953). The LD,, for nitrite given by stomach tube to rats and swine is 15 to 18 mg. N02-N per kilogram, a value which indicates that these species may be more susceptible to nitrite than the ruminant. Hemoglobin of the newborn human and of the bovine fetus is more easily oxidized to methemoglobin by nitrite than that of the respective adult (Betke, 1958). Both pregnancy and disease have been shown to facilitate the oxidation of hemoglobin to methemoglobin in rats and in humans ( Metcalf, 1962a,b). There is some evidence that in young animals methemoglobin may be reduced more rapidly to functional hemoglobin than in older animals (Spicer and Reynolds, 1949), and this might increase their chances of recovering from methemoglobinemia. Unfortunately, the exact application of these studies to the usual poisoning situation in ruminants is unknown. c. Treatment. Acutely affected animals should be treated intravenously with a solution of methylene blue. A 1 to 4 per cent aqueous solution given at the rate of 2 g. methylene blue per 500 pounds body weight is adequate (Bradley et al., 1940a) although methylene blue in 5 per cent glucose or 1.8 per cent sodium sulfate has been suggested (Case, 1957a). Methylene blue is thought to act as an electron acceptor for the methemoglobin reductase enzyme normally present in blood, thereby speeding the reconversion of methemoglobin to hemoglobin. Precautionary measures can be taken if excess nitrate is thought to have accumulated in the forage. Forages over 0.34 to 0.45 per cent NOS-N should be considered potentially toxic and should be mixed with safer feeds prior to use. The diet should be adequate in carbohydrate and protein, and the animals should not be permitted to get overly hungry. 2. Chronic Toxicity In addition to losses from acute nitrate toxicity that followed a severe drought in 1954, specialists at Missouri (Brady et al., 1955; Case, 1957a) observed losses in beef and milk production that they ascribed to sublethal effects of nitrate. This marked the beginning of nearly all studies on chronic toxicity, which has since received much publicity in popular farm publications in the United States. Lowered growth and milk production, abortions, and impaired vitamin A and iodine nutrition have been included among the sublethal effects. I t is extremely difficult, however, to find clear-cut experimental confirmation of these effects. a. Effects on iodine and vitamin A nutrition. The thyroid gland

NITRATE ACCUMULATION AND POISONING

231

produces a hormone, thyroxine, that governs the general rate of metabolism throughout the body. A reduced production of thyroxine slows the rate of enzyme-catalyzed reactions, retards growth of muscle and bone, and lowers production of milk and wool. Iodine is a constituent of the thyroxine molecule. The presence of some inorganic anions in the blood may decrease the concentration of iodide in the thyroid gland, indirectly lowering the rate of thyroxine synthesis. These anions are thought to interfere with the collection, and perhaps also the retention, of iodide by the thyroid, and their effect may be overcome by increasing the iodine concentration in the blood. Wyngaarden et al. (1952) studied the ability of twenty different anions, including nitrate and nitrite, to displace iodide from, or to block iodide uptake by, the thyroid of rats. Eight of these compounds, namely perchlorate, chlorate, hypochlorite, periodate, iodate, biiodate, nitrate, and thiocyanate, inhibited collection and/or retention of iodide by the thyroid. Of these, perchlorate was the most potent inhibitor and nitrate was the least potent. Rats injected intraperitoneally with 1.4 mg. NOS-N developed slight thyroid hyperplasia ( increased number of thyroid cells ) and showed a slight reduction in thyroid iodine concentration. Although Wyngaarden was not concerned with nitrate toxicity, his data have been widely used as evidence that nitrate interferes with thyroid function. The poor production of animals eating sublethal quantities of nitrate has been explained in this manner. Since 1952, a number of abstracts (Welsch et al., 1961, 1962; Yadav et al., 1962) and papers (Bloomfield et al., 1961, 1962a) have been published in which dietary nitrate has been shown to reduce iodine uptake in the thyroid of rats, the implication being that these data applied also to ruminant animals. But thyroid uptake of and concentration of plasma protein-bound II3* was increased in sheep fed a diet containing 0.21 per cent NO,-N (Bloomfield et al., 1961, 1962a), contrary to expectations based on the data obtained with rats. Bloomfield et al. (1962b) later observed that an increase in plasma protein-bound I"' may not be a valid measure of thyroid activity. There are no experimental data at the present time indicating a decreased activity of the thyroid gland of ruminants fed nitrate. Further work on nitrate-thyroid interrelationships in the ruminant appears to be needed. Many vitamin A deficiencies occurring in feedlot cattle that depended on carotene in forage to meet their needs for vitamin A have been blamed on the nitrate contained in the forages being fed. Carotene, a plant pigment that possesses no vitamin A activity, is normally converted into vitamin A in the intestinal walls of animals. On the basis of the previous speculation that dietary nitrate lowered production of thyroxine by the

232

MADISON J, WRIGHT A N D KENNETH L. DAVISON

thyroid, and on the observation that an active thyroid increases the intestinal absorption and/or conversion of carotene to vitamin A in rats (Johnson and Bauman, 1947), it was speculated that the nitrate could act through the thyroid to produce vitamin A deficiency in the cattle. The following evidence suggests that the thyroid plays an important role in vitamin A and carotene metabolism of ruminants. When a thyroid hormone, triiodothyronine, was given to cattle the plasma carotene was increased in those animals receiving diets supplemented with carotene, and the plasma vitamin A and liver storage of vitamin A were increased in those given either carotene or vitamin A as a supplement (Jordan et al., 1963). When carotene was added to homogenized tissue from the anterior of the small intestine of calves with abnormally active thyroids, the conversion to vitamin A occurred at a normal rate; but tissue from calves with subnormal thyroid activity converted carotene to vitamin A at a subnormal rate (Reddy and Thomas, 1962). Liver storage of vitamin A is a better indication of the vitamin A status of an animal than plasma vitamin A and carotene, so the effect of nitrate on vitamin A storage in this organ has been studied quite extensively. Carotene is not stored in significant quantities in the liver. Depletion of liver vitamin A storage has been reported in sheep fed nitrate (Goodrich et al., 1962) or nitrite (Holst et al., 1961). But dietary nitrate did not deplete liver stores of vitamin A in numerous other experiments with cattle (Hale et aZ., 1961; Jordan et al., 1961, 1963; Weichenthal et al., 1961; Smith et al., 1962; Davison et al., 1962) and sheep (Cline et al., 1962; Sokolowski et al., 1961; Smith et al., 1962). The weight of evidence strongly suggests no interference with liver storage of vitamin A in the ruminant. In experiments at Cornell (Davison et al., 1962; Davison and Wright, 1963) no thyroid abnormalities (thyroid weight or histological characteristics) have been observed in cattle fed nitrate, nor has there been any indication of impaired vitamin A nutrition. Since some of these animals were fed a barely sublethal daily dose of nitrate for many months, it appears that a ruminant would probably die of toxicity before the level of nitrate or nitrite in any part of its body could reach a concentration great enough to interfere with iodine or vitamin A nutrition. Some studies with swine have suggested that dietary nitrate interferes with vitamin A and carotene metabolism (Garner et al., 1958; Koch et al., 1963), while other studies have not (Tollett et al., 1960). The addition of nitrate salts to swine rations will lower voluntary feed consumption, which automatically lowers vitamin A consumption if the vitamin is fed as a fixed proportion of the ration. Although the above reports indicated a reduced feed consumption, corrections for the resulting lower level of

NITRATE ACCUMULATION AND POISONING

233

vitamin A intake were not indicated. Therefore, the relationship of nitrate to vitamin A metabolism in swine is still unclear. An impaired liver storage of vitamin A, and a reduced conversion of carotene to vitamin A has been indicated in studies with rats fed nitrate or nitrite (Garner et al., 1958; O’Dell et al., 1960; Yadav et al., 1962). Vitamin A and carotene were shown to be rapidly destroyed by nitrite in acid solutions at a pH similar to that of the stomach (Pugh et al., 1962; Olson et al., 1963). Nitrite feeding did not alter liver storage of vitamin A when the vitamin was injected under the skin (Emerick and Olson, 1962). Therefore the oxidation in the stomach of vitamin A or carotene by nitrite may partially explain the impaired vitamin A metabolism in the rat. Olson et al. (1963) observed that carotene was fairly stable when incubated in vitro with abomasal juice ( p H 3 ) from sheep, but when nitrite was added the carotene disappeared in about 15 minutes. There are no data on concentrations of nitrite in the abomasum of the living animal with which Olson’s observations can be compared. The concentration of 110 p.p.m. N02-N used in his study is much greater than the approximately 20 p.p.m. N02-N reported as a maximum in the rumen (Holtenius, 1957). It is doubtful that much nitrite passes from the rumen into the abomasum, since nitrite is rapidly absorbed from the rumen, and rapidly disappears upon entering the blood. Destruction of carotene during reduction of nitrate in the rumen has also been investigated. Studies with rumen microorganisms in vitro have shown that these organisms can rapidly destroy carotene, but the addition of nitrate in physiological quantities has not hastened this process (Olson et al., 1963; Davison and Seo, 1963). Since gaseous oxides of nitrogen are produced during fermentation of silage, Emerick and Lievan (1963) studied the effects of nitrogen dioxide on stability of carotene. Carotene was rapidly destroyed by relatively pure nitrogen dioxide gas, and the resultant product, while nontoxic, was shown to have no vitamin A activity in the rat. The rate of destruction of carotene during an artificial fermentation of corn silage was more rapid when nitrate was added (Olson et aZ., 1963). If this last observation is confirmed under natural ensiling conditions, it would be of great economic importance to the animal industry. Corn Belt cattle have been wintered on silage with the assumption that its carotene content was adequate to meet their needs. b. Efects on growth. The earliest experiments with prolonged sublethal dosages of nitrate were conducted by Bradley et al. (1940b), who did not observe a loss in weight or other abnormalities in cattle over a two-month period. They concluded that chronic toxicity did not exist.

234

MADISON J. WRIGHT AND KENNETH L. DAVISON

The South African workers (Clark and Quin, 1951) fed to mature sheep diets containing 4 per cent sodium and ammonium nitrate with very favorable results as a part of their search for nitrogen sources that could serve as supplements to diets of “poor quality” grass hay and molasses. Maintenance of body weight and feed consumption were the only measurements reported. Chronic toxicity was not expected nor indicated. Feed consumption declined greatly when diets containing 6 and 8 per cent ammonium nitrate and sodium nitrate were offered, so these diets were not continued. The results of some recent experiments are more confusing. Rate of gain has been reduced in some cases but not in others. Additions of 1 per cent potassium or sodium nitrate to grain reduced feed consumption and rate of gain of fattening cattle (Hale et aZ.,1961; Weichenthal et d., 1961), but in other experiments additions of 1 to 2 per cent potassium nitrate (dry matter basis) to corn silage or hay did not reduce gains of cattle, nor did additions of 4 per cent potassium nitrate to corn silage affect gains of sheep (Smith et al., 1962). If feed consumption was not reduced in Smith‘s experiments (data were not given), this may explain the apparent contradiction. In other experiments, in which fattening lambs were fed up to 4 per cent potassium nitrate in mixed rations that were high in concentrate, rates of gain were either not affected (Cline et al., 1962) or were reduced (Sokolowski et aZ., 1961). In the experiments of the latter, increasing the sulfur content of the diet nullified the effect of nitrate on rate of gain, but the significance of this observation is not clear. Nonpregnant and pregnant dairy heifers have been fed oat hay containing 0.28 per cent N03-N (Crawford, 1960), or u p to 150 mg. NOB-N per kilogram body weight daily as the sodium salt on chopped hay (Davison et al., 1963) without reducing body weight gains even though some heifers in the latter study died of acute toxicity. Studies of nitrate effects on the growth of swine are not numerous. A reduction in feed consumption and rate of gain has been observed only when the level of nitrate nitrogen exceeded 0.34 per cent of the diet (Tollett et al., 1960; Koch et al., 1963). In nitrate feeding experiments with cattle, sheep, or swine where feed consumption was reported, rate of gain has not been reduced unless there was also a reduction in feed consumption. Animals seem to accept a higher percentage of nitrate in forage than in grain mixtures. Since grain ordinarily contains less nitrate than the vegetative portion of the plant, studies in which nitrate was added to the roughage resemble more closely agricultural practice with ruminant animals. c. E ~ e c t s on reproduction. There is less evidence to indicate sublethal effects of nitrate on reproduction than there is on growth. There

NITRATE ACCUMULATION AND POISONING

235

is no doubt that abortions in the case histories cited by Bradley et al. (1940b) or Garner (1958) were caused by excessive nitrate, but the level of nitrate presumably consumed was far above the 0.2 to 0.4 per cent N03-N range that has been described in the popular literature as dangerous to pregnant animals. Abortions have been produced experimentally by placing 14 g. NOB-N, as potassium nitrate, directly into the rumen of pregnant heifers (Simon et al., 1959a; Asbury, 1963). Asbury noted that the effect of the nitrate was more severe in the presence of other stresses, such as a sudden feed change or addition of a large quantity of antibiotics. These reviewers have fed nitrate on chopped hay to 40 heifers at daily levels of 0, 100, and 150 mg. NOs-N per kilogram ( LD50of 160 to 224 mg./kg. ) beginning before breeding and continuing through gestation. The estrous cycle remained unchanged, but decreased conception rate was indicated in those fed the higher level of nitrate. One abortion occurred in those fed the lower level of nitrate, and two abortions and two deaths occurred in those fed the higher level. An organism, Pseudomonas aeruginosa, was isolated from one fetus of the latter group. Although this organism has been observed in other aborted fetuses, its exact significance is unclear ( McEntee, 1963). The gestation periods, the placentas, and the birth weight and appearance of the resulting calves were normal for all remaining heifers. The reproductive tracts obtained at the end of the experiment were normal. Similar studies are being continued with both cattle and sheep. The effect of nitrate on reproduction has been studied to a lesser extent in sheep than in cattle. Abortions were not observed in sheep fed oat hay containing 0.11 per cent N03-N (Eppson et al., 1960), nor in sheep drenched with nitrate at levels equal to 0.07, 0.14, and 0.28 per cent Nos-N in the diet (Setchell and Williams, 1962). Four of 5 ewes drenched with the higher level of nitrate were killed after 41 to 74 days of drenching. At the present time, the level of nitrate that will cause abortion or other reproductive difficulties in the ruminant cannot be distinguished from the level that will cause acute toxicity. Nitrate does not present a problem in reproduction of swine. Abortions have been reported in swine that were grazing high-nitrate rape (Case, 1957b), or were drinking high-nitrate water (Case, 1963), but experimentation has discounted the role of nitrate. Swine have been fed as much as 0.28 per cent NOs-N without causing abortion (Garner et al., 1958). Tollett et al. (1960) could find no effect of nitrate on number of corpora lutea, percentage of implantation of embryos, or on the weight of ovaries, embryos, placentas, and thyroids of swine.

236

MADISON J. WRIGHT AND KJ3NNETH L. DAVISON

d. Eflects on milk production. The first experiments on the effect of nitrate on milk production were conducted at Missouri (Muhrer et al., 1956; Stewart and Merilan, 1958). Varying levels of nitrate were administered to dairy cows in their grain, by stomach tube, or by capsule. It was concluded that milk production was decreased, but the minimum level of nitrate required was not specified, nor was it evident from the data presented. Two of the cows died, one during and one after the experimental period. In two larger-scale experiments with cows of higher milk production, no such depressive effect was observed. In one experiment, Crawford (1960) fed roughage containing 0.53 per cent N03-N to ? cows without lowering milk production; and in the second experiment, Davison et al. (1963) fed sodium nitrate on chopped hay at daily levels of 100 and 150 mg. N03-N per kilogram without lowering milk production during the first 30 days post-calving. Many of the heifers in the latter case were fed nitrate at near-lethal levels for over 350 days prior to calving. They were all sacrificed after 30 days of lactation for thorough postmortem examination, but no nitrate-associated lesions were found. It is realized that dairy cattle will maintain lactation at a high rate during the first few days after calving regardless of diet by sacrificing body weight if necessary, but abnormal weight loss was not observed. From these observations, it appears that nitrate is unlikely to lower milk production unless it is fed at a level that keeps the cows constantly on the threshold of collapse, or at a level that lowers feed consumption,

HAZARDS TO MAN 1. Silo Gases Nitric oxide (NO) and nitrogen dioxide ( NO2) gases may be evolved D. Porn-

by enzymatic action of plant and bacterial origin on ensiled forages, even though the nitrate content of the material ensiled was low enough to be fed with safety. Nitrogen dioxide is extremely toxic to man and animals and is a menace in both agricultural and industrial activities. A summary of the toxicity of various levels of nitrogen dioxide, as compiled by the U. S. Public Health Service (1962b) is given in Table V. Upon contact with water, nitrogen dioxide forms nitric acid which reacts with the lung tissues. High concentrations of these gases cause immediate throat irritation and coughing, but lower concentrations, even though dangerous, may be breathed with no immediate discomfort. Hours later the patient may develop edema of the lungs, coughing, and shortness of breath. The thickening of the lung tissues and the 3uids produced may eventually stop oxygen transfer into the blood, causing death by suffocation. Damaged lung tissue may also be progressively replaced

” R A T E ACCUMULATION AND POISONING

237

with fibrous scar tissue, causing death days after exposure (Lowry and Schuman, 1956). This syndrome is called “silo-fillers disease” when the gases responsible are generated in a silo. There is no specific treatment for individuals suffering from contact with nitrogen dioxide aside from placing them in an oxygen tent to assist their breathing while the lung tissues heal. Adrenocorticosteroids ( adrenal hormones) were suggested by Lowry and Schuman (1956), but the effectiveness of these compounds has not been studied thoroughly. It is extremely important to prevent secondary infections from occurring during the healing process. TABLE V Toxicity of Nitrogen Dioxide Gas to Mana Concentration (w . m .1

Physiological response ~~

5 62 75-1 00 100 100-150 200-700

700 0

Odor threshold, maximum allowable concentration for 8 hours Least concentration causing immediate throat irritation Visual threshold Least concentration causing coughing 3/2-1 hour exposure dangerous Rapidly fatal for short exposure Fatal in 30 minutes or less

U. S. Public Health Service (1962b).

Since its recognition as a potential hazard in agriculture about 1949, nitrogen dioxide has been observed in silos in many states. Peterson et al. (1958) observed concentrations as high as 100,000 p.p.m. nitrogen dioxide in a sump that collected the drainage from a University of Wisconsin silo. Scaletti et al. (1960) reported that nitrogen dioxide was present in concentrations of 4 p.p.m. or greater in 42 per cent of 332 silos surveyed in Minnesota, clearly indicating that toxic concentrations of this gas are not uncommon on farms. As there is no specific treatment for “silo-fillers disease” at the present time, prevention cannot be overemphasized. Silo gases are heavier than air and will hover near the surface of the silage, or flow down drain pipes or silo chutes. It is advised that all silos be ventilated with a blower or fan before entry during a period of 7 to 10 days after filling. During this period air from the silo chute should not be allowed to discharge into tightly closed buildings, nor should animals be housed near the base of the chute. 2. Water Cyanosis of unknown origin, sometimes called idiopathic cyanosis, has been observed in human infants (“blue babies”) for a number of years. The absence of the enzyme methemoglobin reductase, or its

238

MADISON J. WRIGHT AND KENNETH L. DAVISON

failure to function properly if present, is the cause of one form of the cyanosis. Another form of cyanosis was discovered by Comly (1945) in infants that were fed formulas containing nitrate-contaminated well water, Comly reported that the wells were dilapidated and the water was badly contaminated with coliform organisms presumably capable of reducing nitrate to nitrite, but only traces of nitrite were found. Infant cyanosis caused by nitrate in water has since been reported in a variety of countries and states, Cornblath and Hartmann (1948) noted that while infants were suffering from methemoglobinemia, adults drinking the same water remained unaffected. Noting also that cyanotic infants sometimes had diarrhea and that most nitrate-contaminated waters also contained coliforms, they investigated the importance of the organisms to the toxiciy. Traces of nitrite were sometimes detected in suspected water, but not in concentrations sufficient to explain the troubles. Several infants were fed nitrate in their formulas without becoming cyanotic, but if nitrate-reducing organisms were added some of the infants did develop cyanosis. Further investigations revealed that these organisms grow best at a p H between 5 and 7 and that in the newborn, and in infants previously affected with methemoglobinemia or diarrhea, the pH of the stomach was above 4 ( abnormally high), Nitrate-reducing organisms were isolated from the saliva and gastric contents of infants suffering with methemoglobinemia. These authors concluded that the affected infants acquired nitratereducing organisms from improperly sterilized water, and lack of acidity in their stomachs permitted these organisms to grow while the higher acidity in the stomachs of older infants and adults inhibited bacterial growth. Generally confirmatory is the absence of reports of methemoglobinemia in human adults who drink nitrate-contaminated water even though nitrate contents as high as 630 mg. N03-N per liter have been reported (Campbell et al., 1954). In his review of the literature relating to infant methemoglobinemia caused by water contaminated with nitrate, Walton (1951) found that no cases had been reported where water contained less than 10 p.p.m. NOs-N, but the cases became progressively more frequent as the level of nitrate increased. The U. S. Public Health Service (1962a) has relied upon the Walton survey in recommending that drinking water contain no more than 10 mg. N03-N or 45 mg. NOa- per liter. The major source of nitrate contamination appears to be animal and human wastes. Nitrate concentrations are higher in dug and shallow wells than in drilled wells, and are higher in wells with broken casings and poor covers (Borts, 1949). The nitrate concentration in wells has been shown to vary markedly from day to day.

239

NITRATE ACCUMULATION AND POISONING

3. Food Products Many leafy vegetables are known to accumulate large quantities of nitrate, and nitrate contents have been reported by several research workers (Wilson, 1949; Gilbert et al., 1946; Kilgore et al., 1963), but toxicity to humans from this source has not been reported. Many prepared meat products such as corned beef, weiners, bologna, and sausages have both nitrate and nitrite added to preserve color and retard spoilage. Nitrate is reduced by bacterial action to nitrite and the nitrite reacts with water and muscle hemoglobin (myoglobin) to produce a more permanent pink substance called nitric oxide myoglobin. Nitrite in excess of that required to produce nitric oxide myoglobin will retard TABLE VI Nitrate Residues in Milk and Meat of Cattle Fed Sodium Nitrate on Their Haya Level of NO3-N fed ( mg./kg. ) 0

100 150 a

Milk N03-N ( Pg./ml. ) 1.1 2.0 4.5

Meat NO,-N ( Fg./g. ) 0.2 2.3 2.5

Davison and Wright, 1963.

microbial growth, but this excess may not exceed a maximum level of 200 p.p.m. nitrite ion (60 p.p.m. NO,-N) in the finished product, a level established by the Meat Inspection Division of the U. S. Dept. Agr. (1960). All meat processing plants are subject to this regulation even though the plant itself may not be Federally inspected. Toxicities from consuming meat products high in nitrite are rare, but have occurred in spite of governmental testing and regulation (Orgerson d al., 1957). Donahoe ( 1949) suggested the possibility of infant methemoglobinemia from nitrate present in milk, but Nebraska health officials (Nebraska Dept. Health, 1950) were unable to detect nitrate in a qualitative test of the milk of cows fed 114 p.p.m. N03-N in their water supply. Because the possibility of nitrate accumulation in the tissues of animals fed this material was never investigated, studies were conducted by the authors to measure quantitatively the nitrate content of both milk and meat from animals that were fed 0, 100, and 150 mg. NO3-N per kilogram per day for periods of from 6 to 12 months before slaughter. Although the nitrate content of both milk and meat was higher for animals fed this material than for controls (Table V I ) , the levels are far below currently accepted standards for meat or water; there are no specific regulations for maximal level of nitrate permissible in milk. Nitrate contamina-

240

MADISON J. WRIGHT AND KENNETH L. DAVISON

tion of food by this means is not expected to be high enough to be a hazard to human health. In some European countries nitrate is used in the manufacture of cheese to stop microbial growth at a predetermined level of fermentation by liberation of nitrite, which also serves as a preservative. Nitrate is not used in cheese processing in the United States, and toxicities of this nature, if they occur, were not reviewed. VI. Conclusions

Prediction of nitrate accumulation in the field remains less than certain despite several decades of field experience. An association of high nitrate accumulation with certain species, nutrient excesses or deficiencies, and meteorological conditions has been demonstrated, but exceptions to these generalized relationships can almost always be found. The impression one gains from the literature is that a high degree of confidence in predicting accumulation must await a fuller understanding of the cellular processes responsible. Sweeping changes in agronomic practices and recommendations do not appear to be needed. From a practical viewpoint, there appears to be enough information readily available to guide extension specialists, veterinarians, and farmers safely through the hazards presented by acute nitrate toxicity. Recent experiment station publications from California (Tucker et al., 196l), Iowa (Hanway et al., 1963), Missouri (Garner, 1958), New York (Crawford and Kennedy, 1960), and Wisconsin (Sund et al., 1963a, 1963b) provide both a general review of the several aspects of the problem and lists of precautionary and remedial measures. Public health authorities are increasingly aware of the nitrate content of water supplies. In regard to silo gases, however, many and perhaps most farmers are dangerously ignorant of the risks involved, and not all physicians serving rural areas know that these poisons are encountered in agriculture. The fact that dairy cattle fed barely sublethal levels of nitrate did not pass the substance into the milk, or accumulate it in their tissues, in more than trace amounts, is reassuring to the human consumer. The least satisfactory aspect of the entire problem is the matter of chronic, or sublethal, effects. Research workers disagree as to whether there are in fact sublethal effects of nitrate. Such disagreement is hardly surprising, for the effects under study are not clearly defined and may often be attributed to agents other than nitrate. It is probably significant that the sublethal effects of nitrate were first given serious attention at a time when both cropping practices and

“ R A T E ACCUMULATION AND POISONING

241

husbandry programs were feeling the full impact of a technology that urged the performance of plants and animals to unprecedented levels. Subtle physiological changes induced by the new practices have surely escaped detection in many cases, but their undesirable manifestations may require immediate attention. In the cases discussed here, an increase in nitrate content was a conspicuous change that was rather consistently associated with a group of maladies. Unfortunately, the tentative indictment of nitrate that was made pending the demonstration of a causal relationship has been widely publicized, and as disseminated it has in many cases gained an air of certainty unwarranted by present evidence. An example of the troubles that have been rather freely attributed to nitrate is abortion in cattle. Fully 70 per cent, and perhaps as much as 80 per cent, of the abortions that now occur in cattle herds at U. S . agricultural colleges cannot be ascribed to any known agent despite careful supervision and exhaustive testing. It is assumed that an even higher proportion of cases on farms are not readily and positively diagnosable. The experimental data suggest that abortion is unlikely to occur from nitrate alone until the mother is near death, and in a high percentage of cases does not occur even then. At present relatively little is known of the chemical changes that attend the accumulation of nitrate in plants used as feeds. If animals fed these plants are adversely affected but their symptoms are not reliably reproduced by administration of nitrate, further experimentation under closely controlled conditions is likely to be required for an unequivocal implication of nitrate. The difficulty of this experimentation is increased if the offending diet is heterogeneous. REFERENCES Ackerson, C. W. 1963. Feed Age 13, 32-34. Alexander, M. 1961. “Introduction to Soil Microbiology.” Wiley, New York. ap Griffith, G. 1960. J. Sci. Food Agr. 11, 626-629. ap Griffith, G., and Johnston, T. D. 1960. J. Sci. Food Agr. 11, 623-626. Asbury, A. C. 1963. Personal communication. Barker, A. V. 1962. Ph.D. Thesis, Comell Univ., Ithaca, New York. Bamett, A. J. G., and Bowman, I. B. R. 1957. J . Sci. Food Agr. 8, 243-248. Bathurst, N. O., and Mitchell, K. J. 1958. New Zeahnd J. Agr. Res. 1, 540-552. Becker, M., and Oslage, W. 1955. Landwirtsch. Forsch. 8, 100-110. Berg, R. T., and McElroy, L. W. 1953. Can. J . Agr. Sci. 33, 354-358. Bemheim, F., and Dixon, M. 1928. Biochem. J. 22, 125-134. Berthelot, M. 1884. Compt. Rend. 99, 1506-1511. Betke, K. 1958. Folia Haematol. 75, 243-251. Bloomfield, R. A,, Welsch, C. W., Gamer, G. B., and Muhrer, M. E. 1961. Science 134, 1690. Bloomfield, R. A,, Welsch, C., Gamer, G. B., and Muhrer, M. E. 1962a. J . Animal Sci. 21, 988 (abstr.).

242

MADISON J. WRIGHT AND KENNETH L. DAVISON

Bloomfield, R. A., Welsch, C. W., Gamer, G. B., and Muhrer, M. E. 196213. Proc. SOC. Exptl. Biol. Med. 111, 288-290. Bloomfield, R. A,, Hersey, J. R., Welsch, C. W., Gamer, G. B., and Muhrer, M. E. 1962c. J . Animal Sci. 21, 1019 (abstr.). Bollard, E. G . 1957. Australian J. Biol. Sci. 10, 292-301. Bollard, E. G . 1960. Ann. Rev. Plant Physiol. 11, 141-166. Borts, I. H. 1949. Am. J. Public Health 39, 974-978. Bould, G., and Hewitt, E. J. 1963. In “Plant Physiology” (F. C. Steward, ed.), Vol. 3. Academic Press, New York. Bradley, W. B., Beath, 0. A., and Eppson, H. F. 1939a. Science 8, 365. Bradley, W. B., Eppson, H. F., and Beath, 0. A. 1939b. J. Am. Vet. Med. Assoc. 94, 541-542. Bradley, W. B., Eppson, H. F., and Beath, 0. A. 1940a. J. Am. Vet. Med. Assoc. 96, 41-42. Bradley, W. B., Eppson, H. F., and Beath, 0. A. 1940b. Wyoming Agr. Expt. Sta. Bull. 241. Brady, D. E., Comfort, J. E., Lasley, J. F., and Pfander, W. H. 1955. Missouri, Uniu., Agr. Expt. Stu., Sta. Bull. 652, 19-20. Breniman, G. W., Neumann, A. L., Smith, G. S., and Jordan, H. A. 1961a. J. Animal Sci. 20, 684 (abstr. ). Breniman, G. W., Neumann, A. L., Smith, G. S., and Zimmerman, J. E. 1961b. 1. Animal Sci. 20, 926 (abstr.). Burstrom, H. 1946. Kgl. Lantbrukshogskol. Ann. 13, 1-86. Burt, R. F. 1963. M. S. Thesis, Cornell Univ., Ithaca, New York. Campbell, E. G . 1924. Botun. Guz. 78, 103-116. Campbell, J. B., Davis, A. N., and Myhr, P. J. 1954. Can. J. Comp. Med. Vet. Sci. 18, 93-101. Candela, M. I., Fisher, E. G., and Hewitt, E. J. 1957. Plant Physiol. 32, 280-288. Case, A. A. 1957a. J. Am. Vet. Med. Assoc. 130, 323-329. Case, A. A. 195713. Sheep Breeder Sheepman 77(7), 12-13. Case, A. A. 1963. Nutl. Hog Farmer 8 ( 3 ) , 29-35. Clark, R., and Quin, J. I. 1951. Onderstepoort J. V e t . Res. 25, 93-103. Cline, T. R., Hatfield, E. E., and Carrigus, U. S. 1962. J. Animal Sci. 21, 991 (abstr.), Comly, H. H. 1945. J. Am. Med. Assoc. 129, 112-116. Cook, R. L. 1930. J. Am. SOC. Agron. 22, 393-408. Comblath, M., and Hartmann, A. F. 1948. J. Pediat. 33, 421-425. Crawford, R. F. 1960. Ph.D. Thesis, Cornell Univ., Ithaca, New York. Crawford, R. F., and Kennedy, W. K. 1960. Cornell Uniu. Agr. Expt. Stu. Misc. Bull. 37. Crawford, R. F., Kennedy, W. K., and Johnson, W. C. 1961. Agron. J. 53, 159162. Davidson, W. B., Doughty, J. L., and Bolton, J. L. 1941. Can. 1. Comp. Med. Vet. Sci. 5, 303-313. Davison, K. L., and Wright, M. J. 1963. Unpublished data. Davison, K. L., and Seo, J. 1963. J . Dairy Sci. 46, 862-864. Davison, K. L., Hansel, W., Wright, M. J., and McEntee, K. 1962. Cornell Nutr. Conf. Feed Manuf. Proc. Cornell Uniu., Zthuca, New York 1962 pp. 44-51. Davison, K. L., Hansel, W., McEntee, K., and Wright, M. J. 1963. J . Animal Sci. 22, 835 (abstr.).

NITRATE ACCUMULATION AND POISONING

243

Diven, R. H. 1963. Personal communication. Dodd, D. C., and Coup, M. R. 1957. New Zealand Vet. J . 5, 51-54. Donahoe, W. E. 1949. Pediatrics 3, 308-311. Doughty, J. L., and Warder, F. G. 1942. Sci. Agr. 23, 233-236. Emerick, R. J., and Lievan, V. F. 1963. J. Nutr. 79, 168-170. Emerick, R. J., and Olson, 0. E. 1962. J. Nutr. 78, 73-77. Eppson, H. F., Glenn, M. W., Ellis, W. W., and Gilbert, C. S. 1960. J. Am. Vet. Med. Assoc. 137, 611-614. Fertig, S. N. 1952. Proc. 6th Ann. Meeting Northeastern Weed Control Conf. New York, New York 1952 pp. 13-19. Finnell, H. H. 1932. Oklahoma Agr. Expt. Sta. Bull. 203. Flynn, L. M., Gehrke, G. W., Muhrer, M. E., Smith, G. E., and Zuber, M. S. 1957. Missouri Univ. Agr. Expt. Sta. Res. Bull. 620. Frank, P. A., and Grigsby, B. H. 1957. Weeds 5, 206-217. Gamer, G. B. 1958. Missouri Univ. Agr. Expt. Sta. Bull. 708. Garner, G. B., ODell, B. L., Radar, P., and Muhrer, M. E. 1958. J. Animal Sci. 17, 1213 (abstr.). Gilbert, C. S., Eppson, H. F., Bradley, W. B., and Beath, 0. A. 1946. Wyoming Agr. Erpt. Sta. Bull. 277. Goodrich, R. D., Emerick, R. J., and Embry, L. B. 1962. J . Animal Sci. 21, 997 (abstr.). Greene, I., and Hiatt, E. P. 1954. Am. 1. Physiol. 176, 463-467. Grigsby, B.H., and Fanvell, E. D. 1950. hlich. State Uniu. Agr. Expt. Sta. Quart. Bull. 32, 378-385. Gul, A., and Kolp, B. J. 1960. Agron. J. 52, 504-506. Hageman, R. H. 1963. Personal communication. Hageman, R. H., and Flesher, D. 1960. Plant. Physiol. 35, 700-708. Hageman, R. H., Flesher, D., and Gitter, A. 1961. Crop Sci. 1, 201-204. Hageman, R. H., Cresswell, C. F., and Hewitt, E. J. 1962. Nature 193, 247-250. Hageman, R. H., Zieserl, J. F., and Leng, E. R. 1963. Nature 197, 263-265. Hale, W. H., Hubbert, F. J., and Taylor, R. E. 1961. 1. Animal Sci. 20, 934 ( abstr.). Hanway, J. J. 1962. Agron. J. 54, 222-229. Hanway, J. J., and Engelhorn, A. J. 1958. Agmn. J. 50, 331-334. Hanway, J. J., and Moldenhauer, W. C. 1964. Zowa Agr. Expt. Sta. Res. Bull. (in press). Hanway, J. J., Herrick, J. B., Willrich, T. L., Bennett, P. C., and McCall, J. T. 1963. Iowa State Ext. Serv. Spec. Rept. 34. Harker, K. W., and Kamau, A. K. 1961. E. African Agr. Forestry J . 27, 57-60. Helwig, D. M., and Setchell, B. P. 1960. Austrulian Vet. J . 36, 14-17. Hiatt, E. P. 1940. Am. J. Physiol. 129, 597-609. Hoener, 1. R., and DeTurk, E. E. 1938. J. Am. Soc. Agron. 30, 232-243. Holley, K. T., Dulin, T. G., and Pickett, T. A. 1931. Georgia Expt. Sta. Bull. 169. Holst, W. O., Flynn, L., Gamer, G. B., and Pfander, W. H. 1961. J. Animal Sci. 20, 936 (abstr. ). Holtenius, P. 1957. Acta Agr. Scand. 7, 113-163. Hueper, W. C., and Landsberg, J. W. 1940. A.M.A. Arch. Pathol. 29, 633-648. Hvidsten, H., Odelien, M., Baerug, R., and Tollersrud, S. 1959. Acta Agr. Scand. 9, 261-291. Jacobson, W. C., and Wiseman, H. G . 1963. J. Dairy Sci. 46, 617-618 (abstr.).

244

MADISON J. WRIGHT AM) KENNETH L. DAVISON

Jainudeen, M. R., Hansel, W., Davison, K. L., and Wright, M. J. 1963. J . Dairy Scd. 46, 634-635 (abstr.). Jamieson, N. D. 1958. Nature 181, 1601-1802. Johnson, R. M., and Bauman, C. A. 1947. J. Bid. Chem. 171, 513-521. Jones, D. I. H., ap Griffith, G., and Walters, R. J. K. 1961. J. Brit. Grassland SOC. 16, 272-275. Jordan, H. A., Neumann, A. L., Smith, G. S., Zimmerman, J. E., and Vatthauer, R. J. 1961. J . Animal Sci. 20, 937-938 (abstr.). Jordan, H. A., Smith, G. S., Neumann, A. L., Zimmerman, J. E., and Breniman, G. W. 1963. J. Animal Sci. 22, 738-745. Kearley, E. O., Gamer, G. B., Bloomfield, R. A., and Muhrer, M. E. 1962. I. Animal Sci. 21, 1002 (abstr.). Kendrick, J. W., Tucker, J,, and Peoples, S. A. 1955. J. Am. Vet. Med. ASSOC.126, 53-56. Kilgore, L., Almon, L., and Gieger, M. 1959. J . Nuts. 69, 39-44. Kilgore, L., Stasch, A. R., and Barrentine, B. F. 1963. J . Am. Dietet. Assoc. 43, 39-42. Knipmeyer, J. W., Hageman, R. H., Earley, E. B., and Seif, R. D. 1962. Crop Sci. 2, 1-5. Koch,-B. A., Parrish, D. B., and Sukhonthasornpa, S. 1963. 1. Animal Sci. 22, 840 ( abstr. ) . Kretschmer, A. E., Jr. 1958. Agron. J. 50, 314-316. Leeper, G. W. 1941. 1. Australian Znst. Agr. Scl. 7 , 161-162. Lewis, D. 1951a. Biochem. J . 48, 175-180. Lewis, D. 1951b. Biochem. J . 49, 149-153. Lowry, M. W., Huggins, W. C., and Forrest, L. A. 1936. Georgia Agr. Expt. Stu. Bull. 193. Lowry, T., and Schuman, L. M. 1956. J . Am. Med. Assoc. 162, 153-160. Lycklama, J. C. 1963. Actu Botan. N e e d 12, 361-423. McEntee, K. 1963. Personal communication. McIntosh, I. C., Nielson, R. L., and Robinson, W. D. 1943. New Zealand J . Agr. 66, 341-343. McKee, H. S. 1962. “Nitrogen Metabolism in Plants.” Oxford Univ. Press (Clarendon), London and New York. MacKenzie, A. J., Spencer, W. F., Stockinger, K. R., and Krantz, B. A. 1963. Agron. J. 55, 55-59. Mayo, N. S. 1895. Kansas Agr. Expt. Stu. Bull. 49. Metcalf, W. K. 1962a. Phys. Med. B i d . 6, 427-435. Metcalf, W. K. 1962b. Phys. Med. BioZ. 6, 437-444. Monson, W. G. 1963. Unpublished data. Mulder, E. G. 1948. Plant Soil 1, 94-119. Muhrer, M. E., Gamer, G. B., Pfander, W. H., and O’Dell, B. L. 1956. J . Animd Sci. 15, 1291 (abstr.). Nason, A. 1962. Bacterol. Rev. 26, 16-41. Nebraska Department of Health. 1950. Better HeuZth, Ofic. Bull. State Nebraska Dept. Health 9 ( 3 ) , 1. Newsom, I. E., Stout, E. N., Thorp, F., Jr., Barber, C. W., and Groth, S. H. 1937. J. Am. Vet. Med. Assoc. 90,66-75. Nightingale, G. T., Schermerhom, L. G., and Robbins, W. R. 1930. New Jersey Agr. Expt. Sta. Bull. 499.

NlTMTE ACCUMULATION AND POISONING

245

Nowakowski, T. Z. 1961. J. Agr. Sci. 56, 287-292. O’Dell, B. L., Erek, Z., Flynn, L., Gamer, G. B., and Muhrer, M. E. 1960. J. Animal Sci. 19, 1280 (abstr.). Ogata, G., and Caldwell, A. C. 1960. Agron. J. 52, 65-68. Olofsson, S. 1962. Kgl. Lantbrukshogskol. Statens Lantbruksforsok, Statens Jordbruksforsok Medd. No. 135. Olson, 0. E., and Moxon, A. L. 1942. J . Am. Vet. Med. Assoc. 100, 403-406. Olson, 0. E., and Whitehead, E. I. 1940. Proc. S. Dakota Acad. Sci. 20, 95-101. Olson, 0. E., Nelson, D. L., and Emerick, R. J. 1963. J. Agr. Food Chem. 11, 140-143. Orgerson, J. D., Martin, J. D., Caraway, C. T., Martine, R. M., and Hauser, G. H. 1957. Public Health Rept. 72, 189-192. Peterson, W. H., Burns, R. H., Sant, R., and Little, H. N. 1958. 1. Agr. Food Chem. 6, 121-126. Pfander, W. H., Gamer, G. B., Ellis, W. C., and Muhrer, M. E. 1957. Missouri Uniu. Agr. Expt. Sta. Res. Bull. 637. Pugh, D. L., Gamer, G. B., Bloomfield, R. A., and Muhrer, M. E. 1962. J. Animal Sci. 21, 1009 (abstr.). Rath, M. R., and Krantz, J. C., Jr. 1942a. J. Pharmacol. Exptl. Therap. 76, 27-32. Rath, M. R., and Krantz, J. C., Jr. 1942b. J. Phamucol. Exptl. Therap. 76, 33-38. Reddy, B. S., and Thomas, J. W. 1962. J . Animal Sci. 21, 1010 (abstr.). Riggs, C. W. 1945. Am. J. Vet. Res. 6, 194-197. Rimington, C., and Quin, J. I. 1933. Onderstepoort J . Vet. Sci. Animal Ind. 1, 469-489. Robinson, W. D. 1942. New Zealand J. Agr. 65, 199-202. Sapiro, M. L., Hoflund, S., Clark, R., and Quin, J. I. 1949. Onderstepoort J. Vet. Sci. Animal Ind. 22, 357-372. Savage, A. 1949. Can. 1. Comp. Med. Vet. Sci. 13, 9-10. Scaletti, J. V., Gates, C. E., Briggs, R. A., and Schuman, L. M. 1960. Agron. J. 52, 369-372. Schimper, A. F. W. 1888. Botan. Z . 46, 65-69, 81-89, 97-107, 113-123, 129-139, 145-153. Schwarte, L. H., Eveleth, D. F., and Biester, H. E. 1939. Vet. Med. 34, 648-651. Sessions, A. C., and Shive, J. W. 1933. Soil Sci. 35, 355-374. Setchell, B. P., and Williams, A. J. 1962. Australian Vet. J . 38, 58-62. Simon, J., Sund, J. M., Wright, M. J., Winter, A., and Douglas, F. D. 1958. J. Am. Vet. Med. Assoc. 132, 164-169. Simon, J,, Sund, J. M., Douglas, F. D., Wright, M. J., and Kowalczyk, T. 1959a. J . Am. Vet. Med. Assoc. 135, 311-314. Simon, J,, Sund, J. M., Wright, M. J., and Douglas, F. D. 1959b. J. Am. Vet. Med. ASSOC. 135, 315-317. Smith, D. 1960. Agron. J. 52, 637-639. Smith, G. S. 1963. Personal communication. Smith, G. S., Hatfield, E. E., Durdle, W. M., Neumann, A. L. 1962. J. Animal Sci. 21, 1013-1014 (abstr.). Sokolowski, J. H., Garrigus, U. S., and Hatfield, E. E. 1960. J. Animal Sci. 19, 1295 ( abstr. ) . Sokolowski, J. H., Garrigus, U. S., and Hatfield, E. E. 1961. J. Animal Sci. 20, 953 ( abstr. ). Sorensen, C. 1960. Acta Agr. Scand. 10, 17-32.

246

MADISON J. WRIGHT AND KENNETH L. DAVISON

Sorensen, C. 1962. Acta Agr. Scand. 12, 106-124. Spicer, S. S., and Reynolds, H. 1949. Am. J. Physiol. 159, 47-56. Stahl, A. L., and Shive, J. W. 1933. Soil Sci. 35, 375-399. Stahler, L. M., and Whitehead, E. I. 1950. Science 112, 749-751. Stewart, G. A., and Merilan, C . P. 1958. Missouri Univ. Agr. Expt. Sta. Res. Bull. 650.

Steyn, D. G. 1934. The toxicology of plants in South Africa, together with a consideration of poisonous foodstuffs and fungi. Central News Agency, Ltd., Johannesburg. 631 pp. Steyn, D. G. 1951. Farmer's Weekly 82, 53, Oct. 10. Stillings, B. R., Bratzler, J. W., Keck, E., and Marriott, L. F. 1961. J. Animal Sci. 20, 198 (abstr.). Stormorken, H. 1953. Proc. 15th Intern. Vet. Congr., Stockholm 1953 Vol. 1, pp. 501-505. Stout, P. R., and Meagher, W. R. 1948. Science 108, 471-473. Stoy, V. 1956. Biochim. Biophys. Acta 21, 395-396. Sund, J. M. 1959. Unpublished data. Sund, J. M., and Mao, C.-H. 1958. Unpublished data. Sund, J. M., and Wright, M. J. 1959. Down Earth 15, 10-13. Sund, J. M., Wright, M. J,, and Simon, J. 1957. Agron. J . 49, 278-279. Sund, J. M., Niedermeier, R. P., and Burris, R. H. 1963a. Wisconsin Agr. Expt. Sta. Ext. Circ. 624. Sund, J. M., Burch, C. W., Murdock, J. T., and Werner, G. M. 196313. Wisconsin Ext. Circ. Agr. Erpt. Sta. 627. Theron, A. A. 1957. W.D. Thesis, Cornell Univ., Ithaca, New York. Tollett, J. T., Becker, D. E., Jensen, A. H., and Terrill, S. W. 1960. J. Animal Sci. 19, 1297 (abstr.). Tottingham, W. E., Stephens, H. L., and Lease, E. J. 1934. Plant Physiol. 9, 127142. Tucker, J. M., Cordy, D. R., Berry, L. J., Harvey, W. A,, and Fuller, T. C. 1961. Calif. Agr. Expt. Stu. Ext. Circ. 506. Ulrich, A. 1961. Am. Inst. Biol. Sci. Publ. 8, 190-211. Ulrich, A., Hills, F. J., Rine, D., George, A. G., and Morse, M. D. 1959. Culif. Agr. Expt. Sta. Bull. 766. U. S . Department of Agriculture. 1960. Regulations Governing the Meat Inspection of the U. S. D. A., A. R. S., Meat Inspection Division. Regulation 18.7 ( k ) . U. S. Govt. Printing Office, Washington, D. C. U. S. Public Health Service. 1962a. U. S. Public Health Sew. Publ. 937. U. S . Public Health Service. 1962b. U. S . Public Health Sew. Publ. 956. Walton, G. 1951. Am. J . Public Health 41, 986-996. Wang, L. C., and Burris, R. H. 1960. J . Agr. Food Chem. 8, 239-242. Wang, L. C., Garcia-Rivera, J., and Burris, R. H. 1961. Biochem. J . 81, 237-242. Wanntorp, H., and Swahn, 0. 1953. Proc. 15th Intern. Vet. Congr. Stockholm 1953 VOI. 1, pp. 496-501. Webb, L. J. 1952. J . Australian Inst. Agr. Sci. 1 8 ( 3 ) , 164-167. Weichenthal, B. A., Emerick, R. J., Embry, L. B., and Whetzal, F. W. 1961. J. Animal Sci. 20, 955 (abstr.). Welch, C. D., and Bartholomew, W. V. 1963. Agron. J. 55, 441-443. Welsch, C. W., Bloomfield, R. A., Garner, G. B., and Muhrer, M. E. 1961. J. Animal Scl. 20, 981 (abstr.).

NITRATE ACCUMULATION AND POISONING

247

Welsch, C. W., Bloomfield, R. A., Gamer, G. B., and Muhrer, M. E. 1962. J. Animal Sci. 21, 1032 (abstr.). Whelan, M. 1935. Biochem. J. 29, 782-787. Whitehead, E. I. 1961. S . Dakota Farm Home Res. 12, 8-9. Whitehead, E. I., and Moxon, A, L. 1952. S . Dakota Agr. Expt. Sta. Bull. 424. Whitehead, E. I., and Olson, 0. E. 1941. Proc. S. Dakota Acad. Sci. 21, 67-72. Whitehead, E. I., Viets, F. G., and Moxon, A. L. 1948. S . Dakota Agr. Expt. Sta. Tech. Bull. 7. Whitehead, E. I., Kersten, J., and Jacobsen, D. 1956. Proc. S. Dakota Acad. Sci. 35, 106-110. Willard, C. J. 1950. Proc. 7th Ann. Meeting N . Central W e e d Cont. Conf. Milwaukee, Wisconsin, 1950 p p . 110-112. Williams, C . H., and Hines, H. J. G. 1940. Australion Vet. J . 16, 14-20. Wilson, J. K. 1943. J. Am. Soc. Agron. 35, 279-290. Wilson, J. K. 1949. Agron. J. 41, 20-22. Winter, A. J. 1962. Am. J . Vet. Res. 23, 500-505. Wolfer, H. 1941. Bodenk. Pfianzenemaehr. [ N . S.] 23,226-259. Wright, N., and Trautman, R. J. 1962. Agron. J. 54, 363-364. Wright, N., Trautman, R. J., and Streetman, L. J. 1960. Agron. 1. 52, 671. Wyngaarden, J. B., Wright, B. M., and Ways, P. 1952. Endocrinology 50, 537-549. Yadav, K. P., Garner, G. B., O’Dell, B. L., Flynn, L. M., Bloomfield, R. A., and Muhrer, M. E. 1962. J. Animal Sci. 21, 1017 (abstr.). Zacharias, E. 1884. Botan. Z. 42, 65-73. Zieserl, J. F., and Hageman, R. H. 1962. Crop Sci. 2, 512-515. Zieserl, J. F., Rivenbark, W. L., and Hageman, R. H. 1963. Crop Sci. 3, 27-32.

This Page Intentionally Left Blank

CHARACTERIZING SOIL OXYGEN CONDITIONS WITH A PLATINUM MICROELECTRODE L. H. Stolzy and J. Letey University of California, Riverside, California

I. Introduction ................................................ 11. Polarography ................................................ ........ A. Development . . . . . B. Principle of the Platinum Electrode ........................ C. Theory and Calculations . . . . . 111. Problems Associated with the Use of A. Standardization of B. Soil Moisture . . . . C. Temperature . . . . . . ........................ D. Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Relationships between Oxygen Diffusion Rates and Biological Responses A. Plant .................................................. .................... B. Microorganisms . V. Results of Field Meas ................................ VI. Summary . . . . . . . . . . . . . ....................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Page 249 250 250 251

257 258 258 258 269 272 275 277

I. Introduction

The evaluation of life processes occurring in soils is a continual challenge to a large number of research groups. Old tools are revised and new ones developed to measure the endless activities taking place in soils. Oxygen is required by all aerobic forms of life. An understanding of oxygen requirement and mechanism for supply to an organism is an essential factor in understanding its life processes. For air-breathing animals, the essential factors are open passage to the lungs and an adequate concentration of oxygen in the air that they breathe. A measurement of oxygen concentration in the air is sufficient to characterize the animal’s oxygen environment. Most fish extract oxygen from the water by means of gills. Since fish are mobile, a measurement of the oxygen concentration dissolved in water would characterize the oxygen environment for fish. Plant physiologists have long known that plant roots respire and grow 249

250

L. H. STOLZY AND J. LETEY

in solution if the solution is agitated with a fresh supply of oxygen. Because of the simplicity of maintaining oxygen conditions at various levels in solution cultures, a vast amount of information on soil-oxygenplant relationships has been obtained under these conditions. These data do not completely provide information applicable to plants and microorganisms growing in the soil because the mechanism for oxygen supply is not adequately accounted for. Plant roots and microorganisms do not have their surfaces continuously swept by solution in which oxygen is dissolved. The water films around the respiring surface and in the soil micropores are rather stationary. This means that living cells in soils must obtain their supply of oxygen by the process of diffusion through a moisture film. The diffusion coefficient of oxygen through solution is about 10,000 times slower than diffusion in the gaseous phase. It is likely that the limiting factor for oxygen suppIy is diffusion through the moisture film, and therefore this is the factor which should be measured to characterize soil oxygen conditions. This paper will discuss the application of polarographic principles to measure oxygen diffusion rates in the soil water system. II. Polarography

A. DEVELOPMENT The term polarography was coined by Jaroslav Heyrovsky in 1922 to describe a method of chemical analysis. This method is based on the unique characteristics of the current-voltage curves obtained when solutions of electrooxidizable or electroreducible substances are electrolyzed in a cell in which the dropping mercury electrode is used (Kolthoff and Lingane, 1952). Polarography has since crossed over into many fields of research with various adaptations of the method to fit conditions peculiar to the specific conditions encountered. Substitution of the platinum microelectrode for the dropping mercury electrode has been common in several fields of research, Most of the research in polarography, however, has dealt with the dropping mercury system. Theory and practical adaptations of the dropping mercury electrode are thoroughly described in various books on polarography. Research on the use of the stationary platinum electrode is more limited. Although both the dropping mercury and platinum electrode can be used to measure many elements in solution at different applied potentials, the measurement of dissolved oxygen in electrolytic solutions has been important. The reason for its importance is that there are few satisfactory methods for the determination of dissolved oxygen. It should be pointed out that the polarographic method is usually used in a well-defined

CHARACTERIZING SOIL OXYGEN CONDITIONS

251

system. However, in some cases, the stationary platinum electrode is not used in a well-defined system. For example, Blinks and Skow (1938) used a platinum microelectrode for the determination of oxygen during photosynthesis by leaves. Another specialized use of a stationary platinum electrode was in the study of Davies and Brink (1942) on 0, tension in animal tissues. Lemon and Erickson (1952) introduced the use of the platinum microelectrode to the field of soil science by measuring the oxygen diffusion rate in soils. Karsten (1939) had previously studied oxygen in soil suspensions with a dropping mercury electrode. In his study, he used six different soils of widely varied texture and organic matter. The only substance in the soil that was reduced at the electrode was oxygen, and this only at certain potentials. B. PRINCIPLE OF

THE

PLATINUM ELECTRODE

When a certain electrical potential is applied between the platinum electrode inserted in the soil and a reference electrode, oxygen is reduced at the platinum surface. The general reaction taking place at the platinum microelectrode surface in the reduction of oxygen is in two steps involving two electrons in each step (Kolthoff and Lingane, 1952). The reactions in two different media are: Acid medium

0,

+ 2H+ + 2e- + H 2 0 2

H20, + 2H+ + 2e- + 2H,O Neutral OT alkaline medium 0 , + 2H,O + 2e- + H 2 0 , 2 0 H H,O, + 2e- + 2 0 H -

+

An electric current flows between the two electrodes and is proportional to the rate of oxygen reduction. The rate of oxygen reduction is in turn related to the rate at which it diffuses to the electrode. Poel (1960a) presented a current-voltage curve obtained in a soilwater suspension. At low potential (low will be used in the absolute sense) the current is low. As Davies and Brinks (1942) point out, the reaction rate at low potential is controlled by what could be called intrinsic factors. The energy barrier is limiting the rate of transfer of charge. The energy barrier is overcome by the increased field strength, and the reaction rate increases at higher potentials. The increase of current with potential continues until it is limited by the rate at which 0 2 can diffuse from the body of the solution to the electrode. So at the higher potentials, the reaction rate is limited by an extrinsic factor and the current becomes somewhat independent of the potential, as is indicated by the plateau shown by Poel (1960a) in his current-voltage

252

L. H. STOLZY AND J. LFTEX

curve. A second reaction starts at a potential of 0.7 or 0.8 volt. The second rise is caused by the reduction of ionic hydrogen to molecular hydrogen. Because the current in the plateau region is limited by the maximal rate at which oxygen can diffuse to the platinum microelectrode surface, the calculation of oxygen movement becomes a problem in classical diffusion theory. C. THEORY AND CALCULATIONS Since the general procedure in using the platinum microelectrode to characterize soil oxygen conditions consists of measuring a current which is proportional to the oxygen diffusing to the electrode, it is necessary to consider diffusion theory in order to completely understand what is actually being measured. Fick's law of diffusion for a linear system is

where f is the flux (quantity per unit area per unit time), D is the diffusion coefficient, and 6c/6x is the concentration gradient. Lemon and Erickson (1952) used the term oxygen diffusion rate (O.D.R.) to denote flux when measurements are made in soil. The O.D.R. for a linear system, therefore, includes the oxygen diffusion coefficient and concentration gradient at the electrode surface. Letey et al. (1964a) point out that a cylindrical coordinate system is more appropriate when a platinum wire electrode is used. The general diffusion equation in cylindrical coordinates is

6% -+--=o 6T2

1 6c T 8T

Applying the steady state boundary conditions

c = 0 when T = a (electrode radius) c = c2 when T = r1 the integration of Eq. ( 2 ) results in

-6 C_-

c2

(3) a(lnrl-lna) The flux is equal to the diffusion coefficient times the concentration gradient and is Dc2 fr=a (4) a(ln r1 -1n a ) The value r1 is the distance from the electrode center to where the concentration changes very slowly from c2 (the initial uniform concen6r

CHARACTERIZING SOIL OXYGEN CONDITIONS

253

tration). In some cases of unsaturated soil, r1 corresponds to the distance to the edge of the water film surrounding the electrode. The factors included in the O.D.R. value are seen from Eq. (4) to be the oxygen concentration, oxygen diffusion coefficient, a film thickness factor, and the electrode radius. The electrode radius cannot be factored out from the O.D.R. value without knowledge of the value for rl, which is not readily obtained. Unfortunately, and important to remember, is that a different O.D.R. value may be determined for identical oxygen conditions if electrodes of different radii are used to make the measurement. The factor measured is actually the electric current. The electric current is related to the flux by

iX = nFAf (5) where i is the current in microamperes, n is the number of electrons required to reduce one molecule of oxygen and has the value of 4, F is the Faraday constant, A is the platinum area, and f has the units of moles of oxygen per second per square centimeter. It has been traditional to report the O.D.R. (identical to flux) in units of grams of oxygen per square centimeter per minute. The O.D.R. calculated from Eq. 5 can be converted to the traditional units by multiplying by the appropriate conversion values. 111.

Problems Associated with the Use of Platinum Microelectrodes in Soils

The interest and application of polarography techniques to various fields of research have helped in understanding measurements made in soils. However, as was pointed out previously, most of these studies were conducted in well-defined two-phase water-solid systems. The classical oxygen polarogram (current-voltage curves) with a definite plateau in soil-water suspension (Lemon and Erickson, 1952; Poel, 1960a) does not occur in a three-phase soil-water-air system (Birkle et al., 1964). Figure 1 illustrates the relationship between current and potential which occurs in a three-phase system. The applied potential to the electrode becomes critical if O.D.R. results from various investigations are to be compared. A. STANDARDIZATION OF METHODS

In the discussion of oxygen diffusion rates in relation to plant responses, it will become very evident that a certain amount of empiricism exists in the O.D.R. measurement. This then requires standardization of the method if in the future investigators wish to compare results.

254

L. H. STOLZY A N D J. LETEY

Choice of the same applied potential to the platinum microelectrode should be one of the first considerations. The current-voltage curves of Lemon and Erickson (1955) and Poel (1960a) would indicate that an applied potential of 0.80 volt is the extreme limit before hydrogen reduction. Birkle et at. (1964) has shown this to be the case also in soils. They have shown that any potential between 0.55 and 0.75 volts loo

1

90 I 7 .-

25 GAUGE ELECTRODE ,H4

80

E

?' 70

/"'

E

0

*0

J

60

fl

~Slope=100

x

50

Ir 40

d o 30 20

-.50

-.60

-.70

-.80

-.go

Potential, Volts

FIG.1. Oxygen diffusion rate (O.D.R.) as a function of applied potential as measured with a 25-gauge wire electrode. Each curve represents measurements made in a particular environment (Letey et al., 1964a).

can be used for a standard recommendation. Inasmuch as many published results have been measured at 0.65 volt and there seems to be no strong reason for using another potential, it was recommended that 0.65 volt be adopted when measuring soil oxygen diffusion rates. Another factor that seems to vary considerably among investigators taking oxygen diffusion measurements in soils is the time assumed after the potential is applied for steady state conditions. Lemon and Erickson (1952) used 5 minutes for a steady state reading. They also indicated

CHARACTERIZING SOIL OXYGEN CONDITIONS

255

that in many cases the current had become nearly steady within 3 minutes. Birkle et al. (1964) studied the relationships between current and time in soils with two electrode sizes (22- and =-gauge platinum wire, 4 mm. long) and in bentonite mixture with 25-gauge electrodes. It was found that the decrease in current with time becomes very small after 4 or 5 minutes. In a system where 10 electrodes (Letey and Stolzy, 1964) have potential applied at the same time, approximately 45 seconds will be required to measure and record all values. The recommendation is that at least 3 minutes should be allowed, and preferably 4 minutes for steady state conditions, before taking readings (Birkle et al., 1964). It has been pointed out that one would expect a different O.D.R. to be measured with electrodes of different radii. Also the flux cannot conveniently be corrected for different electrode sizes. The choice of electrode size depends on the type of study being investigated. The 25gauge wire electrode has been used continuously in greenhouse studies. This wire is too fragile for use under field conditions. Platinum microelectrodes of 22-gauge have been successfully used by several investigators. Although results with the two different-size electrodes are not easily comparable, the advantage of accurately measuring O.D.R. over a greater moisture range with the 25-gauge electrode in controlled experiments seems to be more important at this time. B. SOIL MOISTURE In a discussion of limitations and errors involved with the platinum microelectrode method, Lemon and Erickson (1955) indicated that the most serious limitation was the wetting of the electrode when making in situ measurements in the soil. In order to obtain quantitative measurements, the area of the electrode must be known and it is essential to have the whole electrode surface covered by a moisture film. It is only when the entire electrode is wet that the total surface area functions as a reducing surface. Theoretical data by Letey et al. (1964a) show that the O.D.R. is expected to increase as the soil moisture decreases. However, measurements taken as a function of soil moisture by Lemon and Erickson (1955) and Lemon and Kristenson (1960) showed an increase in O.D.R. as moisture content decreased to a point beyond which O.D.R. decreased with decreasing moisture content. A detailed study was made by Birkle et al. (1964) on factors influencing oxygen diffusion rates in soil as measured by the platinum microelectrode. One aspect studied was how soil moisture affected O.D.R. measurements. In their studies, a ceramic double-walled pot was used to maintain different soil suction levels and moisture contents in sands. Oxygen diffusion measurement with 22- and 25-gauge electrodes showed

256

L. H. STOLZY AND J. LETEY

that at high moisture contents similar O.D.R. were measured with both electrodes. As the moisture content decreased, the O.D.R. measured by the larger electrode (22-gauge) dropped off first and at a higher moisture content than the 25-gauge (Fig. 2 ) . This information plus other studies were used to conclude that the decrease in O.D.R. with decreasing soil moisture is caused by incomplete electrode wetting. The authors used the same ceramic double-walled pot equipment described by Birkle d al. (1964) to study the relationship between soil suction and O.D.R. in a loamy sand and a Krilium-treated silt loam *O

1

0

I

I

.I0 .20 .30 .40 VOLUMETRIC WATER CONTENT

I

.50

FIG. 2. Oxygen diffusion rate at various moisture contents in sand for two platinum microelectrode sizes ( Birkle et al., 1964 ).

(Fig. 3 ) . The loamy sand had very low O.D.R. from 0 to 20 millibars (mb.) of suction. As the soil suction was increased from 20 to 80 mb., O.D.R. increased to a maximum; and from 80 to 240 mb., rapidly decreased. Oxygen diffusion rate in the silt loam rapidly increased with increasing soil suction between 0 and 20 mb. suction. At 80 to 300 mb., g. cm.-2 min.-l. Lemon and the O.D.R. was between 80 and 90 x Erickson ( 1955) discussed preliminary investigations of the effects of soil suction and texture on O.D.R. In coarse-textured soil, as soil suction increased, moisture films contracted away from the surface of the platinum microelectrode surface. They felt this was greatly influenced by the size of soil particles in the region of the electrode. With finer-textured soils, a greater soil suction was required to rupture the liquid films wetting the platinum surface. Lemon and Erickson, in their discussion, concluded that the electrode could be successfully used over most of the available moisture range in loams and finer-textured soils. The data presented in Fig. 3 show that the platinum microelectrode can be used

257

CHARACTERIZING SOIL OXYGEN CONDITIONS

over a much larger range of soil suction in the silt loam than in the loamy sand, which has a very restricted range. Lemon and Kristenson (1960) suggested that the decrease in O.D.R. as the soil suction increased or moisture content decreased was due to a buildup of OH- concentration formed by the reaction because of slow diffusion away from the electrode. Birkle et al. (1964) studied the effect of OH- concentration on the current by measuring the current in

L

0

I

I

5

10

I

I

I

25

30

I

15 20 SUCTION,

1

35

I

40

CB

FIG.3. Oxygen diffusion rate (O.D.R.) at different soil suctions for a Hanford loamy sand and Krilium-treated Yo10 silt loam with a 25-gauge platinum microelectrode.

saturated silica sand treated with various concentrations of NaOH. Their data showed that the O.D.R. dropped off very slowly with increasing OH- concentration up to 0.6 N Concentration. By calculating the grams of OH- produced per second per microampere (1.76 X 10-lo) and using values of current often encountered in making measurements, it was concluded that it was unlikely that the OH- concentration would be high enough to significantly modify the O.D.R. results.

C. TEMPERATURE The oxygen diffusion rate is temperature dependent. An increase in temperature decreases the solubility of oxygen and increases the diffusion coefficient through both gas and liquid. The diffusion of oxygen through water increases in the range of 3 to 4 per cent per degree centigrade (Carlson, 1911; Millington, 1955; Lammann and Jessen, 1929). The

258

L. H. STOLZY A N D J. LETEY

solubility of oxygen decreases approximately 1.6 per cent per degree centigrade. Letey et d. (1962a) estimated that oxygen diffusion rates should increase in a range of 1.4 to 2.4 per cent per degree rise in soil temperature for a given oxygen concentration in the soil pores. Oxygen diffusion rates measured in the study by Letey et al. (1962a) at soil temperatures of 23” and 13” showed that O.D.R. increased approximately 1.8 per cent per degree rise in soil temperature, The measured result compared favorably with theoretical calculations.

D. POISONING The use of the word “poisoning” in connection with platinum microelectrode denotes a change in the surface of platinum that alters the sensitivity of the electrode to reduction of oxygen. Sawyer et al. (1959) indicated that sulfur compounds commonly poison platinum. Sawyer and Interrante (1961) showed that oxidized platinum surfaces will give different readings than the nonoxidized or prereduced surface. Birkle et al. (1964) showed that platinum microelectrodes left in the soil 2 weeks gave significantly different O.D.R. measurements than those freshly inserted, Their data indicated that poisoning could occur with the buried platinum electrodes but is prevented merely by reinserting the electrodes into the soil. The abrasive action of the soil appeared to remove whatever chemical deposit caused the decrease in O.D.R. In that study, the comparison was made between electrodes left intact in the soil and electrodes reinserted. It is possible that the differences in O.D.R. by the two sets could have been caused by differences in water geometry associated with electrodes left in place and those replaced. This factor requires more investigation. IV. Relationships between Oxygen Diffusion Rates and Biological Responses

A. PLANT The rate of oxygen supply through the soil to the plant roots has been emphasized by several investigators (Cannon, 1925; Hutchins, 1926; Raney, 1950; Taylor, 1949). The importance of evaluating conditions at the interface between the root surface and the soiI system as an index of soil aeration on plant response was suggested by Russell (1952). The use of a platinum microelectrode to measure oxygen diffusion rate through the liquid phase of a soil-water-air system was first studied by Lemon and Erickson (1952,). These diffusion values were then related to plant responses.

CHARACTERIZING SOIL OXYGEN CONDITIONS

259

1. Roots Root elongation is one of the plant functions that can be closely correlated with oxygen diffusion rates. Stolzy et al. (1961a) used a special technique to vary the oxygen concentration over the soil surface. These treatments influenced the depth of rooting, as illustrated in Fig. 4.During a 17-day treatment period, downward progression of roots was marked on the cylinders. Roots in the cylinders with a soil surface

FIG. 4. Root growth achieved in relation to Oxygen concentration at the soil surface. The numbers indicate 0, concentration maintained at the soil surface. The markings represent progressive root development with time (Stolzy et al., 1961a).

environment of 21 per cent oxygen had grown the full depth of soil. Plants with an oxygen treatment of 15 per cent had a few roots near the lower end of the cylinder wall. The other treatments caused progressively less root penetration with decreasing amounts of 0 2 at the soil surface. Oxygen diffusion measurements taken through ports in the sides of the cylinders at different depths were correlated with depth of rooting (Fig. 5 ) . I t was concluded that depth of rooting and O.D.R. were related to the concentration of 0 2 at the soil surface. Root initiation was reduced or stopped when oxygen diffusion rates in the soil were in the range of 18 to 23 x lo-* g. min.-l. Diffusion rates above 23 x 10W8 g. min.-l permitted relatively good root growth.

L. H. STOLZY AND J. LETEY

280

The complexity of determining whether or not a soil is poorly aerated is illustrated by the data of Fig. 5, where no single diffusion rate characterizes the aeration condition of the entire soil volume. It is quite possible that aeration may be limited for root growth at a given soil depth, but it is conceivable that under proper management practices the volume of soil in which aeration is relatively good may be sufficient to produce OXYGEN TREATMENTS:

*

' *

21Y. 0148% ml13X

38Y1

020% 00.7%

*82%

z 2 cn

40-

3

IL

LL 0

z W

tB

c

20-

0

I

,

4.5

1

10 19 30 DEPTH I N CENTIMETERS

38

FIG.5. Oxygen diffusion rate as a function of depth for various oxygen concentrations at the soil surface. The arrows indicate the depth of root penetration (Stoky et al., 1961a).

good shoot growth. Although it is not possible to assign a specific diffusion rate to a given plant shoot response, further examination of the literature shows that O.D.R. can be used as an index to predict where root growth will or will not occur. A study by Letey et d. (1961a) showed the effects of soil temperature (23°C. and 31°C) on oxygen diffusion and the effects of both O.D.R. and temperature on root growth of sunflowers and cotton. Oxygen diffusion rate measurements taken in soils with temperatures of 31°C. were higher than those taken in soil at 23°C. This is in accordance with early discussions on the influences of temperature on O.D.R. In the above study, O.D.R. of 20 x g. cm,-2

CHARACTERIZING SOIL OXYGEN CONDITIONS

261

min.-l which was found to essentially stop root growth for snapdragons (Stolzy et @I., 1961a) represents a limiting diffusion rate for cotton and sunflowers. In a second study by Letey et al. (1962a) with sunflowers at different soil and air temperatures as well as different oxygen conditions, g. cm.-2 min.-l. roots ceased to grow under O.D.R. less than 20 x The results seem to indicate that soil temperature has little effect on the oxygen diffusion rate which will stop root growth. However, studies by Berry and Norris (1949) and Jensen (1960) showed that increased temperature causes an increase in respiration rate. Studies with O.D.R. and soil temperature indicate that temperature factors causing changes in O.D.R. are in the same direction as those affecting the root. Cannon (1925) attributed the need for higher oxygen concentrations in the soil atmosphere to maintain normal root growth at higher temperatures to the decreasing solubility of oxygen in the soil solution as the temperature increases. It was pointed out in the discussion on temperature that solubility is only one of the factors that determine the rate of oxygen supply to the root surface. A study of Newport bluegrass showed that an O.D.R. of 20 x g. cm.-2 min.-' is required for root growth (Letey et al., 1964b). The optimum, however, was an O.D.R. of 40 x loY8. The question could be asked whether all plant species stop root growth at O.D.R. of 20 x 10P8 g. cm.-2 min.-l. Cannon (1925) studied root growth in relation to oxygen supply of 30 species of plants. Growth stopped in all species when oxygen was entirely removed from the soil. Most of the species maintained a certain slow growth-rate in 0.5 per cent oxygen for a limited period of time. Cannon (1925), however, pointed out that the rate of supply rather than the concentration of oxygen influences the rate of growth. One of the plant species studied by Letey et at!. (1962b) that seemed to have a lower O.D.R. critical for root growth was barley. This plant grew slightly in a soil area where the diffusion rate was 17 x 10W8 g. cm.-2 min.-l but appeared to be stopped when O.D.R. values were less than 15 x lop8. A study by Stolzy et al. (1963) on well established citrus seedlings showed that O.D.R. between 21 and 32 x g. cm.-2 min.-l caused some decay and rotting of roots during the winter when plants were less active. However, during the summer, an active period of the year for or slightly less were critical plant growth, O.D.R. values of 20 x and O.D.R. of 31 x allowed some root growth. One possible explanation for roots growing at a lower O.D.R. in the summer is that the actively growing plant internally supplies the root with oxygen. Cannon (1925) suggested that plants, in soil with very low partial pressure of oxygen, may have a slow downward movement of oxygen into the root

262

L. H. STOLZY AND J. LETEY

from the shoot and finally into the soil from the root itself. Rather rapid diffusion of oxygen through the roots of barley and rice was demonstrated by Barber et al. (1962) by using 015-labeled oxygen around the tops of plants. Bertrand and Kohnke (1957) in a study on corn plants using bulk density and distances to a water table as variables showed that an O.D.R. of 20 to 30 g. x min.-l was detrimental to root growth. Their study showed that dense subsoils may act as effective barriers to normal root penetration of corn plants. The effect of this barrier was not entirely mechanical because of the decrease in oxygen. As they pointed out, if it had been only mechanical, the wet soil, being more pliable would allow root penetration more readily than the drier soil. This was not the case. A study (Wiersma and Mortland, 1953) on sugar beets revealed that whenever the oxygen diffusion rate is higher within a treatment, the sugar beets are longer. Oxygen diffusion rates of 20 to 30 X g. cm.-* min.-l at 4-inch soil depth were critical in the growth of sugar beets. In the same experiment when calcium peroxide fertilizer was used, the level of production was increased over that of normal fertilizers in soils with very low O.D.R. Van Diest (1962) studied the effect of soil aeration on corn. The experiment was designed in two blocks so that one block had pots of unfertilized soil and the other block had pots of fertilized soil. Within each of the two blocks, three aeration treatments were imposed. The three consisted of soil packed at two bulk densities and soil packed in a special container to allow oxygen diffusion from the top and bottom. Oxygen diffusion rates were taken at a 3-inch soil depth following the daily water adjustment to 20 per cent by weight. Significant differences between aeration treatments in root dry weight were found in the unfertilized plots. The least number of roots were in the compacted soil with O.D.R. of 10 to 13 X loF8 g. min.-I. The plants in the normal packed container had an intermediate amount of root growth with the O.D.R. of the soil 29 to 32 X whereas the aerated pots had the most roots with O.D.R. of 36 to 41 x 10-8 g. rnin.-I. The plants in the fertilized plots had more root growth than plants in the unfertilized plots. The aeration treatments were reversed in that the fertilized plants in the compacted soil had the most roots. Hanan and Langhans (1963a) measured the effects of O.D.R. on the growth of snapdragons. In their studies, values of less than 30 x g. cm.-* min.-' usually resulted in root death. For many plants, O.D.R. of 20 to 30 in soils will inhibit or retard root growth. Wiegand and Lemon (1958),from laboratory information on oxygen requirement of roots, used a theoretical approach to deter-

CHARACTERIZING SOIL OXYGEN CONDITIONS

263

mine critical oxygen concentrations at the root surface. In a field study on Miller clay and Amarillo fine sandy loam, comparisons were made of ( a ) oxygen content of the soil atmosphere (gaseous oxygen), ( b ) oxygen diffusion rates (moisture film), and ( c ) calculated oxygen concentration at root surface (dissolved oxygen) at soil depths of 4, 8, and 12 inches. It had been previously shown that the critical concentration of oxygen at the root surface is 12 per cent. At field capacity in both soils, the gaseous compositions were above 12 per cent. However, calculation of the oxygen concentration at the root surface resulted in values less than 12 per cent at both the 8- and 12-inch soil depths for Miller clay. The Amarillo fine sandy loam was well above the critical value at all for the three soil depths. It had O.D.R. well above 20 x depths. The O.D.R. in Miller clay was approximately 20 and less than 20 X lo-* at 8- and 12-inch depths, respectively. These results lend support to the critical values of O.D.R. on root growth which were empirically determined. The results are in agreement with calculations based upon respiration rates. In a soil such as Miller clay, there was a long period following irrigation or rainfall in which the plant would not produce at a maximal rate. The production potential of this soil would be limited by insufficient root oxygen. These comparisons are based on revision by Wiegand and Lemon (1963). Phillips and Kirkham (1962) studied soil compaction in relation to corn growth. The treatments were normal, moderate, and severe compaction. Although they did not measure O.D.R., they did sample the gaseous oxygen concentration in the soil pores. Oxygen contents of the gas-filled soil pores were 10 to 15 per cent for the three compactions. On the basis of past reports, they concluded that this concentration would be adequate for plant growth. They indicated that the yield reductions caused by compaction may be due to mechanical impedance if the 10 to 15 per cent level of oxygen in the soil was adequate for plant growth. In accordance with Weigand and Lemon’s (1958) theoretical approach to determining critical oxygen concentrations at the root surface, the more compacted soils could very likely have had limiting oxygen at the root surface. It was pointed out by Phillips and Kirkham (1962) that the primary difficulty is that of separating out the effects of compaction, poor aeration, and excess moisture on plant growth. 2. Tops Oxygen conditions in the rhizosphere adversely affect the plant top in many interrelated ways. As was stated previously, it is not easy to assign a specific O.D.R. value to a shoot response, as was the case for roots. Studies by Letey et al. (1961a,b, 1962a,b, 1964b) show a pro-

L. H. STOLZY AND J. LETEY

264

nounced effect of reduced soil oxygen on top growth. In these studies, oxygen was varied independently of other factors. Oxygen diffusion measurements were taken with the platinum microelectrode to characterize and evaluate the oxygen status of the soil. The plant top responses in relation to reduced oxygen over the soil surface are shown in Fig. 6. The relation between percentage of oxygen at the soil surface and O.D.R. in the soil column for these plants is shown in Fig. 5. It is apparent that the change in plant height is less affected than the weight of the plant. The general tendency for plants under adverse conditions to lengthen but not to fill in is commonly observed under field conditions.

-

a

1.6

> a

0 1.4

a

0

I-

+ z

1.2

1.0

-I

a

0.a 0.6 0 0

I L

5

OXYGEN TREATMENT

-

PER,CENT

FIG. 6. The change in height and total dry weight of snapdragons grown with different oxygen concentrations over the soil surface.

Letey et a,?.(1962b) using relatively small containers in a study on the influence of O.D.R. on sunflower growth at various soil and air temperatures, showed an average O.D.R. of 40 x g. cm.-2 min.-l was optimal for shoot growth. Low O.D.R. (less than 30 to 40 X was more detrimental to shoot growth at high air or soil temperatures than at lower temperatures. In studies on the relation of roots to aeration of soil, Cannon (1925) mentioned the interrelation of temperature and the supply of oxygen in soil on plant responses. Letey et al. (1961a) found a general similarity between diffusion rates and shoot growth curves for cotton and sunflowers grown at 23" and 31°C. This provided further indication that solution oxygen-diffusion rates are useful and, at present, probably the best measurable characteristic devised for characterizing soil-aeration conditions.

CHARACTERIZING SOIL OXYGEN CONDITIONS

265

Lemon and Erickson (1952) suggested an O.D.R. of 30 to 40 X g. cm.-2 min.-l at an 8-inch soil depth as critical for tomato plants. Although it was not stated in the article, this is assumed to be an observation for top growth. Peas (Erickson and Van Doren, 1960) did not measurably respond to low O.D.R. at early stages of growth. However, 40 to 50 days after planting, growth was increased as the O.D.R. increased from 15 to 72 x 10-8 g. cm.-2 min.-l. Green weight of the and 15 X shelled peas was 16.7 g. and 4.6 g. at O.D.R. of 46 x g. cm.-2 min.-l, respectively. Bertrand and Kohnke (1957) showed that top growth of corn increased as O.D.R. increased to about 25 x g. cm,-2 min.-l. These studies were conducted in containers where the soil in the bottom was compacted to various bulk densities and had different water levels. The surface soil was compacted uniformly for all treatments. Hanan and Langhans (1963b) studied the effects of O.D.R. on the growth of snapdragons for cut flowers. Oxygen diffusion rates of less than g. min.-l lowered the quality of the cut flower, fresh 80 x weight, and stem length. Oxygen diffusion measurements were taken to complement the data in a study on the effect of soil moisture on forage grasses and legumes (Finn et al., 1961). Ranges of O.D.R. at different soil suctions were: and 0 mb., 4 to 5 x 10-$25 mb., 5 to 9 x 40 mb., 8 to 18 X min.-l. In general, the grasses at field capacity 23 to 37 x 10-8 g. were more tolerant to flooding than were the !egumes. The order of tolerance in descending order was: reed canarygrass. timothy, bromegrass, birds-foot trefoil, Ladino clover, and alfalfa. Yields of grasses tended to increase with increased moisture while the yields of legumes decreased with increased moisture. Alfalfa growth was impaired by O.D.R. less than 18 g. cm.-2 min.-'. When using O.D.R. values in relation to top responses, many aspects can be considered. Erickson and Van Doren (1960) found that reduced O.D.R. in soils for one day just prior to flowering caused a reduction in the yield of peas. Stolzy et al. (1961b) applied a low oxygen treatment on soil supporting established tomato plants. After 40 hours of low oxygen treatment, the plant tops were fumigated with air-borne oxidants of either ozone or peroxyacetyl nitrate (PAN). The effect of the reduced soil oxygen treatment as compared to check on plant top protection to ozone is shown in Fig. 7. Oxygen diffusion measurements were taken in the soil after the fumigation treatments. Plants growing in soil with O.D.R. of 16 to 24 x min.-' were not damaged by PAN g. at a concentration of 10 p.p.m. or ozone at concentrations of 0.45 and 0.34 p.p.m. Whereas plants growing in soils with O.D,R, of 34 to 90 x

x

266

L. H. STOLZY AND J. LETEY

were moderately to severely damaged. The reasons for this change in plant top susceptibility to air-borne oxidants by reduced soil oxygen conditions were investigated more completely by Stolzy et al. (1964) in a later study. The physiological changes in plant leaves following short periods of critical soil O.D.R. were investigated. A reduction in soil O.D.R. for 8.5 hours caused 50% decrease in C 0 2 fixation by plants, but a shorter period ( 3 hours) had only a slight effect on apparent photo-

FIG.7. Ozone damage to most recently matured leaves of tomato plants growing in soil where different partial pressures of oxygen (as indicated by numbers) were maintained over the soil surface. Leaves horizontally from letters H and L were from plants exposed to ozone for 3 hours at atmospheric concentrations of 0.45 p.p.m. and 0.34 p.p.m., respectiveIy (Stolzy et nl., 1961b).

synthesis. However, the short period of anaerobic treatment (N2 gas over the soil surface) followed by normal soil aeration for 20 to 24 hours reduced the susceptibility of the plant to ozone damage by as much as 50 per cent. Carbohydrate analyses of leaf tissue from similarly treated plants showed a 50 per cent reduction in sucrose and 33 per cent reduction in starch. It is hypothesized that the anaerobic pretreatment caused a change in the root membrane permeability or a shift in metabolism that promoted excretion of certain carbohydrates from the root which in turn resulted in a lowering of the concentration of carbohydrates in

267

CHARACTERIZING SOIL OXYGEN CONDITIONS

the leaves. This could also be the reason why Erickson and Van Doren (1960) found that tomato plants were stunted by periods of oxygen deficiency when the plants were still very young. These data tend to indicate that soil aeration is generally more of a problem than previously believed. 3. Mineral Accumulution in Tops The uptake and translocation of certain ions from root to tops is much dependent on an oxygen supply to the roots, as numerous liquid-

!

0.5

qL

Na I

0

I

I

I

,

Clipping height I inch 2 inch

0-4

c

0.3

lo.2: 0.1

2 5 10 20 0 2 5 10 20 Oxygen t r e a t m e n t , p e r c e n t

FIG. 8. The concentration of certain nutrients in grass clippings in relation to different ogygen concentrations over the soil surface. Data on nutrients were from treatments in which grass was clipped at heights of 1 and 2 inches.

culture studies have shown. In soils, as Russell (1952) points out, uptake of nutrients is one of the most important physiological functions of living plants because it represents the connecting link between soil conditions and plant growth. An example of the effects of different environmental oxygen concentrations over the soil surface on the accumulation of K, N, Na, and P in grass blades are shown in Fig. 8. The grasses were clipped at two different heights. In general, N, P, and K concentration increased with increasing oxygen whereas Na was concentrated in the tops at lower soil-oxygen levels. Results from this grass study are in agreement with

268

L. H. STOLZY AND J. LETEY

similar studies on mineral concentration in the tops of different plant species due to various oxygen conditions (Letey et al., 1961a,b, 1962b, 196413). The one exception was Na in the study on barley (Letey et al., 1962b). The concentration of Na was considerably higher in barley than in the shoots of other plant species with very little effect of reduced soil O.D.R. Leggett and Stolzy (1961) showed that an anaerobic pretreatment of 1 hour activated sodium transport to barley tops for 15 hours after the plant was returned to aerobic conditions. The accumulation of cations by higher plants is considered to be an aerobic process. The study on sodium uptake indicated that two or more reactions are involved in the uptake of sodium by the root and shoots. The uptake of potassium and phosphate and possibly other ions differs from sodium accumulation with respect to anaerobic treatment. Potassium and phosphate accumulation is immediately suppressed by anaerobic conditions and returns to normal with a change from nitrogen treatment to air. The study by Cline and Erickson (1959) on O.D.R. and applied fertilizer in relation to chemical composition of pea plants showed that N, P, and K accumulation in tops at different O.D.R. was markedly affected by fertilizer treatments. The study by Letey et al. (1962a) indicated that 30 to 40 x loe8 g. cm.-2min.-1 was critical for the accumulation of K and P. Oxygen diffusion rates above 30 to 40 x showed only slight increases in concentration of these two nutrients. Sodium on the other hand increased as the O.D.R. decreased from 30 X lo-*. Labanauskas et al. (1964) studied the absorption and translocation of macro- and micronutrients by C i t w sinensis var. BESSIE at three different oxygen concentrations over the soil surface. Although O.D.R. values were not reported, the oxygen treatments of <0.3, 13, and 152 mm. Hg partial pressure had O.D.R. values of <20, 20 to 30, and > 60 X g. cm.-2 min.-l. In general, the total uptake by plants of the 12 minerals studied was decreased with a decrease in O.D.R. Translocation to the top of N, P, K, Ca, Mg, Fe, and B was increased whereas the translocation of Zn, Cu, and Mn to the tops was not significantly changed by the aeration treatment. Stolzy et al. (1963) in a study with Citrus sinensis var. HOMASSASA found that leaf concentrations of P, K, Ca, Mg, Fe, Mn, and B were reduced at diffusion rates below 33 x g. cm.-2 min.-' compared with leaf concentrations of plants grown in soil with O.D.R. above 62. Leaf concentrations of Cl were maximum at O.D.R. of less than 22 X lo-* and dropped to a minimum at O.D.R. of around 30 x Van Diest (1962) found a significant decrease in N of corn tops in

CHARACll?,RIZINGSOIL OXYGEN CONDITIONS

269

unfertilized plots at O.D.R. of 10 x cm,-2 min.-l as compared to O.D.R. of 30 and above. For certain minerals and plants species, there is substantial agreement on the fact that an increase or decrease in concentration of these minerals in the tops occurs with an increase in O.D.R. There are very few data on the micronutrients in relation to soil aeration and plant species. Another void in the literature is the comparison of top and root as to total nutrients taken up and translocated in respect to given aeration condition. Much more research is needed on the relation of plant nutrition and O.D.R. B. MICROORGANISMS Studies in the field of soil microbiology have dealt with many aspects of soil environment and response. To the authors' knowledge, only a few studies have been conducted where measurements of oxygen diffusion have been used to characterize the soil environment. 1. Nematodes Van Gundy et d. (1962) studied the influence of oxygen supply on survival of plant parasitic nematodes in soil. Oxygen diffusion rates in soils containing nematodes were changed by controlling the concentration of oxygen over the soil surface. There was a significant increase in the survival of Xiphinemu umericanum as O.D.R. increased from 4 to 41 x g. cm.-* min.-l. In general, X . umericanum was more sensitive than Tylenchulus semipenetrum to low O.D.R. The hatching of eggs of T . smipenetram was inhibited at 30 x Stolzy et a,?.(1960) in a study on oxygen tolerances of four plant parasitic nematodes took samples for nematode counts at 3, 7, and 10 days after oxygen treatment. Meloidogyne incognita, TricMorus christiei, and Tylenchulus semipenetrans were significantly reduced by the < 1% oxygen over the soil surface whereas no reduction occurred under the 2,7, and 21% oxygen treatments. The number of X . americunum was progressively reduced with decreasing oxygen treatment over the entire oxygen range. The effect of soil oxygen on the development of Melaidugyne javunjcu with tomato plants acting as host was reported by Van Gundy and Stolzy ( 1961). The lowest oxygen concentration over the soil surface that allowed development of the host and the nematode was 3.5%. Although not stated, this would be an O.D.R. in the range of 30 x g. c m . 3 min.-l. Below this level the plant root growth, size of developing females, and production of nematode eggs were reduced. Nematode activity as measured by the number of nematode galls on the roots of treated plants was sharply reduced at an O.D.R. below 40 X

270

L. H. STOLZY AND J. LETEY

Sweet orange seedlings (Citrus sinensis) were infected with the citrus nematode (Tylemhulus semipenetruns) 3 months prior to oxygen treatment over the soil surface (Stolzy et al., 1963). Oxygen diffusion rates between 24 and 33 x min.-l were critical for plant growth. g. There was very little reproduction of the citrus nematode below O.D.R. of 33 X The optimal reproduction occurred between 80 and 100 x g. min.-l. The host plant roots of the citrus nematode were more tolerant of lower soil oxygen conditions than were the nematodes

30-

OJ,

0.3

5 I5 LO9 Partial P r r a a u r s ( m m . H9)

;L 0 152

FIG. 9. Comparisons of citrus root weight and counts of Tylenchulus semipenetrans with different oxygen partial pressures over the soil surface (Stolzy et al., 1963).

in their reproductive and parasitic functions (Fig. 9). The citrus nematode population which has a generation cycle of from 42 to 56 days, was largely due to survival of larvae, production of eggs, and development of these eggs under conditions imposed by the treatments. Wallace (1958) concluded that nematode movement in sand is optimal at the soil suction where pores just drained. Since this is also the point where air diffuses back into the sand, increased oxygen could have been responsible for the increased nematode movement. Van Gundy and Stolzy (1964) used cellulose sponges in place of sand to study the effect of moisture and oxygen on nematode movement. Oxygen diffusion rates in the sponge at a given moisture suction were regulated by

CHARACXTRIZING SOIL OXYGEN CONDITIONS

271

changing the concentration of oxygen in the pore space. Movement of

Meloidogyne javanica in the sponge strips decreased as O.D.R. decreased. Movement of Hemicycliophora arenaria was inhibited only at the lowest oxygen diffusion rate. It appears from the studies on M . juvunicu larvae that a good supply of oxygen, as well as the proper moisture, was essential for optimal nematode movement. The effect of O.D.R. on movement may also explain the reduction in secondary invasion of M . javanica in tomato roots exposed to different oxygen concentrations (Van Gundy and Stolzy, 1961). In the case of H . arenuria, a nematode with a thick cuticle and heavy annules, there was little effect of moisture or oxygen on movement. Nematodes vary widely in their requirements for oxygen, a fact that helps to explain distribution, numbers, and parasitic functions in different soil environments. More studies on the relationship between soil oxygen and nematode activity are needed. The platinum microelectrode is useful to characterize the soil oxygen status in these studies. 2. Root-Rotting Fungi Oxygen conditions at various soil depths affect the location and distribution of Phytophthora purm'tica, P. citrophthoru, and Thielaviopsis basicoh. Studies by Stolzy and Klotz (1962) on citrus seedlings with three levels of soil aeration showed a reduction of two Phytophthoru spp. g. cm.r2min.-l, as compared when O.D.R. were less than 12,x to O.D.R. of 22 X Klotz et ul. (1963) studied oxygen requirement of the three root-rotting fungi previously mentioned. Spore germination and growth of T. bassicola had a higher oxygen requirement than the two Phytophthmu spp. This correlates with results of isolation of the fungi from several soil depths in two citrus orchards with different soil types. Thielauiopsk basicola and the two Phytophthora populations were determined at 3-inch depth increments down to 24 inches in 6 locations. Thielaviopsis bmkohz was recovered from 15 samples taken at 3- and cinch depths and from only 4 samples taken at depths greater than 6 inches. Of the 18 samples taken in the top 6 inches of soil, 9 yielded Phytophthora while of the 36 samples taken below the 6-inch level, 33 yield Phytophthora. Oxygen diffusion rates taken with depth showed a decrease with depth. A cultivation pan occurred at 8 to 10 inches. In a second orchard sampled in seven locations, 12 of 21 soil samples from the depth of 0 to 6 inches yielded T. basicoh while 15 of the 42 samples at depths of 9 to 24 inches yielded this fungus. Pythium was found in 60 out of 63 of these soil samples. Oxygen diffusion rates made at 4 locations with depth were higher at the deeper depths than at the surface in two of the locations.

272

L. H. STOLZY AND

J, LETEY

The use of the platinum microelectrode to evaluate soil oxygen conditions in pathological and microbiological studies seems feasible. Such studies would show a direct relationship between O.D.R. and response, as did roots. Other responses of plants, such as top growth, are much more difficult to correlate with a single oxygen diffusion value. V. Results of Field Measurements

The platinum microelectrode method of measuring soil oxygen conditions has several advantages over other methods that make it a valuable tool for field studies. One important feature is its portability as well as its inexpensiveness and simple electronic circuit. Erickson and Van Doren (1960) reported a number of field studies where the platinum microelectrode was used to measure oxygen availability to plants. In a study on tomatoes, O.D.R. was varied by types of tillage, compaction, and different amounts of irrigation water. Yield of tomatoes from individual plots were plotted against the average oxygen digusion rate measured 24 hours after irrigation. Measurements were taken between the 26th and 43rd days after transplanting. Yields beyond increased with increasing O.D.R. up to a value of 40 x which there was no increase. They reported a study with corn in which O.D.R. of 35 X did not seem to be deficient. The results on corn are similar to observations made by us in California. Oxygen diffusion to a depth rates of soils where beans were growing were 25 to 27 x of 30 cm. (Table I ) . The beans became very yellow following the first irrigation 6 weeeks after planting. Corn planted on this soil in previous years did very well. Several studies on sugar beets showed that they are sensitive to soil oxygen conditions as reported previously in this paper. Wiersma and Mortland (1953) in a greenhouse study showed that 20 to 30 x at a 4-inch soil depth was critical for sugar beets. Erickson and Van Doren (1960) reported the O.D.R. in the low thirties was deficient for maximum field sugar beet production. As much as 50 per cent reduction in yield occurred because of wet conditions at planting time. This reduction in yield was related to a sugar beet emergence study on two varieties. One variety (2,16X 216) increased in emergence with increasing O.D.R. up to 42 X lo-* while the other variety (30B 3-0) did not respond to increases in O.D.R. over 35 x g. cm.-2 min.-l. Oxygen diffusion rates in the low thirties would reduce emergence and thus yield. They also reported that there was a very heavy rain early in July of Yields that year 1957 which reduced diffusion rates to around 30 x were reduced by 50 to 60 per cent as compared to 1955. We investigated

CJ3ARACTEREZNG SOIL OXYGEN CONDITIONS

273

“sprangling” of sugar beet roots in a beet-growing area of California. The sandy loam soil had a hard pan at 18 to 20 inches. The field was well fertilized with a good stand of beets. After the beets reached a certain soil depth, well along in the growing season, the main tap root stopped growing and several other growing tips were initiated (called sprangling) with a subsequent reduction in yields. The O.D.R. in Table I under sugar beets would indicate that, because of the hard pan, irrigaTABLE I Oxygen Diffusion Measured in Field Soils under Different CroD Covers

Plant growing Broccoli Lettuce Beans

Hours after irriga- SUCSoil tion tion type 48 - Loam 24 - Silt loam 96 16 Loam

O.D.R. (g. x 108 cm.-2 min.-l) Depth in cm.

I

-

10 53 49 68 27

20 30 31 38 26 32 45 43 27 25

Sugar beets

2

-

Loam

58 60

Strawberries

-

4-8 7-11

Sandy loam

36 32 54 42

-

0

Clay loam

-

8

7 24

Sandy loam

64 45

Cotton

Citrus

- 2-4

9 11

Remarks 40 50 10 - Very good growth 38 26 Good growth 45 31 - - Chlorotic plants after irrigation do not fully recover 16 - - Sprangling of tap root 34 58 - Chlorosis early in 42 37 season ( March-April) Furrow Chlorotic after irRidge rigation. Recovered before next irrigation 39 43 - Rapid root growth

tion water stops moving downward, reduces O.D.R. rate, and stops the sugar beet tap root growth. Erickson and Van Doren (1960) reported on the cylic nature of O.D.R. results in field soils, They feel that a good measure of aeration would be the rate of increase in O.D.R. after irrigations or rains. Such a study was made where 4 inches of water were applied to corn plots under two different rotations. Both were very deficient in oxygen (O.D.R. of 19 to 22 )( one hour after the water was added. Plots which had been in alfalfa for 8 out of 20 years had an oxygen deficient (<40 x period of one day whereas plots with intensive rotation had 2.5 days of deficient oxygen. A greenhouse study by Letey et al. ( 1 9 6 2 ~ )found that for cotton,

L. H. STOLZY AND J, LETEY

274

sunflower, and green beans, low soil oxygen is most detrimental during early stages of growth following germination. Also an oxygen-deficient period causes a time lag in recovery of root growth and, because of reduced transpiration of plants under waterlogged conditions, plants cannot help to correct the condition. Wiersum ( 1960) compared oxygen diffusion measurements in fallow soil and a reed-covered plot to evaluate the water removal from virgin clay soil (Netherlands). The reed plants removed the water and improved O.D.R. down to 90cm. because of the heavy soil cracks that penetrated the deeper layers. Wheat seedling emergence was measured on soils of different textures, bulk density, and moisture (Hanks and Thorp, 1956). In this study, TABLE I1 Oxygen Diffusion Measured in Soils of Putting Greens Green appearance Good

Bad

Good, but getting worse

Depth (em.)

(g. x 108 cm.-2 min.-1)

O.D.R.

5 10 30 5 10 5 10 20 30

51.6 27.1 3.8 10.0 10.2 13.6 14.7 7.5 4.1

O.D.R. of about 75 to 100 x 10-8 g. cm.-2 min.-l was limiting to emergence of wheat. Erickson and Van Doren (1960) reported that O.D.R. of 24 x less prevented the emergence of potatoes. At O.D.R. of 26 to 35 X than 20 per cent of the potatoes emerged whereas O.D.R. above 61 X did not influence emergence. Investigations have been conducted on golf course putting greens. Letey (1961) reported on various factors that may cause trouble in management of the greens. Oxygen diffusion measurements taken on different greens with a visual evaluation of the condition are presented in Table 11. One can compare these data with data from controlled greenhouse experiments (Letey et al., 1964b) where root growth is limited at O.D.R. g. cm.-2 min.-l. The area that looked good was the only of 20 x one where O.D.R. was optimal at the shaIIow depth of 2 inches. The good area did not have optimal growth at 4 inches. Obviously since grass was there, roots must have made some growth. These measurements (Table I ) were taken the same day that irrigation water was applied,

CHARACTERIZING SOIL OXYGEN CONDITIONS

275

which was during the summer months. Later in the day or the next morning, O.D.R. could have been higher, at least in the very surface. Several other investigations by the authors on various crops are reported in Table I. The conditions in soil where broccoli grew well caused yellowing of bean leaves shortly after irrigation, but the soil conditions changed and permitted greening before the next irrigation. The same yellowing condition following irrigation is observed in certain cottongrowing areas of California. Lettuce and other crops grew well in the soiI where O.D.R.'s were taken (Table I). Citrus trees grown in soiIs contained in large cylinders very quickly grew roots to the bottom (42 inches) of 3-ft. tiles, sealed on the side and bottom. Strawberries grown in a sandy loam under field conditions became very chlorotic in these soils during the early part of the growing season when water use was low and soil temperature cool. With these O.D.R. values in soil one is better able to judge conditions that could contribute to plant responses. Poel (1960a,b, 1961) used O.D.R. to characterize soil conditions contributing to various types of plant communities in noncultivated areas of the British Isles. Plant communities were associated with O.D.R. in the following order: Pteridietum > Juncus a c u t i f l m d a r e x ~~~CCU-HOZCUS lanutus > Juncetum acutiflori > Juncetum conglomerati ( waterlogged). The range of O.D.R. was 5 to 25 x g. cm.-* min.-l. In another area, O.D.R. of soils in 13 plant communities on a hill grazing area were compared. Molinia and Nardus areas had the highest O.D.R. (17 x and occurred in elevated areas but seldom in marshes. Transitions from Pteridium aquiliniurn colonies (bracken ) which require good conditions to Juncetum mutiflori colonies which grow in less well drained soils were investigated. Oxygen diffusion rates in soils with Pteridium colonies were 17 x 10-8 and above, while Jun~etum acutijlori colonies were below 14 X lo-* g. min.-'. Poel's plant ecology studies in relation to O.D.R. form one speciaIized application of the platinum microelectrode method. VI. Summary

The classical theory of polarography has been modified to apply to the stationary solid electrode used to characterize the soil oxygen status in soils. The rate of oxygen diffusion to a platinum wire which could represent a plant root is the factor measured. The method requires further research for a complete understanding of the factors that could influence the measurement and interpretation of results, but sufficient

276

L. H. STOLZY AND J. LETEY

data have been obtained to indicate that the technique can successfully be used to measure soil oxygen conditions. The platinum microelectrode technique fails to properly characterize the soil oxygen conditions of relatively dry soils. This problem is most likely associated with failure of the electrode to be completely wet. The range of moisture over which the method can adequately be used depends upon soil texture and pore size relations. Since oxygen is deficient because of high moisture, measurements are usually desired when the soil is wet. The failure of the method at low soil moisture is not therefore a serious problem as long as the investigator is aware of this limitation. Temperature influences the oxygen diffusion rate. The results of measurements at various soil temperatures indicate that differences in O.D.R. are associated with the effect of temperature on oxygen solubility and diffusion coefficient. These factors would also influence the rate of diffusion to a root or organism. It is interesting to note that increased temperature results in increased respiration and also an increase in oxygen supply rate. Oxygen diffusion rates in relation to root growth for several plant species have been investigated. Substantial agreement exists between different investigators on ranges of O.D.R. that are deficient for root growth. Critical O.D.R. values in soils into which roots of several plant g. cm.-2 min.-l. Roots species will not grow is approximately 20 X A theorespond to differences in O.D.R. in the range of 20 to 30 x retical aproach to determine critical oxygen concentrations at the root surface agreed with critical values empirically determined by measurement and correlation with plant roots. Results of studies in which plant top responses are related to O.D.R. vary widely depending on the plant responses considered. Oxygen diffusion rates of 40 X and above can generally be considered to be optimal for vegetative growth. The O.D.R. under which plants fail to survive is much lower than these values. Studies where other plant functions such as flowering or fruit production were investigated, indicated that higher O.D.R. values are required for maximal production. Grasses have been found to be more tolerant to reduced oxygen conditions than the legumes. Alfalfa is more susceptible to low soil oxygen conditions than Ladino clover. Physiological changes in plant leaves due to short periods of critical oxygen were pronounced. Low O.D.R. for 8.5 hours caused 50 per cent decrease in COz fixation by tomato plants. Reduced oxygen for a 3-hour period shows the effects as much as 24 hours later in a reduction in the susceptibility of the plant leaves to ozone damage as well as a significant reduction in leaf carbohydrate. The effects of low O.D.R. under field

CHARA(;TERIZING SOIL OXYGEN CONDITIONS

277

conditions following rains or irrigations for limited periods of time may account for the level of plant production in certain soils. The nutrition of a plant is altered in several ways by O.D.R. of soils. min.-’ reduce the Oxygen diffusionrates below 30 to 40 x g. concentration of the more important macro- and micronutrients and cause increased concentration of certain undesirable minerals such as sodium. Manganese, however, increased in the tops of citrus plants with a decrease in O.D.R., while iron decreased. Nematode activity and survival were successfully correlated with oxygen diffusion rates in the soil. The O.D.R. critical for the 5 nematodes min.-l. The hatching of eggs of tested was around 30 x g. Tylenchulus smipenetram was inhibited at 30 x Studies of soil oxygen condition effects on Meloidegyne j m n i c a show that development of females, production of eggs, and numbers of galls formed are affected. Reproduction of citrus nematodes is minimal at O.D.R. of < 33 x lops g. cm. - 2 min. - l, while optimal reproduction occurred between 80 and 100 X Nematodes vary widely in their requirements for oxygen. This explains their distribution, number, and parasitic functions in many soils. Laboratory and field studies show a difference in the oxygen requirement of three root-rotting fungi. Phytophthora parasitica and P . citrophthora have lower oxygen demands for germination and growth than does Thielauiopsis basicoh. Oxygen diffusion rates of < 12 x in soils would tend to suppress the concentration of the two Phytophthora. The use of the platinum microelectrode method in measuring oxygen conditions in pathological studies is indicated. The usefulness of the platinum microelectrode to evaluate soil oxygen conditions in the field is evident from various studies with different crops. Oxygen diffusion measurements taken in the field and correlated with plant responses are generally in agreement with greenhouse data of this nature. Under wet soil conditions where diffusion rates are less than 20 X g. min.-l, narrow ranges of O.D.R. can determine the type of plant communities in well established noncultivated areas. A certain amount of empiricism exists in making oxygen diffusion measurements. The authors, therefore, strongly recommend a standardization of procedure so that the results of various investigators will be comparable. Application of 0.65 volt potential and allowance of 4 minutes are recommended for approximately steady state conditions. REFERENCES Barber, D. A., Ebert, M., and Evans, N. T. S. 1962. J . Exptl. Botany 13,397-403. Berry, S. L., and Norris, W. E. 1949. Biochim. Biophys. Acta 3, 593-599. Bertrand, A. R., and Kohnke, H. 1957. Sod Sci. SOC. Am. PTOC.21, 135-140.

278

L. H. STOLZY AND J. LETEY

Birkle, D. E., Letey, J., Stolzy, L. H., and Szuszkiewicz, T. E. 1964. Hilgardfu In press. Blinks, L. R., and Skow, R. K. 1938. Proc. Natl. Acad. Sci. U . S. pp. 24, 413, 420. Cannon, W. A. 1925. Carnegie Inst. Wash. Publ. 368, 1-168. Carlson, T. 1911. J. Am. Chem. SOC. 33, 1027-1032. Cline, R. A., and Erickson, A. E. 1959. Soil Sci. SOC. Am. Proc. 23, 333-335. Davies, P. W., and Brink, F. 1942. Reo. Sci. Instr. 13, 524-533. Erickson, A. E., and Van Doren, D. M. 1960. Trans. 7th Intern. Congr. Soil Sci. Madison, Wisconsin, 1960 Vol. 3, pp. 428-436. Finn, B. J., Bourgot, S. J., Nielsen, K. F., and Dow, B. K. 1961. Can. J . Soil Sci. 41, 16-23. Hanks, R. J., and Thorp, F. C. 1956. Soil. Sci. SOC. Am. Proc. 20, 307-310. Hanan, J. J., and Langhans, R. W. 1963a. N . Y . State Flowers Growers Bull. 213, 1-4. Hanan, J. J., and Langhans, R. W. 1963b. N . Y . State Flowers Growers Buff. 210, 3-6. Heyrovsky, J. 1922. Chem. Listy 16, 256. Hutchins, L. M. 1926. Plant Physiol. 1, 95-150. Jensen, G. 1960. Physiol. Plantarum 13, 822-830. Karsten, K. S. 1939. Am. J. Botany 26, 855-880. Klotz, L. J., Stolzy, L. H., and DeWolfe, T. A. 1963. Phytopathology 53, 302-305. Kolthoff, I. M., and Lingane, J. J. 1952. “Polarography,” 2nd ed., Vol. 1, pp. 3, 18-29; Vol. 2, pp. 552-558. Wiley (Interscience), New York. Labanauskas, C. K., Stolzy, L. H., Klotz, L. J., and DeWolfe, T. A. 1964. Soil Sci. SOC. Am. Proc. In press. Lammann, G., and Jessen, V. 1929. 2. Anorg. Allgem. Chem. 179, 125-144. Leggett, J. E., and Stolzy, L. H. 1961. Nature 192, 991-992. Lemon, E. R., and Erickson, A. E. 1952. Soil Sci. SOC. Am. Proc. 16, 160-163. Lemon, E. R., and Erickson, A. E. 1955. Soil. Sci. 79, 382-392. Lemon, E. R., and Kristensen, J. 1960. Trans. 7th Intern. Congr. Soil Sci. Madison, Wisconsin, 1960 Vol. 1, pp. 232-240. Letey, J. 1961. Calif. Turfgrass Culture 11, 17-21. Letey, J,, Stolzy, L. H., Blank, G. B., and Lunt, 0. R. 1961a. Soil Sci. 92, 314-321. Letey, J., Lunt, 0. R., Stolzy, L. H., and Szuszkiewicz, T. E. 1961b. Soil Sci. SOC. Am. Proc. 25, 183-186. Letey, J,, Stolzy, L. H., Valoras, N., and Szuszkiewicz, T. E. 1962a. Agron. 1. 54, 316-319. Letey, J., Stolzy, L. H., Valoras, N., and Szuszkiewicz, T. E. 1962b. Agron. J. 54, 538-540. Letey, J,, Stolzy, L. H., and Blank, G. B. 1962c. Agron. J. 54, 34-37. Letey, J., and Stolzy, L. H. 1964a. Hilgardfu In press. Letey, J., Stolzy, L. H., Lunt, 0. R., and Younger, V. B. 196413. Plant and Soil. 10, 143-145. Millington, R. J. 1955. Science 122, 1090. Phillips, R. E., and Kirkham, D. 1962. Agron. J . 54, 29-34. Poel, L. W. 1960a. J . Ecol. 48, 165-173. Poel, L. W. 1960b. 1. Ecol. 48, 733-736. Poel, L. W. 1961. J . Ecol. 49, 107-111. Raney, W. A, 1950. Soil Sci. SOC. Am. Proc. 14, 61-65. Russell, M. B. 1952. “Soil Physical Conditions and Plant Growth” (B. T. Shaw, ed.), Vol. 2, pp. 253-301. Academic Press, New York.

CHA'RACTERIZING SOIL OXYGEN CONDITIONS

279

Sawyer, D. T., and Interrante, L. V. 1961. J . Anal. Chem. 2, 310-327. Sawyer, D. T., George, R. S., and Rhodes, R. C. 1959. Electroanul. Chem. 31, 2-5. Stolzy, L. H., and Klotz, L. J. 1962. Unpublished data. Stolzy, L. H., Van Gundy, S. D., and Letey, J. 1960. Phytopathology 50, 656. Stolzy, L. H., Letey, J., Szuszkiewicz, T. E., and Lunt, 0. R. 1961a. Soil Sci. SOC. Am. Proc. 25, 463-467. Stolzy, L. H., Taylor, 0. C., Letey, J., and Szuszkiewicz, T. E. 1961b. Soil Sci. 91, 151-155. Stolzy, L. H., Van Gundy, S. D., Labanauskas, C. K., and Szuszkiewicz, T. E. 1963. Soil Sci. 96, 292-298. Stolzy, L. H., Taylor, 0. C., Dugger, W. M., and Mersereau, J. D. 1964. Soil Sci. SOC. Am. Proc. In press. Taylor, S. A. 1949. Soil Sci. SOC. Am. Proc. 14, 55-61. Van Diest, A. 1962. Agron. J . 54, 515-518. Van Gundy, S. D., and Stolzy, L. H. 1961. Science 134, 665-666. Van Gundy, S. D., and Stolzy, L. H. 1964. Nature 200, 1187-1189. Van Gundy, S. D., Stolzy, L. H., Szuszkiewicz, T. E., and Rackham, R. L. 1962. Phytopathology 52, 628-632. Wallace, H. R. 1958. Ann. Appl. B i d . 46, 74-85. Wiegand, C. L., and Lemon, E. R. 1958. Soil Sci. SOC. Am. Proc. 22, 216-221. Wiegand, C. L., and Lemon, E. R. 1963. Soil. Sci. SOC. Am. Proc. 27, 714-715. Wiersma, D., and Mortland, M. M. 1953. Soil Sci. 75, 355-360. Wiersum, L. K. 1960. Neth. J. Agr. Sci. 8, 245-252.

This Page Intentionally Left Blank

SOME PARAMETERS OF POPULATION VARIABILITY AND THEIR IMPLICATIONS IN PLANT BREEDING

. .

.

R W Allard and P. E Hansche University of California. Davis. California

I. Introduction ................................................ I1. The Genetics of Predominantly Self-pollinated Populations .......... A . Variability in Relation to Breeding ......................... B. Factors Affecting Genetic Variability ........................ C . Analyses of Marked Chromosome Segments .................. D . Analyses of Measurement Characters ........................ I11. The Exploitation of Exotic Variability .......................... A . The Species as a Gene Pool ............................... B. The Management of Hybrid Gene Pools ..................... C . Single Gene Cases ....................................... D . Multilocus Cases ......................................... E . Mass Reservoirs for the Exploitation of Exotic Variability . . . . . . IV . Variability within Agricultural Varieties ......................... A. Genetic Diversity and Stability ............................ B. Genetic Control of Buffering .............................. C . Individual Buffering ...................................... D . Populational Buffering .................................... E . Practical Utilization of Genetic Diversity .................... V . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References .................................................. 1

.

Page 281 282 282 284 285 296 302 302 303 305 307 312 313 313 316 316 318 320 323 324

Introduction

Plant breeders are generally committed to the idea that genotypes differ in “value. and it might be said that plant breeding revolves around continuing efforts to develop genotypes that have ever greater value for economic purposes . It is implicit in this idea that progress in plant breeding depends on variability because superior genotypes obviously cannot be selected from homogeneous populations . Also implicit in the above idea is the notion that there is some genotype that will have greatest value. and this has led to the notion that uniform popdations are an ideal to be sought whenever possible. It follows from these arguments that high genetic variability is a necessity in the breeding process itself ”

281

282

R. W.ALLARD AND P. E. H A N S C H E

but that homogeneity is a desiratum in the final product, the agricultural variety. The purposes of this review are to examine these arguments in the light of recent evidence concerning the genetic and environmental factors that influence variability in populations and to consider whether better use can be made of variability, both in the breeding process itself, and in the varieties to which successful breeding leads. Emphasis will be on self-pollinated species. II. The Genetics of Predominantly Self-Pollinated Populations

A. VARIABILITY IN RELATION TO BREEDING It is a postulate of population genetics that adaptation (fitness to a given environment ) and adaptability ( flexibility or capacity for change in fitness) are antagonistic. Success in improving adaptation requires that the population under selection be genetically variable. But high variability tends to be inadaptive because not all the genotypes in variable populations can be optimally fit and the presence of inferior genotypes is expected to reduce the immediate fitness of the population. Thus successful breeding, insofar as it achieves even higher adaptation, reduces genetic variability and also reduces capacity for change. Simmonds ( 1962) has presented evidence that rapid technological gains in advanced plant agricultures during the past century have increased adaptation at the expense of serious losses in adaptability. This evidence suggests that the materials of many plant breeding programs are too narrowly based genetically to provide for optimal rates of advance, i.e., that many breeders are dealing with essentially plateaued populations. Because the solution may be related to causes, it is worthwhile to consider briefly the reasons for this situation. The inadequate variability which is available appears to have arisen from a combination of factors related to the original sample of variability, accidents of sampling associated with small population size, and the intense selection which is commonly practiced in breeding programs. Most of our present-day crop varieties adapted to some particular geographical region are based largely on the materials that were available in the earliest days of cultivation in that region. These materials are known in many well-documented cases to have represented a limited and haphazard sample of the total range of variability of the species ( e.g., Harlan, 1957). Similarly, subsequent introductions that from time to time over the centuries may have enriched local gene pools with exotic genes must also have been haphazard samples because plant introduction as a deliberate and well-organized activity is only a recent development. On grounds of sampling alone it therefore seems unlikely that anything approaching all the potentially

SOME PARAMETERS OF POPULATION VARIABILITY

283

useful genes and gene combinations were ever available in any one place. It is still more unlikely that all useful genes have ever been assembled into any group of locally adapted stocks. There were probably two main consequences as breeding occurred within the locally adapted types in different geographical regions. First, there was no doubt steady advance in performance as selection sorted out the better-adapted segregants that occurred in the progeny of hybrids between locally adapted types. The stuff of which this progress was made was the recombinational potential within the limits of the genetic variability of the local sample. As this recombinational potential was exploited progress probably slowed periodically until the genetic base was again broadened by introduction of exotic types. Second, the very advances in adaptation that were achieved almost certainIy made it more and more difficult to broaden the base and provide the continuing supply of novel locally adapted types that are necessary for sustained progress. The reasons for ever increased difficulty in making use of exotic variability seem clear enough. When an adapted and unadapted variety differ by only a few genes, and particularly if there are some unfavorable linkages, the probability that transgressive segregants will occur in their hybrid progeny is small. Hence, crosses between adapted and unadapted types usually fail, sometimes no doubt because the ill-adapted parent has nothing to contribute, but probably more often because the small population sizes, intense selection for the general characteristics of familiar varieties, and the isolation between families associated with ordinary breeding programs, drastically reduce the chance of success. These are probably the main reasons why the large collections of genetic variability that are now available in germ plasm banks have for the most part been regarded as sources of “characters,” especially disease resistance, to be transferred to locally adapted backgrounds through rigorous selection, or by a series of backcrosses. Many such transfers have been performed, and they have made available to breeders a slow stream of “major” genes. However, standard breeding methods appear inadequate to explore the range of useful variability for complexly inherited characters because they apparently do not allow numbers and the repeated intercrossing necessary for the required recombinations to take place. At any rate the world collections of variability, which often contain thousands of entries, have not contributed nearly as much as might be hoped toward broadening and enriching the genetic base of locally adapted materials. The argument can be summarized and generalized as follows. First, the very successes of plant breeding within ecological areas have often been achieved at substantial cost to the materials of future change. Sec-

284

R. W. ALLARD AND P. E. HANSCHE

ond, although breeding from locally adapted strains has led to progress, there is a limit to achievement on this basis and in many cases progress now appears to have slowed down, or at least is lower than might be possible if the genetic base were broadened. Third, the large collections of types that have been assembled in germ plasm banks provide modern breeders with great opportunities for progress; the problem is to isolate desirable genes and gene combinations in usable form from these collections. Fourth, conventional breeding methods must be rejected for this purpose on the operational considerations that the combinations are too numerous, the number of progenies required are too large, and the probability of success is too small in any single hybrid to justify the effort and expense involved. Fifth and finally, some method is needed which will permit the variability which now goes to waste to be explored and exploited. In principle, so called “mass reservoirs” might provide the needed technique. Mass reservoirs are readily set up on a broad genetic base and are then maintained by mass-propagation methods, thus allowing very large populations to be handled with small effort and cost. However, such mass-propagated populations will give rise to a continuing supply of novel and useful variants only if two conditions are satisfied: fist, the population must constitute a dynamic recombination system over many generations; second, either survival in such populations must be positively correlated with agricultural value or the maintenance of agriculturally desirable types in the population by simple and inexpensive selection procedures must be possible. Whether these two conditions are satisfied depends on the numerical values assumed by the population parameters specifying such factors as mating system, selective coefficients of various genotypes, and other factors that influence amounts of genetic variability and the course of genetic change in populations. We shall therefore next consider factors that affect population structure and then consider the predictions that numerical estimates of these factors permit regarding the long-term recombinational potential of broadly based mass reservoirs and the possible utility of mass reservoirs as sources of locally adapted recombinants.

B. FACTORS AFFECTINGGENETICVARIABILITY Populations of predominantly self-fertilized species are generally considered to be highly uniform and hence to lack the genetic flexibility necessary to respond to long-term changes in environment. Yet a large proportion of flowering plants practice mixed selfing and outcrossing, and among these species many, both natural and agricultural, are both versatile and successful. The success of predominant self-fertilization as

SOME PARAMETERS OF POPULATION VARIABILITY

285

a mating system has commonly been attributed to the genetic uniformity which it presumably encourages in populations. Under selfing a population presumably consists entirely or largely of homozygotes and the effect of selection is postulated to favor adapted genotypes at the expense of the less adapted, leading to populations that consist of one or a few highly fit homozygotes. The favored individuals are expected to produce off spring genetically like themselves, and the population should therefore show high agreement with the optimal phenotype. Various aspects of evolution in self-pollinated species have been reviewed by Dobzhansky (1941), Stebbins (1950,1957), Darlington and Mather (1949), Grant (1958), Baker (1959), and Morley (1959). It is only recently that detailed studies of variability have been undertaken on populations of predominantly self-pollinated species and that precise estimates have been made under population conditions of parameters specifying such factors as reproductive methods, selective coefficients of various genotypes, and fluctuations in the values of parameters in various seasons and generations. The populations from which these estimates have been made have a variety of histories: some were derived from hybrids between two homozygous parents, others were synthesized by compositing hybrids representing many parents, and still others by mechanical mixing of two or more lines. In most cases the populations were maintained after synthesis without conscious selection, but in others selection was practiced regularly or sporadically for various characteristics. The most precise estimates of the relevant population parameters have come from analyses of genotypic frequencies of marked chromosome segments in experimental populations of barley and lima beans; however, additional evidence has been provided in these and other species by studies of changes in various measurement characters such as time of maturity, seed size, yield, and crossover percentages.

C. ANALYSESOF MARKEDCHROMOSOME SEGMENTS

1 . The Basic Model Estimation of genetic parameters depends on models in which observed changes are stated in terms of some function of the parameters that specify factors affecting genotypic frequencies. A population model developed by Hayman (1953) fits the experimental conditions under which the barley and lima bean populations we shall consider were maintained. In this model, genetic change is assumed to be solely a function of the parameters specifying the amount of selfing versus outcrossing and the relative viabilities of the three genotypes at a diallelic locus. Among other factors that might affect genotypic frequencies, migration, mutation,

286

R. W. ALLARD AND P. E. HANSCHE

and accidents of sampling appear to be most important. The experimental conditions make it reasonable to assume that neither mutation nor migration would have much effect on genotypic frequencies in these populations. Random drift is also likely to be unimportant in populations of the size studied ( several thousands of individuals per generation). Thus, Hayman's model appears to take into account the main factors which might affect genotypic frequencies. The recursions relating genotypic frequencies in two successive generations ( n and n 1 ) in terms of mating system and selection parameters are

+

AA Aa

fln+l

oc

wl{ [s(fl"+;f2")]

WP

fZ"+I

+t[fln+$f2q2}

{y1 + 2t [f + -21 [ + y1 } 8 fi"

fin]

In

f3n

f2"]

( 1)

where fdn and are frequencies of genotypes AA, Aa, and aa in generations n and n 1, and wl, w2,and w3 are the respective selective values of the genotypes. The proportion of selfing and random outcrossing are denoted by s and t , respectively (s t = 1). The proportionalities (1)are equalities if w1= wz = ws. If the selective values are not equal, these proportionalities can be made equalities by dividing the sum of terms to the right of the proportionality sign for each genotype by the sum of the right-hand terms for the three genotypes. If w2 is set equal to unity the proportionalities ( 1 ) involve three unknowns, whereas census data for any two consecutive generations provide only two degrees of freedom. Census data alone therefore do not permit simultaneous estimation of wl, w, and t and, consequently, an independent estimate of one of these parameters is a prerequisite to estimating the others. Because independent and accurate estimates of outcrossing are easier than direct estimates of selective values, t has usually been estimated first in actual experiments. We shall therefore consider a method of estimating t and then turn to methods by which this estimate can be utilized to determine selective values.

+

+

2. Estimation of Outcrossing

The amount of outcrossing that occurs under population conditions can be estimated as follows (Allard and Workman, 1963). A sample of plants of recessive phenotype is drawn from the population and progeny grown from these plants under conditions allowing maximum survival.

287

SOME PARAMETERS OF POPULATION VARIABILITY

The progeny are scored for numbers of individuals of dominant ( A a ) and recessive phenotypes (aa)and, since individuals of dominant phenotype can arise only through outcrossing, such data provide an estimate of the proportion of A pollen grains in the pool of pollen grains. This is, however, an underestimate of the amount of outcrossing since outcrosses due to a pollen grains are not detected. If, however, census data are available, gene frequency can be computed for the generation from which the sample was drawn and an appropriate correction applied. Thus t can be estimated as t =H/p (2) where H is the proportion of Aa individuals among the progeny of the sample of au individuals drawn from the harvest in generation n and p is the frequency of A in that generation. The standard error (S.E.) of this estimate of t is given by

S.E.t =

[($,

ZH(1-H)

H

+@ (

N1

I'

p(1-p) NP

)I

* (3)

where N1 is the size of the census sample from which the popuIation gene frequency was estimated in generation n, and N z is the total number of progeny (aa Aa) classified.

+

3. Estimation of Selective Values When an independent estimate of t is available, s and t can be treated as constants in the proportionalities (I), and maximum likelihood estimators of the selective values can be derived (Allard and Workman, 1963). Denoting the observed proportions of AA, Aa, and aa in the SUCcessive generations n and n 1 by P, R, Q and 01,0 2 and 0 3 , respectively, and setting 202 = 1, these estimators are

+

[i +

01 201

=

SR

0 2 [ s (P

0 3 w3

2t (P

+

+

+

R)

[ ~ s R 2t (P

R) (Q

+

R

)]

+ t (F' + f R ) ' ]

+

R ) (Q

+

(4) R)]

= O2[(Q+:R)

+t(Q+;R)I]

The variances of w1 and w3 and the covariance of can be and taken from the information matrix of expected values of second partial

R. W. ALLARD AND P. E. HANSCHE

288

derivatives of the log likelihood expressions. The methods used to extend these results to cases involving two or more loci are formally identical to those used in the single locus case (Workman and Allard, 1962; Allard and Workman, 1963). Because the results would require considerable space to reproduce they will not be given here. Once the population reaches equilibrium, i.e., genotypic frequencies no longer change from generation to generation, estimates of the fitness values can also be obtained in two other ways. The first method involves solving the proportionalities ( 1 ) for w1 and w3; this can be done because values are now available for all the other variables. In the second method we note that for a population in which the proportion of outcrossing is t, the equilibrium inbreeding coefficient is F = (1- t ) / (1 t ) . The expected equilibrium genotypic frequencies ( AA:Aa:aa) are then (1- F ) q 2 Fq:2(1 - F ) q ( l - q ) : ( 1 - F ) (1- q ) 2 F ( 1- q ) (5) and fitness values can be estimated from the ratio of numbers observed to numbers expected (Haldane, 1956). This method is strictly accurate only when there is no selection, and it becomes quite inaccurate when selection is intense. Before considering precise estimates of selective values, it will be worthwhile to examine general patterns of change in genotypic frequencies for some typical one locus cases (Fig. 1 ) . Note that genotypic frequencies changed rapidly in the first 3 or 4 generations, and that the population point in each of the populations thereafter fluctuated aimlessly about apparent equilibrium points. The patterns of change €or D / d and S / s in populations 20 and 75 suggest that the chromosome segments marked by D and S were adaptively superior to the alternative segments marked by d and s, whereas the behavior of S/s in population 65 suggests near equality in fitness of the two homozygotes. A substantial and fairly constant proportion of heterozygotes occurred in later generations for these three marked segments, as well as for several other segments for which the population point is plotted in this figure for only the last available generation. Similar results have been observed in various other populations in lima beans (Allard and Hansche, in press), in experimental populations of barley into the nineteenth generation (Jain and Allard, 1960), and in natural populations of wild oats (Imam and Allard, unpublished). The rapidity with which marked segments have approached apparent equilibrium points indicates the forces operating on the segments are very strong. The fluctuations observed during the approach to equilibrium, and also about the equilibrium points, suggest further that the forces influencing genotypic frequencies vary from season to season. Precise numerical estimates of the relevant population

+

+

+

SOME PARAMETERS OF POPULATION VARIABILITY

289

parameters are therefore of particular interest for the light they may throw on the future genetic composition of these populations. Estimates of mean selective values for some representative single ‘locus” cases and for some two locus cases appear in Table 11. The values in these tables indicate that heterozygotes nearly always contribute more offspring to the next generation than homozygotes. For example, in lima bean population 75 (Table I and Fig. 1) the S / s genotype left 100 offspring on the average to 53 contributed by S/S and 35 by s/s. From 0

\ POPULATION 20

0 0

I ?

FIG. 1. Trilinear diagram showing observed frequencies of homozygotes ( ala,, a2a2) and heterozygotes (a,a,) for the S/s, D/s, and S / s loci in three lima bean populations. Note that populations 20 and 65 were of hybrid origin while population 75 was synthesized by mixing equal proportions of two homozygotes. Triangles show population points in the tenth to fifteenth generations for additional marker genes in other lima bean populations. The percentage of outcrossing in these lima bean populations was approximately 5 per cent on the average. (After Allard and Hansche, in press.)

290

R. W.ALLARD AND P. E. HANSCHE

Table I1 it can be seen that single heterozygotes were consistently more fit than homozygotes; further, double heterozygotes consistently had a selective advantage over single heterozygotes.

4. Fluctuations in Selective Values

It is well known that genotypes do not necessarily behave alike in different environments. Consequently it is expected that the selective values of various genotypes in a population will vary relative to one an-

aIOl

FIG.2. Computer simulations of the populations of Fig. 1. In the simulations it was assumed that 5 five per cent outcrossing occurred ( a = 0.95), that mean selective values were as given in Table I, and that season-to-season fluctuations in selective values were normally distributed about mean values with am = 0.10. Population size was assumed to be 200 per generation (actual population size was 3000 or larger). A single typical computer run is given for each population. (After Allard and Hansche, in press.)

291

SOME PARAMETERS OF POPULATION VARIABILITY

other in different years and may also vary in response to changes in genotypic constitution of the populations over generations. Quantitative evidence on the magnitude of such effects is given in Table 111. It is apparent from these results that the selective values do in fact fluctuate widely TABLE I Estimated Selective Values of the Two Homozygotes Relative to the Heterozygote for Marked Segments of Chromosome in Lima Bean and Barley Populations Locus and population Lima beana s / s (75) D / d (20) S/s (65) Barleyb B/b (C.C.V) s/s (C.C.V) G/g (C.C.V) E/e (C.C.V) Bl/bl (C.C.V) R / T (C.C.V) B t l b t (C.C.V) Sh/sh (C.C.V) a 2,

Mean percentage of outcrossing

w1

w2

4 5 3

0.53 0.58 0.66

1.00 1.00 1.00

0.35 0.47 0.66

2 2 2 2 2 2 2 2

1.06 0.81 1.04 0.47 0.61 0.82 0.96 0.71

1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00

1.31 0.96 0.82 0.59 0.54 0.68 1.06 0.63

Selective values

After Allard and Hansche (in press). After Jain and Allard (1960). TABLE I1 Estimated Selective Values for Three Pairs of Marker Loci in Lima Bean Population 54a Marker loci

Genotype

Cc/Dd

Vv/Dd

PPIRT

AABB ( w l ) AaBB ( w 2 ) aaBB ( w 3 ) AABb ( w 4 ) AaBb ( w 5 ) aaBb ( w , )

0.595 0.811 0.535 0.784 1.000 0.724 0.577 0.794 0.504

0.663 0.821 0.663 0.842 1.000 0.841 0.644 0.803 0.644

0.386 0.593 0.299 0.568 1.000 0.591 0.293 0.606 0.401

AAbb

(w7)

Aabb ( w y ) aabb ( w , ) a

After Allard and Hansche (in press)

from year to year. For example, in population 53, the value of the genotype S S varied over a period of 10 years from 0.30 to 1.36 and that of ss varied from 0.36 to 1.42. Averaged over an three populations the mean selective value of the homozygotes was 0.82 and the standard error of

TABLE I11 Year-to-Year Fluctuations in Selective Values of the Genotypes SS and ss (Ss z 1) in Three Lima Bean Populationsa

Population

53

w1 w3

59

w1

65

1951

1952

1933

1954

1955

1956

1957

1958

1959

1960

Mean

Standard errors about mean

0.92 0.97

0.72 0.71

1.36 1.00

0.52 0.58

0.50 0.68

1.25 1.42

0.33 0.36

0.87 0.88

0.88 0.87

0.30 1.08

0.76 0.86

0.34 0.28

-

-

w3

-

-

2.30 0.95

0.81 0.85

0.52 0.74

0.81 0.76

0.52 0.58

0.87 0.75

0.30 0.92

1.58 0.22

0.97 0.72

0.62 0.21

w1 w3

-

1.51 1.44

0.93 1.00

0.58 0.79

0.72 0.65

0.80 0.79

0.52 0.50

1.13 1.00

0.59 0.54

0.59 0.47

0.82 0.80

0.31 0.31

0.85 0.80

0.42 0.27

Mean W1 WI a

After Allard and Workman ( 1963).

3

B +

3

? M

SOME PAFWMETERS OF POPULATION VARIABILITY

293

the variations about this mean was 0.35 In other words, in 1 year out of 3, selective values are expected to fall outside the range 0.82 & 0.35 (0.47-1.17) and in 1 year out of 20 they are expected to fall outside the range 0.82 _t 0.70, i.e., below 0.12 or above 1.52. Part of this variability in selective values may be due to changes in the genetic composition of the unit itself, or to changes in the composite population genotype. However, the sharp reversals in viabilities that occurred in successive generations, and the lack of obvious trends in relative viabilities over generations, point to year- to-year differences in environment as the major contributor to fluctuations in selective values. In this connection it should be emphasized that these lima bean populations were grown under irrigation so that available moisture was relatively constant from year to year. Competition from other species (weeds, pests), population density, and other factors of the environment were held highly constant. In addition the summer climate at Davis, California, where the populations were grown, appears to be much the same from year to year, so it seems likely that these populations were exposed to more uniform conditions than is usual in field plantings. The observed fluctuations in adaptive values, even though large, are more likely to be underestimates than overestimates of fluctuations in selective values in most crop species. These results demonstrate the dangers of estimating the mean values of population parameters from a limited number of seasons or generations, and they suggest that parameters of variability in themselves may be important parameters of population change. An interesting feature of the selective values in Table I11 was the tendency for w1 and w3 to vary together in the same year. When seed yields were high, indicating that the environment was favorable, homozygotes and heterozygotes tended to contribute more or less equal numbers of offspring to the next generation. However, when seed yields were low, as in 1954, 1955, and 1957, w1 and to3 tended to be much smaller than wz. Apparently the advantage of heterozygotes over homozygotes is particularly associated with stress environments.

5. Mean Population Fitness If it is assumed that fitness values, wc such as those given in Tables I and I1 are independent of genotypic frequencies, f4, the average fitness of a population for the locus or loci concerned can be computed as

w

= ZfiWi. (6) The mean fitness of a population so defined will obviously vary with genotypic constitution. It will be low if poor genotypes are frequent, high when the most fit genotypes predominate, and it can be shown (Wright,

294

R. W. ALLARD AND P. E. HANSCHE

1942, 1949; Li, 1955; Lewontin, 1958b; Lewontin and White, 1960) that gene and/or genotypic frequencies will change so as to maximize mean population fitness. W c a n be put in the form of a fitness curve for single locus cases and, when two loci are considered simultaneously, it can be put in the form of an adaptive surface of topography in which each horizontal dimension represents gene frequency for one locus and vertical distance above the base surface represents mean fitness. Although only one adaptive peak is possible for one locus, three are possible with two loci (Moran, 1963), and with many loci an indefinite number of peaks can exist. Thus an “adaptive landscape” is formed with peaks, valleys, ridges, and saddles, and in any generation the population can be represented by a point corresponding to the gene frequencies. The change in gene and genotypic frequencies will be such that the population point will progress up the slope of local influence until it comes to rest at the nearest summit, At this time Aqk Af$, and 6w/6qc and 6w/6fi will all be zero, i.e., the population will be in equilibrium. Thus populations might come to rest at equilibrium genotypic frequencies that do not correspond to the maximum fitness for the entire topography. Wright (1963) has discussed conditions under which populations might escape from such equilibrium points and progress to higher adaptive peaks. When fitness curves were computed for single-locus cases such as those given in Table I, it was found that the population point in all cases moved to or near the adaptive maximum predicted from the selective values of the three genotypes within 3 or 4 generations (Fig. 2 ) . Thereafter the population point fluctuated from generation to generation, but the population point and the adaptive peak were seldom far apart (Allard and Hansche, in press). In studies of numerous two-locus cases, Allard and Hansche (in press) found that adaptive landscapes in which there was a single prominent peak were most common. Interestingly, observed population points corresponded closely to the adaptive peaks in these cases. Further, in computer simulations based on stochastic models in which the selective values were made random variables, the trajectory of the population point closely followed observed changes in the populations. When discrepancies occurred between actual populations and the simulations the discrepancies were usually in directions which suggested ( a ) that the degree of heterozygote advantage and/or the amount of outcrossing had been slightly underestimated or ( b ) that selective values were not entirely independent of genotypic frequencies. In the latter case any frequency dependency must have been such that the disadvantage of the inferior homozygote became less as its frequency decreased in the population.

295

SOME PARAMETERS OF POPULATION VARIABILITY

Adaptive landscapes representing cases in which 2 unlinked loci interacted with one another were also common. The topography for the P/pR/r loci (Table 11, Fig. 3 ) illustrates such a case. There is a ridge of high fitness running from the upper left to the lower right-hand corner of the topography. This ridge is the result of the relatively high fitness values of the PPRR and pprr homozygotes. On either side of the ridge

0

1

2

.3

.4

5

.6

.7

8

9

1.0

FREQUENCYOFR

FIG.3. Adaptive topography for the P l p and R / r chromosome segments in a lima bean population. Computations were based on the mean fitness values given in Table 11. The triangle represents the observed population point and the trajectory represents a typical 50-generation computer iun. (After Allard and Hansche, in press.)

the surface falls away steeply because of the low fitness values of the P P w and p p R R homozygotes. In most generations the estimated fitness values were such (Table 11) that the middle of the ridge was slightly higher than the ends. However, in some generations the estimated selective values of heterozygous genotypes were slightly lower than the values given in Table 11. In such generations the middle of the ridge was lower than the ends and the low adaptive peak was replaced by a shallow saddle. The population point in the actual population, and in all computer runs based on the mean selective values given in Table 11,

296

R. W. ALLARD AND P. E. HANSCHE

were coincident with the position of the peak. However, computer runs based on the slightly different estimates of selective values which led to the saddle configuration frequently approached fixation by generation 50. Apparently this population is in precarious balance respecting equilibrium for the P/p-R/r segments such that slight decrease in either the amount of outcrossing or in the advantage of heterozygotes might lead to fixation. The nature of the interaction between these marked segments therefore appears to jeopardize maintenance of heterozygosity for these segments because the fitness values of P / p and R / r considered separately appear to take numerical values that are expected to permit stable nontrivial equilibria. Genetic interrelationships are clearly complex with respect to the effect on fitness of marked chromosomal segments.

6. Conclusions from Analyses of Marked Chromosome Segments Analyses of the behavior of marked chromosome segments indicate that the genetic composition of predominantly self-pollinated populations approximates to what one expects if both the amount of outcrossing that occurs and the selective values of the various genotypes in the population are taken into account. The high proportion of cases in which the heterozygote has striking selective advantage over both homozygotes deserves special emphasis. Other factors, which appear to have lesser effects on the structure of such populations, are year-to-year fluctuations in selective values and in amounts of outcrossing, interactions between loci, and gene-frequency dependent selection. Thus, even though inbreeding is a powerful force in determining the course of genetic change in predominantly self-pollinating populations, it is by no means the only important factor affecting the structure of such populations, as has so commonly been supposed.

D. ANALYSES OF MEASUREMENT CHAF~ACXERS The results of the previous section indicate that considerable heterozygosity of marked chromosome segments is maintained even in populations where the level of selfing exceeds 95 per cent. In this section we shall consider whether variability in quantitative characters also is greater in such populations than consideration of mating system alone might predict. 1. General Patterm of Change The literature on variability in predominantly self-pollinated species includes studies of many different kinds of populations in various species (e.g., Adair and Jones, 1946; Atkins, 1953; Harlan and Martini, 1938; Laude and Swanson, 1943; Suneson, 1949, 1956; Allard and Jain, 1962; Allard and Workman, 1963; Akemine and Kikuchi, 1958). Because the

297

SOME PARAMETERS OF POPULATION VARIABILITY

pattern that emerges is highy consistent it will suffice for the present purpose to give a single example drawn from extensive studies of populations of rice by Akemine and Kikuchi (1958). One of the crosses studied was between zumo and NOREN 20, which are, respectively, the earliest and latest varieties grown in Japan. Bulk populations were grown from F2 to FE generations at 20 rice experiment stations scattered through-

41 NOREN20

c EN 20

21

F6

SAPPORO

63.

FUJISAKA

1

I

KONOSU 36' HIRATSUKA 35.

CnlKUGO

33' MlYAZAKl 31 a

'*

lLrl l-

56 70 64 98 112

6

U

68 82 96 110 124 138

L

54 68 82 9

DAYS TO HEADING

FIG.4. Histograms showing changes in a hybrid rice population grown at various latitudes in Japan. The histograms represent the distribution of heading dates of plants grown in central Japan (Hiratsuka) from random seeds taken from each population. (After Akemine and Kikuchi, 1958.) out Japan. Each year random samples of seed were drawn from these bulk populations, grown at Hiratsuka in central Japan, and measured for various characters. The effect of natural selection differed considerably for the various locations (Fig. 4). The plants from seed grown in northern locations were generally early whereas those from southern locations showed the reverse tendency. The amount of variation also differed with location. It was very large for centrally located stations, and, although smaller for southern and northern stations, it remained

298

R. W. ALLARD AND P. E. HANSCHE

much larger than for the parents. Akemine and Kikuchi also noted strong correlations between heading time and other characters. Thus plants from populations maintained in northern locations not only were early, but they tillered profusely and had short culms and compact panicles; the opposite was the case for plants from southern locations.

2. Between- and Within-Family Variability The genetic basis of such variability was analyzed by Allard and Jain (1962) in their studies of various metrical characters in Composite

-I F.4

F8

FIZ

F16

F20

GENERATION

FIG.5. Means and total phenotypic variability for heading time computed from measurements of individual random plants drawn from barley Composite Cross V in various generations. (Adapted from Allard and Jain, 1962.)

Cross V, one of Suneson’s barley populations. This population had been synthesized from intercrosses among 31 varieties and thereafter carried without conscious selection. Results for heading time, a typical metrical character, are illustrated in Figs. 5-7. In generations F3 to FIB, the mean heading time of the population shifted steadily in the direction of earliness (Fig. 5 ) , indicating that this trait was gradually adjusted through directional selection to fit the environmental circumstances. Frequency distributions for various generations showed that there was steady elimination of individuals from both tails of the curve, an indication that stabilizing selection also occurred for this character. The effect

299

SOME PARAMETERS OF POPULATION VARIABILITY

of the combined directional and stabilizing selection was a gradual change in mean and decrease in variance, but it can be seen from Fig. 5 that the population remained highly variable into the Fls generation. To determine the basis of these changes in mean and variance, AlIard and Jain grew progenies derived from a random sample of plants drawn from the population in various generations. Estimates of between- and within-family variances for these families, and for the 31 original homo-

mt

PARENTS

RANDOM SAMPLE FROM POPULATION

F,

Fa

F1z

F16

F20

GENERATION

FIG.8. Mean between-family variability for heading time for the 31 homozygous parents of barley Composite Cross V, and for progenies of random plants drawn from the population in various generations. (Adapted from Allard and Jain, 1962.)

zygous parents of the population, are given in graphical form in Fig. 6 and 7 . The between-family variance was high for the original parents; in the population itself the between-family variance was high in early generations, and it decreased steadily with increase in generation. The distribution of family means indicated that, as the population became subdivided into families under continued inbreeding, there was selection against the more extreme families. However, the between-family variance remained high in the F1s generation, a fact which, as Allard and Jain remarked, provides quantitative support for their observation that a

300

R. W. ALLARD AND P. E. HANSCHE

vast number of different genotypes remained in the population after 18 generations of exposure to natural selection. It can be seen from Fig. 7 that within-family variance decreased steadily from F3 to Fle but remained larger in F19 than in the original homozygous parents. This excess of variability over that of the parents

3.7

-

3.1

-

3.0

-

5i 2.9

-

*4

ew

=

2.0

-

z0

2.7

-

2.6

-

2.5

-

2.4

-

2.3

-

F 4

RANDOM SAMPLE FROM POPULATION

PARENTS

1

F4

F8

Fl2

FI6

F20

GENERATION

FIG.7. Mean within-family variability for heading time for the 31 homozygous parents of barley Composite Cross V and for progenies of random plants drawn from the population in various generations. (Adapted from Allard and Jain, 1982.)

was explained on the basis of segregation resulting from heterozygosity of genes governing heading time. Another possibility is that selection favored genotypes which, owing to poor buffering, were more variable than equally homozygous but well-buffered genotypes. No direct evidence on this point is available for this particular barley population. However, studies of another population (Jana, unpublished) indicate that selection tends to favor well-buff ered genotypes, providing indirect support for the explanation based on heterozygosity.

SOME PARAMETERS OF POPULATION VARIABILITY

301

3. Changes in Fitness and Yield

The experiment with Composite Cross V also provided information regarding changes in fitness of the population, measured as number of seeds produced per plant in various generations. Mean values showed steady increase in fitness from F3 to- Flo. Frequency distributions (Fig. 8 ) indicate that this increased fitness was achieved mostly by elimina-

FIG.8. Frequency histograms for number of seeds produced by progenies of random F, and F,, plants drawn from barley Composite Cross V in various generations. (Adapted from Allard and Jain, 1962.)

tion of inferior genotypes. However it is significant that some genotypes of very high fitness occurred in later generations. The character of greatest interest from the standpoint of the plant breeder is, of course, yield itself. Observed changes in yield of several representative populations, all of which were maintained without conscious selection, are given in Fig. 9. In early generations yields were conspicuously inferior to those of standard locally adapted varieties. This is not surprising because each of the populations was based on a conglomerate of adapted and unadapted parents. The fact that yields improved rapidly in all cases provides evidence that natural selection is a powerful force in eliminating unadapted genotypes.

302

R. W.ALLARD AND P. E. HANSCHE

Suneson (1956) has used the characteristics of individual lines selected in different generations as a gauge of the changes which have been induced by natural selection in barley populations. Among 356 selections made in the F12 generation of Composite Cross I1 none combined generally good agronomic type with ability to outyield a standard locally adapted variety in replicated yield trials. In the F20 generation, when the mean performance of the population was superior to that of the standard variety, two among 50 lines were considered to be outstanding, Sixty-six selections were taken from the F24 generation. Like the selections made in the F20 generation, all were good in yielding ability, at least acceptable in agronomic type, and a few lines were judged exceptional. Suneson concluded that superior types make up a greater and greater proportion of the population as the number of generations increases.

4. Conclusions Regarding Measurement Characters Studies of measurement characters lead to much the same conclusions as studies of marked chromosome segments, namely, that the structure of predominantly self-pollinated species is far from static. Instead populations of even such heavily inbreeding species as barley (99 per cent selfed ) appear to have a “coadapted” population genotype. Individuals of these populations share a common gene pool in a manner differing only in degree from individual members of full-fledged random mating Mendelian populations. The recombinational system appears to be adequate to permit the formation of new and original variants for an indefinitely large number of generations. The basis of change appears to be the entire pool of genes of the population and the pattern of change is such that the populations come to be made up of an ever increasing proportion of superior types as selection tests the steady stream of novel genotypes formed by continuous reassortment of the genes of the original base materials. 111. The Exploitation of Exotic Variability

A. THESPECIESAS

A

GENEPOOL

The hereditary materials of an economic species as a whole can be regarded as a gigantic pool of genes, and the task of the plant breeder can be regarded as that of assembling from this pool those gene combinations that will give optimal performance in his particular environment. The breeding structure of an economic species is such that there is an almost continuous compartmentation of this species gene pool into

SOME PARAMETERS OF POPULATION VARIABILITY

303

smaller and smaller units with varying degrees of isolation from one another. At the base of the pyramid are local breeding stocks. We have argued earlier that such local populations frequently represent a narrow sample of genes and that breeding would often benefit if the wealth of gene combinations in these local gene pools could be enriched by introgression from other similar gene pools. We have also argued that standard breeding methods are inadequate to this task and that mass reservoirs may be useful in exploring and exploiting variability that now goes to waste. We shall now examine these arguments in greater detail, taking into account the parameters of genetic variability discussed in the previous section.

B. THE MANAGEMENT OF Hmm GENEPOOLS When attempts are made to exploit exotic variability, a standard procedure is to hybridize an adapted variety with an introduced variety to establish a hybrid gene pool. The population is then divided into a large number of families and selection is practiced between and within families for a number of generations. The best surviving family is then crossed with a sibling family or with a surviving family from a similar hybrid or with a second locally adapted variety to start the next cycle. Many variations of this procedure are possible depending on the duration and the intensity of selection between and within families, the frequency of intercrossing, the basis of the selection (e.g., phenotype of the selected individuals versus performance in test crosses ) , and so forth. Another option is to cross one or a few locally adapted types to a number of exotic types and combine the hybrids into a single gene pool. This pool can then be separated into a large number of families and selection practiced between and within families in one of the manners listed in the previous paragraph. Alternatively, the hybrid gene pool can be managed as a mass reservoir. Since only the simplest types of artificial selection are practiced (or none at all) the mass reservoir technique generally allows much larger populations to be handled than pedigree procedures. With the pedigree procedures the size of individual families must of necessity be small if large numbers of families are to be accommodated. Thus families of size 10 to 50 individuals are common in many crop species whereas population sizes of several hundreds or thousands of individuals are usually possible with mass reservoirs. In a hybrid gene pool formed by crossing many exotic types to an adapted variety, gene frequency for desirable alleles might vary from 0 to 1. It would obviously be 0 if the most desirable allele in the species

304

R. W. ALLARD AND P. E. HANSCHE.

is not represented in either the adapted or unadapted parents involved, and it might be 1for alleles generally important to fitness in the species. But in many cases the most desirable alleles at a locus would be represented in only one or a few of the exotic parents, in which case its 120

110

100

90

n J

w>

80

-------

70

POPULATION 43 (LIMA BEANS)

---

COMPOSITE CROSS 5 (BARLEY)

---a-

POPULATION 22 (LIMA BEANS) COMPOSITE CROSS 2 (BARLEY)

60

50

I

I

I

I

I

I

F4

Fa

F,,

Fl6

F20

FZ,

I F28

GENERATION

FIG.9. Yield in various generations in four representative populations. Yield of each population is expressed as percentage of the yield of a standard adapted commercial variety. (After Allard and Hansche, in press.)

frequency would be low in the hybrid pool. The problem is to find the most efficient method of managing the hybrid gene pool so as to maximize the probability of obtaining optimal gene combinations. We shall examine this problem first in terms of the fate of single genes in the hybrid pool and then turn attention to multilocus cases.

SOME PARAMETERS OF POPULATION VARIABILITY

305

C. SINGLEGENECASES Many desirable alleles are expected to be in low frequency in hybrid gene pooh of the type described above; to illustrate their fate, we shall concentrate on alleles the original frequency of which is assumed to be 0.1. If effective selection cannot be practiced for the ( + )allele ( w1 =

RANDOM MATING (51% LOST) 95% SELFING (65% LOST)

,1

.2

.3

.4

.5

.6

.7

.8

.9

1.0

ALLELEFREQUENCY

FIG.10. Graph demonstrating the drift of allele frequencies in the fifth generation among populations of size N = 10 ( w l = wg = w3 = l, initial allele frequency = 0.1 ), Dispersion among random-mating families was determined using generation matrix techniques. Dispersion among 95 per cent selfing populations reflects the results of 300 Monte Carlo simulations.

= w3) in populations of size 10, the most likely result is rapid fixation of the more frequent ( - ) allele (Figs. 10 and 11).For example, with random mating, fixation of the ( - ) allele will have occurred in approximately one-half of families by the fifth generation, in 68 per cent of families by the tenth generation, and in 90 per cent of families when the dispersive process has run its full course. Assumption of 95 per cent selfing only slightly increases the rate of dispersion. Figure 12 illustrates that moderate selection for the ( +) allele ( w1 = 1.0, w2 = 1, w3 = 0.9) w.2

R. W. ALLARD AND P. E. HANSCHE

306

alters this result very little. Clearly the effect of mating system and moderate selection on the determination of gene frequencies within any line is small relative to the effect of random drift in such small populations. Thus it is likely that desirable alleles in low frequency will be lost in small populations unless very high selection pressures are maintained. .22 .20

.ia Z

.16

w

2. .14 w K IL

z .12

0

2 1

3

2

.10

.oa .06

RANDOM MATING (68% LOST)

.04

95% SELFING(78% LOST)

.02 .I

.z

.3 .4 .5 .6 .7 ALLELE FREQUENCY

.a

.9

1.0

FIG. 11. Graph demonstrating random dispersion of allele frequencies in the tenth generation among populations of size N = 10 ( w1 = w2 = w3 = 1, initial allele frequencies = 0.1 ) . Dispersion was determined as in Fig. 10.

Figures 13 and 14 make it clear that mating system and selection pressure increase in importance relative to random drift when population size in increased. When N is increased to 20 (Fig. 13), any one of the three factors can be the major determinant, depending on the values of the mating system or selection parameters. When N is increased to 50, drift is relegated to secondary importance in the determination of genotypic frequencies within any single family for most values of the other parameters. Nevertheless, even in populations of this size, drift can be an important factor in determining the probability of increase of a (+) allele within a single family if selection pressures are small and initial gene frequencies are low.

307

SOME PARAMETERS OF POPULATION VARIABILITY

These results can be given a multilocus interpretation for loci that are independently inherited. The probability of fixing N desirable alleles within any one population is the product of the probabilities of fixing the ( +) allele at each locus. Since the individual probabilities are low for infrequent alleles in small populations, it follows that drift can

z

0

2 .10 -I

3 & .08-

RANDOM MATING (41% LOST)

n

95% SELFING (59% LOST)

.06 -

.02-

.04

.I

.2

.3

.4

.5

.6

.7

.8

.9

0

ALLELE FREQUENCY

FIG. 12. Graph demonstrating the effect of random dispersion of allele frequencies in the fifth generation among populations of size N = 10 ( w l = 1.0,w2 = 1.0, wg = 0.9, initial allele frequency = 0.1). Dispersion was determined as in Fig. 10.

seriously reduce the probability of incorporating several plus alleles into any single breeding line.

D. MULTILOCUS CASES To obtain more precise information about multilocus cases, we have written a computer program that takes into account the following: ( a ) initial composition of the hybrid gene pool; ( b ) amount of selfing versus random outcrossing; ( c ) population size; ( d ) selective values at each

308

R. W. ALLARD AND P. E. HANS-

locus; ( e ) linkage between loci; and ( f ) random environmental effects on mating systems and fitness values. This program was used to simulate genetic change in various hypothetical populations in which the relevant population parameters were given values suggested by the analyses of actual data reviewed earlier. Samples of the results are given in Figs. 15 and 16.

17% LOST)

LOST)

.02

t-uJhLL .I

.2

.3

.4

.5

.7

.b

.a

.9

1.0

ALLELE FREQUENCY

FIG. 13. Graph demonstrating the effect of random dispersion of allele frequencies in the fifth generation among populations of size N = 20 ( w l = 1.0,w2 = 1.0, w2 = 0.9, initial allele frequency = 0.1). Dispersion was determined as in Fig. 10.

We have assumed that the hybrid gene pool was made up by compositing hybrids between a locally adapted type and 10 generally unadapted exotic types. The genotype of the adapted type was assumed to be

1010101010 1010101010 and the genotypes of the ten unadapted types

1ooooo0ooo o1oooooooo 1ooooo0ooo’ 0100000000’ *

OOOOOOOOO1 a

*

OooOOOOOOl

.2b .24 .22

-

.I

.2

.3

.4

.5

.b

.7

.0

.9

0

ALLELE FREQUENCY FIG. 14. Graph demonstrating the effect of random dispersion of allele frequencies in the fifth generation among populations of size N = SO ( w1 = 1.0, w, = 1.0, w3 = 0.9, initial allele frequencies = 0.1). Dispersion was determined as in Fig. 10.

unadapted parent ) . Population size was assumed to be constant at either 10 or 500 and the probability of outcrossing was made 5 per cent or were assigned random mating was assumed. Fitness values ( w1:wz:w3) to represent moderate directional selection without heretozygote advantage ( 1:1:O.g) or moderate directional selection with heterozygote advantage (0.85:1:0.80). The probability of crossing over between the loci was set at 0.50 (independent segregation) or 0.10 (moderately tight linkage).

310

R. W. ALLARD AND P. E. HANSCHE

For populations of size N = 10 the results followed the same general pattern almost irrespective of specifications regarding mating system, selective values, and linkage. This pattern is illustrated in Fig. 15, from which it is clear that separation of a hybrid gene pool into a large number of small families limits the opportunity for improvement within any single family. The procedure of separating hybrid gene pools into small units has at least three major disadvantages in combining desirable genes

.1

1

0

2

1

4

6

1

8

10

1

12

14

1

16

18

1

20

22

1

24

26

1

28

30

~

32

34

~

36

GENERATION

FIG. 15. Graph demonstrating the effect of random dispersion of fitness values among five populations of size 10 (95 per cent s e k g ; w1 = 1.0, w2 = 1.0, w 3 = 0.9; probability of crossing over = 0.5). Initial allele frequency at 5 loci = 0.55 and at the other 5 loci = 0.05. Note that only one line remains unfixed at generation 30. See text.

which are dispersed among many different genotypes. First, it places emphasis on family selection, which is both expensive and laborious, at a time when mean fitness of the population is low and individual selection should be adequate to distinguish between the better and the poorer genotypes in the population. Second, some of the potential of the population may be lost owing to drift in small families, even though intensive seIection is practiced. Third, frequent cycIes of intercrossing between families are necessary to provide the variability required for sustained progress. This requirement of frequent cycles of intercrossing limits the amount of material that can be carried and hence the potential for progress.

~

311

SOME PARAMETERS OF POPULATION VARIABILITY

The change in mean fitness of populations of size 500, but with specifications otherwise the same as above, is illustrated for generations 1 to 36 in Fig. 16. For populations of this size steady increase in fitness occurred irrespective of specifications regarding mating system and link1.oo

W

1 Rondom Mating, w 1 = w 2 = 1, W J = .9, Crossing Over = .5 2 Rondorn Moting. w 1 = w 2 = 1, w 3 = .9, Crossing Over = . 1 3 95% Selling, w 1 = 3 5 , w 2 = 1 , W, = .80, Crossing Over = . I 4 95% Selling, w , = .85,w 2 = 1 . w 3 = .80,Crossing Over I .5 5 95% Selfing, w 1 = w 2 = 1 , w 3 = .3, Crossing Over = .5

.35 .30 25

6 9 5 1 Selfing,

w1

=

w2

= 1,

w3

= 9, Crossing Over = .I

.20 .IS

.1I

I

I

2

4

I 6

I

I

I

I

8

10

12

14

I 16

I 18

I

I

20

22

I 24

I 26

I 28

I 30

I

32

I 34

GENERATION

FIG. 16. Graph demonstrating the effect of recombination and selection on fitness of population of size 500. The steady increase in fitness demonstrates that selection and recombination have continued and that few if any favorable alleles have been lost due to drift.

age. These hypothetical cases are certainly much less complex than real populations and it seems reasonable to expect that differences between small and large populations would become even greater as genetic complexity increases. Consequently these results point to the mass-reservoir technique as an effective method of managing broadly based hybrid

36

312

R. W. ALLARD AND P. E. HANSCHE

gene pools with minimum danger of losing infrequent desirable alleles due to drift.

E. MASSRESERVOIRS FOR THE EXPLOITATION OF EXOTIC VARIABILITY It should be emphasized at the outset that mass reservoirs are not a substitute for conventional methods of breeding. Instead they should be regarded as an additional breeding tool especially suited to exploring and exploiting populations of broad genetic base. Their primary function is not to develop new highly adapted varieties directly suitable for use in commercial production, but to provide a continuing supply of reasonably adapted genotypes that can serve as parents in conventional breeding programs. The key features which make mass reservoirs particularly suited to this purpose are: (1) the ease and simplicity with which they can be synthesized, together with the low cost with which they allow many thousands of individuals to be handled per generation; ( 2 ) the opportunity that such populations provide for effective selection and continuing recombination. In many species mass reservoirs can be synthesized by simple physical mixture of the parental strains. However, when the level of outcrossing is very low it will usually expedite the formation of an effective recombinational system to synthesize the population from interstrain hybrids. In such cases various stratagems are available to lessen the labor of making hybrids between large numbers of parents. For example, recessive marker genes can be used to identify natural hybrids; the progeny of such hybrids are then bulked to initiate a mass reservoir. Male-sterility genes are also useful for this purpose, as demonstrated by Suneson and Wiebe (1962) in developing a mass reservoir involving 6200 barley strains. Once a mass reservoir has been synthesized it can usually be handled thereafter by mass propagation methods, thereby keeping costs low. The usefulness of mass reservoirs in exploring and exploiting a wide range of variability rests ultimately on the opportunity they provide for continued recombination and effective selection. It has generally been assumed that the breeding plans appropriate to mass reservoirs are closely dependent on the mating system of the species. Thus, while a physical mixture of strains has been deemed adequate for outbreeders, and a single round of crossing for often cross-pollinated species such as cotton and sorghum, polyallele or cyclical crosses have been recommended for heavy inbreeders to prolong the period of recombination. An interesting implication of the results reviewed here, and the results of Allard and Hansche (in press), is the doubt they throw on the necessity for laborious intercrosses for the maintenance of a dynamic recombinational

SOME PARAMETERS OF POPULATION VARIABILITY

313

system, even in species such as barley in which outcrossing rarely exceeds 1per cent. This is an important factor in a system where low cost and low labor requirements are major factors. However, should it be desirable to impose additional outcrossing on the population, male steriles or recessive marker genes can be used to lessen labor and expense. For example, in certain lima bean populations additional rounds of crossing have been accomplished by selecting large numbers of recessives for various major genes and reconstituting the population from F1 hybrids between these individuals and dominant individuals with which they had outcrossed in the previous generation. Some breeders have expressed concern that many of the survivors in mass populations are not the best types when grown in pure stands. It is true that numerous odd and peculiar variants of little or no agricultural value occur in late generations in mass populations and that many among the more or less normal variants have deficiencies which would preclude their use as commercial varieties, This is, however, unimportant relative to the main issue, namely that an ever increasing number of types with superior agronomic characteristics and yield appear as selection tests the constant supply of new variants which appear in mass populations. It is clear from studies of broadly based populations of both barley and lima beans that enough agriculturally superior types appear within a few generations to make such populations rich sources of parents to be used in conventional breeding schemes. Since, in fact, numerous successful pure line varieties have been selected directly from mass populations (examples in Harlan, 1956), it also appears that ability to survive in competition is not necessarily inimical to performance in pure stand. Further, should it become apparent at any time that aggressive but agriculturally unacceptable types are becoming predominant, it is usually possible to devise inexpensive artificial selection procedures to guide the population in the desired direction. We therefore conclude that the mass reservoir technique is particularly suited among breeding methods for surveying the unrealized potential of the large world collections of germ plasm which are now available for all major crop plants. The method is clearly unsuitable for many species but should be especially valuable in grains and similar seed crops. IV. Variability within Agricultural Varieties

A. GENETICDIVERSITY AND STABILITY Thus far attention has been restricted to the basis and methods of exploiting variability in the development phases of breeding programs during which the breeder attempts to select superior genotypes from

314

R. W. ALLARD AND P. E. HANSCHE

genetically variable populations. In this section we hope to show that variability can sometimes be better utilized than it has in the past in the end product itself, the agricultural variety. In 1950 Frankel stated: “From the early days of plant breeding, uniformity has been sought after with great determination. For this there are many reasons-technical, commercial, historical, psychological, aesthetic.” He added that the concept of purity “has not only been carried to unnecessary lengths but that it may be inimical to the attainment of highest production” since it is “concerned with characters which are readily seen but often of little significance.” In the following discussion we assume that the technical demands for uniformity are so compelling in some crops that no relaxation of standards of uniformity is possible. We also assume that, while a large-scale increase in variability is unwarranted and unnecessary in other crops, given acceptable uniformity for basic agronomic and commercial characteristics, there is no obvious reason why varieties heterogeneous in other respects should not be grown provided the mean yield benefits from doing so. The importance of stability in yield has been emphasized by Finlay and Wilkinson (in press), who concluded from a study of 277 barley varieties grown in different environments that ability to produce high mean yields depends on broad adaptation. They defined an ideal variety as one that combines maximal potential in the best environment with maximal stability. Stated in another way an “ideal” variety shows low genotype-environment interaction through consistent high performance. In the discussion to follow we assume that high performance and stability of performance under fluctuating environmental conditions are the mark of desirable varieties. In this connection it is important to distinguish among the environmental causes of genotype-environmental interactions. Variations in environment can be divided into two general types, variation that is predictable and variation that cannot be predicted. The first category includes permanent characters of the environment such as general features of the climate and soil. It also indudes those aspects of environment associated with agricultural practice which can be fixed more or less at will, such as planting time, fertilizer practice, previous cropping history, and so forth. The second category includes fluctuations in rainfall, temperature, and other aspects of weather, fluctuations in prevalence of diseases and pests, and so forth. It may also include variations in agricultural practice which are not held constant on inefficient farms. The distinction between these two categories is not always clear cut, and the characteristics included will vary from place to place and from crop to crop. Nevertheless, the qualifications that can be applied should not be allowed to mask the essential distinction between the categories. This is

SOME PARAMETERS OF POPULATION VAFlIABILrrY

315

because they have different implications regarding operational procedures during the breeding program itself on the one hand and on the sort of genetic system which is the goal of the breeding program on the other hand. Predictable or fixed differences related to conspicuously different environments, such as those associated with oceanic versus continental climate, are usually difficult neither to recognize nor to appreciate. Such differences often lead to sharp reversals in the performance of varieties which are reflected in different mean yields in the contrasting environments and in large variety x location interactions in interregional yield trials. Similarly large interactions between varieties and treatments ( e.g., sowing dates, fertility levels) indicate that the treatments induce special environments. Many times, however, similar interactions occur when varieties are tested over seemingly homogeneous areas where agronomic practices are apparently uniform. Despite inability to identify the basis of the interaction the conclusion must be that the region includes a number of different and special environments whenever such interaction occurs. The solution is obvious: the breeding program should allow for the development of a number of varieties, each particularly adapted to one of the special environments. Such a course of action is usually feasible because there seems to be no end to the variability available for the development of plants adapted to specific situations. The practical daculties are also obvious. Although the chances of identifying genotypes adapted to special environments increase as the number of test locations is increased, the difficulties of testing selection materials in early generations at many locations are formidable. There is need for better methods of exploring the potential of variable populations during the developmental stages of breeding programs, and it is possible that the mass methods described earlier may be useful for the purpose of producing genotypes with specific adaptation. The implications of variety X year interactions in field performance trials are different from variety x location or variety x treatment interactions because the former are associated with environmental circumstances that cannot be predicted in advance. The breeder can hardly aim his program at developing varieties suited to special circumstances he cannot foresee, nor, if he had a series of varieties with the requisite special adaptations, would there usually be a way of knowing which variety to recommend prior to sowing. This is an important problem in plant breeding since, in variety trials, it is common for interactions containing variety x year terms to be large. Again the solution is obvious: what is needed are varieties adapted to withstand transient fluctuations of environment, i.e., well buffered varieties able so to adjust their life

316

R. W. ALLARD AND P. E. HANSCHE

processes as to maintain high productivity under the entire range of circumstances to which they are likely to be exposed. This raises a series of questions: Is such stability under genetic control? Are some types of genetic systems more likely to be stable than others? If so, can ways be devised to exploit such genetic systems to practical ends? These are key questions in plant breeding which deserve detailed examination.

B. GENETICCONTROL OF BUFFERING The stability with which we are concerned does not imply general constancy of phenotype over a range of environmental conditions. It implies constancy only with respect to those aspects of phenotype relevant to fitness for agricultural purposes, especially yield, although quality of product, and characters such as time of maturity which influence field operations may also be included. Such stability may in fact require that certain aspects of physiology and morphology vary widely if the “fitness” characters are to be held reasonably constant. Varieties that can adjust their phenotypic or genotypic state in response to environmental stimulus in ways to give near maximal economic return for the place and year can be called “well buffered.” This term is therefore equivalent to “homeostatic” in the sense of Lewontin (1958a) but is preferred owing to the controversy in the literature regarding the meaning of the latter term. There are two obvious ways by which a variety can achieve stability. First, the variety can be made up of a number of different genotypes each adapted to a somewhat different range of environments. Second, the individuals themselves can be well buffered so that each member of the population is capable of adjusting its phenotype to the requirements of the particular environment encountered. Populations which are highly homogeneous, such as clonal, pure-line and single-cross varieties, must obviously depend heavily on “individual buffering” to stabilize productivity. However, both “individual buffering” and “populational buffering” are open to heterogeneous populations. C. INDIVIDUAL BUFFERING There is a great variety of evidence that buffering is a property of specific genotypes. Thus every cereal breeder can cite varieties that perform reasonably well under favorable as well as unfavorable conditions, and can also cite varieties the performance of which is erratic. The same is true for other inbreeding crop species and also for clonal varieties in field and horticultural crops. Indeed the entire machinery of varietal testing, which provides for trials repeated in locations and years, is geared to identify those genotypes that are able to produce high mean

SOME PARAMETERS OF POPULATION VARIABILITY

317

yields through reliability of performance. The situation seems to be much the same among homozygous inbred lines in outcrossing species. For example, Shank and Adams (1960) found differences in buffering ability among inbred lines of maize, showing that here also buffering is a feature of specific genotypes. There is also a great deal of work which indicates that high buffering is conspicuously a feature of heterozygotes. Dobzhansky and Levene ( 1955) found in Drosophila pseudoobscura that individuals homozygous for second chromosomes were less well buffered respecting viability than comparable heterozygotes in environments representing food and temperature differences. They concluded that “such homozygotes do quite well in a restricted range of environments, but they lack the resilience necessary to maintain their fitness in other environments. By contrast, the heterozygotes are more often many-sided and versatile in their adaptiveness; hence they are able to live successfully in a broader range of environments.” In a general summary of work with animals, Lerner ( 1954) reached the conclusion that “adaptedness, the attribute of individuals to be fit in the Darwinian sense to their immediate environment, is mediated by heterozygous advantage in buffering ability.” It is commonly accepted by plant breeders that much the same situation prevails in outbreeding plants. For example, the good individual buffering of certain clonal varieties is often attributed to high heterozygosity and that of F, hybrids relative to inbred lines is almost always associated with the heterozygous state. Experiments with maize such as those of Shank and Adams (1960) and Rowe and Andrews (1963) provide quantitative data on the extent of this buffering. Once allowance is made for some regression of variance on means, these experiments make it clear that inbreds as a group are more variable over environments than hybrids as a group, showing that good buffering is a feature of heterozygosity. Additional evidence that individual buffering is associated with heterozygosity comes from the work of Clausen and Hiesey (1958) and Hiesey (1963), who compared the growth of individual races of Mimulus and Potentilk with the growth of F1 and F.2 interracial hybrids at three altitudes in California. They found that the races themselves survived only in their own environments whereas F1 hybrids were as vigorous as each parent in its optimal environment. In the Fz generat’ion some individuals occurred which equaled or even surpassed the F1 hybrids. This constitutes remarkable buffering because the three altitudinal stations have climates that range from Mediterranean to subalpine. Understanding of the part that heterozygosity plays in buffering in inbreeding species is only now beginning to emerge. It is difficult to ob-

318

R. W. ALLARD AND P. E. HANSCHE

tain large numbers of hybrids in most such species and, as a result, experiments have for the most part compared parents and F1 hybrids in only a single environment and at spacings wider than ordinary for commercial plantings. Also the characters studied have usually been morphological features whose relationship to fitness is not clear. There is, however, some recent evidence that bears on buffering for fitness. Some good evidence comes from the work of Langridge and Griffing (1959) and Griffing and Langridge (1963) on Arubidopsis t h a l i ~ n ~ . When different homozygous ecotypes were compared with hybrids between ecotypes over a range of temperatures it was found that hybrids exceeded parents in mean growth. The greater mean growth resulted partly from superiority of hybrids over parents in the lower and medium temperatures, but especially from their superiority at the highest temperatures. They postulated that the homozygotes cease growth due to deficiencies in particular enzymes differentiating different homozygous ecotypes and that the greater stability of the heterozygotes is a result of the combination in hybrids of alleles which produce different thermostable products. Additional evidence comes from studies of homozygous lines of lima beans and their F1 hybrids grown under field conditions (unpublished data). Observations over many years indicate that parents and F1 hybrids differ little in number of seeds produced when the environment is favorable, as evidenced by high seed yields. But in unfavorable years the F1 hybrids may yield twice as much as their better homozygous parent. Further evidence in lima beans comes from estimates of the selective values of the two homozygotes and the heterozygote for “marker genes” under population conditions. In studies conducted over ten successive years it was found that homozygotes and heterozygotes tended to contribute more or less equal numbers of progeny to the next generation when seed yields were high. But in poor years the heterozygotes sometimes contributed more than twice as many offspring to the next generation as corresponding homozygotes (Allard and Workman, 1963). These results thus tend in the same direction as those of Griffing and Langridge : under optimal conditions homozygotes and heterozygotes differ little in fitness, but under stress conditions the advantage of heterozygotes is often striking.

D. POPULATIONAL BUFFERING Populational buffering arises in interactions among different cohabiting genotypes within populations; i.e., it is buffering above and beyond the individual buffering of specific constituents of the population, Populational buffering has been widely recognized in natural species

SOME PARAMETERS OF POPULATION VARIABILITY

319

and discussed under a variety of terminologies (review in Lewontin, 1958a). The most precise information on populational buffering in crop plants comes from comparisons between pure line varieties grown singly and in mixture. Simmonds (1962) reviewed the literature on this topic and found that stability in yield is in fact commonly associated with mixtures. For example, in wheat coefficients of variability over years were about two-thirds as large for mixtures (7.3 per cent) as for homogenous populations (11.6 per cent). There was some suggestion that greater stability was not a general phenomenon but was characteristic of particular combinations. In mean yielding ability the mixtures held an average advantage of the order of 3 to 5 per cent over the means of their components, but some mixtures outyielded the highest component. The mixtures surveyed were compounded more or less randomly from good local varieties whose “combining ability” was unknown. It would be extraordinarily interesting to know what magnitudes of gain in yield and stability mi& result from mixtures deliberately compounded of components known to “nick” well. Information on populational buffering in heterozygous materials comes from both outbreeding and inbreeding species. In maize Sprague and Federer (1951) found that both variety x location and variety X year interactions were smaller in yield trials for double crosses than for single crosses, indicating that double crosses are more stable in productivity. Jones (1958) also compared the stability of single and double crosses in maize in an extensive survey of yield trials and found that coefficients of variability were smaller for double crosses (12.3 per cent) than for single crosses (21.4 per cent). Jones attributed this stability of double crosses to populational buffering and suggested that it is stability that allows double crosses to make high yields averaged over many seasons, even though the highest yield in any one place and year is likely to be obtained from some particular single cross. The work of Finlay on barley and Allard on lima beans shows that advanced generation hybrid populations in self-pollinated species are often highly buffered. Finlay (in press) tested 10 barley varieties and their 45 F.2hybrids in the variable environment of South Australia and found that the F2 populations not only outyielded their parents substantially, but were markedly superior in stability of productivity. Finlay emphasized that much of the advantage of the Fz populations resulted from their good performance in poor environments. In lima beans Allard (1961) found that three unselected F7 populations outyielded their parents by 7 per cent as an average over 16 environments. The F7 populations achieved superiority through steady good performance while the parental varieties, although highly efficient in some environments,

320

R. W. ALLARJI AND P. E. HANSCHE

were inefficient in others, The extent of individual versus populational buffering cannot be assessed in either of these studies because the populations studied were heterozygous to some extent as well as genetically heterogeneous. E. PRACTICAL UTILIZATION OF GENETICDIVERSITY There seems to be little doubt that genetic diversity, whether it occurs within genotypes ( heterozygosity ) , or between coexisting genotypes, often improves ability to prosper over a range of environmentaI conditions. This brings up the question whether individual and population buffering can in combination be put to practical use. The genetic diversity associated with heterotic combinations of different alleles in heterozygous individuals has been widely recognized and exploited in outbreeding species, Experiments such as the ones reviewed earlier, together with commercial experience with F1hybrid v?rieties (especially in the sorghums) suggest that the individual buffering commonly associated with heterosis has substantial contributions to make in increasing and stabilizing yield in self-pollinating species as well. Current intensive activity directed toward development of the technology required for economical production of F1 hybrid seed in various self-pollinated species (e.g., in wheat) provides further evidence of the growing recognition of the value of genetic diversity. The other aspect of genetic diversity, i.e., populational buffering, is much less widely recognized, and there have been few conscious attempts to exploit its possible advantages in increasing and stabilizing performance. Thus, although populational buffering is widely exploited in maize through the production of heterogeneous double-cross hybrids, double crosses were originally adopted not for their consistency in performance, but to overcome the handicaps that single crosses have in seed production. Whether further practical use can be made of populational buffering depends on the biological and agricultural properties and economic feasibility of the various types of populations that might exploit such variability. In cross-pollinated species a number of types of population in addition to double crosses seem feasible. These include deliberately compounded mixtures of single crosses, mixtures of double crosses, and synthetic varieties. Unfortunately, there is little evidence on the performance of such populations, except some suggestion that synthetic varieties of maize may be steadier in production on the margins of the corn belt than single crosses or double crosses. In self-pollinated species there are also a number of possibilities. Evidence was presented earlier that random mixtures of good local pure-

SOME PARAMETERS OF POPULATION VARIABILITY

321

line varieties frequently led to increased stability and that certain combinations “nicked to produce yields higher than the better component in blends consisting of equal proportions of two varieties. The question of varying the proportions of components and of the potential of multiline blends apparently remain entirely open. The value of deliberately compounded blends of two or more F1 hybrids is a closely related question. In this case evidence is available from the studies of Kramer and Loden (1963) on blends of forage sorghum hybrids. In the great plains of the United States the management of forage sorghums varies widely from farm to farm and year to year in response to fluctuations in weather and agricultural exigencies. Kramer and Loden found that certain blends of hybrids yielded nearly half again more than the best hybrid under a range of management practices designed to simulate those often used in commercial practice. One of the blends is now grown commercially on approximately 500,000 acres. Still another possibility in self-pollinated crops is the utilization of advanced generation hybrid populations as commercial varieties. The examples which follow illustrate the sort of thing that can be done. The first example comes from breeding experience with ACALA 4-42, the variety which has been grown since 1949 in the one-variety cotton district of the San Joaquin Valley of California (Turner, 1963). This variety is resynthesized annually from 3 to 10 “families,” among which no single family is outstanding in all important traits. These families are maintained by essentially a pedigree method program. Each year elite plants from the better progenies within each family are chosen and their progeny are tested the following year. Data from these tests are used to determine which progenies and the proportion of each to include in family tests at several locations in the following year. The urgency of altering specific traits in the variety to meet agricultural and industrial requirements is taken into account in determining the proportion of seed of each family to use in making up the appropriate composite to be increased for commercial production. This increase is carried out over a three-year period, and since outcrossing is of the order of 5 to 30 per cent, the final planting seed represents a “synthetic variety” consisting of a blend of the component families and their hybrids and hybrid products. Miravalle et al. (1962) studied the composite after various generations of exposure to natural hybridization and found that yield and quality were not influenced by extra generations of multiplication. It was concluded that “populational buffering” was more important that heterosis in this breeding and seed increase program. The second example comes from the lima bean populations mentioned earlier ( Allard, 1961). Three unselected F7-F9 populations devel-

322

R. W. ALLARD AND P. E. HANSCHE

oped from hybrids between the three best commercial varieties crossed in pairs were conspicuously more stable in performance and 7 per cent higher yielding than the parents when tested in 16 different environments ( 4 years at 4 locations). These populations generally did not make the highest yields in any one environment, but they were only slightly less likely to make exceptional yields than the best pure-line varieties. Significantly, these populations had no “obvious deficiencies either in agronomic characteristics or quality factors that would rule them out as commercia1 varieties, even in their unselected state.” Another example in lima beans is provided by the work of Sanchez and Tucker (unpublished) on a population synthesized from hybrids among 4 rather similar pure-line varieties in 1950 and carried thereafter without conscious selection. In the Flo generation one hundred plants were selected from which 11 derived lines were subsequently retained for further testing on the basis of standard yield and quality trials. Two features stood out in comparisons among the unselected F13-F14 population, the standard commercial variety (VENTURA), and the 11 selected lines grown in pure stand and in ll-way mixture. First, the unselected population and the ll-way mixture gave higher mean yields than the commercial variety or any of the 11lines grown singly. Second, the mixed populations were more stable in yield than the standard variety or the lines grown in pure stand. It has been argued that use of F1 hybrids cannot be justified in self-pollinated species on the basis of lack of a consistent pattern of environmental stability for pure lines versus F1 hybrids. It was noted earlier, however, that most evidence concerning stability in self-pollinated crops has been with morphological characters whose relationship to yield is not clear cut. It has also been argued that commercial utilization of heterosis in self-pollinated species can be justified only until homozygous lines which equal or surpass the F1 are isolated by inbreeding and selection. The difficulty is that when yield or general desirability are the primary concern, the isolation of such pure lines may not be easy. The number of genes determining such characters is almost certainly large and even in the absence of tight repulsion linkages, large population sizes and numerous generations will be required for the necessary recombinations to take pIace. The success of Fl hybrid sorghum varieties illustrates this point. A half century of intensive hybridization and selection to isolate desirable homozygous lines failed to produce pure-line varieties that could stand the competition of the hybrid varieties once they were available. It can also be argued that homogeneous populations can be produced which will cope with unpredictable fluctuations in environment

SOME PARAMETERS OF POPULATION VARIABILITY

323

as well as heterogeneous populations. This may be the case, but in the meantime populations in which there is an appropriate compromise between the demands for uniformity and the advantages of diversity appear to have much to offer in terms of improving and stabilizing performance. V. Summary

1. Recent evidence concerning the numerical values assumed by parameters specifying various factors that affect population variability are reviewed in relation to plant breeding. 2. Studies of experimental populations of predominantly self-pollinated crop plants show that chromosome segments marked by major genes often have large selective values and that these values can fluctuate violently in different environments. In a high percentage of cases segmental heterozygotes have a striking selective advantage over homozygotes. Predictions based on estimates of relevant population parameters indicate that stable nontrivial equilibria exist for many such chromosome segments. 3. Studies of measurement characters in predominantly self-pollinated populations indicate great variability between families in advanced generations. This was the case both for populations of hybrid origin and populations made up by mixing homozygous lines. Within-family variability was larger than that of homozygous parents for progeny of random individuals drawn from such populations. This excess of variability over that of normal pure lines presumably results from segregation of genes governing quantitative characters. 4. Populations of even such heavily self-pollinated species as barley cannot adequately be characterized as arrays of homozygous familial isolates that are independent of one another in reproduction. Instead, individual members of these populations share in a common gene pool in a manner apparently differing only in degree from members of fullfledged Mendelian populations. The recombinational system appears to permit the formation of new and original variants for an indefinitely large number of generations. 5. The yield of broadly based populations maintained without conscious selection improves rapidly and within 10 to 15 generations after synthesis approaches or equals that of good locally adapted varieties. Agronomically superior types make up a greater and greater proportion of such populations in later generations. 6. The hereditary materials possessed by an economic species as a whole are viewed as a vast pool of genes and the task of the plant breeder that of assembling from this pool the gene combinations which

324

R . W. ALLARD AND P. E. HANSCHE

will give optimal performance in his environment. In this task recombination is regarded as occupying a central role. 7. It is suggested that among available breeding methods mass reservoirs provide the best opportunity for the required recombination to take place. The method allows large populations to be carried inexpensively. Hence it can be used to survey, recombine, and maintain variability on a scale not possible with conventional techniques. Such surveys could also provide information about the value as parents of particular items in “world collections” and also about geographical areas that are especially useful sources of variability. 8. Mass reservoirs are considered as a supplement to, rather than a substitute for, conventional breeding methods. Since they place emphasis on individual rather than family selection they should be especially useful so long as the general level of adaptation in the population is low. Once the frequency of desirable gene combinations has been increased to the point where family selection is required to identify generally small differences between genotypes, conventional methods may be more efficient in further exploitation of the components of mass reservoirs. 9. Evidence relating to the biological significance of genetic variability within the end product of successful breeding, the agricultural variety, is also reviewed. This evidence indicates that genetically diverse populations are frequently higher yielding over a range of environments than genetically homogeneous populations, and hence that optimal yield may depend on breeding varieties in which the appropriate compromise is found between the demands for uniformity and the advantages of diversity. REFERENCES Adair, C. R., and Jones, J. W. 1946. 1. Am. SOC. Agron. 38, 708-718. Akemine, H., and Kikuchi, F. 1958. In “Studies on the Bulk Method of Plant Breeding” (K. Sakai, T. Takahashi, and H. Akemine, eds.), pp. 89-105 (in Japanese). Yokendo Press, Tokyo. Allard, R. W. 1961. Crop Sci. 1, 127-133. Allard, R. W., and Hansche, P. E. PTOC.11th Intern. Congr. Genet., The Hague, 1963 Pergamon Press, New York. In press. Allard, R. W., and Jain, S. K. 1962. Evolution 14, 90-101. Allard, R. W., and Workman, P. L. 1963. Evolution 17, 470-480. Atkins, A. E. 1953. Agron. 1. 45, 311-314. Baker, H. G. 1959. Cold Spring Harbor Symp. Quant. B i d . 24, 177-191. Clausen, J., and Hiesey, W. M. 1958. Carnegie Inst. Wash. Publ. 615. Darlington, C. D., and Mather, K. 1949. “The Elements of Genetics.” Allen & Unwin, London. Dobzhansky, T. 1941. “Genetics and the Origin of Species,” 2nd ed. Columbia Univ. Press, New York.

SOME PARAMETERS OF POPULATION VARIABILITY

325

Dobzhansky, T., and Levene, M. 1955. Genetics 40, 797-808. Finlay, K.W. 1st Intern. Barley Genet. Symp., Waginingen. In press. Finlay, K. W., and Wilkinson, G. N. Australian 1. Agr. Sci. In press. Frankel, 0. H. 1950. Heredity 4, 89-102. Grant, V. 1958. Cold Spring Harbor Symp. Quant. Biol. 23, 337-363. Griffing, B., and Langridge, J. 1963. In “Statistical Genetics in Plant Breeding” (W. D. Hanson and H. F. Robinson, eds.), Publ. 982, pp. 368-394. Natl. Res. Council-Natl. Acad. Sci., Washington, D.C. Haldane, J. B. S. 1956. J. Genet. 54, 294-296. Harlan, H. V. 1957. “One Man’s Life with Barley.” Exposition Press, New York. Harlan, H. V., and Martini, M. L. 1938. J. Agr. Res. 57, 189-199. Harlan, J. R. 1956. Brookhaven Symp. Biol. 9, 191-208. Hayman, B. 1. 1953. Heredity 7, 185-192. Hiesey, W. M. Proc. 11th Intern. Congr. Genet., Tlre Hague, 1963 Pergamon Press, New York. In press. Iman and Allard, unpublished data. Jain, S. K., and Allard, R. W. 1960. PTOC.Natl. Acad. Sci. U.S. 46, 1373-1377. Jana, S. Unpublished data. Jones, D. F. 1958. Am. Naturalist 92,321-328. Kramer, N. W., and Loden, H. D. 1963. Agron. Abstr. p. 84. Langridge, J., and Griffing, B. 1959. Australian 1. Biol. Sci. 12, 117-135. Laude, H. H., and Swanson, A. F. 1943. J . Am. Soc. Agron. 34, 270-274. Lerner, I. M. 1954. “Genetic Homeostasis.” Oliver & Boyd, London. Lewontin, R. C. 1958a. Cold Spring Harbor Symp. Quant. Biol. 23, 395-408. Lewontin, R. C. 1958b. Genetics 43, 419-434. Lewontin, R. C., and White, M. J. D. 1960. Eoolution 14, 116-129. Li, C. C. 1955. Am. Naturalist 87, 257-261. Miravalle, R. J., Turner, J. H., and Lehman, M. 1962. Calif. Agr. 16, 2-3. Moran, P. A. P. 1963. Australian J. Biol. Sci. 16, 1-5. Morley, F. H. W. 1959. Cold Spring Harbor Symp. Quant. Biol. 24, 47-56. Rowe, P. R., and Andrew, R. H. 1963. Agron. A b ~ t r .p. 89. Sanchez, R. L., and Tucker, C. L. Unpublished data. Shank, D. B., and Adams, M. W. 1960. J. Genet. 57, 119-126. Simmonds, N. W. 1962. Biol. Rev. Cambridge Phil. SOC. 37, 422-465. Sprague, G. F., and Federer, W. T. 1951. Agron. 1. 43, 535-541. Stebbins, G. L., Jr. 1950. “Variation and Evolution in Plants.” Columbia Univ. Press, New York. Stebbins, G. L., Jr. 1957. Am. Naturalist 41, 337-354. Suneson, C. A. 1949. Agron. J . 41, 459-461. Suneson, C . A. 1956. Agron. I . 48, 188-190. Suneson, C. A., and Wiebe, G. A. 1962. Crop. Sci. 2, 347-348. Turner, J. H. 1963. Crop Res., Agr. Res. Ser. Publ. 34-51, 13 pp. Workman, P. L., and Allard, R. W. 1962. PTOC. Natl. Acad. Sci. U.S. 48, 13181325. Wright, S. 1942. Bull. Am. Math. Soc. 48, 223-246. Wright, S. 1949. In “Genetics, Paleontology and Selection” (G. L. Jepson, ed.), pp. 365-389. Princeton Univ. Press, Princeton, New Jersey. Wright, S. 1963. In “Statistical Genetics and Plant Breeding” (W. D. Hanson and H. F. Robinson, eds.), Publ. 982, pp. 368-394. Natl. Res. Council-Natl. Acad. Sci., Washington, D.C.

This Page Intentionally Left Blank

AMORPHOUS INORGANIC MATERIALS IN SOILS B. D. Mitchell, V. C. Farmer, a n d W. J. McHardy The Macaulay Institute for Sail Research, Aberdeen, Scotland

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Nature and Occurrence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Silica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Aluminum Oxides . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Iron Oxides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Allophane . . . . . . . . . . . . . . . . . . . . . . . . ................ 111. Methods of Detection and Estimation . . . . . A. Pretreatment . ........... B. Applications of ............ IV. Origin of Amorphous Material in Soil . . . . . . . . . . . . . . . . . . A. Weathering ....................... B. Silica ................................... C. Aluminum Oxides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Iron Oxides . . . . . . . . . ....................... E. Allophane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Relationships between Amorphous Inorganic Material and Specific Physical and Chemical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Morphological Proporties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Chemical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Page 327 328 329 333 335 338 341

368 369 370 372 372 373 375 375

1. Introduction

The clay fraction of the soil contains the finest and therefore the most reactive particles, be they organic or inorganic. In consequence, many of the properties of the soil are determined by the nature of this fraction, even though the amount present may be only a few per cent. During the late nineteenth century and for the first part of this century most soil scientists, doubtless greatly influenced by the extensive studies of Van Bemmelen, regarded the clay fraction as a colloidal complex and completely amorphous although many mineralogists, as pointed out by Mackenzie (1903), appreciated that clays could have a considerable amount of crystalline material. The mixed-oxide hypothesis (Van Bemmelen, 1910) or the association of the colloidal state with the concept of 327

328

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

noncrystallinity persisted until Hendricks and Fry ( 1930), using X-ray methods, produced the first important evidence establishing the presence of crystalline components in soil clays, At about the same time Kelley et al. (1931) made a similar observation. It is somewhat surprising that the amorphous concept prevailed until this late date since it was well known that microcrystalline materials may be highly colloidal and furthermore that microcrystalline material may assume the properties of a gel. However, the concept of crystalline clays resulted, in many circles, virtually in the renouncement of ideas regarding the amorphous nature of clays, and researches were directed almost exclusively to the elucidation of the physiochemical reactions of minerals in the clay fraction in terms of crystal architecture, A great deal has been learned from these studies and much of it has been successfully applied to soil clay investigations (Gieseking, 1949; Hauth, 1951) despite this emphasis on crystallinity. Mattson, in a long and distinguished series of papers on the colloidal chemistry of soil, never lost sight of the importance of inorganic amorphous soil components-a balanced view which is now becoming more generally accepted (Rich and Thomas, 1960). Soil clay can therefore best be regarded as consisting of the weathering products of primary minerals which may be classified as crystalline clay minerals, accessoq minerals, and amorphous material. This review deals with amorphous inorganic material in the clay fraction. Recent developments in the study of such material are considered in relation to its nature and occurrence, The physical and chemical methods employed for the detection and estimation of amorphous material are reviewed. Finally the possible modes of formation of amorphous constituents and their effect on soil properties are tentatively assessed. II. Nature and Occurrence

The upper size limit of soil clay particles is normally considered to be 2~ equivalent spherical diameter, and the lower size limit is probably comparable with moIecular dimensions. Because of this large range of particle size and because soil clays frequently consist of assemblages of many different and, at times, ill-defined constituents, accurate assessment of the nature of a particular component is often difficult, if not impossible. A further complication is the fact that a clay particle may well be an aggregate of still smaller particles. A considerable volume of infonnation has, however, been obtained on soil clays by various methods which allow the unambiguous identification of a number of well-defined crystalline clay minerals in soils (Rich and Thomas, 1960). On the other hand, comparatively little information is available concerning the nature

AMORPHOUS INORGANIC MATERIALS IN SOILS

329

of such amorphous components as hydrous oxides and organomineral complexes. The fact that amorphous materials are difficult to characterize and often occur in relatively small amounts in soils no doubt explains why these materials have to some extent been neglected. Their presence, however, cannot be ignored in clays which contain less than 5 per cent of crystalline material (Birrell and Fieldes, 1952). The principal forms of amorphous inorganic materials which occur in soils are the oxides, or more usually, the hydrous oxides of iron, aluminum, manganese, and silicon, either separately or combined. It is likely that phosphate can also enter these gels. The nomenclature of such material is nebulous, but a few names such as allophane for a mixed gel of alumina and silica (see Ross and Kerr, 1934) are commonly accepted and perhaps justifiable. More specific naming is possible for amorphous mineral deposits (Brown, 1955): these include opaline silica ( SiO2.nHz0), limonite ( FezOR.nHzO),kliachite ( A1203.nH20),wad (MnOp*nHZO), allophane ( A1203* 2Si0,. nH20) , hisingerite ( Fe20a2SiO2. nHtlO ) , evansite ( AI3PO4(OH)s.nH20), and azovskite (Fe,PO,( O H ) s . n H ~ O )Inter. mediates between allophane, hisingerite, and opaline silica can be indicated by the prefixes ferro-, ferri-, alumino-, and silico-. In many soil studies it is customary to determine, usually by extraction methods, the amount of so-called “free oxides,” and these are often assumed to be amorphous. This term is in some ways unfortunate, because it is generally taken to refer only to iron oxide and alumina, and is not necessarily related to the crystallinity or otherwise of the extracted material. Indeed, the extracted oxides are probably normally crystalline, if they exist as such, but equally may not necessarily be free in the soil, since they may be combined with silica in allophanic materials. In current Japanese literature the terms “active oxides” and “inactive oxides” have been revived to distinguish the readily extractable forms of the oxides from the less readily extractable or nonextractable forms. Here again, active oxides are generally assumed to be amorphous to X-rays and the inactive crystalline.

-

A. SILICA Alpha quartz is perhaps the most abundant accessory mineral in clays and is undoubtedly the commonest form of silica in soils. Silica may, however, be present in clays in hydrated noncrystalline forms (Si0p.nHpO),occurring either as opaline silica, or as a colloidal gel (Foster, 1953). Clays containing an abnormal amount of amorphous silica have been reported by Peters (1962) and termed opalinous. Opaline silica differs from quartz, and from the less commonly occurring microcrystalline form chalcedony, by being noncrystalline and by having a lower density

330

B. D. MITCHELL, V.

C. FARMER, AND W.

J. MCHARDY

and refractive index. Zabelin (1962) has studied the structure of silica in opals, opaline rocks, and synthetic silica hydrogels, and Wey and Siffert ( 1961) have quantitatively examined the transformation of amorphous silica to quartz and cristobalite. The occurrence of amorphous silica in soil clays appears frequently to be associated with the parent material; for instance, it is commonly observed in New Zealand and Japanese soils of volcanic ash origin. Fieldes and Williamson ( 1955) have produced evidence of amorphous silica in many New Zealand soil clays. The clay from the A2 horizon of a Kaieri podzol was especially interesting since it was shown to consist essentially of silica, a large proportion of which was amorphous. This tended to aggregate into very thin sheets in contrast to the clustered aggregates which formed when amorphous silica was associated with amorphous aluminum and iron oxides. Amorphous hydrous silica has also been observed in immature Japanese soils developed on pumice (Matsui, 1959; Kanno, 1959); the morphology of this silica has not, however, been described. Amorphous silica is not exchsive to immature soils on recent volcanic materials. Hoyos and Pino (1958) determined free silica in Spanish soils derived from granitic rocks, and VAmos (1961) has carried out an extensive investigation of its formation and accumulation in degraded alkaline soils. Glenn et al. (1960), investigating the weathering of layersilicate clays in loess-derived Tama silt loam, found that although SiO2:A1203 molar ratios generally increased in the profile with increasing particle size and depth, the highest SiO2:AI203ratio (8S:l) occurred in ) the surface horizon. Mitchell and Farmer the coarse clay ( 2 to 0 . 2 ~ of (1962) also reported siliceous clays in the organic surface horizons of certain well-drained Scottish soils, Van Rummelen (1953) described isotropic silica occurring in Indonesian soils and considered it to have been deposited originally round rootlets or in old root channels. The high content of opaline silica in Japanese soils and grasses, particularly in the A horizons of soils developed on volcanic ash, is believed by Kanno and Arimura (1958) to be of biological rather than volcanic origin-the occurrence of plant opal they claim is related directly to the high humus content of these soils. The association of amorphous silica with biological activity is by no means a new concept and the fact that small particles of opaline silica found in soil are derived from plants has been recognized by Russian scientists for a considerable time. According to Tyurin (1937), Ruprecht in 1866 described the morphology of opaline silica particles, termed phytoliths, which he found in the grass Stipa pennata and in the surface horizon of a chernozem soil. Studies such as that of Parfenova and Yarilova (1956),

AMORPHOUS INORGANIC MATERIALS IN SOILS

331

are indicative of continued Russian interest in this form of silica. Accounts of the occurrence in British soils of plant opal particles (Smithson,

1956a,b, 1958) suggest that they are widespread and are derived mainly from grasses. Opaline sponge spicules have also been noted (Smithson, 1959). An interesting observation by Riquer (1960) is that a soil type developed on basalt on the Isle of Reamine (East Africa) has a horizon 5 to 30 cm. thick, resembling the A2 horizon of a podzol, and consisting entirely of opaline silica phytoliths. Phytoliths have been found in podzols of temperate regions and also in iron-rich tropical soils.

B. ALUMINUM OXIDES Hydrated aluminum oxides and alumina gels are of considerable pedological significance. Free alumina attains a maximum concentration in highly leached tropical soils and is found in varying, but normally small, amounts in soils throughout the temperate region: consequently its presence is used by many pedologists as an index of weathering within the profile. Generally free alumina is present in the form of gibbsite [y Al(OH)3]: this has been confirmed by the study of laterites by Alexander et al. (1956)) who also observed that boehmite [y AlO.OH] could predominate in some. A great deal of work has been carried out on synthetic alumina gels because of the effect of aging upon their absorptive capacities for enzymes and viruses. Electron microscopy has been widely applied in the work and an excellent review of such morphological studies has been prepared by Moscou and van der Vlies (1959). The thermal behavior of pure alumina gels has been discussed by Mackenzie (1957b) who with Meldau (Mackenzie and Meldau, 1959; Mackenzie et al., 1962) continued the investigation of these gels using a combination of thermal, electron-optical, and infrared techniques in order to provide information of value in pedological studies. In the gel aged at pH 5 for 60 days there was little or no evidence of crystallization, whereas at pH 10 aging led to a considerable increase in the degree of crystallization associated with changes in the morphology of the crystals. There are few accounts in the literature of the occurrence of free amorphous alumina in soils although laboratory studies on the aging of these materials indicate that there is negligible crystallization under mild acid conditions. Moreover, Tamura and Jackson ( 1953) , and Fieldes ( 1955), while proposing different mechanisms for the silicification of alumnia, postulate amorphous hydrous alumina as the starting material. Mackenzie (1957b) considered that aluminum, in contrast to iron, may not form particulate amorphous oxides, but from differential thermal analysis evidence Fieldes and Williamson (1955) claim that small

332

B. D. MITCHELL,

V. C.

FARMER, AND W. J. MCHARDY

amounts of amorphous hydrous oxides of aluminum are present in the early weathering stages of many New Zealand soils. Kanno (1959) noted amorphous alumina in red-yellow soils derived from granitic rocks, and Matsui (1959) in young soils developed on pumice. In a mineralogical investigation of the bottom clays of the North Bohemian coal basin, Zemlicka ( 1958) reported gibbsite, and locally a metastable amorphous aluminum hydrate was fixed to the clays of the bottom layer. Gradusov and Dzyadevich ( 1961) found aluminum hydrophilic compounds in strongly podzolized soils, and Wicklund and Whiteside (1959) observed alumina accumulation in the B horizon of all the podzolic soils they examined. Evidence of amorphous alumina in the clay fraction of soils developed on igneous rocks has been obtained by Fridland (1961), and Gradusov and Targulyan ( 1962). Expanded three-layer minerals containing A1-OH groups in the interlayer space are apparently of widespread occurrence; the term “intergradient” has been proposed for these minerals (Jackson, 1959). Since consideration should be given to the possible presence of interlayered alumina in studies aimed at determining the nature of alumina in clays it is appropriate that in this review reference be made to recent investigations on these minerals. The aluminum of the aluminous interlayers is not exchangeable (Rich and Obenshain, 1955; Klages and White, 1957). The initial material may be aluminum ions, originating from the H saturation of the clay, and these may be subsequently hydrolyzed, polymerized, and finally fixed. It was proposed by Klages and White (1957) that a continuous series of minerals may exist ranging from vermiculite to dioctahedral chlorite depending upon the degree of hydrolysis of the interlayered Al. Theory certainly does not preclude complete alumination to form chlorite, but usually the amount of aluminum in the interlayer spaces of clays is small compared with that required for the formation of a complete gibbsite sheet (Dixon and Jackson, 1959; Rich, 1960). Aluminous interlayers are usually associated with the surface horizons of mildly acid leached soils such as brown podzolic or brown forest soils (see, for example, Tamura, 1958; Tamura et al., 1959; Avery et al., 1959; Sawhney, 1960; Scheffer et al., 1961). In a study of the clay mineralogy of a soil profile developed on glacial till, Quigley and Martin (1963) noted an iron-aluminum interlayer material at a depth of 55 inches, which is considerably deeper than hitherto recorded. Unlike aluminous interlayers, which are usually restricted to the upper horizons of acid soils, exchangeable aluminum frequently increases with depth. Coleman et al. (1959), and Rich et al. (1959) presented evidence indicating that in the basal horizons of acid soils aluminum may be the dominant cation, and Fiskell et al. (1958) stated that cal-

AMORPHOUS INORGANIC MATERIALS IN SOILS

333

careous and montmorillonitic soils overlying calcareous rocks contained appreciable amounts of exchangeable aluminum. The interactions between soil organic matter, absorbed cations and sesquioxides involve numerous factors but treatment of this extensive and very important subject is outside the scope of this review. C. IRONOXIDES Iron oxides are probably the most common of the accessory minerals in clays, Carroll (1958) quoting 10.3 per cent as the average ferric oxide content of 162 soils from various parts of the United States. Iron oxides occur as discrete particles or as surface coatings (Fripiat and Gastuche, 1952; Rich and Thomas, 1960; Gradusov and Dzyadevich, 1961; Oades, 1962; Sumner, 1963) on the clay particles and account to a large extent for the color of the soil. Frequently they are associated with organic matter and organisms (Aleksandrova and Nad, 1958; Wurman et al., 1959; Kuron et al., 1961; Duchaufour, 1963). The monohydrates of ferric oxide, goethite, and lepidocrocite and the anhydrous oxide, hematite, are the usual crystalline forms found in soils. Hydrated ferric oxide gels, analogous in some respects to the gelatinous precipitate formed by the addition of alkali to ferric salts, and anhydrous iron oxides are also known to occur (Mackenzie, 1957b; Schwertmann, 1959). Because amorphous coatings and crystals of free iron oxides function as cementing agents, their removal is frequently desirable before proceeding with certain analyses of soils and clays. Such removal not only results in a more effective dispersion of the soil separates and improves X-ray diffraction patterns, but also aids differential thermal analysis, electron-optical studies, cation-exchange capacity, and specific surface area determinations. Much attention has been given to the investigation of techniques aimed at the efficient removal of iron oxides and numerous methods have been proposed, some of which will be discussed later. Unfortunately, until recently, little consideration has been given to the fact that iron oxides, or the other sesquioxides for that matter, may exist in soils both in the crystalline and amorphous states. Gorbunov et al. (1961) in a study principally concerned with methods of determining amorphous and crystalline iron oxides in soils and clays, indicated that the relative distribution of these forms within the soil profile can have a most important bearing on the physicochemical properties of the soil, on the interpretation of soil genesis, and on studies connected with the intensity and conditions of weathering. The determination of the crystalline and amorphous iron oxide content of soil can, however, be extremely difficult, because, as Taylor (1959) comments, neither X-ray diffraction nor differential thermal analysis, the principal

334

B. D. MITCHELL, V.

C. FARMER, AND

W. J. MCHARDY

techniques used to identify crystalline iron oxides, is particularly sensitive, and consequently appreciable amounts of such materials may go undetected, Identification is further complicated by the fact that iron oxides may be finely particulate and poorly crystalline. The stability of a particular form of iron oxide depends upon the environment (Carroll, 1958), and there is a certain amount of evidence to suggest that soil-forming processes are also involved. Fieldes and Swindale (1954) observed important amounts of amorphous iron oxides in the early stages of weathering of the zonal soils of New Zealand and Kanno (1959) also recorded amorphous iron oxide in young soils of Japan developed on pumice and volcanic ash. Amorphous iron oxides are not, however, restricted to immature soils, since they have also been observed along with goethite and hematite in laterites (Alexander et al., 1956). The type of free iron oxide in terra rossa soils may depend on their origin: Muiioz Taboadela (1953) found hematite in Spanish soils, whereas Taylor (1959) found that goethite accounted for all the free iron oxides in a Barbados terra rossa. Following the observation of Gheith ( 1952) that precipitated iron hydroxides recrystallize relatively rapidly at normal temperatures, Taylor ( 1959) considers that amorphous iron oxides would rarely occur in appreciable amount in soils. Gorbunov (1959), however, concluded that amorphous iron oxides predominated in podzolic soils although the iron oxides in red earths and laterites were primarily crystalline. Earlier, Kawaguchi and Matsuo ( 1957) associated amorphous iron oxide with an unstable environment and intense weathering conditions. They separated by a magnesium reduction method (Kawaguchi and Matsuo, 1954) active and inactive iron oxides from the soil profile, cIaiming the active to be amorphous and the inactive to be combined in crystalline primary and secondary minerals. Hoyos and Pino (1958), investigating freely drained silty soils developed on granite in the Canary Islands and Spanish Guiana, found, in addition to goethite, highly hydrated iron gels. Fridland ( 1961), also examining freely drained soils but developed on basic rocks, recorded large amounts of what he termed free or unstable iron oxides. Further evidence of the association of amorphous iron oxide with intense weathering in highly leached soils is provided by the studies of Gradusov and Dzyadevich ( M l ) , Kaurichev and Nozdrunova ( 1961), Gradusov and Targulyan (1962), and Oades (1962). Most work on free iron oxides has been primarily concerned with their removal rather than with their nature, and consequently, relatively few attempts have been made to ascertain which oxides exist in pedological features which by their color, or perhaps structure, suggest a local concentration of iron oxides. An exception is the accumulation of sesquioxides

AMORPHOUS INORGANIC MATERIALS IN SOILS

335

followed by induration, which is a widespread pedogenic phenomenon reaching maximum expression in tropical climates ( D’Hoore, 1954) and on which a considerable amount of information is available. In an investigation of the colloid chemistry of black earth soils, Arkhangelskaya (1955) obtained a correlation between the high amount of microaggregates in the subsurface horizon and the nonsilicate iron content which was 45 to 63 per cent of the total iron. Schwertmann (1959) has carried out a detailed investigation of the forms of iron oxide in specific morphological features of a number of soil profiles. He has examined mottlings, concretions, weathering crusts, hard pan, and bog iron formations. Poorly crystallized goethite predominated in all these, and appreciable amounts of hydrated iron oxides amorphous to X-rays were also present. The amorphous oxides could be separted almost completely from the crystalline material by treatment with ammonium oxalate at pH 3. Schwertmann (1959) also showed from experiments with synthetic goethite that it formed through the aging of amorphous ferric oxide gel which had been precipitated from ferric solution by hydroxyl ions or by oxidation of ferrous carbonate. However, differential thermal analysis suggested that the amorphous hydrated iron oxides in the soils were not identical with hydrated ferric oxide precipitates.

D. ALLOPHANE The identification and separation of amorphous silica, alumina, and iron oxide in soil clays present many difficulties even when the most modern methods and techniques are available for their examination. The “free” oxides of silicon, aluminum, and iron probably represent only a small part of the amorphous inorganic material frequently encountered in soil, the bulk consisting of poorly defined combinations of silica and sesquioxides. The mineralogical definition of allophane has already been referred to, but many pedologists use the same term to denote extremely variable noncrystalline silica-containing materials that occur in soils. The composition of allophane has been defined as Al2O3.2SiO2.nH20 (Brown, 1955), but most soil mineralogists follow Ross and Kerr (1934) in considering that the name should apply to all mutual solutions of silica, alumina, and water with minor amounts of other bases, although Harrassowitz (1926) also included simple mixtures of alumina and silica in his definition. White (1953), appreciating the heterogeneity of amorphous inorganic material in soil, defined allophane as any amorphous substance which may be present in clay materials and which has indefinite composition. This definition recognizes that ferric oxides may also be involved in such combinations (Brown, 1955; Jackson, 1956), but does not emphasize sufficiently that silica is an essential component.

336

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHAml’

It has been wideIy recognized that allophane possesses different degrees of order and different stability to specific reagents, and this has led to much confusion in nomenclature. Fieldes ( 1955) distinguished two forms of allophane: allophane B in which amorphous silica was discrete and the particle size very small, and allophane A in which the silica and alumina were randomly combined and the particle size large. Jackson (1956) also recognized two forms of allophane; that which was soluble in dilute acids and alkalis he termed “unstable allophane,” and that which was only slightly soluble in these he called “stable allophane.” Yoshinaga and Aomine (1962a,b) in a study of Ando soils were able to separate from the crystalline minerals of the clay fraction two mineral colloids. One of these colloids remained dispersed in both acid and alkali, while the other flocculated in alkali. The former was amorphous to X-rays and considered to be allophane, and the latter, which possessed a low degree of crystallinity, was termed imogolite, since it was first observed in the Imogo soil-a brownish yellow soil developed on volcanic ash. The nature and properties of soil allophanes have been to a great extent determined from studies on the clay fraction of soils developed on recent volcanic deposits in New Zealand and Japan. In these soils allophane is usually the principal clay mineral and frequently can be isolated in a comparatively pure state. Following Birrell and Fieldes ( 1952), Gradwell and Birrell ( 1954) determined the physical properties of certain clays of volcanic ash origin, while Dixon (1954) examined the surface properties of similar soils and Birrell and GradweIl (1956) their exchange capacity. The reactions of amorphous soil colloids with ions in solution have been investigated by Birrell ( 1958). Fieldes and Schofield ( 1960) have studied the mechanism of ion adsorption by inorganic colloids in New Zealand soils, and Birrell ( 1961a,b) has investigated ion fixation by allophane and also (Birrell, 1962) the acidity of subsoil clays, The value of electron microscopy in the characterization of allophanic clays and the fact that infrared absorption spectra and differential thermal analysis curves can provide direct evidence of the presence of allophane and of amorphous hydrous oxides of silicon and aluminum in clays has been demonstrated by Fieldes and his co-workers (Fieldes and Williamson, 1955; Fieldes et aZ., 1958; Fieldes, 1957). Recent volcanic deposits constitute the parent material of many Japanese soils, and in the past decade the amorphous aluminosilicate content of these soils has been, as in New Zealand, the focus of numerous investigations (see, for example, Sudo, 1954, 1956; Harada, 1955; Kanno, 1955, 1959; Sasaki and Ishizuka, 1957; Ishii and Mori, 1959). Matsui

AMORPHOUS INORGANIC MATERIALS IN SOILS

337

(1959) states that, in general, Japanese soil clays are kaolinitic except those developed on volcanic ash which are allophanic. Allophane associated with halloysite appears to predominate in the younger soils of areas of recent volcanic activity in Chile ( Garcia-Vicente and Besoain, 1961; Besoain and Garcia-Vicente, 1962), and high concentrations of allophane have also been noted in the clays from weathered ash deposits in Hawaii (Bates, 1960). Whittig et al. (1957) and Robertson (1963) found allophane in soils from regions in Oregon which had been affected by volcanic activity. DeMumbrum (1960), and DeMumbrum and Bruce (1960) reported amorphous aluminosilicates in recent deposits on the Mississippi coastal terrace, the deposits being of volcanic or metamorphic origin. The clay mineral composition of some desert lakes in Nevada, California, and Oregon has been examined by Droste (1961), who noted that several contained sediments rich in amorphous silicates and that those which contained most amorphous material were in areas where recent volcanization had been extensive. There was no evidence of kaolinite in these deposits, but montmorillonite was observed and thought to be formed from the volcanic glass. On the other hand, allophane is not invariably a constituent of young soils on volcanic material: for example, Fazzini and Olivieri (1961) did not observe it in the volcanic sands which they examined. Furthermore the occurrence of allophane is not restricted to recent volcanic deposits. Kanno et al. (1956) noted amorphous aluminosilicates in all samples of red yellow soils derived from Pleistocene sediments which they examined, and earlier Tamura et al. (1953) found allophane in latosols. Hosking et al. (1957) appreciated that amorphous aluminosilicates were not confined to volcanic ash soils and concluded that amorphous clay minerals were probably more widespread than was generally accepted. Sanchez-Calvo ( 1961) reported allophane and other amorphous colloids in the Braunlehm soils of the West Canaries. The Braunlehm (Harrassowitz, 1926) is a fossil soil which according to Kubiena (1953) is characterized by highly mobile silica rich clay and peptized ferric hydroxides. Russian soil clay mineralogists have shown interest in colloidal amorphous inorganic material. Thus, Lomonovich ( 1955), studying the origin of loess, found that each size fraction into which he divided the material had the same mineralogical composition, the various microaggregates being simply an agglomeration of fine particles with amorphous inorganic colloids acting as cementing agents. Gorbunova (1961) observed amorphous colloids including allophane in the bottomland soils of Kazakhstan, and Yarilova and Parfenova (1960) identified similar materials in other Russian soil types. According to Kashiwagi and Yokoi (1952) colloids in the top soil

338

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

are frequently in the soil solution, whereas in the subsoil, they are in the gel state. There is also much evidence that amorphous materials are concentrated in the surface horizon. Glenn et a,?. (1960) found that the fine clay ( < 0.08 p) from the A12 horizon of a Tama silt loam soil contained 27 per cent amorphous material while the same fraction from the C horizon had 14 per cent amorphous siliceous material. Stefanovits (1959) noted an increase in amorphous material in the surface horizon compared with that in the basal layer, and he related this directly to the organic matter content, concluding that very little clay mineral formation occurs in the organic horizons of Hungarian soils. Examination of clays from the A and C horizons of Scottish soils (Mitchell and Farmer, 1962) showed that they contained a proportion of highly hydrated amorphous material resembling allophane in its properties and this component was particularly high in clays from the surface horizons with the greatest organic matter content. 111.

Methods of Detection and Estimation

Knowledge of the nature and distribution of amorphous colloids in soil clays is far from complete and even with the aids of modern highly refined instrumentation little direct evidence is obtainable so that the pedologist has to base his concepts of this material largely upon implication. Various physicochemical techniques are currently being applied to the clay fraction of soils to ascertain the nature of their inorganic colloids and to give some estimate of the amounts present. X-ray diffraction, thermal analysis, infrared absorption, and electron microscopy are the principal methods used, and the information given by each of these is essentially complementary to that of the others. Cation-exchange capacity determinations, specific surface area measurements, classical optical methods, and silicate analyses provide valuable supplementary data. Chemical treatments, to varying degrees specific for the removal of free iron oxides, aluminum oxides, and interlayered alumina have been in use for some time, and currently alkali differential-dissolution techniques are being investigated. An accurate assessment of amorphous material in soil clays probably depends upon the further development of these chemical methods and their correlation with precise physical measurements. A. PRETREATMENT Ideally the pedologist prefers to examine the soil sample in as close to the natural condition as possible, but, in general, before any physicochemical measurements can be carried out on the inorganic constituents of a soil a certain amount of chemical pretreatment is necessary, Before

AMORPHOUS INORGANIC MATERIALS IN SOILS

339

discussing the contribution of instrumental techniques to the characterization of amorphous material in soil clays, consideration will be given to pretreatment of the sample. In the first place it usually has to be divided into the recognized soil separates, namely coarse sand, fine sand, silt, and clay. Organic matter nearly always has to be removed and for some purposes it is also desirable to remove the “free” iron oxides. It is inevitable that such treatment will result in some alteration of clay properties; indeed Harward and Theisen (1962) have pointed out that the X-ray identihation of clay minerals in a given sample can be dependent on such things as specimen carriers, dispersion reagents, method of iron removal and the cation saturating the clay. Amorphous material by its very nature will be most subject to attack or alteration and it is important that the utmost care be taken in pretreatment of the sample for such investigations. The usual method of separating a soil clay is by sedimentation, after dispersing the soil in either dilute NaOH or NHlOH at about pH 10. Dilute acids have been used but it is generally accepted that clay minerals are more readily attacked by acids. Ostrom (1961), for example, used dilute acetic acid and HCl to separate clay minerals from carbonate rocks and found that well-crystallized minerals showed considerable resistance to attack but that as the degree of crystallinity decreased, the clay became more susceptible to decomposition by acid. Soils containing relatively large amounts of amorphous oxides may be difficult to disperse, especially if these oxides are involved in cementing the crystalline clay minerals, and several methods proposed for dispersion involve the deliberate dissolution of amorphous material ( Jackson et al., 1950; Hashimoto and Jackson, 1960). The factors affecting the dispersion of volcanic ash soils containing principally allophane and amorphous oxides have been discussed in some detail by Birrell and Fieldes ( 1952), who concluded that the dispersion of soils containing such materials appears to be practicable only if the free sesquioxide content is low and the pH of the medium is at least 10. They did not recommend acid pretreatment on account of possible attack on allophane. To separate allophane from imogolite effectively, Yoshinaga and Aomine (1962a) used alkali dispersion followed by dilute acid dispersion. Where it has been found desirable to remove organic matter from soils and soil clays in order to facilitate X-ray, infrared absorption, and differential thermal analysis studies on the inorganic constituents, the method most commonly employed involves treatment with hydrogen peroxide. Farmer and Mitchell ( 1963) have shown that water-soluble complex oxalates of aluminum and iron, and to some extent insoluble

340

B. D. MITCHELL,

V. C.

FARMER, AND W. J. MCHARDY

oxalates, are formed by the peroxidation of organic matter in soil clays. They confirmed that the effects of peroxidizing soils and clays of high organic matter content could be considered as similar to that of extracting with acid oxalate under ill-defined conditions. For this reason Farmer and Mitchell (1963) do not recommend peroxidation before dispersion since this would involve oxidation of even greater amounts of organic matter and the corresponding dissolution of greater amounts of aluminum and iron. Futhermore, under certain conditions, the complex oxalates formed may decompose, leading to the precipitation of amorphous aluminum and iron hydroxides which would be recovered with the clay fraction. During the past forty years many methods have been proposed for the removal and quantitative determination of iron in soils. Some of these methods have been reviewed and compared by Deb (1950), Aguilera and Jackson (1953), Gorbunov et al. ( 1961), and Harward et al. (1962). Much of this work has been directed toward “cleaning up” samples for further mineralogical analysis, the aim usually being to ensure the removal of coatings of “free” oxides, frequently mixtures of amorphous and crystalline iron oxides. Nearly all the methods at present in use are based upon the reduction and mobilization of iron by sodium dithionite. The most recent procedures for the selective removal of iron oxides from soils and clays are, however, based upon Tamm’s (1922) acid oxalate method. Jeffries and Johnson (1961) advocate boiling with a solution of potassium oxalate and oxalic acid, and de Endredy (1963) found that the efficiency of the acid ammonium oxalate extraction was considerably improved by irradiation by light in the near ultraviolet region. In addition to removing free iron oxides, all the methods remove some aluminum oxides and silica. The various methods examined by Mehra and Jackson (1960) all dissolved a considerable amount of silica, particularly from soils high in amorphous material; for example, Mehra and Jackson’s dithionite-citrate-bicarbonate method extracted 5.0 per cent iron oxide, 5.6 per cent alumina, and 15.0 per cent silica from a Japanese Ando soil rich in allophane. Mehra and Jackson (1960) also studied the destructive effect of various methods of iron removal on iron-containing silicate minerals in soils, using as their criterion the change in cation-exchange capacity. The tendency was for the cationexchange capacity to decrease with iron oxide removal, the marked decrease brought about by some methods being considered to indicate attack of the crystalline minerals. Deb (1950) made a similar study, finding in some instances, however, an increase in exchange capacity after “free” iron oxide removal, and this he attributed to the removal

AMORPHOUS INORGANIC MATERIALS IN SOILS

341

of FeO+ groups from exchange positions or the removal of Al( OH)2+ from the outer edges of the crystal lattice. It would thus seem desirable in the investigation of amorphous inorganic material in soil clays to keep chemical pretreatment to a minimum. One must consider appropriate techniques. Kashiwagi ( 1961 ) described a method of isolating the clay fraction of soils by means of air flow: although chemical pretreatment is thus avoided, it is most unlikely that small aggregated clay particles would be dispersed. Vasileva (1958) has used ultrasonic oscillations to disperse soils and claims that this technique is more efficient than conventional methods. Because particle size is a factor that influences most physical measurements on clays, and also the extent of chemical attack during pretreatment, it must therefore be taken into consideration. In this connection prolonged grinding of a sample is to be avoided since not only does this reduce particle size, but it has been shown to lead to considerable alteration in clay mineralogy (Mackenzie et al., 1956; Gorbunov and Sharina, 1958; Takahashi, 1959; Yamaguchi and Sakamoto, 1959). In general, however, it appears that little alteration of clay properties will be induced if the soil can be effectively dispersed by NH40H or NaOH at pH 10. Much more serious is the question of peroxidation, and it is difficult to see how this step can be avoided. Farmer and Mitchell (1963) suggest that a critical appraisal of oxidizing procedures, including the hypobromite method of Troell (1931) is required to determine the optimal conditions and technique. Anderson (1963) recommends the use of sodium hypochlorite solution, claiming that the organic matter left in samples was commonly less than that remaining after hydrogen peroxide treatment and that sesquioxides were neither dissolved nor complexed. Furthermore, stable suspensions result from the use of sodium carbonatesodium bicarbonate solution to wash and sodium-saturate the samples. It is unlikely that any reagent will be found that will be specific either for the removal of free iron oxide alone or for separating crystalline forms of this oxide from amorphous forms. In any event where a study of the amorphous material is the principal objective of an investigation, it is most probable that iron oxide will be a constituent. Methods of iron extraction, therefore, should be considered only in conjunction with other chemical methods aimed at extracting amorphous materials as a whole. AND CHEMICAL TECHNIQUES B. APPLICATION OF PHYSICAL

Most investigators of amorphous inorganic material in soil have used initially a combination of X-ray diffraction and differential thermal methods, tentatively assigning anomalies or discrepancies in the quantita-

342

B. D. MITCHELL, V.

C. FARMER, AND W. J. MCHARDY

tive determination of the crystalline material by these methods to amorphous constituents. Infrared absorption and electron microscopy are being used increasingly to provide supplementary evidence ( Fieldes and Williamson, 1955; Fieldes et al., 1956; Mitchell and Farmer, 1962; Yoshinaga and Aomine, 1962a,b; Robertson, 1963). Each of these methods provides a certain amount of information. X-ray patterns reflect the type and regularity of the structural arrangement of atoms in three-dimensional space, providing specific data on a number of components in the soil clay. Consequently, the X-ray method is valuable for materials with relatively well-ordered structures provided the particle size is not much smaller than 0 . 1 ~ .In contrast, the infrared spectrum of a soil clay is determined by the atomic masses and the pattern of strong interatomic forces within a structure. Absorption bands are associated with vibrations of the structure. Amorphous substances absorb as strongly as crystalline compounds of similar composition, but their absorption bands are more diffuse. Many minerals give characteristic infrared absorption spectra in the 2.5 to 25 p range. Differential thermal analysis, which involves recording the temperature difference between a sample and an inert material as the two are heated side by side, shows all the energy changes occurring during heating irrespective of whether these are associated with loss of material from or structural rearrangement within the specimen. The peak temperature is roughly indicative of the energy required to initiate the reaction, whereas the area enclosed by the peak provides a measure of the energy change associated with the reaction. Thermogravimetry and differential thermogravimetry give quantitative information only upon the loss and gain in weight during a reaction. The particle size of the sample is important in differential thermal analysis inasmuch as the smaller the particle the more readily is the heat liberated: consequently the thermal peak may be sharper and the area greater. Particle size effects operate when the reaction has its rate controlled by the surface area of the reactants or active material, as in oxidation reactions. Because most soil clay particles are too small to be examined by classical optical methods and since they usually consist of a varied assemblage of minerals, the advantages of utilizing electronoptical methods, thereby obtaining a picture of the exact shape and size of the particles, is obvious. Kinter et aZ. (1952) examined a range of soil clays and showed most conclusively the value of this method. Modern electron microscopes allow electron-diffraction patterns to be obtained from individual clay particles. Because the wavelength associated with the electron beam is shorter than that of the X-ray beam, eIectron-diffraction patterns permit the recognition of shorter-range order. It is important, therefore, that

AMORPHOUS INORGANIC MATERIALS IN SOILS

343

the method of investigation should be stated when describing a material as amorphous (Brown, 1955). 1. X-Ray Diffraction Although not applicable directly to amorphous material, X-ray diffraction methods are essential in studying the crystalline portion of the soil clay, lack of information on which could lead to invalid deductions. These techniques have been treated extensively elsewhere ( e.g., Brown, 1961) and will not therefore be reviewed here. However, the occurrence of aluminum and iron hydroxides in the interlayer position of expansible minerals has been found, in recent years, to be fairly common, and the value of X-ray diffraction in the study of such minerals will be briefly considered. As a result of pedological weathering, aluminum, iron, and magnesium hydroxides may be precipitated in the cleavage space of expanding layer silicates (Jackson, 1960; Dixon and Jackson, 1959, 1960) forming incomplete interlayers which are thermally unstable in the 400 to 500°C. range when the clays collapse from 14 A. to 10 A. These interlayers are more stable than free gibbsite but less stable than the interlayers in true chlorites. Intergradient minerals do not expand to 18 A. with glycerol, nor collapse to 10A. at 300" following K-saturation. The removal of the interlayer material can be effected by NaOH digestion (Dixon and Jackson, 1959) or by digestion with sodium citrate (Tamura, 1958; Sawhney, 1960). However, the latter procedure can on occasion result in the use of relatively large amounts of citrate which are extremely difficult to displace completely and which can be very inconvenient when such techniques as differential thermal analysis and infrared absorption spectroscopy are to be applied to the treated sample. The thermal instability of the intergradient clays in the 400" to 500°C. range resulting simultaneously in the loss of the 7 A . peak and the collapse of the 14 A. spacing to 10 A. makes precise resolution of kaolinite and halloysite from 2:1 and 2:2 intergradient minerals difficult if not impossible. Garrett and Walker ( 1959), Wada ( 1961 ) , and Jackson (1962), among others, have been involved in the development of the intersalation technique by means of which K+ and NH4+ salts are introduced between the layers of both halloysite and kaolinite. These techniques represent a considerable advance in the characterization of clays by X-ray diffraction. 2. Diferential Thermal Analysis Undoubtedly X-ray diffraction is the most valuable single method for clay mineralogical investigations. Thermal methods have generally been

344

B. D. MITCHELL, V. C. FARMER, A N D W. J. MCHARDY

used to supplement results obtained by X-ray examination. It has been shown, however (Mackenzie, 1957a), that thermal methods have reached a stage of development which permits a number of valid mineralogical assessments to be made by these alone. Indeed, in the detection and estimation of amorphous materials in soil clays thermal methods have a major role. The hydroxyl groups of the hydrous silicates and oxides, which make up the clay fraction of soils, on heating release water, the evolution of which produces an endothermic effect on the differential thermal curve. The strength with which the hydroxyl groups are retained within a structure determines the temperature at which this release occurs. The temperature of the peaks on the differential thermal curve resulting from the evolution of this water are, therefore, characteristic of the materials present. Apart from free water wetting the clay, water is also absorbed on the external and internal surfaces of clay particles. The differential thermal curves of silica gel and allophane (Fig. 1,curves a and f ) feature a very broad endothermic peak between 100" and 200°C. which is due to the loss of water from the internal and external surfaces and probably also some of the hydroxyl water. The rate at which this water is lost depends upon the particle size and degree of development of structure (Fieldes, 1957); there are thus variations in peak temperature on the differential thermal curve. Most glassy opals do not, however, exhibit an endothermic effect in the 100" to 200°C. region (Jones et al., 1963). The loss of structural hydroxyl groups occurs rapidly in most of the crystalline clay minerals and is reflected on the differential thermal curve by a well-defined endothermic peak in the 500" to 700°C. range ( Fig. 1,curve g ) . Generally no distinctive peak corresponding to the loss of structural hydroxyl groups of amorphous aluminosilicates is found on their differential thermal curves, but imogolite, which has a certain degree of order, exhibits a small endothermic peak at about 425" to 435"C, when its diffraction lines also disappear (Yoshinaga and Aomine, 1962b). Differential thermal analysis curves of soil clays with large endothermic peaks in the 100" to 200°C. range and small endothermic effects between 500°C. to 700°C. indicating, respectively, large amounts of hygroscopic moisture and little or no sharply defined dehydroxylation are regarded as evidence of the presence of substantial amounts of amorphous silicates in the clay always provided montmorillonite is absent ( Fieldes, 1957; Kanno, 1959; Matsui, 1959; DeMumbrum, 1960; Sanchez-Calvo, 1961; Aleixandre-Ferrandis et al., 1962; Gradusov and Targulyan, 1962). Vermiculites containing polyvalent exchangeable ions give a somewhat similar thermal curve, but confusion can be avoided by saturating the clay with NH4+ (Mitchell and Farmer, 1962).

345

AMORPHOUS INORGANIC MATERIALS IN SOILS

.

.

200

.

.

.

.

,

400 600 TEMPERATURE %.

I

800

1c 0

FIG.1. Differential thermal curves: a, synthetic silica gel (100 mg.); b, synthetic alumina gel (25 mg. ); c, gibbsite (25 mg. ); d, synthetic hydrated ferric oxide gel ( 100 mg.); e, goethite ( 100 mg.); f , allophanic soil clay ( 100 mg.); g, kaolinite ( 100 "8.).

346

B. D. MITCHELL, V, C. FARMER, AND W. J. MCHARDY

The differential thermal curve for alIophane B (Fieldes, 1955; Fieldes et al., 1956), in which the hydrous silica and alumina are thought to be discrete, has a large peak due to dehydration below 150°C. and a very small endothermic peak ascribed to nascent amorphous hydrous alumina about 200°C. This 200°C. peak is absent on the thermal curve for allophane A, in which the silica and alumina are certainly in combination. There is really no definite evidence that amorphous alumina gels occur in nature although minerals supposedly of this type, for example, kliachite (Doelter and Cornu, 1909), have been reported. A number of differential thermal curves of synthetic alumina gels (Fig. 1, curve b ) have been published (Houldsworth and Cobb, 1923; Weiser and Milligan, 1942; Souza Santos et al., 1953; Mackenzie, 1957b; Mackenzie et al., 1962), but none bear evidence of a 200°C. endothermic peak. Gibbsite, the most common naturally occurring hydrate, shows a marked endothermic effect in the 300" to 350°C. range (Fig. 1, curve c ) , and the peak temperature is independent of particle size ( Mackenzie, 1957b). The differential thermal curves of allophane feature an exothermic effect between 800" and 1000°C. This may correspond to the exothermic effect shown by the kaolin minerals in this temperature range which has been ascribed to the nucleation of mullite and y-alumina (Holdridge and Vaughan, 1957). Fieldes (1955) observed that with increasing order of structure from allophane B + allophane A + metahalloysite + kaolin ( well-crystallized) the temperature and intensity of the high temperature exothermic peak increased. Yoshinaga and Aomine (1962b) noted that the exothermic peak on the curve for imogolite was 20" to 40°C. higher and more intense than that on the curve of the allophane from which it was separated, and they suggested that this reflected differences in the inner structure of these minerals. Mitchell and Farmer (1962) examined, up to 1050"C., four samples of allophane from different localities, only one of which showed an exothermic effect by this temperature (Fig. 1, curve f ) . The influence of organic matter on this high-temperature recrystallization reaction has been referred to by Oades (1962), and Mitchell and Farmer (1962) have noted that it was absent from the thermal curves of soil clays high in organic matter when determined in a nitrogen atmosphere. It is possible that the residue from earlier pyrolysis prevents nucleation. Study of the characteristics of the high temperature exothermic peak on the differential thermal curves of soil clays has been somewhat neglected, and it would appear from these results that a detailed investigation of this effect might be rewarding. Heating to 600°C. is known to cause the collapse of expanding layer silicates with consequent loss of interlayered water (Brown, 1961), but it was observed by Mitchell and Farmer (1962) that allophane re-

AMORPHOUS INORGANIC MATERIALS IN SOILS

347

hydrated to a considerable extent after this treatment, whereas the commonly occurring clay minerals, apart from illite, did not. The following general trends were noted for soil clays treated in this manner: (1) the areas of the dehydration peaks of heat-treated clays high in amorphous colloids were 60 to 70 per cent of those of the corresponding untreated clays, and ( 2 ) the size of the dehydration peaks of the pretreated clays containing a higher proportion of crystalline material were 30 to 50 per cent of those of the untreated clay. Illite was present only as a minor component in the soil clays examined by Mitchell and Farmer (1962), and the pronounced dehydration peaks given, especially by the heattreated clays from the surface horizons of highly organic soils, was in agreement with the conclusion that these clays contained a high proportion of amorphous inorganic material. Numerous differential thermal curves are available for the hydrous oxides of iron (Mackenzie, 1949, 1952, 1957b; Kulp and Trites, 1951; Taylor, 1959; Oades, 1962). The differential thermal curves for goethite (Fig. 1, curve e ) and lepidocrocite show endothermic peaks between 300" and 400°C. The peak temperature for goethite is, however, markedly influenced by particle size, and peak temperatures as low as 200°C. have been reported for finely particulate goethite in soil clays (Mackenzie, 1958). Amorphous hydrated ferric oxide precipitates exhibit exothermic peaks between 300" and 500°C. (Fig. 1, curve d ) . The exothermic peak temperatures increase with the temperature of precipitation and with final pH and are considerably modified by the nature of the ions present during the precipitation. The occurrence of hydrous ferric oxide gels in nature, denoted by a sharp exothermic peak at 350" to 400"C., has been reported by Kurnakov and Rode (1928) and Mackenzie (1949). From their investigation of bog-iron ores Kurnakov and Rode (1928) concluded that ferric oxide gels are absent in old deposits and found only in deposits in dynamic equilibrium. The observations on soil clays and laboratory studies on the aging of such gels (Mackenzie, 1957b) agree with this conclusion. The necessity for distinguishing between crystalline and amorphous hydrous oxides of aluminum and iron in pedological studies has been stressed by Gorbunov (1961). However, the resolution of complex admixtures of hydrous iron oxides in soils cannot, as yet, be carried out with complete satisfaction by a combination of X-ray and differentia1 thermal methods. For example, samples of goethite which give identical X-ray patterns give a variety of differential thermal curves. Also the sensitivity of the X-ray method may be such that a material may have a clearly defined X-ray pattern but a thermal curve lacking in characteristic peaks. Mackenzie (1957b) decribed soil clays containing up to 6.7 per

348

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

AMORPHOUS INORGANIC MATERIALS IN SOILS

349

cent free iron oxide, as determined by the dithionite method of Mitchell and Mackenzie (1954), which gave no differential thermal peaks or X-ray or electronoptical pattern that could be associated with iron oxide. He concluded that some amorphous iron oxides may be undetected by normal methods and suggested that these materials may occur on the clay particles as a coating which, although too thin to be detected by electron diffraction, is sufficient to color the soil. 3. Infrared Spectroscopy

The general features of the infrared absorption of amorphous hydrous oxides are best seen in spectra of synthetic preparations, recorded at the Macaulay Institute. The spectra of coprecipitated silica-alumina gels (Fig. 2 a to c ) show progressive changes with increasing A1:Si ratio. The broad absorption due to OH stretching frequencies of SiOH, AlOH, and of associated adsorbed water increases markedly in intensity while the maximum shifts from about 3.0 p to near 2.87 p. Some estimate of the contribution of adsorbed water to this absorptiton can be made from the intensity of the 6.15 p absorption band (De Kimpe et al., 1961), to which only water contributes. Simultaneously, absorption due to Si-0 stretching vibrations weakens and broadens, and the maximum shifts from 9.07 p, for pure silica gels, toward longer wavelengths before its identity is lost at low Si:Al ratios in the general absorption of highalumina gels. This shift in maximum can be ascribed to the formation of a mixed polymer in which Si-O-Al'linkages increasingly replace Si-0-Si linkages (Launer, 1952). Other features of silica gels include a band at 10.55 p due to SiOH groups (Benesi and Jones, 1959) and at 12.45 p, due to Si-0-Si linkages. These also broaden and shift to longer wavelengths (to about 11.4 p and 14.0 p respectively) with increasing alumina content, and finally merge into the high level of absorption characteristic of alumina gels in this region. Pure alumina gels (Fig. 2,c) have rather featureless absorption rising from 8 p to a plateau with ill-defined maxima near 10.6 p and 14.5 p. Touilleaux et al. (1960) have noted the steadily rising absorption near 10.5 p in such gels with increasing alumina content. Comparison with the spectrum of crystalline gibbsite (Kolesova and Ryskin, 1959) indicates that absorption in the 8 to 12p region arises FIG. 2. Infrared spectra of synthetic hydrated oxides and soil clays at the concentrations indicated (milligrams in 0.5 inch diameter KBr disks). Hydroxyl absorption (2.65 to 3.65 p ) recorded after drying in the disk at lOO"C., a, Synthetic silica gel; b, synthetic silica-alumina gel (atomic ratio %:A1 = 1:l);c, synthetic alumina gel; d, soil clay containing a highly hydrated amorphous compound absorbing at 2.89 p, and highly siliceous amorphous material absorbing at 9.2 p; e, soil clay containing poorly crystalline goethite absorbing at 3.12, 11.05, 12.55, and 14.8p.

350

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

principally from AlOH bending vibrations, and that in the 12 to 16 p region from both AlOH bending and A1-0 stretching vibrations, together with absorption due to adsorbed water. Consistent with this interpretation there is a marked drop in the overall intensity of absorption, but particularly in the 9 to 12 p region, when high alumina gels are heated to dehydration. Ferric oxide gels are distinguished from alumina gels by weaker absorption lying at longer wavelengths: thus their OH stetching absorption has its maximum near 3.17 p, and absorption due to FeOH bending and Fe-0 stretching appears as steadily rising absorption in the 10 to 16 p region. Infrared spectra of allophanes from soils and mineral deposits ( Adler, 1951; Fieldes, 1955; Fieldes et al., 1956; Kanno, 1959; Kanno et al., 1960; Yoshinaga and Aomine, 1962a,b) resemble those of synthetic silicaalumina gels. As pointed out by Fieldes (1955) the absence of the absorption bands of pure silica gels in their spectra conclusively shows that most allophanes are not simple mixtures of discrete silica and alumina gels, although free amorphous silica is indicated in the spectra of allophanes found in some young soils (Fieldes, 1955; Fieldes et al., 1956). Allophanes showing S i - 0 stretching absorption with maxima at wavelengths as long as 10.5 p have been reported (Kanno et al., 1960). This indicates a high degree of depolymerization of the silica component, approaching isolated Si-0 tetrahedra in an alumina matrix (Launer, 1952). Others give broad maxima in the 9.0 to 10.0 p range, suggesting a range in the degree of polymerization of the Si-0 tetrahedra. Generally the 10.5 to 1 1 . 4 ~ band found in synthetic silica-alumina gels is absent or weakly developed in natural allophane, indicating a lower proportion of SiOH groups in the latter. In mixed clays, the absorption of allophane in the 9.5 to 10.5 p region is overlain by that of the crystalline clay silicates, the spectra of most of which have now been recorded (Van Der Marel, 1961; Lyon, 1962; Moenke, 1962). The presence of high-silica gels, with absorption maxima in the 9 to 9.5 p region can, however, be easily detected in the presence of many crystalline clays, such as illite, vermiculite, and montmorillonite (Fig. 2, d ) . Due allowance must be made for absorption arising from kaolin and quartz, both of which have absorption at 9.06~1,but the presence of which will be recognized from other characteristic bands in their spectrum. The presence of allophane or amorphous alumina in mixed clays is most clearly indicated by their strong hydroxyl absorption near 2.9 p. Interlayer water in expanding clay minerals also absorbs here, but this water is lost at 100°C. if the minerals are saturated with a monovalent ion, whereas the hydroxyl of allophane is not, and is readily distinguished by this means (Beutelspacher and Van Der Marel, 1961;

AMORPHOUS INORGANIC MATERIALS IN SOILS

351

Mitchell and Farmer, 1962). The ionic-hydroxyl absorption diagrams of most crystalline clay minerals have maxima in the 2.70 p to 2.80 p range, but chlorites have an additional band at 2.90 p (Tuddenham and Lyon, 1959). This chlorite band is narrower than those of allophanes and can be further distinguished from them by observing the spectrum after heating the clay from 300" to 400°C. under conditions that prevent rehydration. The hydroxyl absorption of allophanes is then largely lost (Mitchell and Farmer, 1962) whereas chlorites are stable up to 600 to 700°C. Crystalline gibbsite also absorbs in this region, but can be readily distinguished from allophanes and amorphous alumina by its sharp absorption bands, best seen with the resolution of a lithium fluoride prism or a grating ( Frederickson, 1954). An impure, poorly crystallized gibbsite is unlikely to be distinguished from a mixture of amorphous alumina with better crystallized material. In addition to contributing this 2.90 p band, hydrated alumina and allophane also raise the level of general absorption, at wavelengths longer than 11p, above that given by crystalline clay minerals. Hydrated iron oxides can also be detected in clays by their hydroxyl absorption with maxima lying in the 3.1 to 3.2 p region, but as the absorption of the crystalline forms, geothite and lepidocrocite, lie in this region, and are also rather featureless, they cannot readily be distinguished from amorphous material. Distinctive absorption bands of these crystalline forms occur in the 8 to 16 p region, but these bands are weak, and become diffuse in poorly crystallized material (Fig. 2, e ) , so that they cannot easily be seen in mixtures with crystalline clay silicates. The presence of anhydrous or weakly hydrated amorphous oxides cannot, in general, be so readily recognized in mixed clays as the more highly hydrated forms, although the spectrum of opaline silica differs little from that of more hydrated forms, and can be detected under the same conditions. Amorphous iron and aluminum oxides can be expected to give broad regions of general absorption in the 15 to 2 5 and ~ 13 to 20 p regions, respectively: i.e., regions in which crystalline forms have their principal absorption bands. The absorption of weakly hydrated amorphous aluminosilicates, if present, will be overlain by that of crystalline minerals. The principal contribution of infrared spectroscopy, therefore, is in indicating the degree of polymerization of the silica phase in allophanes and in recognizing the presence of allophanes and hydrated amorphous oxides in mixed clays. The hydrated forms are likely to be the most reactive, and infrared spectroscopy provides the possibility of studying the involvement of hydroxyl groups in such reactions and the nature of the reaction products. In this field Uytterhoeven et al. (1959) and Fripiat et al. (1960) have studied the action of organic reagents on

352

8. D. MITCHELL, V. C. FARMER, AND W. J . MCHARDY

FIG. 3a

FIG.3b

FIG. 3c

AMORPHOUS INORGANIC MATERIALS IN SOILS

353

silica-alumina gels, and numerous investigators ( Basila, 1962; Fripiat et al., 1962) have explored the physical and chemical adsorption of gases on amorphous oxides and the nature and distribution of their hydroxyl groups.

4. Electron Microscopy The degree of order in clay structures varies considerably, mica and aluminosilica gels being the extremes. Because clays are finely particulate, the electron microscope is an ideal instrument with which to examine their diverse morphology. This is clearly brought out by Bates (195S), who reproduced some excellent electron micrographs of clays and other fine-grained minerals. The appearance of clays can be influenced, sometimes considerably, by the method of sample preparation, and in the examination of electron micrographs such pretreatment and its possible effect on the minerals under investigation should always be kept in mind (Bates, 1958). In order to achieve complete dispersion of soil-clay particles, treatment with hot alkali reagents is frequently employed since these will dissolve all but the most resistant amorphous cementing material. Jackson et al. (1950) assessed the efficiency of treatment with hot dilute sodium carbonate solution as an aid to dispersion by taking electron micrographs of the sample before and after treatment. This use of electron microscopy to follow the cleaning-up procedures for clay particles is now fairly common practice. Fieldes and Williamson (1955) carried out an electron microscopic study of clays from New Zealands principal soil groups, describing in dktail the form of crystalline and amorphous clay minerals, and relating this to structure. Their observations, particularly those on the amorphous materials, are most signscant. Thin sheets of amorphous silica were noted in all clays containing the secondary silica chalcedonite, and small aggregates of amorphous silica with perhaps alumina were observed in the electron micrographs of most clays. They also noted that if amorphous alumina was present in excess of silica it tended to coat the surfaces of clay minerals. Amorphous alumina appeared as a cloud of finely particulate material, incapable of resolution by electron optics, when present in greater amount than the negative colloids. From electronoptical data Fieldes ( 1955) distinguished the finely particulate alloFIG.3. Electron micrographs: ( a ) allophanic clay from Foula, Shetland. ( b ) Allophane from Woolwich, Kent; although there is some degree of order, this material was amorphous to X-rays and electrons. ( c ) Finely particulate allophane occurring as a coating on an illite crystal.

354

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

phane B and the coarser allophane A. Yoshinaga and Aomine (1962a,b) found imogolite to have a fibrous structure as distinct from allophane which was composed of minute particles in aggregated masses. In one Japanese soil imogolite accounted for 20 per cent of the clay fraction. Occasional particles in imogolite preparations were seen to be fluffy ( Aomine and Yoshinaga, 1955; Yoshinaga and Aomine, 1962b). Robertson ( 1963), examining an allophanic soil from Oregon, noted spongelike, presumably amorphous, masses and very thin ribbons which in places seemed to be coalescing into very thin crinkly sheets. Electron micrographs of amorphous aluminosilicates obtained at the Macaulay Institute (Fig. 3a,b, and c ) show that whereas such material may be completely amorphous to X-rays and electrons it may nevertheless possess some degree of organization. The authors of these various electronoptical studies of the amorphous inorganic fraction of soil clays are careful to stress the tentative nature of their interpretation of the results. The real value of these results, however, does not lie in these initial interpretations, but in the indication that they give of the potential of electron microscopy in the investigation of noncrystalline constituents of soil clays.

5. Surface Area By analogy with the high specific surface areas characteristic of amorphous oxides used for catalytic purposes, it is reasonable to suppose that amorphous material in soil would have a high specific surface area and that this characteristic could be of some value in its estimation. The most widely used method of measuring the surface area of fine particles is by the adsorption of inert gases at temperatures near to their condensation point: for example, nitrogen at the temperature of liquid nitrogen or oxygen. When the adsorption data are analyzed according to the Brunauer-Emmett-Teller ( B.E.T. ) theory of multimolecular adsorption ( Brunauer et al., 1938), they yield a value for the volume of gas required to form a monolayer on the surface of the particles. With an appropriate value for the cross-sectional area of the molecule and assuming close packing, this can then be converted to a surface area. The technique is widely described in the literature, and many modifications and simplifications have been suggested for the procedure, for example, by Bugge and Kerlogue (1947), Starkweather and Palumbo ( 1957), Birrell and Packard ( 1958), Lippins and Hennans ( 1961), and Amiel et al., (1961). Two instruments are now available commercially for the measurement of specific surface area, one based on the continuous flow method of Nelsen and Eggertson (1958); the other is the

AMORPHOUS INORGANIC MATERIALS IN SOILS

355

Atlas Betograph depending on the principle enunciated by Schlosser (1959). Emmett et al. (1938) were first to apply the B.E.T. theory to the measurement of the surface areas of soil colloids, and Makower et al. (1937) explored the use of specific surface area measurements as a means of calculating the amount of colloid in the soil. Nelson and Hendricks (1943) also concluded that the colloid content of soils could be determined from gas sorption measurements. They found that the average particle sizes of a number of clay minerals, soils and soil colloids were in essential agreement with values obtained from electron micrographs. They also found that the particle sizes of colloids from the surface horizons of five soils of widely different type were greater than those of the B and C horizons. Burford et al. (1964), applying the B.E.T. method to soil clays, obtained very low values which increased markedly after treatment with hydrogen peroxide, and this observation has been confirmed in the authors’ laboratory. Burford and co-workers concluded that “combined organic material present in naturally occurring clayorganic complexes prevents access of nitrogen to some of the clay surfaces. Dyal and Hendricks (1950) introduced a method for determining the surface area of clays based on the retention of a monomolecular layer of ethylene glycol. A weighed sample of clay was wetted with an excess of glycol which was subsequently removed by vacuum distillation over anhydrous calcium chloride. Establishment of a monomolecular layer was assumed when the rate of loss of glycol became very low. Martin (1955) modified the procedure by introducing a source of free glycol vapor so that equilibrium was established more quickly. He did not attempt to relate the “glycol retention” values obtained under these conditions to specific surface area. Further refinements were made by Bower and Goertzen (1959), who equilibrated the sample with glycol vapor at the equilibrium pressure of a calcium chloride-glycol solvate. In addition to ethylene glycol, glycerol retention has been proposed by Diamond and Kinter (1958) as a relatively simple and rapid method of measuring the surface areas of both expanding and nonexpanding clays. Regarding the applicability of glycol or glycerol retention to amorphous materials, Gradwell and Birrell (1954) claimed that allophane resembled montmorillonite in possessing an apparently high cation-exchange capacity and a large total surface area as measured by glycol retention. Unlike montmorillonite the surface area of allophane remained virtually unchanged after heating to 65OOC. Aomine and Yoshinaga (1955) found that, although allophane was the predominant clay mineral in the welldrained soils on volcanic ash which they studied, the ethylene glycol retention of the clays was reduced after heating to 60O0C.,but not to

356

B. D. MITCHELL, V. C. FARMER, A N D W. J. MCHARDY

the same extent as montmorillonitic soil clays, Kinter and Diamond (1960) critically examined the 600°C. heat treatment of minerals and found that it reduced the external surface area of some minerals while increasing that of others. They advocated saturation with the triethylammonium cation as an alternative to heat treatment. They found that glycerol retention of an allophanic Japanese Ando soil was almost as high as that of montmorillonite but remained unchanged after saturation with triethylammonium, whereas that of the montmorillonite decreased. The 600°C. pretreatment, in contrast, resulted in a reduction of retention to approximately half that of the untreated or triethylammonium-saturated samples. In general, therefore, the conditions for the formation of a monolayer or duolayer with such adsorbates as glycol and glycerol are not well defined for the variety of surfaces found in soils. Greenland and Quirk (1964) have proposed the adsorption of cetylpyridinium bromide as a method of determining the total specific area of soils, retaining the lowtemperature adsorption of nitrogen as the most reliable means of estimating external surface area. They have dealt with advantages and limitations of the cetylpyridinium bromide method, and comparison has been made with ethylene glycol retention data obtained by the method of Bower and Goertzen (1959). One of the disadvantages of the use of cetylpyridinium bromide pointed out is that it does not form a complete monolayer on materials with a low surface charge density. This, however, could be of advantage in the investigation of amorphous material and there is reason for thinking that cetylpyridinium bromide adsorption, together with low-temperature nitrogen adsorption data, could be of considerable value in such a study.

6. Chemical Analysis Doubtless because of the development of powerful instrumental aids for the examination of clay minerals and the realization that clays are invariably mixtures of minerals, the value of chemical analysis in their study has for many years been given little consideration. Mackenzie ( 1960), Mackenzie and Robertson ( 1961), and Robertson ( 19f3) have indicated that from chemical analysis of a clay, and a certain amount of basic mineralogical data, it is possible to evaluate the composition of the principal clay minerals and also to determine accurately the mineralogical constitution of the clay. Because this has become possible only with recently acquired knowledge of clay minerals and their interrelationships, it is conceivable that chemical analysis will have an increasing function in clay mineralogy.

AMORPHOUS INORGANIC MATERIALS IN SOILS

357

7. Diflerential Dissolution Because no naturally occurring clays, and certainly no soil clays, have been found to be monomineralic and free from accessory material, studies based upon the relative stability of a component or group of components in clays to specific chemical reagents seems an obvious approach to the problems of identification and estimation of the constituents of such complex systems. Chemical degradation procedures designed to improve dispersion of the soil and to clean up the clay, basically in order to identify more accurately by instrumental techniques the better-ordered structures within the clay, have been discussed under pretreatment. They, however, represent only a few of the selective or differential dissolution procedures which have been used by clay mineralogists in attempts to isolate components of the clay fraction and to determine their structure and properties. Treatment with inorganic acids has been used for the selective dissolution of crystalline material-for example, Brindley and Robinson (1951) have used digestion in warm dilute HC1 to dissolve chlorites. Earlier, Pask and Davies (1945) observed that the amount of aluminum extracted from minerals by digestion with sulfuric acid was related to the temperature to which the mineral had been heated. Preheating of clay minerals to their dehydroxylation temperature modifies their solubility in dilute acids. Brindley et al. (1951) found that chlorite after heating to 500°C. dissolved easily in HCl, and Steger (1953) reported that kaolinite behaved in a similar way. Gastuche (1959b) investigated the effect of various chemical reagents (for example, 2N HCI, NaCl, MgClz solution ) on kaolinite using principally electron microscopy to observe alterations. Oberlin and Tchoubar (1961) also used electronoptical techniques in their study of the effect of sulfuric acid on kaolinite, concluding that the pH of the solution and the presence of soluble salts such as magnesium sulfate largely determined the form of the product. A report on the preliminary study of the structure of glauconite based upon the effects of acid dissolution has been given by Cloos et al. (1961). The action of mild acid solutions on clays and gels was followed by Fripiat (1960), who states that such treatment provides indications of the relationship between aluminum located at the surface of the particle and the charge. Gastuche et al. (1960), from an examination of the solubility kinetics of kaolinite and of a cracking catalyst in HCl found that no activation energy was involved in the initial extraction of aluminum and also that the aluminum content was related to the magnitude of the cation-exchange capacity and independent of the nature of the saturating cations.

358

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

For many years the usual method of determining soluble silica in soils and minerals has been treatment with hot Na2C03solution (Salvetat, 1851), and, in general, alkaline reagents have been used to dissolve amorphous aluminosilicate, free silica, and alumina, while acids have been employed for the selective dissolution of crystalline material. Boiling with dilute Na2C03 solution removes the finely divided amorphous siliceous material which may be cementing the clay particles (Jackson d al., 1950). Hashimoto and Jackson (1960) reported the effect of treatment with hot NazCo3 on a highly weathered ferruginous humic latosol with a high content of amorphous aluminosilicate stating that only small amounts of silica and alumina were dissolved. They concluded that digestion with NazCo3 solution may not remove completely the amorphous aluminosilicates, and suggest that this may be due to the slow rate of reaction and to saturation of the extracting solution. Foster (1953), in order to determine the free silica and alumina in montmorillonites, digested specimens with 0.5 N NaOH for 4 hours. This procedure removed completely the free silica and alumina and, although it attacked crystalline clays to some extent, was considered superior to digestion with 5 per cent NaaC03 solution. Tests showed that, with this reagent, solution of opal was never complete, irrespective of the amount present or the time of digestion. A method of distinguishing goethite from gibbsite in soil clays, proposed by Muhoz Taboadela ( 1953), involved treatment with 5 per cent NaOH on a steam bath for 20 minutes; this is effective in removing gibbsite but brings about considerable destruction of the crystalline clay minerals such as halloysite. It has been noted above that preheating clays to their dehydroxylation temperature affected the amount dissolved by acid treatment, and Hislop (1944) found that dehydroxylated kaolinite was stable to Na2C03 treatment. Hashimoto and Jackson ( 1960) investigated the differential dissolution of clays with NaOH solution and found that substantial amounts of allophane, free silica, and alumina were brought into solution by boiling for 2.5 minutes with 0.5 N NaOH solution. To avoid reprecipitation of silica they stated that the ratio of clay to caustic solution should always be less than 100 mg. to 100 ml., and noted that prolonged boiling brought about marked destruction of the crystalline components. The iron released by this procedure was removed by the dithionite-citratebicarbonate method of Mehra and Jackson (1960). Allophane and the free iron oxides were therefore removed by these two procedures. Hashimot0 and Jackson ( 1960), however, went further, finding that dehydroxylated kaolinite and halloysite were dissolved by boiling with caustic solution for 2.5 minutes whereas chlorite and montmorillonite heated to

AMORPHOUS INORGANIC MATERIALS IN SOILS

359

500°C. were virtually unaffected by this digestion technique; interlayered alumina was rendered soluble to 0.5N NaOH by heating to 400°C. for 4 hours. They were thus able to produce a flow sheet for the selective dissolution of allophane, gibbsite, amorphous silica, and interlayered alumina, kaolinite, and halloysite. Mitchell and Farmer (1962) in an investigation of amorphous inorganic material in brown forest soils of low base status and gray-brown podzolic soils found that successive digestions on the steam bath with 5 per cent Na2C03 solution effectively removed highly hydrated amorphous inorganic material. A New Zealand allophanic soil clay was also largely dissolved by this treatment. Studies in progress in the authors’ laboratory have shown that digestion with hot carbonate solution also removes interlayered aluminum from some soil clays. Moreover, it was found that finely particulate crystalline minerals could dissolve as easily as amorphous inorganic material, and consequently, after removal of free iron, they have adopted a procedure of exhaustive extraction with cold 5 per cent Na2C03 solution followed by digestion on the steam bath with 5 per cent Na,2C0,. This treatment with Na2C03 solution does not in every instance remove all the amorphous material, a result suggesting that the more resistant amorphous material is somewhat better organized. It has previously been mentioned that digestion with sodium citrate solution may be used to remove interlayered alumina and the number of digestions required depends upon the stability of the particular interlayered material. Like alkaline extractants, sodium citrate will probably attack allophane and amorphous alumina. The acid oxalate solution of Tamm (1922) is the classical extractant for free alumina in soils, and it must also dissolve aluminum from labile allophanes. It is surprising that, so far as can be ascertained, this solution has not been used in the investigation of interlayered alumina. A few of the numerous methods proposed for the removal of hydrous iron oxides from soil were considered under pretreatment (Section 111, A ) , and it was noted that the methods apparently varied considerably in their efficiency. This varying ability of the methods to remove hydrous oxides from soils may, however, reflect the selectivity of a reagent for a particular form of hydrous oxide. Gorbunov et al. (1961) used a combination of five methods to distinguish between amorphous (“free”), and crystalline sesquioxides. The five methods were (1) acid ammonium oxalate (Tamm, 1922); ( 2 ) 0.2 N HC1 (Kirsanov, 1958); ( 3 ) aqueous solution of complexone plus potassium chloride (Stefanovits, 1955); ( 4 ) sodium dithionite (Deb, 1950); (5) sodium dithionitecitrate-bicarbonate (Mehra and Jackson, 1960). They found that none of the methods dissolved the hydrous oxides in a single treatment, that

360

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

Tamm’s or Mehra and Jackson’s method dissolved most of the amorphous sesquioxides, and that Deb’s was specific for hydrous iron oxide. The methods of Kirsanov and Stefanovits removed appreciably less amorphous sesquioxides than the others. Although the findings of this investigation are not altogether conclusive, they nevertheless further serve to indicate the value of selective dissolution techniques in assessing the stability of soil clay constituents. The stability of such constituents to a chemical reagent cannot be directly equated to the inherent degree of order since other factors, such as particle size, may be involved but the ease with which a component in the clay is removed by a reagent may be a reasonable guide to its activity and mobility in the soil. It would appear, therefore, that the development of selective dissolution techniques represents one of the most promising approaches to soil clay studies.

8. Cation-Exchange Capacity High cation-exchange capacities of soils containing no detectable smectites, vermiculites, or illites have been attributed to the presence of amorphous hydrous oxides. In their examination of lateritic soils Fieldes et al. (1952) concluded that the organic matter content was not sufficient to account for the cation-exchange capacities found ( u p to 44 meq./ lOOg.). From differential thermal and chemical analyses of the soils and from a study of synthetic amorphous hydrous oxides they concluded that the principal exchange material was amorphous hydrous alumina. Birrell and Fieldes (1952) attributed the high exchange capacity (54 meq./ lOOg.) of a clay from a volcanic ash soil to the presence of allophane, and Aomine and Yoshinaga (1955) obtained comparable exchange capacity results for Japanese soils developed on material of recent volcanic origin. The cation-exchange capacity values for soils containing allophane and amorphous oxides were subsequently found (Birrell and Gradwell, 1956) to vary with ( 1)concentration of the leaching solution, ( 2 ) nature of ion in solution, ( 3 ) the volume and water content of the washing alcohol. Egawa et al. (1959) confirmed these findings, noting also that the effect of these factors on the exchange capacity could be suppressed to some extent by either air-drying the samples or by heating to 100°C. They found that this also applied to measurement of the exchange capacity of silica, alumina, and alumina-silica gels. Wada and Ataka (1958) investigated the dependence of the anionas well as the cation-exchange capacities on the concentration and the pH of the equilibrating solution for allophanic, montmorillonitic, illitic, and halloysitic soil clays. Their cation-exchange capacity results agreed with those of Birrell and Gradwell (1956), and the anion-exchange ca-

AMORPHOUS INORGANIC MATERIALS IN SOILS

361

pacities followed a parallel dependence and was greater or less than the cation-exchange capacity according to the pH of the equilibrating solution. The difference between anion- and cation-exchange capacities showed no appreciable change with concentration of equilibrating solution. Wada and Ataka (1958) distinguished between coulombic and noncoulombic adsorption, both of which contributed to the measured cationand anion-exchange capacities. The former is the normal ion-exchange reaction and is equivalent to the number of positive or negative sites on the mineral lattice, whereas the noncoulombic adsorption is the salt held in the interstitial solution and involves an equivalent amount of cations and anions. Ion uptake by a montmorillonite-illite clay was predominantly by coulombic adsorption, because of the negative charges produced by isomorphous replacements within the crystal lattice, and the anionexchange capacity was only 2 per cent of the cation-exchange capacity and virtually independent of the concentration of the equilibrating sohtion. The halloysitic clay showed noncoulombic adsorption due, it was thought, to the formation of interlayer complexes between halloysite and specific salts. The allophanic soil clay possessed both positive and negative charges even at pH 7, and noncoulombic adsorption accounted for the greater part of the cation- and anion-exchange capacities. The charges carried by allophane, unlike the montmorillonite-illite and halloysitic clays, were greatly affected by the pH of the solution, the cation-exchange capacity being reduced considerably between pH 7 and pH 5. The physical adsorption of ions by allophane and related amorphous oxides is established by the work described above, and it is readily understood when the high content of sorbed water in these materials is considered. The water of hydration will be, in fact, a salt solution the concentration of which depends upon the external equilibrating solution. The salt solution is not readily washed out by alcohol although the use of water greatly reduces the amount of cation physically adsorbed (Birre11 and Gradwell, 1956). The removal of excess salt in cation-exchange capacity determinations was examined extensively by Rich ( 1962), who concluded that the process was largely controlled by the solubility of the salt in the solvent. For clays which had anion retention properties, subsequent salt removal apparently depended upon cation- and anionexchange reactions, these being controlled by the degree of dissociation of the solvent and the diffusion rates of the ions involved. It has been suggested by Wada and Ataka (1958) that the cationexchange capacity for clay minerals, including allophane, may be defined as the negative charges carried by the clay particles due to isomorphous replacement in the lattice, to broken bonds, to dissociation of H + from OH groups and to other possible mechanisms which would

362

B. D. MITCHELL, V. C. FARMER, AND

W.

J. MCHARDY

depend on the pH but not on the concentration of the equilibrating solution, Rich (1962) suggests that the cation-exchange capacity, of anion-retaining soils at least, be determined as the “net negative charge” by saturating with calcium chloride, removing most of the excess salt with two water and three methanol washes, and then determining both the Ca++ and the C1- extracted by another salt solution; the cationexchange capacity would be given by the difference between the Ca+ + and C1- contents. Adsorption phenomena are particularly significant for clays rich in amorphous and poorly crystalline sesquioxides and lead to their possible use in the detection and estimation of such materials. Birrell and Gradwell (1956) showed that the uptake of cation conforms to the BrunauerEmmett-Teller ( B.E.T. ) theory of multilayer physical adsorption. Birrell (1961a) extended the analogy by applying the B.E.T. theory to calculate the amount of surface covered by physically adsorbed salts knowing the effective sizes of the ions. In soil clays suspected of containing amorphous material conventional methods of determining cation-exchange capacities should be supplemented with equilibrium experiments in order to test the significance of the results. Aomine and Jackson (1959) suggested a method of estimating allophanic material based upon what they termed the “cation-exchange capacity delta value.” One sample of a hydrogen peroxide-treated, ironextracted soil was treated with 2 per cent Na2C03solution for 60 minutes and another with boiling sodium acetate at pH 3.5 for 15 minutes. The exchange capacity of the sample was determined after each treatment, and the difference represented the delta value. It was found that the delta value was very large for allophanic clays, the average of several from Japanese Ando soils being 100 meq./lOOg. Large values were obtained for halloysite and montmorillonite, but those for kaolinite, gibbsite and quartz were very small or nil. Aomine and Jackson (1959) claimed success for the use of this property in detecting and estimating allophanic material even in the presence of appreciable amounts of crystalline clay minerals. 9. Physical and Chemical Studies of Clay Organic Complexes

The study of complexes and compounds of organic reagents with clays, by X-ray, differential thermal and infrared techniques, can greatly assist the differentiation of clay components, and provides valuable information on their surface properties. Clay minerals may form two types of complex with organic compounds: (1) ionic complexes in which organic cations. ( e.g., piperidinium ions) replace the exchangeable inorganic cations; and ( 2 ) molecular complexes in which an organic liquid

AMORPHOUS INORGANIC MATERIALS IN SOILS

363

( e.g., glycerol, pyridine, nitrobenzene) replaces the sorbed water. The differential thermal curves of these clay organic complexes show considerable differences, and it has been suggested from time to time that complexing of clay minerals with organic compounds could possibly assist in their identification by differential thermal analysis, but as yet this procedure has not attracted a great deal of attention. Allaway (1948) found that when a piperidine-treated clay was examined by differential thermal analysis a stepwise combustion of the absorbed piperidine was frequently observed and suggested that the temperature at which this occurred might be related to the composition of the clays. Carthew (1955) and Oades and Townsend (1963) used the piperidine saturation techniques as an aid to the identification of crystalline clay minerals by differential thermal analysis, and Greene-Kelley ( 1957) carried out similar experiments with montmorillonites using triethylammonium saturation. Ramachandran et al. ( 1961 ) investigated the mechanism of the thermal decomposition of complexes formed by montmorillonite with piperidine, malachite green, and methylene blue. Sudo (1954), in a study of the alteration of volcanic glasses, used the piperidine saturation technique. He found that the differential thermal curve of a piperidine-saturated allophanic clay from a soil developed on volcanic ash showed a broad, diffuse exothermic peak in contrast to that of piperidine-saturated allophanic material from a soil on pumice, which showed a relatively sharp exothermic effect. It seems possible that with the development of controlled-atmosphere and gas sampling techniques for differential thermal analysis, coupled with the improvements in sensitivity of recording, further work on the thermal characteristics of these clay-organic complexes could prove of value in the identification and semiquantitative estimation of amorphous inorganic material. Organic derivatives of silica gels have been studied by Deuel (1954). Acetyl derivatives of silica gels, aluminosilica gels, and kaolinite have been prepared by Cloos and Fripiat (1958) and Uytterhoeven et al. (1959), using either acetic acid or acetyl chloride. Examination of the acetylated gels by infrared spectroscopy and differential thermal analysis showed that all the OH groups were not on the external surfaces: some were located on the surfaces of the pores or channels in the gels. The OH groups on the internal surfaces were, it was found, not affected by the reagents, and by employing reagents of different reactivity it was possible to ascertain differences in the reactivity of the OH groups present in the gel structures. The surface content of OH groups in amorphous aluminosilicates has been determined by Uytterhoeven and Fripiat (1962) from an examination of methylated aluminosilicates. Using thermogra-

364

B. D. MITCHELL, V. C. FARMER, AND W. ;I. MCHARDY

vimetry and infrared spectroscopy, they have shown, for example, that in an amorphous aluminosilicate containing 13 per cent Al,ZO, the maximum density of surface OH groups per square mp, excluding water of hydration, varies from 8 to 4 depending on the pretreatment temperature. Nuclear magnetic resonance has been used to determine the average distance between hydroxyl groups in layer silicates (Gastuche et aZ., 1963), and this new technique could conceivably prove of value in the study of amorphous silicates. Some of the hydroxyl groups on the surface of aluminosilica gels react with diazomethane, a finding which suggests that the hydroxyl surface is heterogeneous (Fripiat et d.,1954). This heterogeneity depends upon the nature of the cation-hydroxyl bond and also on the distribution of hydroxyl groups. The hydroxyl surface of silica gel, for example, may be regarded as heterogeneous because of interactions between hydroxyls and between hydroxyls and water molecules. The heterogeneity of the silica gel surface has also been demonstrated by following the isotopic exchange OH-OD by infrared absorption (Fripiat et al., 1962). These physicochemical techniques currently being used and developed by Fripiat and co-workers are bringing precision to the measurement of the surface properties of synthetic gels and should prove of considerable value in the comprehensive study of amorphous inorganic material in soils. 10. General Comment It is apparent from the foregoing account that some of the methods referred to have contributed more than the others to our knowledge concerning the nature of amorphous inorganic material in soil. All available methods, however, should be used since, as expressed by Mackenzie ( 1 9 5 7 ~ )“every method, whatever its nature, will add its own quota of information and fill in some of the shadows or emphasize some of the highlights of the picture.” IV. Origin of Amorphous Material in Soil

In the study of soil clay as part of a static system a great deal of effort has been devoted to the identification and classifmation of the constituents. Although such work is of the utmost value attention must also be directed toward the genesis of clays in the soil profile. This involves studying the soil as a dynamic system. The distribution of individual soil constituents in the profile has to be determined, and the possible origins, for example, physicial transportation or chemical alteration, have to be evaluated; clearly the difficulties presented in such studies are considerable. Deuel (1960) commented that soil chemistry was in its infancy as

AMORPHOUS INORGANIC MATERIALS IN SOILS

365

evidenced from the fact that, “in most publications no chemical formulae at all are given and if such formulae are presented, they are in many cases more symbols of our imagination and ignorance than symbols for the arrangements of atoms in the soil.” Of course, in some senses, science is always in its infancy. The inorganic compounds in the soil are derived mainly from the parent material which is present from the start of soil formation whereas the organic compounds are derived from the biological material which is constantly being replenished. Concepts of soil genesis and soil processes have changed over the years in order to accommodate increases in knowledge and understanding of soil properties ( Mattson and Gustafsson, 1937; Wiklander, 1945; Nikiforoff, 1959; Simonson, 1959). The increasing recognition of the importance and distribution of amorphous components in soils must inevitably lead to such changes. In this section, mechanisms which have been suggested, and in some instances established, for the presence of various amorphous soil components are presented. A. WEATHERING Determination of the weathering stage of a soil takes into consideration the concentration and properties of the clay fraction, and any study of soil clay formation must take into account the nature of the parent rocks since variations in these may produce important changes in the parent material and hence in the soil clay. Differential weathering is the process by which different parts of a rock mass weather at different rates, and Leet and Judson (1960) consider that inequality in rates of weathering are related to composition of the rock and to the intensity of weathering. Cady (1960) found that the rate of weathering of silicates in hard igneous rocks and metamorphic rocks was distinct from that in unconsolidated materials like glacial till and volcanic ash. In a study of the differential weathering of volcanic ash and pumice, Aomine and Wada (1962) observed that the relative stability of primary minerals to weathering increases in the following order: volcanic glass < feldspar 5 hypersthene-augite < magnetite. According to Bhattacharya ( 1962) mineral stability in the weathering environment is determined by the crystal structure of the minerals and the degree of isomorphic substitution in the lattice. McKeague and Cline (1963) have considered stability in relation to structure, composition, and particle size. The genesis of clay minerals has been discussed by Gastuche (1959a) and Rich and Thomas (1960). Assessment of available information indicates that although soil clay minerals may be inherited directly, and in some instances quite substantially, from the parent material, they

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

366

may also be altered from minerals of similar structure, or synthesized from dissolved and amorphous products of weathering ( Mitchell, 1955). Chemical weathering can be regarded as a typical solid state surface reaction, the initial reaction being hydration of the surface of the primary mineral. The decomposition reaction is mainly one of hydrolysis and ion transfer ( Frederickson, 1951; Barshad, 1955). Large cations are most readily mobilized and released, and even silica and alumina may pass into solution. Yaalon (1959) represented the weathering reaction as: aluminum silicate water + ions and amorphous intermediate products + clay soluble salts. Dissolution of the solid forms of silica and silicates are affected by the nature of the surfaces ( McKeague and Cline, 1963); although the process requires only water and can occur in neutral and alkaline media, it is usually accelerated under acid conditions. The silica released by weathering which is not involved in clay mineral synthesis is usually removed from the system in solution as undissociated monosilicic acid (Krauskopf, 1956). Bastisse (1960), has, however, shown that silica can occur in drainage waters and in soil solutions as pseudo-complexes with iron and other metals. Very little iron is incorporated in the lattice of clay minerals, and the fate of the excess iron and aluminum depends upon the intensity of leaching and the pH of the soil solution (Carroll, 1958; Yaalon, 1959). Under conditions of intense leaching and provided the soluble products of decomposition of the weathering reaction are continually removed, in time most of the primary minerals will decompose with the formation of clay materials. Crompton ( 1960) emphasizes the necessity for appreciating differences in the relative rates of weathering and leaching. In the weathering of feldspars (Fieldes and Swindale, 1954) the rate of formation of diphormic minerals may be comparable with the rate of mobilization and release of silica and aluminum: consequently little or no amorphous material may be present. More easily weathered parent materials, such as volcanic glass and basic igneous rocks appear, however, to pass through a stage where highly amorphous clays predominate. The rate of leaching in the weathering horizons is difficult to ascertain, as a fundamental feature of soil formation is that the living matter present retains the mobile compounds in amounts which are greater than their annual loss by leaching (Nikiforoff, 1959). Only the material in excess of this biotic requirement is removed in the drainage water or may accumuIate until the steady state is attained. Living matter, by trapping the released ions and building them into its structure, counterbalances the effects of weathering.

+

+

AMORPHOUS INORGANIC MATERIALS IN SOILS

367

B. SILICA Amorphous silica has been observed in young soils developed on volcanic ash (Fieldes and Swindale, 1954; Fieldes and Williamson, 1955; Matsui, 1959), and with age apparently it crystallizes to cryptocrystalline chalcedonite ( Fieldes, 1952) and eventually perhaps to secondary quartz. McKeague and Cline (1963) in a review of silica in soils cite a number of references to secondary quartz in soil and comment that it is not known whether this quartz is formed directly from solution or represents an ordered form of amorphous silica. Mohr ( 1944), examining laterite developed on volcanic ash, noted that silica was precipitated at depth, and Lindqvist (1959) reported chalcedony at the base of a laterite-like profile. Little attention has been given to the silica liberated in the laterization process, it being generally assumed to have been transported from the profile in the leaching solution. Lindqvist suggests that probably alterations in the level of the groundwater table influence the movement and flocculation of colloidal silica. Bates (1960) also comments that silica released from minerals in one part of the profile may be temporarily reconcentrated in another and has found soils in which, 2 to 3 feet below the surface, the concentration of silica was greater than in the horizons above and below. He suggests that resilicification possibly occurs in dry conditions owing to the impounding of silica-rich solutions; such factors as variations in rock composition and texture may impede the attack of leaching solutions. So far, only inorganic mechanisms have been considered for the deposition of silca in soil, but there is considerable evidence to show that the biotic factor is also involved. McKeague and Cline (1963) provide a volume of evidence to substantiate the claim that the action of higher plants provides one of the principal mechanisms for the occurrence of silica in soil. They quote Russell (1950) as stating that perhaps twice as much silica is cycled annually through plants as is lost in drainage water, and they conclude: “It is apparent that biological depositions of silica proceed on a grand scale in soils.” Opaline silica in surface soils may well be derived almost exclusively from plants. Nevertheless inorganic mechanisms with a biotic basis have been postulated for concentration of silica in the upper horizons of the soil profile. VBmos (1961) for example suggested that, when alkaline soils were under water, anaerobic conditions induced by bacteria led to the production of iron sulfides in the upper horizons. Sulfuric acid formed by the subsequent atmospheric oxidation of these sulfides could attack soil minerals, leaving amorphous silica or silica-rich residues. Glenn et al. ( 1960), and Mitchell and Farmer ( 1982) have also ascribed high silica contents in the clay fraction of the surface

368

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

horizons of well-drained soils to siliceous relics left by the mobilization of alumina. In these soils the attacking agents are assumed to be humic acids and organic acids released by plants and soil microorganisms. Silica of biological and inorganic origin should be distinguishable by optical microscopy because plant phytoliths have distinctive forms.

C. ALUMINUMOXIDES Reports of particulate amorphous aluminum oxides in soils are rare, but this may well be due to the difficulties in establishing their presence. Krauskopf (1959) considers that aluminum oxides have a specific affinity for silica in solution, and Rich and Thomas (1960) suggest that the lack of amorphous alumina in soils may be due to it being stabilized by silica in allophane. Fieldes and Williamson (1955) observed that when alumina is in excess of silica it forms a coating upon negatively charged colloids such as layer silicates, and, when present in excess of all negatively charged colloids, occurs in the system in an ultrafine form beyond the resolution of the electron microscope. Precipitation of alumina in the interlayer space of expanding layer lattice minerals leads to the formation of the intergradient minerals showing some of the features of chlorites, and these are of widespread occurrence, particularly in the upper horizons of leached soils such as podzols and brown forest soils, provided the organic matter content is low (Sawhney, 1960). Jackson (1962, 1963) considers that intergradient 2: 1, 2: 2, layer silicates are developed by the interlayer precipitation of hydroxides of aluminum, iron, and magnesium in the course of pedological weathering and probably also during the burial of sediments. Rich and Thomas (1960) have discussed the mechanism of aluminum interlayer formation suggesting that the process is most likely controlled by the production of Al+ + + ions, under acid weathering conditions. The Al+++ ions are then hydrolyzed, polymerized, and finally fixed. They have also shown that the maximum formation of aluminum interlayers is closely related to the organic matter content of the surface horizons. When the organic matter content of the surface soil is low, maximum aluminum interlayer formation occurs in the A horizon, but when the surface layers are rich in organic matter aluminum interlayers are at a maximum in the B horizon. The free alumina and exchangeable aluminum in soils have been discussed in relation to soil acidity by Hattori and Kawaguchi (1959) and Blanchet et al. (1960) and proposed mechanisms for the release of aluminum have been treated by Rich and Thomas (1960) and Tsyurupa (1961). Recently Schwertmann and Jackson (1963) have given a com-

AMORPHOUS INORGANIC MATERIALS IN SOILS

369

prehensive account of the changes involved in the aging of H-montmorillonite to Al-montmorillonite. OXIDES D. IRON In view of the difficulties involved in distinguishing amorphous and crystalline iron oxides in soils, and the few references to the occurrence of amorphous oxides, it has been necessary to include in this section consideration of the origin of the somewhat ill-defined free iron oxides. Until proved otherwise free iron oxides must, in general, be regarded as including a proportion of amorphous iron oxide. Like the oxides of silicon and aluminum, those of iron are either residual from the parent material or weathering products of iron-bearing primary minerals. The mechanisms for their mobilization and deposition may be physical and/or chemical, involving abrasion, comminution, solution, chelation, and reduction. H6nin (1956) has reviewed studies on solution and precipitation of iron in the formation of soil types. The deposition of iron oxide coatings on clay minerals has been investigated by Fripiat and Gastuche (1952), who noted that the morphology of these oxides on the surface of kaolinite depended upon the pH of the environment and the cations present. When the kaolinite was H-saturated the surface was porous and disordered, and absorbed large amounts of iron oxide. If, however, Na, K, Mg, or Ca ions were present, the surface of kaolinite was ordered and nonporous and consequently was quickly saturated with iron oxide. Sumner (1963) studied the effects of iron oxides on the positive and negative charges of clay minerals and soil clays, and found that a considerable portion of the clay was covered with iron oxide. This reduced the negative charge of the clay particles, presumably because of a positive charge in the oxide coating. Carroll (1958) considered the role of clay minerals in the transportation of iron and discussed the nature and stability of iron oxides in relation to environment, concluding that the form of iron oxide gives an indication of rock weathering. The distribution of free iron oxide in the soil profile serves, according to Gorbunov et al. (196l), to indicate the soil processes involved, and the same authors consider it unfortunate that so little attention has been paid to distinguishing amorphous iron oxides from crystalline forms. It has been noted above (Section 11, C ) that accumulation of iron oxides is often found in the upper horizons of intensively weathered and highly leached soils, where they probably arise as residues from primary silicates. Different mechanisms must, however, be involved in the localized accumulations found in such features as rusty mottling, concretions, weathering crusts, and hard pans studied by Schwertmann (1959). This author also noted amorphous iron oxide in addition to crystalline iron

370

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

oxide in fine-textured gley soils in which a fluctuating water table results in rapidly alternating oxidation and reduction conditions. Kamoshita and Iwasa (1959), on the other hand, observed only lepidocrocite in gleyed paddy soils. The organic fraction of the soil has long been associated with the mobilization and transportation of iron within the profile. Recently Martin (1960) has studied the flocculation of humus by ferric and ferrous iron, and Duchaufour (1963) has investigated the role of iron in clay humus complexes. Kaurichev and Nozdrunova ( 1961) have studied iron compounds in forest podzol and meadow sod podzol soils. They noted that under conditions of high acidity the content of mineral iron was high, but a more important observation was that during the wet season iron compounds migrated in these soils in the form of water-soluble organic-iron complexes. The discussion of the vast field of organomineral complexes is outside the compass of this review: brief reference has only been made to indicate that the form and function of iron in soil processes cannot be considered solely in terms of inorganic mechanisms.

E. ALLOPHANE The forms and concentrations of silica and silicates in soils are criteria that have been commonly used for some time to differentiate soils at a fairly high categorical level, and the allophane content of a soil, when high, is an important diagnostic feature. Soils containing large amounts of allophane are generally regarded as being only slightly weathered and on the pedogenic time scale considered to be young soils. It has been shown that allophane frequently predominates in soils developed on recent volcanic deposits, and it is in fact regarded as a logical stage in the weathering of glass often encountered in ash and in the matrix of much volcanic rock. Fieldes (1955) and Kanno (1959) consider that the weathering sequence in soils developed on recent volcanic deposits is: volcanic ash + “allophane B” + “allophane A” + metahalloysite -+ kaolinite. Fieldes found that humic colloids impeded development of “allophane A” from “allophane B.” Aomine and Wada (1962) considered that hydrated halloysite is an end product of the weathering of volcanic ash and pumice, allophane being an intermediate stage. From field observations and laboratory analyses they concluded that differential weathering arose from differences in leaching and biotic activity. It is assumed that chemical weathering of these volcanic materials increases with depth and thus constitutes a measure of the weathering time since deposition. Tsuchiya and Kurabayashi (1958) noted that the clays of the upper layers of the Kanto loam, from Pleistocene volcanic ash, were rich in allophane while those of the lower layers were high

AMORPHOUS INORGANIC MATERIALS IN SOILS

371

in hydrated halloysite. They, however, attributed the differences in clay composition to changes in deposition environments rather than to time of weathering. Allophane is not exclusively supergene in origin and has been found in hydrothermal deposits ( Grim, 1953; Skvortsova and Kopchenova, 1958). Tamura et al. (1953) suggested allophane in latosols was a product of kaolinite weathering. Jackson ( 1‘356) considered that, in soils, “stable” allophane is a weathering relic of halloysite and kaolinite and that “unstable” allophane is a material which forms quickly in the weathering of volcanic glass. Bates (1960) suggested that allophane might be the intermediate stage in the change from halloysite to gibbsite but pointed out that the system was complicated by the probable existence of silica, iron, aluminum, and water in various amorphous to poorly crystalline binary, ternary, and quaternary combinations. Siffert (1962) has studied many reactions of silica in solution, and McKeague and Cline (1963) found that iron, nickel, and cobalt had a marked affinity for silica in solution; Krauskopf (1959), on the contrary, claimed that aluminum had a specific affinity for silica. Beckwith and Reeve (1963) also noted that oxides and hydroxides of iron as well as those of aluminum sorb monosilicic acid from solution. Jackson (1956) states that coprecipitated iron oxide is an important constituent of amorphous aluminosilicates in soils. Laboratory studies on the silicification of rocks have been made by Bisque (1962) and on the silicification of volcanic ash soils by Onikura (1959), The latter used potassium silicate solution and found that part of the silicon added reacted with “active” aluminum and iron producing an amorphous silicate, but most of it reacted with the hydrated halloysite present giving a new mineral with a 16 A. basal spacing, the differential thermal analysis curve of which showed an endothermic effect at 650”. Minerals with 14 and 7 A. spacings were unaffected. It is apparent from this discussion that the mechanisms for allophane formation are probably manifold. Fieldes and Swindale (1954) commented that weathering sequences are of value in predicting the end products of weathering: however, a specific clay mineral may be formed by several different mechanisms. Within the surface horizon of a soil the water content varies most widely, and chemical activity is at a maximum because of the decomposition of large amounts of organic matter. Concomitantly the activity of microorganisms and the root activity of many plants are most intensive in the A horizon so that large amounts of carbon dioxide and organic acids are liberated and nutrients are absorbed. Kashiwagi and Yokoi (1952), Stefanovits (1959), Glenn et al. (1960),and Mitchell and Farmer

372

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

( 1962 ) have found that substantially more amorphous aluminosilicates occur in the clay from the surface horizons, especially when high in organic matter. This has been observed in soils developed on chemically different parent materials. The occurrence of highly amorphous clays in surface soils also seems to depend upon the permeability of the subsoil, since they have been observed, up to the present, only in soils with free internal drainage. The presence of organic chelating compounds derived from plant residues will result in the leaching of aluminum and iron, thus increasing the rate of dissolution of primary and secondary silicates with the formation of amorphous material. Recently, Webley et al. (1960) and Henderson and Duff (1963) have demonstrated that naturally occurring crystalline silicates can be decomposed by such chelating agents as 2-keto gluconic acid, oxalic acid or citric acid produced by soil microorganisms, and that the mineral residues are amorphous to X-rays. Kulia (1962) in a study of the decomposition of aluminosilicates by bacteria in the rhizospheres of forest stands found that the numbers of bacteria capable of attacking silicates were greatest during the maximum growth period of the trees. He obtained a correlation between bacterial number and potassium release in laboratory experiments. Thus the presence of amorphous inorganic material in soil clays may also be attributable to the activity of microorganisms. V. Relationships between Amorphous Inorganic Material and Specific Physical and Chemical Properties

A. MORPHOLOGICAL PROPERTIES Pedologists have produced numerous correlations between particular macromorphological features of the soil profile and the nature and amount of “free” silica, alumina, and iron oxide present. The structure and consistence of a soil are interrelated; the former is the resultant of forces within the natural soil while the latter is concerned with the forces themselves. The degree of structure and of consistence of some soils has been equated with the bonding action of free sesquioxides, principally iron, on the surface of the primary particles ( Arkhangelskaya, 1955; Hosking et al., 1957; Carroll, 1958). Frequently, field textures depart markedly from the resuIts of mechanical analysis, and these anomalies can sometimes be attributed to the cementing of clay minerals with free oxides into silt and even larger-sized aggregates. McIntyre (1956), in his study of soil aggregation and structure, observed that cementation resulted from precipitation and irreversible drying of iron oxide colloids, and that deflocculation was inhibited by iron in solu-

AMORPHOUS INORGANIC MATERIALS IN SOILS

373

tion and by the formation of iron-organic complexes. Prebble and Stirk (1959) emphasize that aluminum could be just as effective as iron in this role. Induration of a soil refers to a handling property of the soil and appears not to be greatly affected by moisture content. McKeague and Cline (1963) discussed a number of soil horizons, for example, duripans and fragipans, the induration of which is attributed, at least in part, to cementation by silica. They point out, however, that although silica may be the cementing agent in these horizons, this conclusion has not been altogether substantiated by experimental evidence. Wurman et al. ( 1959) examined the properties of fine-textured subsoil bands in sandy Michigan soils, noting that the finer-textured material contained more free iron oxide and organic matter than the coarser-textured horizons. Milford et al. ( 1961) investigated indurated horizons in coarse-textured soils and found no difference in the free oxide content of the indurated horizons and of the horizons above and below. They considered that fine and very fine sand produced the induration in association with tillage operations and a wetting and drying cycle. Consideration has always been given to factors that control and alter the available water in soils as this property is used for their agronomical evaluation. Kun-Huang and Tsen-Tuo ( 1959) studied the moistureretaining capacity of clay in relation to free iron oxide content, finding that in the Pinchen clay which they examined the water-retaining capacity was reduced from 10.3 per cent to 7.3 per cent after deferrification. On the other hand, an increase in the water-retaining capacity of soils after removal of free iron with dithionite was noted by El Ashkar et al. (1956). Removal of iron improved the dispersion of the soil particles and increased the volume of the coarser pores and thus the amount of available water.

B. CHEMICAL PROPERTIES Gorbunov et al. ( 1961) emphasized the value of amorphous sesquioxide determinations for the interpretation of soil genetic and agrochemical problems. They comment on the difficulties of distinguishing between crystalline and amorphous free oxides in soils. In the interests of elucidating physicochemical properties of the soil they illustrate the necessity for differentiating between amorphous and crystalline forms by the fact that amorphous iron oxide sorbs 109 times as much phosphate as crystalline iron oxide, and amorphous alumina sorbs 137 times as much phosphate as crystalline alumina. The hydrous oxides which are released by weathering can retain phosphate against leaching and, depending upon the form of the oxide, the phosphate may become fixed, Dixon

374

B. D. MITCHELL, V. C. FARMER, AND

W.

J. MCHARDY

(1958) found that iron and aluminum oxides, in the amorphous form, retain large amounts of phosphate but that their ability to retain phosphate decreases as they age and crystallize, since crystallization involves a reduction in surface area. Dixon noted that the older red loam soils derived from basalt possessed a high phosphate-fixing power and assumed, therefore, that these soils contained active hydrous oxides and consequently that the oxides were amorphous. Colwell (1959) investigated the sorption of phosphate by goethite, hematite, lepidocrocite, and ferric oxide gel, and also by gibbsite, boehmite, and alumina gel. The sorption values for these compounds at pH 4 showed very large differences due to variations in surface area, hydration, and activity associated with differences in particle size. Colwell stated that a direct comparison cannot be made with sorption values of phosphate by sesquioxide-containing soils because of uncertainties regarding particle size and surface properties of the oxides in the soils, and adds that another influence would be complex formation with organic matter. Williams ( 1960), studying phosphorus relationships on acidic surface soils developed on different glacial tills found that pronounced parent material effects could be equated with varying contents of acid oxalate soluble aluminum and iron. Williams concluded that phosphate retention capacity depended principally upon the soluble aluminum, but there were also significant correlations with acid oxalate soluble iron and with loss on ignition and carbon content. The corresponding correlation for soil clays and silts treated with peroxide was very poor (Williams et al., 1958). The high correlation found between aluminum extracted by Tamm’s reagent and carbon content must be the main reason for the high correlation found between phosphate sorption and carbon, since the phosphate does not combine with purely organic groups under the conditions used. The problem arises of distinguishing between iron and aluminum present as organometallic complexes and present in an entirely inorganic amorphous form. Kuron et al. (1961) found that potassium accumulation in the soil profile could be correlated with the content of free iron oxide and also with the organic matter content. LeRiche and Weir (1963) studied the distribution of trace elements in soil fractions and found the oxides extracted under ultraviolet light by ammonium oxalate ( p H 3.3) to be the fraction richest in trace elements and to contain a large proportion of the total Co, Cu, Mn, Pb, and V. They considered that these trace elements were incorporated principally in iron minerals. Isomorphous replacement of iron by aluminum in soil goethite has been observed by Norrish and Taylor (1961). The degree of substitution of aluminum for iron apparently depends upon the weathering conditions in the soil. Since aluminum probably restricts the size of the goethite crystal, this could

AMORPHOUS INORGANIC MATERIALS I N SOILS

375

influence phosphate fixation and other chemical properties. Norrish and Taylor found that finely particulate goethite containing a large amount of aluminum is not removed by dithionite treatment. The reactions of phosphorus, potassium, and lime in acid soils are intimately connected with exchangeable aluminum, which, according to Hattori and Kawaguchi (1959), may be derived from both clay mineral lattices and gibbsite, The role of aluminum in soil acidity has been considered in detail by Coleman et al. (1958) and Rich and Thomas (1960). VI. Summary

In this review an attempt has been made to emphasize the need for greater attention to be given to amorphous inorganic material in soils. The fact that the amorphous compounds of soils are often qualified by the term “active” is indicative of their importance in soil processes and genesis. Many difficulties arise in the determination of the physical and chemical characteristics of such material. By their very nature amorphous substances are difficult to detect and estimate, and frequently their presence is determined by implication rather than by direct measurement. Recent developments and refinements in instrumentation hold out hope of adding considerably to our knowledge of this material. A haphazard approach to sample pretreatment has detracted from the vaIue of the early, and even of some of the more recent, work on amorphous material, confusing interpretations and producing difficulties in correlation. The need for careful pretreatment cannot therefore be overstressed, and, indeed, selective chemical pretreatment is probably the most important prerequisite to the study of the amorphous constituents in soils. A detailed and careful study of the macromorphology, and if possible of the micromorphology, of the soil profile in order that it might be adequately identified and its component parts accurately sampled is, of course, essential for fundamental investigations of soil. REFERENCES Adler, H. H. 1951. “lnfra-red Spectra of Reference Clay Minerals,” Am. Petrol. Inst. Res. Proj. 49, pp. 1-72. Am. Petrol. Inst., New York. Aguilera, N. H., and Jackson, M. L. 1953. Proc. Soil Sci. SOC. Am. 17, 359-364. Aleixandre-Ferrandis, V., Garcia-Vicenti, J., and Aleixandre, T. 1962. Andes Edafol. Agriobiol. (Madrid) 21, 117-158. Aleksandrova, L. N., and Nad, M. 1958. Pochvovedenie No. 10, 21-27. Alexander, L. T., Cady, J. G., Whittig, L. D., and Dever, R. F. 1956. Trans. 6th Intern. Congr. Soil Sci., Paris, 1956 Vol. 5, pp. 67-72. Allaway, W. H. 1948. Proc. Soil Sci. SOC. Am. 13, 183-188. Amiel, J., Figlarz, M., and Songeon, J. 1961. Compt. Rend. 252, 1783-1785.

376

B. D. MITCHELL, V.

C. FARMER, AND W. J. MCHARDY

Anderson, J. U. 1963. Proc. lUth Natl. Conf. Clays, Clay Minerals, 1961 pp. 380388. Pergamon Press, New York. Aomine, S., and Jackson, M. L. 1959. Proc. Soil Sci. SOC. Am. 23, 210-214. Aomine, S . , and Wada, K. 1962. Am. Mineralogist 47, 1024-1048. Aomine, S., and Yoshinaga, N. 1955. Soil Sci. 79, 349-358. Arkhangelskaya, N. A. 1955. Tr. Gorkousk. Sel’skokhoz. lnst. 7, 135-141. Avery, B. W., Stephen, I., Brown, G., and Yaalon, D. H. 1959. J. Soil Sci. 10, 177-195. Barshad, I. 1955. I n “Chemistry of the Soil” (F. E. Bear, ed.), pp. 1-52. Reinhold, New York. Basila, M. R. 1962. J. Phys. Chem. 66, 2223-2228. Bastisse, E. M. 1960. Ann. Agron. A l l , 441-467. Bates, T. F. 1958. Mineral Ind., Penna. State Uniu. Proj. E-27. Bates, T. F. 1960. Mineral Ind., Penna. State Uniu. 2 9 ( 8 ) , 1-6. Beckwith, R. S., and Reeve, R. 1963. Australian J. Soil Res. 1, 157-168. Benesi, H. A., and Jones, A. C. 1959. J. Phys. Chern. 63, 179-182. Besoain, M. E., and Garcia-Vincente, J. 1962. Neues Jahrb. Mineral. Abhandl. 98, 349-366. Beutelspacher, H., and Van Der Marel, H. W. 1961. Actu Uniu. Carolinae Ceol. SUPPI. 1, 97-114. Bhattacharya, N. 1962. Bull. Geol. SOC. Am. 73, 1007-1020. Birrell, K. S. 1958. New Zealand SOC. Soil Sci. Proc. 3, 11-13. Birrell, K. S. 1961a. J. Soil Sci. 12, 307-316. Birrell, K. S. 1961b. New Zealand J. Sci. 4, 393-414. Birrell, K. S. 1962. New Zealand J. Sci. 5, 453-462. Birrell, K. S., and Fieldes, M. 1952. J . Soil Sci. 3, 156-166. Birrell, K. S., and Gradwell, M. 1956. J. Soil Sci. 7, 130-147. Birrell, K. S., and Packard, R. Q. 1958. New Zealand J. Sci. 1, 70-82. Bisque, R. E. 1962. Proc. 9th Natl. Conf. Clays, Clay Minerals, 1960 pp. 365-373. Pergamon Press, New York. Blanchet, R., Perigaud, S., Chaumont, C., and Nadeau, J. C. 1960. Ann. Agron. [N.S.] 11, 621-649. Bower, C. A., and Goertzen, J. 0. 1959. Soil Sci. 87, 289-292. Brindley, G. W., and Robinson, K. 1951. I n “X-Ray Identification and Crystal Structures of Clay Minerals” (G. W. Brindley, ed.), pp. 173-198. Mineralogical Society, London. Brindley, G. W., Oughton, B. M., and Youell, R. F. 1951. Acta Cryst. 4, 552-557. Brown, G. 1955. Clay Minerals Bull. 13, 294-302. Brown, G., ed. 1961. “The X-Ray Identification of Clay Minerals.” Mineralogical Society, London. Brunauer, S., Emmett, P. H., and Teller, E . 1938. J. Am. Chern. SOC. 60, 309-319. Bugge, P. E., and Kerlogue, R. H. 1947. J. SOC. Chem. Ind. (London) 66, 377-381, Burford, J. R., Deshpande, T. L., Greenland, D. J., and Quirk, J. P. 1964. J. Soil Sci. 15, No. 2 (in press). Cady, J. G. 1960. Trans. 7th Intern. Congr. Soil Sci. Madison, Wisconsin, 1960 Vol. 4, pp. 418-424. Carroll, D. 1958. Geochim. Cosmochim. Acta 14, 1-28. Carthew, A. R. 1955. Soil Sci. 80, 337-347. Cloos, P., and Fripiat, J. J. 1958. Pddologie, Gent. 8, 13-24.

AMORPHOUS INORGANIC MATERIALS IN SOILS

377

Cloos, P., Gastuche, M. C., and Croegaert, M. 1961. Proc. 21st Intern. Geol. Congr. 1960 Pt. XXIV, pp. 35-50. Berlingske Bogtrykkeri, Copenhagen. Coleman, N. T., Kamprath, E. J., and Weed, S. B. 1958. Aduan. Agron. 10, 475522. Coleman, N. T., Weed, S. B., and McCracken, R. J. 1959. Proc. Soil Sci. SOC. Am. 23, 146-149. Colwell, J. D. 1959. Australian J. Appl. Sci. 10, 95-103. Crompton, E. 1960. Trans. 7th Intern. Congr. Soil Sci. Madison, Wisconsin, 1960 Vol. 4, pp. 406-412. Deb, B. C. 1950. 1. Soil Sci. 1, 212-220. de Endredy, A. S. 1963. Clay Minerals Bull. 5, 209-217. DeKimpe, C., Gastuche, M. C., and Brindley, G. W. 1961. Am. Mineralogist 46, 1370-1381. DeMumbrum, L. E. 1960. Proc. Soil Sci. SOC. Am. 24, 185-189. DeMumbrum, L. E., and Bruce, R. R. 1960. Soil Sci. 89, 333-337. Deuel, H. 1954. Ber. Deut. Keram. Ges. 31, 1-7. Deuel, H. 1960. Trans. 7th Intern. Congr. Soil Sci. Madison, Wisconsin, 1960 Vol. 1, pp. 38-53. D’Hoore, J. 1954. African Soils 3, 66-82. Diamond, S., and Kinter, E. B. 1958. Proc. 5th Natl. Conf. Clays, Clay Minerals 1956 pp. 334-347. Natl. Acad. Sci. Natl. Res. Council Publ. 566. Washington, D.C. Dixon, J. B., and Jackson, M. L. 1959. Science 129, 1616-1617. Dixon, J. B., and Jackson, M. L. 1960. Proc. 8th Natl. Conf. Clays, Clay Minerals, 1959 pp. 274-286. Pergamon Press, New York. Dixon, J. K. 1954. 1. New Zealand Inst. Chem. 18, 1-15. Dixon, J. K. 1958. New Zealand SOC. Soil Sci. Proc. 3, 4-10. Doelter, C., and Cornu, F. 1909. Z. Chem. Ind. Kolloide 4, 89-92. Droste, J. B. 1961. Science 133, 1928. Duchaufour, P. 1963. Compt. Rend. 256, 2657-2660. Dyal, R. S., and Hendricks, S. B. 1950. Soil Sci. 69, 421-432. Egawa, T., Watanabe, Y., and Sato, A. 1959. Aduan. Clay Sci. (Tokyo) 1, 260-272. El Ashkar, M. A., Bodman, G. B., and Peters, D. B. 1956. Proc. Soil Sci. SOC. Am. 20, 352-356. Emmett, P. H., Brunauer, S., and Love, K. S. 1938. Soil Sci. 45, 57-65. Farmer, V. C., and Mitchell, B. D. 1963. Soil Sci. 96, 221-229. Fazzini, P., and Olivieri, R. 1961. Atti Mineral. Accad. Sci. Lettere Arti Modena 3, 60-81. Fieldes, M. 1952. Nature 170, 366-367. Fieldes, M. 1955. New Zealand I. Sci. Technol. B37, 336-350. Fieldes, M. 1957. New Zealand J . Sci. Technol. B38, 533-570. Fieldes, M., and Schofield, R. K. 1960. New Zealand J . Sci. 3, 563-579. Fieldes, M., and Swindale, L. D. 1954. New Zealand J . Sci. Technol. B36, 140154. Fieldes, M., and Williamson, K. I. 1955. New Zealand 1. Sci. Technol. B37, 314335. Fieldes, M., Swindale, L. D., and Richardson, J. P. 1952. Soil Sci. 74, 197-205. Fieldes, M., Walker, I. K., and Williams, P. P. 1956. New Zealand J . Sci. Technol. B38, 31-43.

318

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

Fiskell, J. G. A., Gammon, N., Yuan, T. L., and Zmeskal, 0. 1958. Proc. Soil Sci. SOC. Am. 22, 339-343. Poster, M. D. 1953. Goechim. Cosmochim. Acta 3, 143-154. Frederickson, A. F. 1951. Bull. Geol. SOC. Am. 62, 221-232. Frederickson, L. D. 1954. Anal. Chem. 26, 1883-1885. Fridland, V. M. 1961. Pochuouedenie No. 12, 57-74. Fripiat, J. J. 1960. Trans. 7th Intern. Congr. Soil Sci. Madison, Wisconsin, 1960 VO~.4, pp. 502-511. Fripiat, J. J., and Gastuche, M. C. 1952. Publ. Inst. Natl. Etude Agron. Congo Belge Ser. Tech. 54. Fripiat, J. J., Gastuche, M. C., and Van Compernolle, G. 1954. Tram. 5th Intern. Congr. Soil Sci. Leopolduille, 1954 Vol. 2, pp. 401-422. Fripiat, J. J., Uytterhoeven, J., Schobinger, U., and Deuel, H. 1960. Heh. Chim. A C ~ 43, U 176-181. Fripiat, J. J., Gastuche, M. C., and Brichard, R. 1962. J. Phys. Chem. 66,805-812. Garcia-Vicente, J., and Besoain, M. E. 1961. Anales Edafol. Agrobiol. (Madrid) 20, 497-550. Garrett, W. G., and Walker, G. F. 1959. Clay Minerals Bull. 4, 75-80. Gastuche, M. C. 1959a. Rev. Questions Sci. No. 1, 64-92. Gastuche, M. C. 195913. Silicates Ind. U, 237-244. Gastuche, M. C., Delmon, B., and Veilvoye, L. 1960. Bull. SOC. Chim. Fronce pp. 60-70. Gastuche, M. C., Toussaint, F., Fripiat, J. J,, Touilleaux, R., and Van Meersche, M. 1963. Clay Minerals Bull. 5, 227-236. Gheith, M. A. 1952. Am. J. Sci. 250, 677-695. Gieseking, J. E. 1949. Aduan. Agron. 1, 159-204. Glenn, R. C., Jackson, M. L., Hole, F. D., and Lee, G. B. 1960. Proc. 8th Natl. Conf. Clays, Clay Minerals 1959 pp. 63-83. Pergamon Press, New York. Gorbunov, N. I. 1959. Pochuouedenie No. 7, 1-13. Gorbunov, N. I. 1961. Soviet Soil Sci. (English Transl.) No. 7, 712-724. Gorbunov, N. I., and Sharina, N. A. 1958. “The Investigation and Utilization of Clays” (in Russian), pp. 108-116. Publ. House Lvov Univ., Lvov. Gorbunov, N. I., Dzyadevich, G. S., and Tunik, B. M. 1961. Soviet Soil Sci. (Englbh Transl.) No. 11, 1252-1259. Gorbunova, Z. N. 1961. Pochvouedenie No. 1, 61-67. Gradusov, B. P., and Dzyadevich, G. S. 1961. Soviet Soil Sci. (English Transl.) NO. 7, 749-756. Gradusov, B. P., and Targulyan, V. 0. 1962. Pochvovedenie No. 11, 24-33. Gradwell, M., and Birrell, K. S. 1954. New Zealand J. Sci. Technol. B36, 108-122. Greene-Kelly, R. 1957. In “The Differential Thermal Investigation of Clays” ( R. C. Mackenzie, ed. ) pp. 140-164. Mineralogical Society, London. Greenland, D. J., and Quirk, J. P. 1964. 1. Soil Sci. 15, No. 2 (in press). Grim, R. E. 1953. “Clay Mineralogy.” McGraw-Hill, New York. Harada, M. 1955. Soil Plant Food (Tokyo) 2, 81-83. Harrassowitz, H. 1926. Fortschr. Geol. 4 ( 1 4 ) , 253-566. Hanvard, M. E., and Theisen, A. A. 1962. Proc. Soil Sci. SOC. Am. 26, 335-341. Hanvard, M. E., Theisen, A. A., and Evans, D. D. 1962. Proc. Soil. Sci. SOC. Am, 26, 535-541. Hashimoto, I., and Jackson, M. L. 1960. P~oc.7th Natl. Conf. Clays, Clay Minerals 1958 pp. 102-113. Pergamon Press, New York. Hattori, T., and Kawaguchi, K. 1959. J . Sci. Soil Manure, Tokyo 30, 57-60.

AMORPHOUS INORGANIC MATERIALS IN SOILS

379

Hauth, W. E. 1951. Iowa State Unio. Eng. Ezp. Sta. Eng. Rept. 7 , 1-20. Henderson, M. E. K.. and Duff, R. B. 1963. J . Soil Sci. 14, 236-246. Hendricks, S. B., and Fry, W. H. 1930. Soil Sci. 29, 457-479. HCnin, S. 1956. Bull. Groupe Franc. Argiles 8, 1-10. Hislop, J. F. 1944. Trans. Brit. Cemm. SOC. 43, 49-51. Holdridge, D. A., and Vaughan, F. 1957. In “The Differential Thermal Investigation of Clays” ( R . C. Mackenzie, ed.), pp. 98-139. Mineralogical Society, London. Hosking, J. S., Neilson, M. E., and Carthew, A. R . 1957. Australian J . Agr. Res. 8, 45-74. Houldsworth, H. S., and Cobb, J. W. 1923. Trans. Brit. Ceram. SOC. 22, 111-137. Hoyos, A., and Pino, C. 1958. Agrochimica 2, 218-235. Ishii, J., and Mori, T. 1959. J . Geol. SOC. Japan 65, 357-383. Jackson, M. L. 1956. “Soil Chemical Analysis-Advanced Course.” Publ. by Professor M. L. Jackson, Madison, Wisconsin. Jackson, M. L. 1959. Proc. 6th Natl. Conf. Clays, Clay Minerals 1957 p p . 133-143. Pergamon Press, New York. Jackson, M. L. 1980. Trans. 7 t h Intern. Congr. Soil Sci. Madison, Wisconsin, 1960 Vol. 2, pp. 445-455. Jackson, M. L. 1962. Proc. 9th Natl. Conf. Clays, Clay Minerals 1960 pp. 424-430. Pergamon Press, New York. Jackson, M. L. 1963. Proc. 11th Natl. Conf. Clays, Clay Minerals 1962 pp. 29-46. Pergamon Press, New York. Jackson, M. L., Whittig, L. D., and Pennington, R. P. 1950. Proc. Soil Sci. SOC. Am. 14, 77-81. Jeffries, C . D., and Johnson, L. 1961. Soil Sci. 92, 402-403. Jones, J. B., Segnet, E. R., and Nickson, N. M. 1963. Nature 198, 1191. Kamoshita, Y., and Iwasa, Y. 1959. J . Sci. Soil Manure Tokyo 30, 185-188. Kanno, I. 1955. Soil Plant Food ( T o k y o ) 1, 1-2. Kanno, I. 1959. Adt;an. Clay Sci. ( T o k y o ) 1, 213-233, Kanno, I., and Arimura, S. 1958. Soil Plant Food ( T o k y o ) 4, 62-87. Kanno, I., Honjo, Y., and Arimura, S. 1956. Soil Plant Food ( T o k y o ) 2, 220-232. Kanno, I., Kuwano, Y., and Honjo, Y. 1960. Acluan. Clay Sci. ( T o k y o ) 2, 355-356. Kashiwagi, H. 1961. Soil Sci. Plant Nutr. ( T o k y o ) 7, 12-17. Kashiwagi, H., and Yokoi, H. 1952. Sci. Rept. Shiga Agr. Coll. Ser. I 1, 84-90. Kaurichev, I. S., and Nozdrunova, E. M. 1961. Dokl. mosk. Sel’skokhoz. Akad. 71, 151-155.

Kawaguchi, K., and Matsuo, Y. 1954. J . Sci. Soil Manure Tokyo 25, 154-158. Kawaguchi, K., and Matsuo, Y. 1957. Soil Plant Food ( T o k y o ) 3, 29-35. Kelley, W. P., Dore, W. H., and Brown, S. M. 1931. Soil Sci. 31, 25-55. Kinter, E. B., and Diamond, S. 1960. Proc. 7th Natl. Conf. Clays, Clay Minerals 1958 pp. 125-134. Pergamon Press, New York. Kinter, E. B., Wintermeyer, A. M., and Swerdlow, M. 1952. Public Roads 27, 89-100.

Kirsanov, A. T. 1958. “Methods for the Agrochemical Analysis of Soils” (in Russian). Publ. House Acad. Sci. USSR, Moscow. Klages, M. G., and White, J. L. 1957. Proc. Soil Sci. SOC. Am. 21, 16-20. Kolesova, V. A., and Ryskin, Y. I. 1959. Opt. i Spektroskopiya 7, 261-263; Opt. Spectry. ( U S S R ) (English Transl.) 7, 165-167. Krauskopf, K. B. 1956. Goechim. Cosmochim. Acta 10, 1-28. Krauskopf, K. B. 1959. Spec. Publ. SOC. Econ. Palaentol. Mineral. 7, 4-19.

380

B. D. MITCHELL, V. C. FARMER, AND W. J. MCHARDY

Kubiena, W. L. 1953. “The Soils of Europe,” p. 220. Murby, London. Kulia, G. A. 1962. Bull. Acad. Sci. USSR Biol. Ser. (English Trarasl.) 27, 915-920. Kulp, J. L., and Trites, A. F. 1951. Am. Mineralogist 36, 23-44. Kun-Huang, H., and Tsen-Tuo, C. 1959. Proc. Soil Sci. SOC. Am. 23, 270-273. Kumakov, N. S., and Rode, E. J. 1928. Z . Anorg. Allgem. Chem. 169, 57-80. Kuron, H., Preusse, H. U., and Fohrenbacher, A. 1961. Z. Pflanzeneruehr. Dueng. Bodenk. 92, 233-245. Launer, P. J. 1952. Am. Minerdogist 37, 764-784” Leet, L. D., and Judson, S. 1960. “Physical Geology.” Prentice-Hall, Englewood Cliffs, New Jersey. LeRiche, H. H., and Weir, A. H. 1963. 3. Soil Sci. 14, 225-235. Lindqvist, B. 1959. Bull. Geol. Inst. Univ. Ups& 38, 89-103. Lippins, B. C., and Hermans, M. E. A. 1961. Powder Met. 7, 66-81. Lomonovich, M. I. 1955. “Problems of the Geology of Asia” (in Russian), Vol. 2, pp. 509-522. Publ. House Acad. Sci. USSR, Moscow. Lyon, R. J. P. 1962. “Minerals in the Infrared.” Stanford Res. Inst., Stanford, California. McIntyre, D. S. 1956. 3. Soil Sci. 7, 302-306. McKeague, J. A,, and Cline, M. G. 1963. Advan. Agron. 15, 339-396. Mackenzie, R. C. 1949. Nature 164, 244. Mackenzie, R. C. 1952. “Problems of Clay and Laterite Genesis,” pp. 65-75. Am. Inst. Mining Met. Engrs, New York. Mackenzie, R. C., ed. 1957a. In “The Differential Thermal Investigation of Clays,” pp. 1-22. Mineralogical Society, London. Mackenzie, R. C., ed. 195713. In “The Differential Thermal Investigation of Clays,” pp. 299-328. Mineralogical Society, London. Mackenzie, R. C. 1957c. Agrochimica 1, 305-327. Mackenzie, R. C. 1958. Proc. Inst. Civil Engrs. (London) 11, 345-348. Mackenzie, R. C. 1960. Silicates Ind. 25 (1, 2 ) , 12-18, 71-75. Mackenzie, R. C. 1963. Proc. 11th Natl. Conf. Clays, Clay Minerals 1962 pp. 11-28. Pergamon Press, New York. Mackenzie, R. C., and Meldau, R. 1959. Mineral. Mag. 32, 153-165. Mackenzie, R. C., and Robertson, R. H. S. 1961. Acta Unio. Carolinae, Geol. Suppl 1, 139-149. Mackenzie, R. C., Meldau, R., and Farmer, V. C. 1956. Ber. Deut. Kemm. Ges. 33, 222-229. Mackenzie, R. C., Meldau, R., and Card, J. A. 1962. Mineral. Mag. 33, 145-157. Makower, B., Shaw, T. K., and Alexander, L. T. 1937. PTOC. Soil Sci. SOC. Am. 2, 101-108. Martin, A. E. 1960. 3. Soil Sci. 11, 382-393. Martin, R. T. 1955. Proc. Soil Sci. SOC. Am. 19, 150-164. Matsui, T. 1959. Advan. Clay Sci. (Tokyo) 1, 244-259. Mattson, S., and Gustaffson, Y. 1937. Lantbruks-Hogskol. Ann. 4, 1-54. Mehra, 0. P., and Jackson, M. L. 1960. Proc. 7th NatZ. Conf. Clays, Clay Minerals 1958 pp. 317-327. Pergamon Press, New York. Milford, M. H., Kunze, G. W., and Bloodworth, M. E. 1961. Proc. Soil Sci. SOC. Am. 25, 511-515. Mitchell, B. D., and Farmer, V. C. 1962. Clay Minerals Bull. 5, 128-144. Mitchell, B. D., and Mackenzie, R. C. 1954. Soil Sci. 77, 173-184. Mitchell, W. A. 1955. 3. Soil Sci. 6, 94-98. Moenke, H. 1962. “Mineralspektren.” Akademie-Verlag, Berrin.

AMORPHOUS INORGANIC MATERIALS IN SOILS

381

Mohr, E. C . J. 1944. “The Soils of Tropical Regions.” Edwards, Ann Arbor, Michigan. Moscou, L., and van der Vlies, G. S. 1959. Kol1oid.-2. 163, 35-41. Muiioz Taboadela, M. 1953. J. Soil Sci. 4, 48-55. Nelsen, F. M., and Eggertson, F. T. 1958. Anal. Chem. 90, 1387-1390. Nelson, R. A., and Hendricks, S. B. 1943. Soil Sci. 56, 285-296. Nikiforoff, C. C. 1959. Science 129, 186-196. Norrish, K., and Taylor, R. M. 1961. J. Soil Sci. 12, 294-306. Oades, J. M. 1962. The nature of iron compounds in various soil types. Ph.D. thesis, Dept. Agr., Univ. Leeds, Leeds, England. Oades, J. M., and Townsend, W. N. 1963. Clay Minerals Bull. 5, 177-182. Oberlin, A., and Tchoubar, C. 1961. Proc. 21st Intern. GeoE. Congr. 1960 Pt. XXlV, 51-59. Berlingske Bogtrykkeri, Copenhagen. Onikura, Y. 1959. 1. Sci. Soil Manure Tokyo 30, 357-360. Ostrom, M. E. 1961. J. Sediment. Petrol. 31, 123-129. Parfenova, E. I., and Yarilova, E. A. 1956. Pochuouedenie No. 4, 38-42. Pask, J. A., and Davies, B. 1945. U. S . Bur. Mines Tech. Paper No. 664, 56-78. Peters, T. 1962. Schweiz. Mineral. Petrogr. Mitt. 42, 359-380. Prebble, R. E., and Stirk, G. B. 1959. Soil Sci. 88, 213-217. Quigley, F. M., and Martin, R. T. 1963. PTOC. 20th Natl. Conf. Clays, C Z y Minerals, 1961 pp. 107-116. Pergamon Press, New York. Ramachandran, V. S., Garg, S. P., and Kacker, K. P. 1961. Chem. Ind. pp. 790792. Rich, C. I. 1960. PTOC.Soil Sci. SOC. Am. 24, 26-32. Rich, C. I. 1962. Soil Sci. 93, 87-94. Rich, C. I., and Obenshain, S. S. 1955. PTOC. Soil Sci. SOC. Am. 19, 334-339. Rich, C. I., and Thomas, G. W. 1960. Aduan. Agron. 12, 1-39. Rich, C. I., Seatz, L. F., and Kunze, G. W. 1959. Southern Coop. Ser. Bull. 61. Riquer, J. 1960. Trans. 7th Intern. Congr. Soil Sci., Madison, Wisconsin, 1960 VOI. 4, pp. 425-431. Robertson, R. H. S. 1963. Clay Minerals Bull. 5, 237-247. Ross, C. S., and Kerr, P. F. 1934. U . S. Geol. SUTU.Profess. Papers 185-G, 135148. Ruprecht, F. 1866. “Geobotanical Investigation on a Chernozem.” Acad. Sci. Russia. Russell, E. J. 1950. “Soil Conditions and Plant Growth,” 8th ed. Longmans, Green, New York. Salvetat, L. A. 1851. Ann. Chim. Phys. 31, 102-116. Sanchez-Calvo, M. C. 1961. Anales Edafol. Agrobiol. (Madrid) 20, 189-208. Sasaki, S., and Ishizuka, Y. 1957. Soil Plant Food (Tokyo) 2 ( 4 ) , 177-184. Sawhney, B. L. 1960. Trans. 7th Intern. Congr. Soil Sci., Madison, Wisconsin, 1960 Vol. 5, pp. 476-481. Scheffer, F., Meyer, B., and Folster, H. 1961. Z . Pflanzenernaehr. Dueng. Bodenk. 92, 201-207. Schlosser, E. G. 1959. Chem.-1ngr.-Tech. 31, 799-805. Schwertmann, U. 1959. 2. Pflanzenernaehr. Dueng. Bodenk. 84, 194-204. Schwcrtmann, U., and Jackson, M. L. 1963. Science 139, 1052-1053. SifFert, B. 1962. Quelques Reactions de la Silice en Solution: La Formation des Argiles. Mem. Carte Geol. Alsace-Lorraine N o . ‘21. Simonson, R. W. 1959. Proc. Soil Sci. SOC. Am. 23, 152-156.

382

B. D. MITCHELL, V. C. FARMER, AND W.J. MCHARDY

Skvortsova, K. U., and Kopchenova, E. V. 1958. Zap. Vses. Mineralog. Obshchestoa 87, 695-698. Smithson, F. 1956a. J. Soil Sci. 7, 122-129. Smithson, F. 1956b. Nature 178, 107. Smithson, F. 1958. J. Soil Sci. 9, 148-154. Smithson, F. 1959. J . Soil Sci. 10, 105-109. Souza Santos, P., Vallejo-Freire, A., and Souza Santos, H. L. 1953. Kolloid-Z. 133, 101-107. Starkweather, F. M., and Palumbo, D. T. 1957. J. Electrochem. SOC. 104, 287-291. Stefanovits, P. 1955. Agrokem. Talaitan 4,265-272. Stefanovits, P. 1959. Agrokem. Talajtan 8, 37-48. Steger, W. 1953. Clay Minerals Bull. 2, 23. Sudo, T. 1954. Clay Minerals Bull. 2, 96-106. Sudo, T. 1956. Sci. Rept. Tokyo Kyoiku Daigaku Sect. C . 5, 39-55. Sumner, M. E. 1963. Clay Minerals Bull. 5, 218-226. Takahashi, H. 1959. PTOC.6th Natl. Conf. CZuys, Clay Minerah, 1957 pp. 279-291. Pergamon Press, New York. Tamm, 0. 1922. Medd. Skogsforsoksanst. Stockholm 19, 385-404. Tamura, T. 1958. J . Soil Sci. 9, 141-147. Tamura, T., and Jackson, M. L. 1953. Science 117, 381-383. Tamura, T., Jackson, M. L., and Sherman, G . D. 1953. PTOC. Soil Sci. SOC. Am. 17, 343-346. Tamura, T., Hanna, R. M., and Shearin, A. E . 1959. Soil Sci. 87, 189-197. Taylor, R. M. 1959. J. Soil Sci. 10, 309-315. Touilleaux, R., Fripiat, J. J., and Toussaint, F. 1960. Trans. 7th Intern. Congr. Sodl Sci., Madison, Wisconsin, 1960 Vol. 4,pp. 460-467. Troell, E. 1931. J . Agr. Sci. 21, 476-483. Tsuchiya, T., and Kurabayashi, S. 1958. J . Geol. SOC. Japan 64,605-615. Tsyurupa, I. G. 1961. Soviet Soil Sci. (English Transl.) No. 4, 443-453. Tuddenham, W. M., and Lyon, R. J. P. 1959. Anal. Chem. 31, 377-380. Tyurin, I. V. 1937. Problems of Soviet Soil Sci. 4, 3-23. Uytterhoeven, J., and Fripiat, J. J. 1962. Bull. SOC. Chim. France pp. 788-792. Uytterhoeven, J., Fripiat, J. J., and Dockx, L. 1959. Bull. Acad. Belg. CZusse Sci. 45, 611-624. VBmos, R. 1961. Agrokem. Talajtan 10, 53-66. Van Bemmelin, J. M. 1910. “Die Absorption.” Steinkopff, Dresden. Van Der Marel, H. W. 1961. Acta Uniu. Carolinae Geol. Suppl. 1, 23-82. Van Rummelen, F. F. F. E. 1953. Landbouw, Buitenzorg 25, 237-246. Vasileva, A. A. 1958. Sb. Nauchn. lssled. Inst. Osnouanii i Podzemn. Sooruzhenii, Akad. Stroit. i Arkhitekt. SSSR 33, 49-58. Wada, K. 1961. Am. Minerologist 46, 78-91. Wada, X.,and Ataka, H. 1958. Soil Plant Food (Tokyo) 4, ( l ) ,12-18. Webley, D. M., Duff, R. B., and Mitchell, W. A. 1960. Nature 188, 766-767. Weiser, H. B., and Milligan, W. 0. 1942. Adoan. Colloid Sci. 1, 227-246. Wey, R., and Siffert, B. 1961. Silicates Ind. 26, 419-423. White, W. A. 1953. Am. Minerologist 38, 634-642. Whittig, L. D., Kilmer, V. J., Roberts, R. C., and Cady, J. G . 1957. Proc. Soil Sci. SOC. Am. 21, 228-232. Wicklund, R. E., and Whiteside, E. P. 1959. Can. J. Soil. Sci. 39, 222-234. Wiklander, L. 1945. Lantbmks-Hogskol. Ann, 13, 87-103.

AMORPHOUS INORGANIC MATERIALS IN SOILS

383

Williams, E. G. 1960. Tram. 7th Intern. Congr. Soil Sci., Madison, Wisconsin, 1960 Vol. 3, pp. 604-611. Williams, E. G., Scott, N. M., and McDonald, M. J. 1958. J . Sci. Food Agr. 9, 551-559. Wurman, E., Whiteside, E. P., and Mortland, M. M. 1959. Proc. Soil Sci. SOC. Am. 23, 135-143. Yaalon, D. H. 1959. J . Chem. Educ. 36, 73-76. Yamaguchi, G., and Sakamoto, K. 1959. Bull. Chem. SOC. Japan 32, 1364-1368. Yarilova, E. A,, and Parfenova, E. I. 1960. Kolloidn. Zh. 22, 237-242. Yoshinaga, N., and Aomine, S. 1962a. Soil Sci. Plant Nutr. (Tokyo) 8 ( 2 ) , 6-13. Yoshinaga, N., and Aomine, S. 196213. Soil Sci. Plant Nutr. (Tokyo) 8 ( 3 ) , 22-29. Zabelin, V. A. 1962. Zap. Vses. Mineralog. Obshchestua 91, 343-350. Zemlicka, J. 1958. Sb. Vysoke skoly Chem. Technol. u Praze Oddil Fak. Anorg. Org. Technol. pp. 269-280.

This Page Intentionally Left Blank

AUTHOR INDEX Numbers in italics indicate the pages on which the complete references are listed.

A Abbe, C., 10, 54 Ackerson, C. W., 198, 211, 241 Adair, C. R., 296, 324 Adams, E. P., 186, 188, 189, 195 Adams, J. R., 75, 100 Adams, M. W., 317, 325 Adler, H. H., 350, 375 Aganval, R. R., 162, 178 Aguilera, N. H., 340, 375 Akeming, H., 296, 297, 324 Alberda, Th., 22, 54 Alciatore, H. F., 11, 54 Alderfer, R. B., 188, 195 Aleixandre, T., 344, 375 Aleixandre-Ferrandis, V., 344, 375 Aleksandrova, L. N., 333, 375 Alexander, D. E., 133, 136 Alexander, L. T., 331, 334, 355, 375, 380 Alexander, M., 221, 241 Alexander, M. W., 124, 136 Allard, R. W., 286, 287, 288, 289, 290, 291, 292, 294, 295, 296, 298, 299, 300, 301, 304, 318, 319, 321, 324, 325 Allaway, W. H., 181, 196, 363, 375 Allison, L. E., 164, 165, 173, 178, 180 Almon, L., 222, 244 Amiel, J., 354, 375 Anderson, J, C., 132, 136 Anderson, J. U., 341, 376 Ando, J., 68, 69, 98 Andrew, R. H., 317, 325 Andrews, W. B., 75, 98 Angots, A., 11, 54 Anthony, J. L., 74, 78, 80, 81, 97, 100 Aomine, S., 336, 339, 342, 344, 346, 350, 354, 355, 360, 362, 365, 370, 376, 383 ap Griffith, C., 204, 210, 211, 213, 241, 244 Apostolakis, C., 78, 99 Arany, S., 149, 178 Archer, J. R., 68, 98

Arimura, S., 330, 337, 379 Arkhangelskaya, N. A., 335, 372, 376 Asana, R. D., 46, 54 Asbury, A. C., 235, 241 Aschoff, F., 61, 100 Ashton, F. M., 4 4 5 4 Ataka, H., 360, 361, 382 Atkins, A. E., 296, 324 Atkinson, R. W., 132, 136 Avery, B. W., 332, 376 Ayers, A. D., 159, 161, 176, 178, 180 Aylesworth, J. W., 186, 195 Azzi, C., 11, 54

B Baerug, R., 211, 243 Baker, A. S., 81, 82, 99 Baker, D. C., 12, 13, 54 Baker, G. O., 164, 172,179,180 Baker, H. G., 285,324 Ballard, L. A. T., 18, 54 Barber, C. W., 199, 244 Barber, D. A,, 38, 57,262, 277 Barber, S. A., 52,56 Barger, G . L., 9, 11,58 Barker, A. V., 202,212,213,241 Barley, K. P., 181, 195 Barnes, E. E., 75, 100 Barnett, A. J, G., 229, 241 Barrentine, B. F., 239, 244 Barshad, I., 366, 376 Bartholomew, W. V., 212, 246 B a s h , M. R., 353, 376 Bastisse, E. M., 366, 376 Bates, T. F., 337, 353, 367, 371, 376 Bauman, C. A., 232,244 Baver, L. D., 184, 188,195 Bear, R. P., 121, 136 Beard, J. B., 9, 10,56 Beath, 0. A,, 199, 203, 213, 214, 215, 216, 219, 220, 224, 227, 230, 233, 235, 239, 242, 243 Beaton, J. D., 84, 98 Becker, D. E., 232, 234, 235, 246 Becker, M., 211, 241

385

386

AUTHOR INDEX

Beckwith, R. S., 371, 376 Beeson, K. C., 74, 91, 100 Beichman, G. A., 52, 58 Belksma, M., 24, 57 Bendixen, R. E., 32, 33, 56 Benesi, H. A,, 349, 376 Bennett, P. C., 219,240,243 Berg, R. T., 218,241 Bemheim, F., 222, 241 Bernstein, L., 156, 157, 159, 161, 162, 174, 175, 178, 180 Berry, L. J., 240,246 Berry, S. L., 261,277 Berthelot, M., 204, 241 Bertrand, A. R., 182, 184, 195, 262, 265, 277 Besoain, M.E., 337, 376, 378 Betke, K., 230, 241 Beutelspacher, H., 350, 376 Bhattacharya, N., 365, 376 Biester, H. E., 224,245 Bingham, F. T., 143,161,178,180 Birkle, D. E., 253, 254, 255, 256, 257, 258, 278 Binell, K. S., 329, 336, 339, 354, 355, 360, 361, 362, 376, 378 Bisque, R. E., 371, 376 Bitters, W. P., 181, 179 Black, J. N., 19,54 Blackman, G. E., 19,54 Blackman, V. H., 18,54 Blair, G. Y.,142, 180 Blake, G. R., 186, 188, 189, 195 Blanchet, R., 368, 376 Blank, G. B., 260, 263, 264, 268, 273, 278 Blinks, L. R., 251,278 Bloodworth, M. E., 373,380 Bloomfield, R. A., 223, 225, 231, 233, 241, 242, 244, 246, 247 Bodman, G. B., 373, 377 Boelter, D. H., 186, 188, 195 Bollard, E. G., 206,242 Bolton, E. F., 186, 195 Bolton, J. L., 200, 208, 215, 224, 228, 229, 242 Bonner, J., 29, 44, 54, 55 Borts, I. H., 238, 242 Boughner, C. C., 11, 35, 54 Bould, G., 212, 242 Bouldin, D. R., 68, 78, 83, 85, 98, 100

Bourgot, S. J., 265, 278 Bower, C. A., 142, 152, 156, 164, 170, 178, 180, 355, 356, 376 Bowman, D. H., 124, 125, 137 Bowman, I. B. R., 229, 241 Boysen-Jensen, P., 19, 54 Brabson, J. A., 68, 85, 87, 94, 96, 97, 98, 99 Bradford, B. N., 94, 95, 98 Bradford, G. R., 148, 178 Bradley, W. B., 199, 203, 213, 214, 215, 216, 219, 220, 224, 227, 230, 233, 235, 239, 242, 243 Bradshaw, G. B., 151, 178, 179 Brady, D. E., 216, 221, 230, 242 Bratzler, J. W., 203, 246 Breniman, G. W., 217,221,232,242,244 Brichard, R., 353, 364, 378 Briggs, G. E., 18,54 Briggs, R. A., 237, 245 Brindley, G. W., 349, 357, 376, 377 Brindley, T. A,, 105, 136 Brink, F., 251, 278 Britten, E. J., 23, 57 Brooks, F. A., 2, 14, 15, 54,55 Brooks, R. H., 165,179 Brosheer, J. C., 91, 98 Brougham, R. W., 18,20,54 Brown, E. H., 94,95,99 Brown, E. M., 26,29,54 Brown, G., 329, 332, 335, 343, 346, 376 Brown, J. G., 157, 158, 179 Brown, J. W., 158, 178, 180 Brown, N. A., 94,100 Brown, S. M., 140, 163, 165, 179, 328, 379 Browne, E. B., 133, 136 Bruce, R. R., 337,377 Brunauer, S., 354, 355, 376, 377 Brunson, A. M.,120,127,136 Buchele, W. F., 189, 196 Bugge, P. E., 354,376 Buie, T. S., 7498 Bunt, A. C., 183, 184, 191,195 Burch, C. W., 240,246 Burch, W. G., 97,98 Burford, J. R., 355,376 Burgess, P. S., 164,178 Buringh, P., 151, 178 Burris, R. H., 220, 221, 224, 225, 237, 240, 245, 246

387

AUTHOR INDFX

Burstrom, H., 198, 242 Burt, R. F., 210, 212, 242 Bushnell, J., 187, 195 Buthurst, N. O., 216,219,241

C Cady, J. G., 331, 334, 337, 365, 375, 376, 382 Caldwell, A. C., 212, 245 Caldwell, A. G., 77, 100 Calpouzos, L., 6, 57 Campbell, E. C.,203,242 Campbell, J. B., 238, 242 Candela, M. I., 209, 217, 242 Cannon, W. A., 258,261,264,278 Caraway, C. T., 239, 245 Cadson, R., 156, 178 Carlson, T., 257,258 Carroll, D., 333, 334, 366, 369, 372, 376 Carthew, A. R., 337, 363, 372, 376, 379 Cartter, J. L., 26,29,55 Case, A. A., 230,235,242 Chapman, H. P., 172, 179 Chaumont, C., 368, 376 Chinoy, J. J., 35, 46, 54 Christiansen, P. D., 140, 178 Christianson, J. E., 172, 179 Clark, F. E., 39,56 Clark, K. G., 70, 72, 98 Clark, R., 224,229,234,242, 245 Clausen, J., 317, 324 Clements, L. B., 78, 79, 81,100 Cline, M. G., 365, 366, 367, 371, 373, 380 Cline, R. A., 184,195,268,278 Cline, T. R., 232,234,242 Cloos, P., 357, 363, 376, 377 Cobb, J. W., 346,379 Coleman, N. T., 332, 375, 377 Colwell, J. D., 374, 377 Comfort, J. E., 216,221,230,242 Comly, H. H., 238,242 Comstock, R. E., 130,136 Cook, R. L., 78, 99, 189, 196, 219, 242 Cooke, G. W., 76,83,98 Cooper, J. P., 32, 54 Cordy, D. R., 240, 246 Cornblath, M., 238, 242 Comu, F., 346, 377 Coup, M. R., 226,243 Crane, P. L., 118,136

Crawford, R. F., 199, 203, 209, 210, 212, 213, 217, 227, 229, 234, 236, 240, 242 Cresswell, C. F., 209,243 Croegaert, M., 357, 377 Crompton, E., 366, 377 Crowther, F., 18, 54

D Daday, H., 30, 32, 54, 56 Darlington, C. D., 111,136,285, 324 Davidson, J. L., 20,54 Davidson, W. B., 200, 208, 215, 224, 228, 229, 242 Davies, B., 357, 381 Davies, P. W., 251, 278 Davis, A. N., 238, 242 Davison, K. L., 211, 225, 226, 229, 232, 234, 236, 239, 242, 244 Deacon, E. L., 5,54 Deatherage, W. L., 121,136 Deb, B. C., 340,359,377 de Endredy, A. S . , 340,377 DeKimpe, C., 349, 377 Delmon, B., 357,378 DeMent, J. D., 78, 79, 81, 83, 84, 94, 95, 98, 100 DeMumbrum, L. E., 337, 344, 377 Denmead, 0. T., 44, 46, 54 De Remer, E. D., 51,54 Deshpande, T. L., 355,376 de Sigmond, A. A. J., 147, 178 DeTurk, 203, 212, 243 Deuel, H., 351, 363, 364, 377, 378 Dever, R. F., 331,334,375 de Wit, C. T., 21,54 De Wolfe, T. A., 268,271, 278 D’Hoore, J., 335, 377 Diamond, S., 355, 356, 377,379 Dicke, F. F., 105, 108,136 Diven, R. H., 215,243 Dixon, J. B., 332, 343, 377 Dixon, J. K., 336,374,377 Dixon, M., 222, 241 Dobzhansky, T., 285, 317,324, 325 Dockx, L., 351, 363,382 Dodd, D. C., 226,243 Doelter, J. B., 346, 377 Doering, E. J,, 151,178 Domby, C. W., 190,196 Donahoe, W. E., 239,243

388

AUTHOR INDEX

Donald, C. M., 22,23, 55 Doneen, L. D., 140, 158, 175, 179, 190, 195 Donnan, W. W., 151,178,179 Dore, W. H., 328,379 Dorsch, R., 61, 100 Doughty, J. L., 200, 208, 212, 214, 215, 216, 219, 224, 228, 229, 242, 243 Douglas, F. D., 228,235,245 Dow, B. K., 265,278 Dregne, H. E., 156,178 Droste, J. B., 337, 377 Drost-Hansen, W. J., 31, 56 Duchaufour, P., 333,370,377 Duff, R. B., 372, 379, 382 Dugger, W. M., 266,279 Dulin, T. G., 208, 243 Dungan, G. H., 107,116,136 Dunn, S., 188, 195 Durdle, W. M., 232, 234, 243 Duvick, D. N., 133, 136 Dyal, R. S., 355,377 Dzyadevich, G. S., 332, 333, 334, 340, 359, 369, 373, 378

E Earley, E. B., 204,217, 244 Eaton, F. M., 140, 142, 163, 171, 172, 179 Ebert, M., 262, 277 Eden, T., 188, 195 Edminister, T. W., 172, 179, 181, 196 Edwards, 0. W., 96,98 Egawa, T., 360, 377 Eggertson, F. T., 354, 381 Eik, K., 79, 100 Ekem, P. C., 42, 55 El Ashkar, M. A,, 373,377 Ellis, W. C., 222,225, 229,245 Ellis, W. W., 235, 243 Embry, L. B., 232, 234,243, 246 Emecz, T. I., 23, 55 Emerick, R. J., 232, 233, 234, 243, 245, 246 Emerson, R., 49, 55 Emmett, P. H., 354, 355, 376, 377 Engelhorn, A. J., 216,243 Engelstad, 0. P., 83, 84, 94, 95, 99, 100 Ensminger, L. E., 71, 76, 83, 84,99, 100 Eppson, H. F., 199, 203, 213, 214, 215,

216, 219, 220, 224, 227, 230, 233, 235, 239, 242, 243 Erek, Z., 233, 245 Erickson, A. E., 184, 195, 251, 252, 253, 254, 255, 256, 258, 265, 267, 268, 272, 273, 274, 278 Erickson, L. C., 185, 196 Evans, D. D., 340,378 Evans, L. T., 21, 55 Evans, N. T. S.,262,277 Eveleth, D. F., 224,245 Ezekial, W. N., 190,196

F Farmer, V. C., 330, 338, 339, 340, 341, 342, 344, 346, 347, 351, 359, 367, 372, 377, 380 Famsworth, R. B., 188, 195 Fanvell, E. D., 218,243 Fazzini, P., 337, 377 Federer, W. T., 319, 325 Fertig, S. N., 218, 243 Fieldes, M., 329, 330, 331, 334, 336, 339, 342, 344, 346, 350, 353, 360, 366, 367, 368, 370, 371, 376, 377 Figarella, J., 21, 58 Figlarz, M., 354, 375 Finlay, K. W., 314, 319, 325 Finn, B. J., 265, 278 Finnell, H. N., 212, 243 Fireman, M. F., 152, 172, 174, 175, 178, 179, 180 Fisher, E. G., 209, 217, 242 Fisher, J. E., 26, 55 Fiskell, J. G. A., 332, 378 Fleming, A. A., 126, 133, 136 Flesher, D., 206, 209,217,243 Fbcker, W. J., 183, 187, 188, 195 Flynn, L. M., 208, 212, 215, 231, 232, 233, 243, 245, 247 Fohrenbacher, A., 333,374,380 Folster, H., 332, 381 Fontaine, E. R., 182, 195 Forrest, L. A., 208, 213, 244 Forristall, F. F., 182, 195 Foster, M. D., 329, 358, 378 Frank, P. A., 217,218,243 Frankel, 0. H., 314, 325 Frederickson, A. F., 366, 378 Frederickson, L. D., 351, 378 Freeman, H. P., 70, 72, 98

389

AUTHOR INDEX

Fresenius, R., 61, 99 Fridland, V. M., 332, 334, 378 Friend, D. J. C., 26, 55 Fripiat, J. J., 333, 349, 351, 353, 357, 363, 364, 369, 376, 378, 382 Fritschen, L. J., 6, 37, 55 Fry,W. H., 328,379 Fuller, T. C., 240, 246 Fullmer, F. S., 175, 179 Funk, C. R., 132,136

G Gaastra, P., 21, 55 Gale, J., 39, 55 Gammon, N., 332, 378 Gapon, E. N., 155,179 Garcia-Rivera, J., 224, 225, 246 Garcia-Vicente, J., 337, 344, 375, 376, 378 Gard, J. A., 331, 346, 380 Gardner, R., 83, 100 Gardner, W. R., 165, 179 Garg, S. P., 363, 381 Garner, G. B., 215, 222, 223, 225, 229, 231, 232, 233, 235, 236, 240, 241, 242, 243, 244, 245, 246, 247 Garrett, J. D., 116, 136 Garrett, W. G., 343, 378 Garrigus, U. S., 229, 232, 234, 242, 245 Gastuche, M. C., 333, 349, 353, 357, 364, 365, 369, 377, 378 Gates, C. E., 237, 245 Gates, C. T., 42,43, 44, 55 Gauch, H. G., 156, 158, 159, 176, 178, 179, 180 Gehrke, G. W., 208, 212, 215, 243 Geiger, R., 185, 195 Center, C. F., 124, 136 George, A. G., 211,246 George, R. S., 258, 279 Geraldson, C. M., 158, 179 Gerlach, 73, 99 Gessel, S. P., 182, 195 Getsinger, J. G., 94, 99 Gheith, M. A., 334, 378 Gieger, M., 222, 244 Gieseking, J. E., 328, 378 Gilbert, B. E., 75,99 Gilbert, C. S., 203, 213, 214, 215, 216, 219, 220, 235, 239, 243 Gill, W. R., 181, 183, 195

Gilliam, J. W., 71, 99 Gilmore, E. C., Jr., 36, 55 Gitter, A., 209, 217, 243 Glenn, M. W., 235,243 Glenn, R. C., 330,338,367,371,378 Goertzen, J. O., 355, 356, 376 Gol'tsberg, I. A,, 2, 3, 55 Goodrich, R. D., 232,243 Gorbunov, N. I., 333, 337, 340, 341, 347, 359, 369, 373, 378 Gorbunova, Z. N., 334, 378 Gordon, N. T., 17, 55 Goss, J. A,, 173, 180 Gough, N. A., 84, 98 Gradusov, B. P., 332, 333, 334, 344, 378 Gradwell, M., 336, 355, 360, 361, 362, 376, 378 Grafius, J. E., 29, 33, 55 Grant, V., 285, 325 Greene, I., 222, 243 Greene-Kelly, R., 363, 378 Greenland, D. J., 355, 356, 376, 378 Gregory, F. G., 18, 32, 55,56 Griffing, B., 30,56, 318, 325 Grigsby, B. H., 217, 218, 243 Grillot, G., 156, 179 Grim, R. E., 371,378 Grissom, P. H., 190, 196 Grobman, A., 132, 136 Grogan, C. O., 118,136, 137 Grogan, R. G., 158, 179 Groth, S. H., 199,244 Grunes, D. D., 52,58 Gul, A., 203, 204, 243 Gurevich, T. V., 5, 7, 10,55 Gustaffson, Y., 365, 380 Guthrie, W. D., 126, 136

H Hagan, R. M., 37, 42, 43, 55, 57, 181, 185, 195, 196 Hageman, R. H., 203, 204, 206,209,217, 243, 244, 247 Haise, H. R., 156, 178, 181, 195 Haldane, J. B. S., 288, 325 Hale, W. H., 232, 234, 243 Hanan, J. J., 262, 265, 278 Hanks, R. J., 274, 278 Hanna, R. M., 332, 382 Hansche, P. E., 288, 289, 290, 291, 294, 295, 304, 324

390

AUTHOR INDEX

Hansel, W., 225, 226, 229, 232, 234, 236, Hill, R. A., 171, 179 Hill, W. L., 63, 65, 78, 99 242, 244 Hills, F. J., 211, 246 Hansen, V. E., 172, 179 Hanway, J. J., 204, 214, 216, 219, 240, Hines, H. J. G., 215,247 Hinkle, D. A., 116, 136 243 Hirst, J. M., 6, 55 Harada, M., 336, 378 Hislop, J. F., 358, 379 Hardesty, J. O., 75, 100 Hoener, I. R., 203,212, 243 Hardin, L. J., 94, 99 Hoffman, J. I., 96, 99 Harker, K. W., 206, 207, 243 Hoffman, W. M., 70,72,98 Harlan, H. V., 282, 296, 325 Hoffmeister, G., 96, 99 Harlan, J. R., 313, 325 Hoflund, S., 224,229,245 Harley, C. P., 158, 179 Holdridge, D. A., 346,379 Harrassowitz, H., 335, 337, 378 Hole, F. D., 330, 338, 367, 371, 378 Hams, F. J., 95, 99 Holland, J. P., 162,179 Harris, F. S., 164, 179 Holley, K. T., 208, 243 Hartmann, A. F., 238, 242 Holmes, R. M., 7, 35,55,57 Harvey, P. H., 130, 136 Holst, W. O., 232, 243 Harvey, w. A., 240, 246 Holtenius, P., 226,229,233, 243 Hanvard, M. E., 339, 340,378 Holzman, B., 24, 57 Hashimoto, I., 339, 358, 378 Honjo, Y., 337, 350, 379 Haskins, H. D., 91, 99 Hatfield, E. E., 229, 232, 234, 242, 245 Hopkins, A. D., 9, 55 Hosking, J. S., 337, 372, 379 Hattori, T., 368, 375, 378 Houldsworth, H. S., 346, 379 Hauser, G. H., 239, 245 Howard, F. D., 187, 195 Hauth, W. E., 328, 379 Howell, R. W., 26,29,55 Hayes, H. K., 123, 126,136 Howes, C. C., 91, 99 Hayman, B. I., 285, 325 Hayward, H. E., 140, 146, 156, 157, 158, Hoyos, A., 330, 334, 379 159, 161, 172, 178, 179, 180 Hubbell, D. S., 190,196 Headley, F. B., 140, 180 Hubbert, F. J., 232,234,243 Heath, 0. V. S., 18,55, 189,195 Huber, L. L., 105, 106, 108, 125, 126, Heinonen, R., 185, 195 136, 137 Helson, V. A., 26, 55 Hueper, W. C., 226,243 Helwig, D. M., 226, 243 Huggins, W. C., 208,213,244 Henderson, D. W., 190, 195 Humbert, R. P., 183,196 Henderson, M. E. K., 372,379 Hutchins, L. M., 258, 278 Hendricks, S. B., 328, 355, 377, 379, 381 Hvidsten, H., 211, 243 Hendrickson, A. H., 182, 185, 191, 196 I Hknin, S., 369, 379 Hermans, M. E. A., 354,380 Iman, 288, 325 Herrick, J. B., 219, 240, 243 Interrante, L. V., 258, 279 Hersey, J. R., 223, 242 Irving, G. W., Jr., 119,136 Hesketh, J. D., 119, 136 Ishii, J., 336, 379 Hewitt, E. J., 209,212,217,242,243 Ishizuka, Y., 336, 381 Heyrousky, J., 250, 278 Israelson, 0. W., 172, 179 Hiatt, E. P., 222, 243 Iwasa, Y., 370, 379 Hiesey, W. M., 317, 324, 325 J Highkin, H. R., 33, 34,57 Jackson, M. L., 330, 331, 332, 335, 336, Hignett, T. P., 67, 68, 71, 85, 87, 99 337, 338, 339, 340, 343, 353, 358, Hilgard, E. W., 140, 151, 164, 179

AUTHOR INDEX

359, 362, 367, 368, 371, 375, 376, 377, 378, 379, 380, 381, 382 Jacob, K. D., 63, 65, 66, 74, 91, 99, 100 Jacobs, C. B., 91,99 Jacobsen, D., 217, 218,247 Jacobson, H. G. M., 186, 196 Jacobson, W. C., 221, 243 Jain, S. K., 288, 291, 296, 298, 299, 300, 301, 324, 325 Jainudeen, M. R., 226,244 Jamieson, N. D., 226,244 Jamison, V. C., 190,196 Jana, S., 300, 325 Jeffries, C. D., 340, 379 Jennings, E. G., 6, 55 Jensen, A. H., 232,234,235,246 Jensen, G., 261, 278 Jessen, V., 257, 278 Johnson, I. J., 123, 136 Johnson, L., 340, 379 Johnson, M. W., 132,136 Johnson, R. M., 232,244 Johnson, W. C., 203, 209, 212, 213, 217, 242 Johnston, T. D., 204,213,241 Jones, A. C., 349, 376 Jones, D. F., 319, 325 Jones, D. I. H., 204,210, 244 Jones, J. B., 344, 379 Jones, J. W., 296, 324 Jones, L. G., 42,43,55 Jones, T. N., 190,196 Jones, W. W., 161, 179 Jordan, H. A., 221, 232, 242, 244 Jordan, J. E., 68, 69,98 Jordon, H. V., 79,99 Josephson, L. M., 108,136 Joulie, H., 61, 99 Judson, S., 365, 380 Jugenheimer, R. W., 120, 123,136

K Kacker, K. P., 363, 381 Kamau, A. K., 206,207,243 Kamoshita, Y., 370, 379 Kamprath, E. J., 375, 377 Kanno, I., 330, 332, 334, 336, 337, 344, 350, 370, 379 Kardos, L. T., 183, 196 Karsten, K. S., 251,278 Kasanaga, H., 20, 55

391

Kashiwagi, H., 337, 341, 371, 379 Katz, Y. H., 3555 Kaurichev, I. S., 334, 370, 379 Kawaguchi, K., 334, 368, 375, 378, 379 Kearley, E. O., 223,225, 244 Keck, E., 203, 246 Keenan, F. G., 66,99 Kelley, W. P., 140, 149, 163, 164, 165, 172, 179, 328, 379 Kelly, C. F., 2, 54 Kemper, W. D., 39,56 Kendall, G. R., 11,35,54 Kendrick, J. W., 202,244 Kennedy, W. K., 203,209,210,212, 213, 217, 229, 240, 242 Kerlogue, R. H., 354, 376 Kerr, P. F., 329, 335,381 Kersten, J., 217, 218, 247 Ketellapper, H. J., 29, 30, 55 Kidd, F., 18, 54 Kiesselbach, T. A., 109, 123,136 Kikuchi, F., 296,297,324 Kilgore, L., 222, 239,244 Kilmer, V. J., 337,382 Kimball, M. H., 14, 15, 55 Kinbacker, E. J., 38, 46,55 King, H. M., 164, 180 Kinter, E. B., 342, 355, 356, 377, 379 Kirkham, D., 183, 184, 186, 194, 196, 263, 278 Kirsanov, A. T., 359, 379 Klages, M. G., 332, 379 Klintworth, H., 171, 179 Klotz, I. M., 31, 55 Klotz, L. J,, 268, 271,278,279 Knipmeyer, J. W., 204,217,244 Koch, B. A,, 232,234,244 Kohnke, H., 182, 184,195,262, 265, 277 Kolesova, V. A., 349, 379 Kolp, B. J., 203, 204, 243 Kolthoff, I. M., 250,251,278 Kopchenova, E. V., 371, 382 Kowalczyk, T., 235, 245 Kozelnicky, G. M., 133,136 Kramer, H. H., 132,136 Kramer, N. W., 321,325 Kramer, P. J., 37, 55 Krantz, B. A., 204,216,244 Krantz, J. C., Jr., 225,226,245 Krauskopf, K. B., 366, 368, 371,379

392

AUTHOR INDEX

Kretschmer, A. E., Jr., 203, 212, 219, 220, 229, 244 Kristensen, J., 184, 196, 255, 257,278 Kubiena, W. L., 337, 380 Kuhn, P. M., 6, 57 Kulin, G. A., 372, 380 Kulp, J. L., 347, 380 Kulp, M. R., 164, 180 Kun-Huang, H., 373, 380 Kunkel, R., 83, 100 Kunze, G. W., 332,373,381 Kunze, R., 61, 100 Kurabashi, S., 370, 382 Kurnakov, N. S., 347, 380 Kuron, H., 333, 374, 380 Kuwano, Y.,350, 379

L Labanauskas, C. K., 261, 268, 270, 278, 279 Lagenverff, J. V., 162, 179 Lammann, G., 257, 278 Lancaster, J. D., 74, 80, 81, 97, 100 Landsberg, J. W., 226,243 Lang, A., 32, 55 Lang, A. L., 107, 116,136 Langhans, R. W., 262,265, 278 Langridge, J., 30, 31, 55, 56, 318, 325 Lasley, J. F., 216, 221, 230, 242 Lathwell, D. J., 77, 100 Latshaw, W. L., 120, 136 Laude, H. H., 296,325 Lamer, P. J., 349, 350, 380 Laurence, B. M., 172,179 Lawton, K., 78, 99 Lease, E. J., 213, 246 Lee, G. B., 330, 338,367, 371, 378 Leeper, G, W., 214,244 Leet, L. D., 365, 380 Legg, J., 156, 178 Leggett, J. E., 268, 278 Lehman, M., 321, 325 Lehr, J. R., 94, 95, 99 Lemon, E. R., 24,56, 184, 196, 251, 252, 253, 254, 255, 256, 257, 258, 262, 263, 265, 278, 279 Leng, E. R., 109, 124, 136, 209, 243 LeRiche, H. H., 374,380 Lerner, I. M., 317, 325 Letey, J., 184, 196, 252, 253, 254, 255, 256, 257, 258, 259, 260, 261, 263,

264, 265, 266, 268, 269, 273, 274, 278, 279 Levitt, J., 32, 56 Lewis, C. M., 49, 55 Lewis, D., 224,225, 229, 244 Lewis, D. T., 82, 99 Lewis, R. D., 103, 137 Lewontin, R. C., 294, 316, 319, 325 Li, C. C., 294,325 Liebig, G. F., Jr., 140, 179 Liebig, J., 60, 99 Lievan, V. F., 233,243 Lilleland, O., 157, 179 Lindner, R. C., 158, 179 Lindqvist, B., 367, 380 Lingane, J. J., 250, 251, 278 Lingle, J. C., 82, 99, 187, 195 Lippins, B. C., 354, 380 Little, H. N., 220, 221, 237, 245 Livingston, B. E., 7, 56 Lloyd, M. G., 6, 56 Loden, H. D., 321,325 Loesch, P. J., Jr., 118,136 Lomonovich, M. I., 337, 380 Lonnquist, J. M., 110, 129, 137 Loomis, R. S., 21, 56 Love, K. S., 355,377 Lowry, M. W., 208,213,237,244 Lucanus, R., 31, 56 Luck, E., 61, 99 Lund, Z. F., 190, 196 Lundell, G. E. F., 96, 99 Lunt, 0. R., 184, 196, 259, 260, 261, 263, 264, 268, 274, 278, 279 Lutz, J. A., 78, 79, 81, 100 Lutz, J. F., 182,196 Lycklama, J. C., 213, 244 Lyerly, P. J., 140, 178 Lyford, W. H., Jr., 188, 195 Lyon, R. J. P., 350, 351, 380, 382

M McCall, J. T., 219, 240,243 McCalla, T. M., 185, 196 McCloud, D. E., 23,24, 31, 56 McCracken, R. J,, 332, 377 McCutcheon, 0. P., 164, 180 McDonald, M. J., 374, 383 McElroy, L. W., 218, 241 McEntee, K., 225, 226, 228, 229, 232, 234, 235, 236, 242, 244

393

AUTHOR INDEX

MacIntire, W. H., 94, 99 McIntosh, I. C., 220, 224,244 McIntyre, D. S., 372, 380 Mack, A. R., 52,56 McKeague, J. A,, 365, 366, 367, 371, 373. 380 McKee, H. S., 198, 208, 213, 244 MacKenzie, A. J., 204, 216, 244 Mackenzie, R. C., 327, 331, 333, 341, 344, 346, 347, 349, 356, 364, 380 McLean, F. T., 1, 56 MacMasters, M. M., 121, 136 Magistad, 0. C., 159, 172, 179, 180 Makower, B., 355, 380 Mangelsdorf, P. C., 132, 136 Maxi, H. C., 94, 99 Mao, C.-H., 213, 246 Marr, J. C., 164, 180 Marriott, L. F., 203, 246 Marsh, A. W., 164, 178 Marshall, T. J., 184, 196 Martin, A. E., 370, 380 Martin, J. C., 164, I80 Martin, J. D., 239, 245 Martin, J. P., 161, 179,180 Martin, R. T., 332, 355, 380, 381 Martin, W. E., 78, 99 Martin, W. P., 186, 188, 195 Martine, R. M., 239, 245 Martini, M.L., 296, 325 Maskell, E. J., 188, 195 Mather, K., 111, 136, 285, 324 Matsui, T., 330, 332, 337, 344, 367, 380 Matsuo, Y., 334, 379 Mattingly, G. E. C., 78, 83, 99 Mattson, S., 365, 380 Mayo, N. S., 198,212,216,244 Meagher, W. R., 208,214,246 Mehra, 0. P., 340, 358,359, 380 Meijer, C., 17, 57 Meldau, R., 331, 341, 346, 380 Menary, R. C., 183, 187, 195 Merilan, C. P., 236, 246 Merriam, C. H., 9, 56 Mersereau, J. D., 266, 279 Metcalf, W. K., 230, 244 Meyer, B., 332, 381 Meyers, M. T., 108,137 Milford, M. H., 373, 380 Miller, R. D., 183, 195

Milligan, W. O., 346, 382 Millington, R. J., 257, 278 Miravalle, R. J., 321, 325 Mitchell, B. D., 330, 338, 339, 340, 341, 342, 344, 346, 347, 349, 351, 359, 367, 372, 377, 380 Mitchell, K. J., 19, 26, 27, 28, 31, 56, 216, 219, 241 Mitchell, W. A., 366, 372, 380 Moenke, H., 350, 380 Mogen, C. A., 172,179 Mohr, E. C. J., 367, 381 Moldenhauer, W. C., 214, 243 Monsi, M., 19,20,55, 56 Monson, W. G., 217,244 Monteith, J. L., 6, 24,25, 55, 56 Moran, P. A. P., 294, 325 Mori, T., 336, 379 MorIey, F. H. W., 29, 32, 56, 285, 325 Morse, M. D., 211, 246 Mortensen, W. P., 77, 78, 81, 82, 99, 100 Mortland, M. M., 262, 272, 279, 333, 373, 383 Moscon, L., 331, 381 Moss, D. N., 119, 137 Moxon, A. I,., 203, 204, 205, 208, 212, 216, 218, 220, 221, 229, 245, 247 Mnhrer, M. E., 208, 212, 215, 222, 223, 225, 229, 231, 232, 233, 235, 236, 241, 242, 243, 244, 245, 246, 247 Mulder, E. G., 202, 214, 244 Mufioz Taboadela, M., 334,358,381 Murdock, J. T., 240,246 Musgrave, R, B., 119, 136 Myhr, P. J.. 238, 242

N Nad, M., 333, 375 Nadeau, J. C., 368, 376 Nason, A., 202,214, 244 Neilson, M. E., 337, 372, 379 Neiswander, C. R., 108,137 Nelsen, F. M., 354, 381 Nelson, D. L., 233, 245 Nelson, R. A., 355, 381 Nelson, W. L., 83, 96, 99 Neubauer, C., 61,99 Neumann, A. L., 217, 221, 232, 234, 242, 244, 245 Newman, J. E., 2, 4, 6, 8, 9, 10, 12, 56, 57

394

AUTHOR INDEX

Newsom, I. E., 199,244 Nickson, N. M., 344, 379 Niedermeier, R. P., 240,246 Nielsen, K. F., 265,278 Nielson, R. L., 220,224, 244 Nightingale, G. T., 202, 213, 219, 244 Nikiforoff, C. C., 365,366,381 Nishida, K., 41, 56 Norris, W. E., 261,271 Norrish, K., 374, 381 Nowakowski, T. Z., 209,213,245 Nozdrunova, E. M., 334,370,379 Nuttonson, M. Y., 15,56

0 Oades, J. M., 333, 334, 346, 347, 363, 381 Obenshain, S. S., 332, 381 Oberlin, A,, 357, 381 Odelien, M., 211, 243 O'Dell, B. L., 231, 232, 233, 235, 236, 243, 244, 245, 247 Oertli, J. J., 39, 40, 56 Ogata, G., 212, 245 Ogden, D. B., 189, 195 Oldham, F. D., 94,99 Olivieri, R., 337, 377 Olofsson, S., 203, 245 Olsen, S. R., 39, 56, 83, 100 Olson, 0. E., 203, 217, 220, 229, 233, 243, 245, 247 Onikura, Y., 371, 381 Oppenheimer, C. M., 31,56 Orgerson, J. D., 239, 245 Oslage, W., 211, 241 Ostrom, M. E., 339, 381 Oughton, B. M., 357, 376 Overstreet, R., 164, 180 Owen, P. C., 42,44, 56

P Packard, R. Q., 354,376 Page, A. L., 143,178 Palumbo, D. T., 354,382 Parfenova, E. I., 330, 337, 381,383 Parker, F. W., 74, 99 Parker, M. W., 25,56 Parrish, D. B., 232,234,244 Pask, J. A., 35'7,381 Pasquill, F., 24, 56 Patterson, F. L., 47, 57

Peak, J. W., 32,56 Pearson, G. A., 159, 161, 173, 180 Pearson, R. W., 76, 83, 84,99,100 Pember, F. R., 75,99 Pendleton, J. W., 107, 116, 136 Pennington, R. P., 339, 353, 379 Penny, L. H., 110,124,137 Peoples, S. A., 202,244 Perigaud, S., 368, 376 Pesek, J. T., 79,100 Petermann, A., 61, 99 Peters, D. B., 37,56,373,377 Peters, T., 329, 381 Peterson, D. F., Jr., 164, 165, 180 Peterson, H. B., 172,180 Peterson, J. B., 184, 196 Peterson, M. L., 32, 33, 42, 43, 55, 56 Peterson, W. H., 220, 221, 237, 245 Petri, A. H. K., 18, 54 Pfaff, H. L., 103,137 Pfander, W. H., 216, 221, 222, 225, 229, 230, 232, 236, 242, 243, 244, 245 Philip, J. R., 20,54 Phillips, R. E., 183, 184, 186, 196, 263, 278 Pickett, T. A., 208, 243 Pillsbury, A. F., 164, 165, 168,180 Pinnell, E. L., 126,136 Pino, C., 330, 334,379 Poel, L. W., 251,253,254,275,278 Ponomarev, B. P., 9, 56 Potts, J. M., 93, 99 Power, J. F., 52,58 Powers, W. L., 164,180 Prebble, R. E., 373,381 Preusse, H. U., 333, 374, 380 Pugh, D. L., 233,245 Puns, D. N., 32,56

Q Quackenbush, F. W., 127, 136 Quick, J,, 78, 99 Quigley, F. M., 332, 381 Quin, J. I., 199, 213, 220, 224, 226, 229, 234, 242, 245 Quirk, J. P., 355,356, 376, 378

R Rackham, R. L., 269,279 Radar, P., 232, 233, 235, 243 Ramachandran, V. S., 363,381

395

AUTHOR MDEX

Ramacharlu, P. T., 185,196 Rameshwar, S., 126, 136 Raney, F. C., 37,57 Raney, W. A., 181,190,196,258,278 Rao, K. S., 185,196 Raschke, K., 38, 56 Rath, M. R., 225,226,245 Ratner, E. I., 161,180 Rawlings, J. O., 127,137 Rea, H. E., 190,196 Reaumur, R. A. F., 9, 56 Reddy, B. S., 232,245 Reeve, R., 371,376 Reeve, R. C., 156, 164, 165, 166, 170, 172, 173, 178,180 Reynolds, H., 230,246 Rhodes, R. C., 258,279 Rich, C. I., 328, 332, 333, 361, 362, 365, 368, 375, 381 Richards, L. A., 146, 152, 180, 184, 196 Richards, S. J., 185, 196 Richardson, J. P., 360, 377 Richey, F. D., 108, 137 Rider, N. E., 7, 57 Riggs, C. W., 220, 224, 245 Rimington, C., 199, 213, 220, 224, 226, 229, 245 Rine, D., 211, 246 Riquer, J., 331, 381 Rivenbark, W. L., 203, 209, 247 Robbins, W. R., 202, 213, 219, 244 Roberts, R. C., 337, 382 Roberts, W. J., 39, 57 Robertson, G. S., 92, 99 Robertson, G. W., 7, 35, 55, 57 Robertson, R. H. S., 337, 342, 354, 356, 380, 381 Robertson, R. N., 33, 34, 57 Robinson, H. F., 130, 136 Robinson, K., 356, 376 Robinson, W. D., 220,224,244,245 Rode, E. J., 347, 380 Rogers, H, T., 71, 76, 84, 99, 100 Rogers, J. S., 36, 55 Romsdal, S. D., 83, 100 Rosanow, S. N., 92, 100 Rosenberg, N. J., 184, 192, 193, 196 Ross, C. S., 329, 335, 381 Ross, W. H., 66, 74, 75, 91, 100 Rowe, P. R., 317, 325 Ruprecht, F., 330, 381

Russell, E. J., 367, 381 Russell, M. B., 37, 57, 184, 196, 258, 267, 278 Russell, S. R., 38, 57 Russell, W. A., 110, 124, 137 Ryskin, Y. I., 349, 379

s Saeki, T., 19, 20, 21, 56, 57 Saini, A. D., 46, 54 Sakamoto, K., 341, 383 Salhuana, W., 132, 136 Salter, R. M., 75, 100 Salvetat, L. A., 358, 381 Sample, E. C., 78, 83, 98 Sanchez, R. L., 322, 325 Sanchez-Calvo, M. C., 337, 344, 381 Sant, R., 220, 221, 237, 245 Sapiro, M. L., 224, 229, 245 Sasaki, S., 336, 381 Sato, A., 360, 377 Savage, A,, 220,245 Saveson, I. L., 190, 196 Sawhney, B. L., 332, 343, 368, 381 Sawyer, D. T., 258, 279 Sayre, J. D., 114, 137 Scaletti, F. V., 237, 245 Scaletti, J. V., 237, 245 Schaal, L. A., 4, 57 Scheffer, F., 332, 381 Schermerhorn, L. G., 202, 213, 219, 244 Schimper, A. F. W., 206, 245 Schlosser, E. G., 355, 381 Schmehl, W. R., 77, 83, 100 Schneider, B. H., 119, 137 Schnelle, F., 9, 57 Schobinger, U., 351, 378 Schofield, R. K., 336, 377 Schuman, L. M., 237, 244, 245 Schwarte, L. H., 224,245 Schwertmann, U., 333, 335, 368, 369, 381 Scofield, C. S., 140, 142, 171, 180 Scott, N. M., 374, 383 Scott, w. c.,93, 100 Scott, w. o., 47, 57 Seath, L. F., 77, 79, 94, 98, 100 Seatz, L. F., 332, 381 Segnet, E. R., 344, 379 Seif, R. D., 204,217,244 Seo, J., 233, 242

396

AUTHOR INDEX

Sessions, A. C., 211, 245 Setchell, B. P., 226, 229, 243, 245 Sevilla, R., 132, 136 Shank, D. B., 317, 325 Sharina, N. A., 341, 378 Shaw, R. H., 2, 4, 6, 8, 37, 44, 46, 54, 55, 56, 57 Shaw, T. K., 355, 380 Shearin, A. E., 332, 382 Sherman, G. D., 337, 371, 382 Shive, J. W., 209, 211, 245, 246 Shultz, R. K., 164, 180 Siegel, M. R., 68, 69, 94, 98, 99 Siffert, B., 330, 371, 381, 382 Silva, S., 21, 58 Silverberg, J., 96, 99 Simmonds, N. W., 282, 319, 325 Simon, J., 212, 218, 228, 235, 245, 246 Simonson, R. W., 181, 196, 365, 381 Skow, R. K., 251, 278 Skvortsova, K. U.,371, 382 Slack, A. V., 93, 94, 100 Sloane, L. W., 190, 196 Smith, D,, 210, 245 Smith, F. W., 189, 196 Smith, G. E., 208, 212, 215, 243 Smith, G. S., 211, 217, 221, 232, 234, 242, 244, 245 Smith, R. L., 51, 54 Smithson, F., 331, 382 Smydzuk, J., 33, 34, 57 Snyder, F. W., 189, 196 Snyder, R. S., 164, 180 Sokolowski, J. H., 229, 232, 234, 245 Songeon, J., 354, 375 Sorensen, C., 204, 206, 211, 214, 217, 220, 245, 246 Souza Santos, H. L., 348, 382 Souza Santos, P., 346, 382 Sprague, G. F., 319, 325 Spector, W. S., 25, 57 Spencer, W. F., 204, 216, 244 Spicer, S. S., 230, 246 Sprague, G. F., 110, 112, 119, 124, 128, 137 Sprague, V. G., 26, 29, 57 Spurr, W. B., 157, 179 Stahl, A. L., 209, 246 Stahler, L. M., 217, 246 Stanberry, C. O., 77, 100

Stanford, G., 83, 99 Starkweather, F. M., 354, 382 Stasch, A. R., 239, 244 Staten, G., 190, 196 Stebbins, G. L., Jr., 285, 325 Stefanovits, P., 338, 359, 371, 382 Steger, W., 357, 382 Stephen, I., 332, 376 Stephens, H. L., 213, 246 Steward, F. C., 49, 57 Stewart, G. A,, 236, 246 Steyn, D. G., 199, 224, 246 Stickler, F. C., 185, 196 Stillings, B. R., 203, 246 Stinson, H. T., Jr., 119, 137 Stinson, J. M., 93, 94, 100 Stirk, G. B., 373, 381 Stockinger, K. R., 204, 216, 244 Stolzy, L. H., 184, 196, 252, 253, 254, 255, 256, 257, 258, 259, 260, 261, 263, 264, 265, 266, 268, 269, 270, 271, 273, 274, 278, 279 Stone, E. C., 6,57 Stormorken, H., 230, 246 Stout, B. A., 189, 196 Stout, E. N., 199,244 Stout, P. R., 208,214, 246 Stoy, V., 217,246 Streetman, L. J,, 212, 214,247 Stringfield, G. H., 103, 108, 115, 116, 124, 125, 126, 136, 137 Striplin, M. M., 93, 94, 100 Strub, J. H., 12, 13, 54 Sudo, T., 336, 363, 382 Sukhonthasornpa, S., 232, 234, 244 Sullivan, J. T., 26, 29, 57 Summer, M. E., 333, 369, 382 Sund, J. M., 203, 211, 212, 213, 218, 228, 235, 240, 245, 246 Suneson, C. A., 296, 302, 312, 325 Suomi, V. E., 2, 4, 5, 6, 8, 56, 57 Swahn, O., 222, 224, 246 Swanson, A. F., 296, 325 Swanson, C., 157, 179 Swanson, C. L. W., 186, 196 Swarmer, L. R., 164, 178 Swerdlow, M., 342, 379 Swindale, L. D., 334, 360, 366, 367, 371, 377 Szeicz, G., 24, 56

397

AUTHOR INDEX

Szuszkiewicz, T. E., 184, 196, 253, 254, 255, 256, 257, 258, 259, 260, 261, 263, 264, 265, 266, 268, 269, 270, 278, 279

T Takahashi, H., 341, 382 Talling, J. F., 21, 57 Tamm, O., 340, 359, 382 Tamura, T., 331, 332, 337, 343, 371, 382 Tanner, C. B., 5, 40, 41, 57 Targulyan, V. O., 332, 334, 344, 378 Taubenhaus, J. J., 190, 196 Taylor, C. F., 6, 57 Taylor, 0. C., 265, 266, 279 Taylor, R. E., 232, 234, 243 Taylor, R. M., 333, 334, 347, 374, 381, 382 Taylor, S. A., 258, 279 Tchoubar, C., 357, 381 Teakle, L. J. H., 146, 180 Teller, E., 354, 376 Terman, C. L., 68, 75, 76, 77, 78, 79, 81, 82, 83, 84, 85, 94, 95, 96, 98, 99, 100 Terrill, S. W., 232, 234, 235, 246 Thatcher, L. E., 115, 137 Theisen, A. A., 339, 340, 378 Theron, A. A., 198, 246 Thesis, T., 6, 57 Thomas, E. E., 164, 179 Thomas, C. W., 328, 333, 365, 368, 375, 381 Thomas, J. W., 232, 245 Thomas, R. P., 68, 77, 98, 100 Thompson, D. L., 127, 137 Thompson, J. C . , 130, 137 Thorne, D. W., 140, 172, 180 Thorne, J. P., 140, 180 Thornthwaite, C. W., 24, 57 Thorp, F., Jr., 199, 244 Thorp, F. C., 274, 278 Tileston, F. M., 164, 178 Timm, H., 187, 188, 195 Tollersrud, S., 211, 243 Tollett, J. T., 232, 234, 235, 246 Tottingham, W. E., 213, 246 Touilleaux, R., 349, 364, 378, 382 Toussaint, F., 349, 364, 378, 382 Townsend, W. N., 363, 381 Trautman, R. J., 212, 214, 216, 247 Trites, A. F., 347, 380

Troell, E., 341, 382 Trouse, A. C., Jr., 183, 196 Tsen-tuo, C., 373, 380 Tsuchiya, T., 370, 382 Tsyurupa, I. G . , 368, 382 Tucker, C. L., 322, 325 Tucker, J. M., 202, 240, 244, 246 Tuddenham, W. M., 351, 382 Tunik, B. M., 333, 340, 359, 369, 373, 378 Turner, J. H., 321, 325 Tyurin, I. V., 330, 382

U Ullstrup, A. J., 104, 132, 136, 137 Ulrich, A,, 210, 211, 246 Upchurch, R. P., 42, 43, 55 Uytterhoeven, J., 351, 363, 378, 382

V Vaadia, Y., 37, 57 Vallejo-Freire, A., 346, 382 Valoras, N., 258, 261, 263, 264, 268, 278 Vrimos, R., 330, 367, 382 Van Bemmelin, J., 327, 382 van Burg, P. F. J., 78, 82, 100 Van Compernolle, C., 364, 378 Van Der Marel, H. W., 350, 376, 382 Van der Veen, R., 17, 57 van der Vlies, G. S., 331, 381 Van Diest, A., 187, 196, 282, 268, 279 Van Doren, D. M., 265, 267, 272, 273, 274, 278 Van Cundy, S. D., 261, 268, 269, 270, 271, 279 Van Meersche, M., 364, 378 Van Oorschot, J. L. P., 24, 57 Van Rummelen, F. F. F. E., 330, 382 Vasileva, A. A., 341, 382 Vatthauer, R. J., 232, 244 Vaughan, F., 346, 379 Veihmeyer, F. J., 182, 185, 191, 196 Veilvoye, L., 357, 378 Ventskevich, G. Z., 3, 57 Verhagen, A. M. W., 23, 57 Vicenti-Chandler, J., 21, 58 Viets, F. G., 208, 220, 247 Vineyard, M. L., 121, 136 Vlamis, J.. 78, 99 Voisey, P. W., 53, 58

398

AUTHOR INDEX

Volkert, E., 9, 57 Vomicil, J. A., 187, 188, 191, 195, 196

W Wada, K., 343, 360, 361, 365, 370, 376, 382 Wadleigh, C. H., 156, 158, 159, 176, 178, 179, 180, 184, 196 Wagner, P., 61, 100 Walker, G. F., 343, 378 Walker, I. K., 336, 342, 346, 350, 377 Wallace, H. R., 270, 279 Walters, R. J . K., 204, 210, 244 Walton, G., 238, 246 Wang, L. C., 220, 221, 224, 225, 246 Wang, J. Y., 9, 11, 12, 36, 58 Wanntorp, H., 222, 224, 246 Warder, F. G., 212, 214, 215,219, 243 Warren Wilson, J., 20, 21, 58 Watanabe, F. S., 39, 56 Watanabe, Y., 360, 377 Watkins, S. C., 96, 99 Watson, D. J., 19, 20, 58 Watson, S. A., 122, 137 Ways, P., 231, 247 Webb, J. R., 68, 77, 79, 83, 85, 100 Webb, L. J., 203, 246 Webley, D. M., 372, 382 Weed, S. B., 332, 375, 377 Weeks, L. V., 185, 196 Weichenthal, B. A., 232, 234, 246 Weir, A. H., 374, 380 Weiser, H. B., 346, 382 Welch, C. D., 212, 246 Welsch, C. W., 223, 231, 241, 242, 246, 247 Went, F. W., 11, 14, 15, 33, 34, 35, 57, 58 Werner, G. M., 240, 246 West, C., 18, 54 West, S. H., 44, 45, 58 Wey, R., 330, 382 Whelan, M., 224, 247 Whetzal, F. W., 232, 234, 246 White, J. L., 332, 379 White, M. J. D., 294, 325 White, W. A., 335, 382 Whitehead, E. I., 203, 204, 205, 208, 212, 216, 217, 218, 220, 221, 245, 246, 247 Whiteside, E. P., 332, 333, 373, 382, 383

Whitney, R. S., 156, 178 Whittaker, C. W., 75, 100 Whittig, L. D., 331, 334, 337, 339, 353, 375, 379, 382 Whyte, R. D., 33, 58 Wicklund, R. E., 332, 382 Widdowson, F. V., 76, 98 Wiebe, G. A., 312, 325 Wiegand, C. L., 262, 263, 279 Wiersma, D., 184, 196, 262, 272, 279 Wiersum, L. K., 183, 196, 274, 279 Wiggans, S. C., 35, 58 Wiklander, L., 365, 382 Wilbanks, J. A., 93, 100 Wilcox, L. V., 140, 142, 143, 163, 164, 165, 166, 171,180 Wilkinson, G. N., 314, 325 Willard, C. J., 217, 218, 247 Williams, A. J., 229, 245 Williams, C. H., 215, 247 Williams, E. G., 374, 383 Williams, P. P., 336, 342, 346, 350, 377 Williams, R. F., 18, 58 Williams, W. A., 21, 56 Williamson, B. F., 190, 196 Williamson, J. T., 74, 100 Williamson, K. I., 330, 331, 336, 342, 353, 367, 368, 377 Willis, W. O., 52, 58 Willits, N. A., 184, 192, 193, 196 Willrich, T. L., 219, 240, 243 Wilson, J. H., 23, 57 Wilson, J. K., 203, 211, 219, 239, 247 Winneberger, J. H., 38, 58 Winter, A. J., 222, 225, 228, 245, 247 Wintermeyer, A. M., 342, 379 Winters, E., 181, 196 Wiseman, H. G., 221, 243 Withrow, A. P., 16, 17, 58 Withrow, R. B., 16, 17, 58 Witts, K. J., 20, 58 Wolfer, H., 211, 247 Wooten, 0. B., 190, 196 Workman, P. L., 286, 287, 288, 292, 296, 318, 324, 325 Wright, B. C., 74, 80, 81, 97, 100 Wright, B. M., 231, 247 Wright, M. J., 203, 211, 212, 218, 225, 226, 228, 229, 232, 234, 235, 236, 239, 242, 244, 245, 246 Wright, N., 212, 214, 216, 247

399

AUTHOR INDFX

Wright, S., 294, 325 Wurman, E., 333, 373, 383 Wursten, J. L., 164, 180 Wyngaarden, J. B., 231, 247

Y Yaalon, D. H., 332, 366, 376, 383 Yadav, J. S . P., 162, 178 Yadav, K. P., 231, 233, 247 Yamaguchi, G., 341, 383 Yarilova, E. A., 330, 337, 381, 383 Yokoi, H., 337, 371, 379 Yoshinaga, N., 326, 339, 342, 344, 346, 350, 354, 355, 360, 376, 383

Youell, R. F., 357, 376 Younger, V. B., 261, 263, 268, 274, 278 Yuan, T. L., 332, 378

Z Zabelin, V. A., 330, 383 Zacharias, E., 206, 247 Zemlicka, J., 332, 383 Zieserl, J. F., 203, 209, 243, 247 Zimmerman, J. E., 217, 232, 242, 244 Zimmerman, R. P., 183, 196 Zmeskal, O., 332, 378 Zuber, M. S., 118, 136, 137, 208, 212, 215, 243

SUBJECT INDEX A Aeration, 184 Agricultural weather station, 2-7 Agrostis tends, 26 Alfalfa, 31, 47, 161, 164, 229 Alkalization, 147-149 Allophane, 339, 340, 346, 350-351, 353, 355, 359, 360, 361-362 nature and occurrence, 329, 335-338 origin of in soil, 370-372 Almond, 157, 161 A l t e m r i a , 117 Aluminosilicate, 344, 363, 364, 372 Aluminum, 339, 340, 343, 346, 347, 353, 372-375 Aluminum oxides, nature and occurrence in soil, 331-333 origin of in soil, 368-369 Amaranthaceae, 203 Amaranthus hay, 220 Ammonia, 66, 74, 75, 81, 224, 225, 234 Ammonium nitrate, 78 Ammonium phosphate, 68, 71 73, 75, 78 Ammonium polyphosphate, 94 Amorphous materials in soil, 327-383 detection and estimation, 338-364 nature and occurrence, 328-338 origin of, 364-372 physical and chemical property relationships, 372-375 Antitranspirant, 39-41 Aphid, 106, 125-126 Apricots, 157 Arabidopsis thaliana, 30, 316 Avocado, 157, 161, 164 Axyris amuranthoides, 218 Azovskite, 329

B Barley, 75, 159, 164, 185, 192, 194, 262, 268, 285, 298-302, 314, 319 Bean, 39, 42, 158, 159, 164, 272, 274 Beet, 158, 161, 164 Bermudagrass, 159 Bicarbonate, 142-143, 145, 146, 158

Biotin, 30 Birds-foot trefoil, 265 Blackberry, 157 Black walnut, 163 Bluegrass, 12, 26 Blue panicgrass, 214, 216 Boehmite, 331, 374 Bone meal, 63 Boron, 141-142, 145, 147, 157, 162-164, 165-166, 268 Bromegrass, 265

C Calcium, 92, 142, 143, 145, 146, 147, 149, 150, 156, 158, 161, 170, 171, 214, 268, 369 Calcium phosphate, 66-68, 74, 78, 83, 84, 85, 91, 95 Calcium polyphosphate, 96 Carbon dioxide, 21, 23-25, 41-42, 51 Carex f7accu, 275 Carotene, 231-233 Carotenoids, 122, 127 Carrot, 158, 159 Cation-exchange capacity, 340, 360-362 Chenopodiaceae, 203 Chemozen, 330 Chloride, 145, 146, 149, 157-158, 176, 222 Chlorophyll, 17, 31, 49 Citrus, 143, 157, 161, 261, 275 Citrus sinensis, 268, 270 Clay genesis, 365-366 Clover, 19, 161 Compositae, 203 Corn, 12, 13, 26, 31, 36, 37, 40, 44-46, 47, 77, 79, 80, 83, 95, 101-138, 159, 164, 182, 184, 186, 198, 203, 206, 208, 209, 210, 213, 214, 215, 216, 234, 262, 263, 265, 269, 272, 273 Corn borer, 104-106, 108, 126-127 Corn improvement, 101-138 culture, 114-119 cytoplasm, 133 exotic germ plasm, 132 hybrid corn and yield, 103-108

400

401

SUBJECT INDEX

industrial use and nutritive value, 119-

122 offense and defense, 108-114 parent stock, 122-132 tetraploid corn, 133-134 Corn smut, 124-125 Cotton, 75, 77, 80, 161, 164, 171, 189-

190, 204, 206, 216, 264, 273, 321 Cmciferae, 203 Cucumbers, 81 Cyanide, 199 Cytidine, 30 Cytoplasm-genotype, 133

D 2,4-D, 217 Dactylis glomerata, 26 Dallisgrass, 158, 161, 219 Denitrification, 220-221 Dew, 6 Diatraea grandiosella, 138 Diplodia maydis, 116 Douglas fir, 182 DTosophila pseudoobscura, 317

E Ensilage, 220-221 Environment characterization, 2-15 Environment, controlled, 47-54 Evansite, 329

F Fertilizers, see also Phosphate fertilizer Field physiology, 1-58 agricultural weather station, 2-7 controlled environments, 47-54 environment characterization, 2-15 phenological observations, 7-11 radiant energy, 15-25 seasons and zones, 11-15 temperature, 25-36 water and humidity, 37-47 Figs, 164 Furrow-basin, 167-168 Fusarium, 117

G Genetics, see aho Corn improvement see also Plant breeding analyses of measurement characters,

296-302

buffering, 316-320 diversity and stability, 313-316, 320-

323 gene pools, 302-304, 308 muhilocus cases, 307-312 population fitness, 293-296 selective values, 287-293 single gene, 305-307 Gibberella zeae, 116 Gibberellin, 32-33 Gibbsite, 331, 332, 349, 358, 371, 374 Glauconite, 357 Goethite, 333, 334, 347, 358, 374 Graminae, 203 Crape, 157, 164 Grapefruit, 163 Grass, 19, 25, 95 Green bean, 192 Growth analysis, 18-19 Guano, 63 Gypsum, 150, 152, 171

H Halloysite, 343, 360, 371 Hay, 199 Heat unit, 35-36 Helminthosporium turcicum, 104 Hematite, 333, 334 Hemicycliophora arenaria, 271 Hemlock, 182 Herbicides, 2 17-218 Hisingerite, 329 Holcus lanatus, 26, 275 Humidity, 51 Hydroxylamine, 225, 226

I Illite, 347, 350, 360 Imogolite, 344, 348 Iodine, 230-231 Iron, 268, 331, 339, 340, 343, 347, 359,

368, 372-374 Iron oxides, nature and occurrence in soil, 333-335 origin of in soil, 369-370 Irrigation, see Salinity in,

J Iuncetum acutiflori, 275 Iuncetum conglomerati, 275 .hncus acutiflorus, 275

402

SUBJECT INDEX

K Kaolinite, 337, 343, 346, 350, 357, 358, 363, 369, 371 Kliachite, 329

L Ladino clover, 42 Laterite, 331 Latosol, 183, 337 Leaf blight, 104 Lemon, 163 Lepidocrocite, 333, 347, 370, 374 Lettuce, 42, 78, 158, 159 Light, 216-217 Lima bean, 285, 319, 321-322 Lime, 75 Limonite, 329 Lithium, 143 Lolium perenne, 26 Lolium sp., 219 L o h m temulentum, 32

M Magnesium, 142, 143, 146, 147, 149, 156, 268, 334, 368, 369 Maize, 319 Manganese, 214, 268 Mangel, 220 Manure, 212 Meloidogyne incognita, 269 Meloidogyne jauanica, 269, 271, 277 Methemoglobin, 225, 227 Millet, 75, 83 Mimulus, 317 Mineral accumulation, 267-269 Molybdenum, 206, 214 Montmorillonite, 350, 355-356, 358, 363, 369 Montmorillonitic soil, 333 MUCOT,117 Muskmelon, 219

N Napiergrass, 206 Nematodes, 269-271 Nitrate, localization, 204-208 metabolic role in plants, 201-202 Nitrate accumulation and poisoning, 197248 external factors governing accumulation, 210-219

internal factors governing accumulation, 202-210 plant accumulation, 201-219 postharvest losses, 220-221 potential hazards to man, 236-240 toxic agent recognition, 198-201 toxicity to animals, 221-240 Nitrate reductase, 30, 209, 215, 217 Nitrate toxicity, 221-240 ruminants vs. nonruminants, 223-226 types of, 226-236 vascular system involvement, 226 Nitrite, 200, 202, 209, 220, 222, 224, 226, 228, 231, 233, 239 Nitrogen, 59, 66, 75, 93, 210-213, 267 Nitrogen dioxide, 236-237

0 Oats, 26, 35, 46, 75, 78, 79, 80, 83, 91, 164, 199, 203, 204, 211, 213, 215, 216, 220 Onion, 159 Opaline silica, 329, 330, 367 Orange, 163 Orchardgrass, 26 Orthophosphoric acid, 93 Ostrinia nubilalis, 104 Oxygen, 51, 184, 194 Oxygen, see also soil oxygen characterization Oxygen diffusion rate, 252, 272, 276, 277 factors affecting, 255-258 field measurement, 272-275 microorganism response, 269-272 plant response, 258-269

P Paniieum antidotale, 216 Paspalum dilatatum, 26, 219 Pea, 26, 29, 35, 164, 265 Peaches, 157 Peat, 183 Pennisdum purpureum, 206 Perennial ryegrass, 204 PhahrC minor, 214 Phenological observations, 7-11 Phosphate, 329, 373-375 Phosphate fertilizers, 59-100 chemical and physical nature, 66-73

403

SUBJECT INDEX

chemical method status, 60-66 crop response prior 1950, 73-76 nonorthophosphate and other fertilizers, 93-96 recent crop response, 77-93 Phosphorus, 214, 267, 268 determination, 61-66 Photosynthesis, 17, 18-25, 33, 42, 44, 49, 217 Phototropism, 17 Phyoliths, 330, 331, 368 Phytochrome system, 17 Phytophthora citrophthora, 271, 277 Phytophthora parasitica, 271, 277 Pineapple, 42 Plant breeding, 281-327 see also Corn improvement exotic variability, 302-313 genetics of predominantly self-pollinated populations, 282-302 variability, 282-285 variety variability, 313-323 Plant-climate zone, 11-15 Plant growth, 42-47 Plant population versus yield, 106-108 Plums, 157 Podzol, 331, 332, 370 Polarography, 250-253 Pole beans, 81 Potassium, 59, 146, 158, 213-214, 234, 235, 267, 268, 369, 374, 375 Potassium phosphate, 94-96 Potato, 26, 35, 77, 81, 83, 164, 187-188, 274 Potentilla, 317 Protein nitrogen, 18 Pseudomonas aeruginosa, 235 Pteridietum, 275 Pumpkin, 164

Q Quartz, 329, 350, 367

R Radiant energy, 38 characterization, 15-18 light and photosynthesis, 18-25 Radiation, 48-50 Radish, 164 Reclamation of salt land, 164-171

Reed canarygrass, 265 Residual sodium carbonate, 142 Rhodesgrass, 158, 161 Rhopalosiphum maidis, 106, 125 Ribonucleic acid, 44 Rice, 159, 161, 263, 297 Russian pigweed, 218 Ryegrass, 20, 26, 219

S Salination, 147, 151-152 Saline-sodic soils, 150 Saline soils, 149-150, 152-156 Salinity in irrigation, 139-180 effect on crops, 156-164 effect on soil, 146-156 irrigation water, 140-146 management practices, 171-177 reclamation, 164-171 Salt tolerance, 158-164 Silica, 340, 346, 350, 351, 353, 358, 366, 372-374 nature and occurrence in soil, 329-331 origin of in soil, 367-368 Smooth brome, 159 Snapdragon, 219, 265 Sodic soils, 150-151, 152-156, 161, 170 Sodium, 141, 142, 143, 145, 146, 150, 151, 153, 154-158, 161-162, 170, 176, 214, 221, 234, 267, 268, 369 Sodium adsorption ratio, 141, 154-156, 170 Soil compaction, 181-196, 263 causes of, 181-182 effect on productivity, 182-185 experimental difficulties, 191-192 mechanistic study of effects on plant growth, 192-194 plant response to, 185-191 Soil moisture, 7 Soil oxygen characterization, 249-280 biological response to diffusion rates, 258-272 field measurement results, 272-275 platinum microelectrode, problems associated with, 253-258 polarography, 250-253 Solanaceae, 203 Solonchak, 149 Solonetz, 150

404

SUBJECT INDEX

Sorghum, 26, 47, 199, 321 Southwestern stalk borer, 128 Soybean, 12, 13, 26, 29, 219, 220 Spring oats, 12 Stalk rot, 118 Stipa pennata, 330 Strawberry, 157, 275 Subterranean clover, 29, 219 Sudangrass, 78 Sugar beet, 20, 22, 24, 42, 159, 188189, 203, 204, 206, 210, 217, 220, 262, 272 Sugar cane, 44, 145, 183 Sulfate, 145, 146, 158 Sunflower, 164, 264, 274 Superphosphate, 60, 66-71, 73-76, 78, 80, 83-85, 89 Sweet corn, 81

T Temperature, 11-15, 38-40, 50-51, 5253, 185 crop production, 25 heat-unit, 35-36 maximum regimes, 29-31 optimum regimes, 26-29 thermoperiodicity, 33-35 Tetraploid corn, 133-134 Thermoperiodicity, 33 Thielaviopsis basicola, 271-277 Timothy, 204, 265 Tomato, 82, 158, 161, 164, 183, 187, 191, 206, 265, 267, 272 Transpiration, 37-42 Trenching, 168-170

Trichoderma, 117 Trichodorus christiei, 269 Trifolium subterraneum, 219 Triticum spp., 133 Turnip, 164, 220 Tylenchulus semipenetrans, 269, 270, 277

U Ustilago maydis, 124 V Vermiculite, 344, 350 Vernalization, 32-33 Vinyl acetate-maleic acid, 188 Vitamin A, 230-233 Vitamin B, 30

W Wad, 329 Water, 373 deficit and plant growth, 42-47 internal plant status, 37-42 salinity of, 140-146 Water and humidity, 37-47 Western red cedar, 182 Wheat, 26, 46, 75, 79, 80, 83, 164, 185, 192, 194, 274 White clover 26, 216

X Xiphinema americanum, 269

Z Zea mays, 102

INDEX OF CONTRIBUTORS Volumes 1-15 A Adair, C. R., Miller, M. D., and Beachell, H. M.: Rice improvement and culture in the United States, XIV, 61 Adams, J. E., et al.: Cotton, 11, 2 Adams, R. S., Jr.: see Ellis, R., Jr., XIII, 197 Addicott, F. T., and Lynch, R. S.: Defoliation and desiccation, IX, 68 Ahlgren, C. H., and Fuelleman, R. F.: Ladino clover, 11, 208 Alexander, L. T.: see Sivarajasingham, S., et al., XIV, 1 Allaway, W. H.: see Smith, C. D., II, 157 Allison, F. E.: The enigma of soil nitrogen balance sheets, VII, 213 Anderson, A. J.: Molybdenum as a fertilizer, VIII, 164 Andrews, W. B.: Anhydrous ammonia as a nitrogenous fertilizer, VIII, 62 Angus, D. E.: see Russell, M. B., XI, 1 App, B. A.: see Garrison, C. S., XII, 41 Army, T. J.: see McCalla, T. M., XIII, 125 Ameman, H. F.: see Stoeckeler, J. H., XII, 127

Beachell, H. M.: see Adair, C. R., XIV, 61 Bennett, C. A.: see Adams, J. E., et al., 11, 50 Berger, K. C.: Boron in soils and crops, I, 321 Bernard, R. L.: see Johnson, H. W., XIV, 149 Bertrand, A. R.: see Parr, J. F., XII, 311 Blaser, R. E., Skrdla, W. H., and Taylor, T. H.: Forage seed mixtures, factors in compounding, IV, 179 Bohart, G. E.: see Garrison, C. S., XII, 41 Bouldin, D. R.: see Terman, G. L., XIV, 265 Broadbent, F. E.: Soil organic fraction, V, 153 Brown, J. C.: Iron chlorosis in plants, XIII, 329 Burton, G. W.: Grasses for the southeastem states, 111, 197 Burton, G . W.: see a k o Russell, M. B., XI, 1

C

Cady, J. G.: see Sivarajasingham, S., et al., XIV, 1 B Cartter, J. L., and Hartwig, E. E.: The management of soybeans, XIV, 359 Baldridge, J. D.: see Henson, P. R., IX, Cheney, H. B., et al.: Field crop pro122 duction and soil management, VIII, Barker, H. D.: see Adams, J. E., et d., 2 11, 56 Barley, K. P.: The abundance of earth- Chepil, W. S.: see Olson, R. V., et d . , X, 56 worms in agricultural land and their possible significance in agriculture, Chepil, W. S., and Woodruff, N. P.: The physics of wind erosion and XIII, 249 its control, XV, 211 Barrons, K. C.: Vegetation control on Childs, E. C., and Collis-George, N.: industrial lands, IV, 305 Soil water, control, 11, 234 Barrows, H. L., and Kilmer, V. J.: Plant nutrient losses from soils by water Christensen, J. J.: see Culbertson, J. O., et al., VI, 161 erosion, XV, 303 405

406

INDEX OF CONTRLBUTORS

Clark, F. E.: Soil microorganisms and plant roots, I, 242 Cline, M. G.: see McKeague, J. A., XV, 339 Cline, M. G.: see also Sivarajasingham, S., et al., XIV, 1 Coile, T. S.: Soil and growth of forests, IV, 330 Coleman, N. T., Kamprath, E. J., and Weed, S. B.: Liming, X, 475 Coleman, N. T.: see also Mehlich, A., IV, 67 Collis-George, N.: see Childs, E. C., 11, 234 Cook, R. L., and Davis, J. F.: The residual effect of fertilizer, IX, 205 Cook, R. L.: see also Lawton, K., VI,254 Coons, G. H., Owen, F. V., and Stewart, D.: Sugar beet improvement, VII, 90 Cope, W. A.: see Henson, P. R., IX, 142 Cowan, J. R.: Tall fescue, VIII, 283 Crafts, A. S., and Harvey, W. A.: Weed control, I, 289 Culbertson, J. O., et al.: Seed-flax improvement, VI, 144

D Danielson, L. L.: see Ennis, W. B., Jr., et al., XV, 161 Davis, J. F.: see Cook, R. L., IX, 205 Davis, M. B.: see McKibbin, R. R., et al., V, 356 Dawson, J. E.: Organic soils, VIII, 378 Dean, L. A.: Fixation of soil phosphorus, I, 391 DeMent, J. D.: see Martin, J. P., et al., VII, 2 Dexter, S. T.: Evaluation of crop plants for winter hardiness, VIII, 204 Doll, J. P.: see Munson, R. D., XI, 133 Donald, C. M.: Competition among crop and pasture plants, XV, 1 Donald, C. M.: see Stephens, C. G., X, 168 Dungan, G. H., Lang, A. L., and Pendleton, J. W.: Corn plant population in relation to soil productivity, X, 436

Dunham, R. S.: see Culbertson, J. O., et al., VI, 168

E Eaton, F. M.: see Adams, J. E., et al., 11, 11 Edminster, T. W., and Miller, H. F., Jr.: Recent developments in agricultural machinery, XI, 171 Ellis, R., Jr., and Adams, R. S., Jr.: Contamination of soils by petroleum hydrocarbons, XIII, 197 Ennis, W. B., Jr.: Weed control in the southern United States, VII, 252 Ennis, W. B., Jr., et al.: Impact of chemical weed control on farm management practices, XV, 161 Ensminger, L. E., and Pearson, R. W.: Soil nitrogen, 11, 81 Ensminger, L. E.: see also Jordan, H. V., X, 408 Erlanson, C. 0.: see Hodge, W. H., VII, 189

F Fergus, E. N., and Hollowell, E. A.: Red clover, XII, 365 Fitts, J. W., and Nelson, W. L.: Soil tests, VIII, 242 Flor, H. H.: see Culbertson, J. O., et al., VI, 152 Foote, W. H.: see Cheney, H. B., et al., VIII, 2 Fuelleman, R. F.: see Ahlgren, G. H., 11, 208 Fuller, W. H., McAlister, D. F., and Metcalfe, D. S.: Agronomy in the Southwest United States, XII, 197

G Gaines, J. C.: see Adams, J. E., et al., 11, 32 Garrison, C. S.: Technological advances in grass and legume seed production and testing, XII, 41 Geddes, W. F.: see Culbertson, J. O., et al., VI, 145

407

INDEX OF CONTRIBUTORS

Gerretsen, F. C.: see Mulder, E. G., IV, 222 Gieseking, J. E.: Clay minerals in soils, I, 159 Gilbert, F. A.: Copper in nutrition, IV, 147 Goulden, C. H.: see McKibbin, R. R., et al., V, 322 Grunes, D. L.: The effect of nitrogen on the availability of soil and fertilizer phosphorus to plants, XI, 369

H Hafenrichter, A. L.: Grasses and legumes for conservation, X, 350 Hagan, R. M.: see Russell, M. B., XI, 1 Haise, H. R.: see Olson, R. V., et al., X, 47 Hardison, J. R.: see Garrison, C. S., XII, 41 Harlan, J. R.: see Olson, R. V., et al., X, 15 Harmsen, G. W., and VanSchreven, D. A.: Organic nitrogen in soil, VII, 300 Harper, H. J.: Soil conservation in the United States, 111, 265 Harrar, J. G.: International collaboration in agriculture, VI, 95 Hartwig, E. E.: see Cartter, J. L., X N , 359 Harvey, W. A.: see Crafts, A. S., I, 289 Hayward, H. E., and Wadleigh, C. H.: Plant growth on saline and alkali soils, I, 1 Heinrichs, D. H.: Creeping alfalfas, XV, 317 Hemwall, J. B.: The fixation of phosphorus by soils, IX, 95 Henderson, D. W.: see Russell, M. B., XI, 1 Henson, P. R., Baldridge, J. D., and Cope, W. A.: Lespedezas, IX, 114 Heyne, E. G.: see Olson, R. V., et al.,

X, 8

Hide, J. C.: see Olson, R. V., et al., X, 23 Hodge, W. H., and Erlanson, C. 0.: Plant introduction, W,189 Hodgson, J. F.: Chemistry of the micronutrient elements in soils, XV, 119 Hollowell, E. A.: see Fergus, E. N., XII, 365 Home, M. K., Jr.: see Adams, J. E., et al., II, 5 Howell, R. W.: The physiology of the 265 soybean, W , Hurlbut, L. W.: Progress in agricultural engineering, V, 185 Hurlbut, L. W.: see ako Russell, M. B., XI, 1

J Jackson, M. L., and Sherman, G. D.: Chemical weathering of soil minerals, v, 221 Jacob, K. D.: Fertilizer production and technology, XI, 233 Joffe, J. S . : Green manuring viewed by a pedologist, VII, 142 Johnson, H. W., and Bernard, R. L.: Soybean genetics and breeding, XIV, 149 Jones, R. J., and Rogers, H. T.: Fertilizers and fertilizer practices, I, 39 Jordan, H. V., and Ensminger, L. E.: Sulfur in soil fertility, X, 408 Justice, 0. L.: see Garrison, C. S., XII, 41

K Kamprath, E. J.: see Coleman, N. T . , x, 475 Keller, W., and Peterson, M. L.: Irrigated pastures, 11, 351 Kelley, 0. J. : Requirement and availability of soil water, VI, 67 Kempthome, 0.: The contributions of statistics, IX, 177 Kennedy, W. K.: see Musgrave, R. B., 11, 274

408

INDEX OF CONTRIBUTORS

Kilmer, V. J.: see Barrows, H. L., XV, Mathews, 0. R.: see Salmon, S. C., V, 3 303 Meek, W. E.: see Adams, J. E., et al., Kirkham, D., and Kunze, R. J.: Isotopes 11, 40 methods and uses in soil physics Mehlich, A., and Coleman, N. T.: Soil research, XIV, 321 colloid type and mineral nutrition, Klingman, D. L.: see Ennis, W. B., Jr., IV, 67 et al., XV, 161 Meijers, P. C.: Soils and land use in Knott, J. E., and Lorenz, 0. A,: Vegethe Netherlands, XI, 333 table production, 11, 114 Melsted, S. W.: New concepts of manageKnowles, P. F.: Safflower, X, 290 ment of Corn Belt soils, VI, 121 Knox, E. C.: see Cheney, H. B., VIII, 2 Metcalfe, D. S.: see Fuller, W. H., W, 127 Kohnke, H.: Soil water, control, 11, 234 Kramer, P. J.: see Russell, M. B., XI, 1 Miller, H. F., Jr,: see Edminster, T. W., XI, 171 Kunze, R. J.: see Kirkham, D., XIV, 321 Miller, M. D.: see Adair, C. R., XIV, 61 L Morley, F. H. W.: Subterranean clover, Lang, A. L.: see Dungan, G. H., X, 436 XIII, 57 Lawton, K., and Cook, R. L.: Potassium Mortland, M. M.: Reactions of ammonia in plant nutrition, VI, 254 in soils, X, 325 Leukel, R. W.: see Salmon, S. C., V, 3 Muir, A,: Podzol and podzolic soils, XIII, 1 Lorenz, 0. A,: see Knott, J. E., 11, 114 Low, P. F.: Physical chemistry of clay- Mulder, E. C., and Cerretsen, F. C.: water interaction, XIII, 269 Soil manganese and plant growth, IV, 222 Lynch, R. S.: see Addicott, F. T., IX, 68 Munson, R. D., and Doll, J. P.: The M economics of fertilizer use in crop McAlister, D. F.: see Fuller, W. H., XIZ, production, XI, 133 197 Musgrave, R. B., and Kennedy, W. K.: McCalla, T. M., and Army, T. J.: StubPreservation and storage of forage ble mulch farming, XIII, 125 crops, 11, 274 McIlvain, E. H., and Savage, D. A.: N Range improvement, VI, 2 McKeague, J. A., and Cline, M. G., Neal, 0. R.: Soil management for conservation and productivity, V, 383 Silica in soils, XV, 339 McKibbin, R. R., et al.: Agronomy and Nelson, L. B.: Mineral nutrition of corn, VIII, 321 horticulture in Canada, V, 320 McMillan, J. A,: Agriculture in England Nelson, W. L., and Stanford, C.: Plant and Wales, 111, 165 nutrient behavior and fertilizer use, X, 68 MacRae, N. A.: see McKibbin, R. R., et al., V, 375 Nelson, W. L.: see also Fitts, J. W., VIII, 242 Martin, J. H.: see Quinby, J. R., VI, 305 Martin, J. P., et tal.: Soil aggregation, Nikitin, A. A.: Technological aspects of trace element usage, VI, 183 VII, 2 Martin, W. P.: see Martin, J. P., et al., Norman, A. G.: The place of microbiology in soil science, VII, 399 VII, 2

409

INDEX OF CONTRIBUTORS

0 Ohlrogge, A. J.: Mineral nutrition of soybeans, XII, 229 Olson, R. V., et al.: Trends and problems in the Great Plains, X, 3 Owen, F. V.: see Coons, G . H., VII, 90

P Page, J. B.: see Martin, J. P., et al., VII, 2 Parks, R. 0.:Irrigated agriculture and soil research, 111, 323 Parr, J. F., and Bertrand, A. R.: Water infiltration into soils, XII, 311 Pearson, P. B.: see Tolbert, N. E., IV, 279 Pearson, R. W., and Yeager, J. H.: Agricultural trends in the Old Cotton Belt, IX, 1 Pearson, R. W.: see also Ensminger, L. E., 11, 81 Pendleton, J. W.: see Dungan, G. H., X, 436 Peterson, M. L.: see Keller, W., 11, 351 Presley, J. T.: see Adams, J. E., et al., II, 26

Ripley, P. 0.: see McKibbin, R. R., et al., v, 339 Rochow, W. F.: Barley yellow dwarf virus disease of small grains, XIII, 217 Rodenhiser, H. A.: see Stakman, E. C., X, 143 Rogers, H. T.: see Jones, R. J., I, 39 Russell, M. B.: Water and its relation to soils and crops, XI, 1

S Salmon, S. C., Mathews, 0. R., and Leukel, R. W.: A half century of wheat improvement, V, 3 Savage, D. A.: see McIlvain, E. H., VI, 2 Shaw, R. H.: see Wilsie, C. P., VI, 199 Shaw, W. C.: see Ennis, W. B., Jr., et nl., xv, 181 Sherman, G. D.: see Jackson, M. L. Simonson, R. W.: see Winters, E., 111, 2 Sivarajasingham, S., et al.: Laterite, XIV, 1

Skrdla, W. H.: see Blaser, R. E., IV, 179 Smith, D. C.: Progress in grass breeding, VIII, 128 Smith, D. D., and Wischmeier, W. H.: Rainfall erosion, XIV, 109 Smith, G. D., Allaway, W. H., and Quinby, J. R., and Martin, J. H.: SorRiecken, F. F.: Prairie soils, II, 157 ghum improvement, VI, 305 Smith, C . D.: see aZso Tavernier, R., IX, 217 R Smith, H. P.: see Adams, J. E., et al., 11, 40 Rampton, H. H.: see Cheney, H. B., Smith, 0.: Potato production, I, 353 et al., VIII, 2 Raney, W. A,: see Martin, J. P., et al., Stakman, E. C., and Rodenhiser, H. A,: Race 15B of wheat stem rust-what VII, 2 it is and what it means, X, 143 Reitemeier, R. F.: Soil potassium, chemisStanford, G.: see Nelson, W. L., X, 68 try, 111, 113 Rhoades, H. F.: see Olson, R. V., et al., Stephens, C. G., and Donald, C. M.: Australian soils, X, 168 X, 36 Rich, C. I., and Thomas, G. W.: Clay Stevenson, T. M.: see McKibbin, R. R., et al., V, 352 fraction of soils, XII, 1 Richmond, T. R.: see Adams, J. E., et al., Stewart, D.: see Coons, G. H., VII, 90 11, 63 Stoeckeler, J. H., and Arneman, H. F.: Fertilizers in forestrv. Riecken, F. F.: see Smith, G . D., 11, 157 ,. XII. 127

Q

~

~~

I

410

INDEX OF CONTRIBUTORS

T

W

Tavernier, R., and Smith, G. D.: Braun- Wadleigh, C. H.: Mineral nutrition of erde (brown forest soil), M,217 plants, VII, 75 Taylor, A. L.: Nematode control in soil, Wadleigh, C. H.: see also Hayward, 111, 243 H. E., I, 1 Taylor, T. H.: see Blaser, R. E., IV, 179 Watson, D. J.: Physiological basis of yield variation, IV, 101 Terman, G. L., Bouldin, D. R., and Webb, J. R.: Evaluation of ferti- Webb, J. R.: see Terman, G. L., XIV, 265 lizers by biological methods, XIV, 265 Weed, S. B.: see Coleman, N. T., X, 475 Thomas, G . W.: see Rich, C. I., XII, 1 Weiss, M. G.: Soybeans, I, 78 "home, W.: Zinc deficiency and its con- White, W. J.: Alfalfa improvement, I, 205 trol, IX, 31 Wiersma, D.: see Russell, M. B., XI, 1 Timmons, F. L.: see Ennis, W. B., Jr., Willard, C. J.: Management of alfalfa et al., XV, 161 meadows, 111, 94 Tolbert, N. E., and Pearson, P. B.: Atomic Wilsie, C. P., and Shaw, R. H.: Crop energy and plant sciences, IV, 279 adaptation and climate, VI, 199 Trumble, H. C.: Grassland agronomy in Winters, E., and Simonson, R. W.: SubAustralia, IV, 3 soil, 111, 2 Wischmeier, W. H.: see Smith, D. D., V XIV, 109 Vaadia, Y.: see Russell, M. B., XI, 1 Woodruff, N. D.: see Chepil, W. S., VanSchreven, D. A.: see Harmsen, G. W., x v , 211 VII, 300 Y Viets, F. G., Jr.: Fertilizers and the Yeager, J. H.: see Pearson, R. W., IX, 1 efficient use of water, XIV, 223 Z Vomocil, J. A,: Soil hulk density and Zimmerman, L. H.: Castorbeans, X, 258 penetrability, M, 159

KEYWORD INDEX Volumes 1-15 A

Clay-water interaction, physical chemistry of, XIII, 269 Clover ladino, 11, 208 red, XII, 365 subterranean, XIII, 57 Coal mine spoils, reclamation of, 11, 318 Copper in nutrition, IV, 147 Corn mineral nutrition of as related to its growth and culture, VIII, 321 plant population in relation to soil productivity, X, 436 Corn Belt soils, new concepts of management of, VI, 121 Cotton, 11, 2 Cotton Belt (Old), agricultural trends in, IX, 1 Creeping alfalfas, XV, 317 crop ( s 1 adaptation and climate, VI, 199 boron in, I, 321 competition among pasture plants and, xv. 1 evaluation for winter hardiness, VIII, 204 field crop production and soil management in Pacific Northwest, VIIJ, 2 production in Sweden, VII, 39 relation of water to soils and, XI, 1

Agriculture engineering, progress in, V, 185 in England and Wales, III, 165 machinery, recent developments in, XI, 171 trends in the Old Cotton Belt, IX, 1 Agronomy and horticulture in Canada, V, 320 in Southwest United States, XU, 197 trends and problems in Great Plains, x, 3 Alfalfa(s) creeping, XV, 317 improvement, I, 205 meadows, management of, 111, 94 Ammonia anhydrous, as a nitrogenous fertilizer, VIII, 62 in soils, reactions of, X, 325 Atomic energy and plant sciences, IV, 279 Australian soils and their responses to fertilizers, X, 168 B Barley yellow dwarf virus disease of small grains, XIII, 217 Boron in soils and crops, I, 321 Braunerde (brown forest soil), concept of, IX, 217

D

C Canada, agronomy and horticulture in, V, 320 Castorbeans: a new oil crop for mechanized production, X, 258 Chemical weathering of soil minerals, v, 221 Chemical weed control, impact on farm management practices, XV, 161 Clay fraction of soils, MI, 1 Qay minerals in soils, I, 159

Defoliation and desiccation: harvest-aid practices, IX, 68 E Earthworms, abundance in land and possible significance in agriculture, XIII, 249 England and Wales, agriculture in, HI, 165 Erosion, see Soil( s ) , Rainfall erosion, Water erosion, and Wind erosion

411

412

KEYWORD INDEX OF TITLES

F Fertilizer( s ) anhydrous ammonia as a, VIII, 62 economics of use in crop production, XI, 133 and efficient use of water, XIV, 223 evaluation by biological methods, XIV, 265 and fertilizer practices, I, 39 in forestry, XII, 127 and lime requirements of soils, VIII, 242 molybdenum as a, VIII, 164 and plant nutrient behavior, changing concepts of, X, 68 production and technology of, XI, 233 residual effect of, IX, 205 responses of Australian soils to, X, 168 Field crop production and soil management in the Pacific Northwest, VIII, 2 Forage crops, preservation and storage, 11. 274 Forage seed mixtures, factors in compounding, IV, 179 Forestry, fertilizers in, XII, 127 Forests, soil and growth of, IV, 330

G Grain, barley yellow dwarf virus disease of, XIII, 217 Grass ( es ) breeding, progress in, VIII, 128 and legume seed testing and production, technical advances in, XII, 41 and legumes for soil and water conservation, X, 350 for southeastern states, 111, 197 Grassland agronomy in Australia, IV, 3 Great Plains, agronomic trends in, X, 3 Green manuring viewed by a pedologist, VII, 142

I International collaboration in agriculture, a pattern for, VI, 95 Iron chlorosis in plants, XIII, 329 Irrigation agriculture and soil research, 111, 323 of pastures, 11, 351

Isotopes methods and uses in soil physics research, XIV, 321

L Ladino clover, 11, 208 Laterite, XIV, 1 Legumes and new grasses for soil and water conservation, X, 350 seed testing and production, XII, 41 Lespedezas, IX, 114 Lime and fertilizer requirements of soils, determination of, VIII, 242 Liming, X, 475

M Manganese, soil, and plant growth, IV, 222 Microbiology in soil science, place of, VII, 399 Micronutrient elements in soils, chemistry of, xv, 119 Microorganisms, soil, and plant roots, I, 242 Mineral nutrition of corn as related to its growth and culture, VIII, 321 of plants as related to microbial activities, VII, 75 and soil colloid type, IV, 67 of soybeans, XII, 229 Minerals, soil, see Soil minerals Molybdenum as a fertilizer, VIII, 164

N Nematode control in soil, 111, 243 Netherlands, soils and land use in, XI, 333 Nitrogen effect on availability of soil and fertilizer phosphorus to plants, XI, 369 mineralization of organic, in soils, VII, 300 in soil, 11, 81 balance sheets of, VII, 213 Nutrition, copper in, IV, 147

0 Organic nitrogen in soil, mineralization of, VII, 300 Organic soils, VIII, 378

KEYWORD INDEX OF TITLES

P Pacific Northwest, soil management and field crop production in, VIII, 2 Petroleum hydrocarbons, contamination of soils by, XIII, 197 Phosphorus fixation by soils, I, 391; IX, 95 soil and fertilizer, effect of nitrogen on availability to plants, XI, 369 Plant introduction as a federal service to agriculture, VII. 189 Plant nutrient( s ) behavior and fertilizer use, changing concepts of, X, 68 losses from soils by water erosion, XV, 303 Plant nutrition, potassium in, VI, 254 Podzol and podzolic soils, XIII, 1 Potassium in plant nutrition, VI, 254 soil, chemistry of, 111, 113 Potato production, I, 353 Prairie soils, 11, 157

R Race 15B of wheat stem rust-what it is and what it means, X, 143 Rainfall erosion, XIV, 109; see also Water erosion Range improvement, progress in, VI, 2 Red clover, XII, 365 Rice improvement and culture in United States, XIV, 61

S Safflower, X, 290 Saline and alkali soils, plant growth on, I, 1 Seed-flax improvement, VI, 144 Silica, in soils, XV, 339 Soil ( s ) aggregation of, VII, 2 Australian, responses to fertilizers, X, 168 horon in, I, 321 braunerde (brown forest soil), concept of, W,217 bulk density and penetrability, measurement of, IX. 159

413

chemistry of micronutrient elements in, xv, 119 clay fraction of, XII, 1 clay minerals in, I, 159 colloid type and mineral nutrition, IV, 67 contamination by petroleum hydrocarbons, XIII, 197 and crops, relation of water to, XI, 1 fertility, role of sulfur in, X, 408 fertilizer and lime requirements of, VIII, 242 and growth of forests, IV, 330 infiltration of water into, XII, 311 and land use in Netherlands, XI, 333 manganese and plant growth, IV, 222 microorganisms and plant roots, I, 242 minerals, chemical weathering of, V, 221 nitrogen, 11, 81 nitrogen balance sheets, enigma of, VII, 213 organic, VIII, 378 organic fraction, V, 153 phosphorus fixation, I, 391 potassium, chemistry, 111, 113 prairie, 11, 157 silica in, XV, 339 water, control, 11, 234 requirement and availability of, VI, 67 Soil conservation grasses and legumes for soil and water conservation, X, 350 in the United States, 111, 265 Soil management for conservation and productivity, V, 383 and field crop production in Pacific Northwest, VIII, 2 Soil science isotopes methods and uses in soil physics research, XIV, 321 place of microbiology in, VII, 399 Sorghum improvement, VI, 305 Southeastern United States, grasses for, 111, 197 Southern United States. weed control in principal crops of, VII, 252

414

KEYWORD INDEX OF TITLES

Southwest United States, agronomy in, XII, 197 Soybean( s), I, 78 genetics and breeding of, XIV, 149 management of, XIV, 359 mineral nutrition of, XII, 229 physiology of, XII, 285 Statistics, contributions of, IX, 177 Stubble mulch farming, XIII, 125 Subsoil, 111, 2 Subterranean clover, XIII, 57 Sugar beet in United States, improvement of, VII, 90 Sulfur, role of in soil fertility, X, 408 Swedish crop production, recent changes in, VII, 39

T Tall fescue, VIII, 283 Trace element usage, technological aspects of, VI, 183

U United States, rice improvement and culture in, XIV, 61 soil conservation in, 111, 265

V Vegetable production, 11, 114 Vegetation control on industrial lands, IV, 305

W Water, see also Irrigation erosion, plant nutrient losses from soils by, XV, 303 fertilizers and the efficient use of, XIV, 223 infiltration into soils, XII, 311 relation to soils and crops, XI, 1 soil, control, 11, 234 requirement and availability of, VI, 67 Water conservation grasses and legumes for soil and water conservation, X, 350 Weathering of soil minerals, V, 221 Weed control, I, 289 chemical, impact on farm management practices, XV, 161 in principal crops of the southern United States, VII, 252 Wheat improvement, a half century of, v, 3 Wind erosion, physics of, and its control, x v , 211 Winter hardiness, evalution of crop plants for, VIII, 204

Y Yield variation, physiological basis of, N, 101

Z Zinc deficiency and its control, IX, 31